breast_cancer_prediction2.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.8"
    },
    "colab": {
      "name": "breast_cancer_prediction2.ipynb",
      "provenance": [],
      "collapsed_sections": [
        "mQQ7iKJY9Hkh",
        "ehY2Sc5h9Hki",
        "qMLchzuU9Hkj",
        "IwxmJhLz9Hkj"
      ],
      "include_colab_link": true
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/pranath/breast_cancer_prediction/blob/master/breast_cancer_prediction2.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3yVGWl3I9Hka"
      },
      "source": [
        "# Breast cancer prediction\n",
        "\n",
        "## 1 Introduction\n",
        "\n",
        "In this project I will look at a dataset of patient data relating to breast cancer, which is available on Kaggle as the [Wisconsin Breast Cancer dataset](https://www.kaggle.com/uciml/breast-cancer-wisconsin-data). \n",
        "\n",
        "The dataset features are computed from a digitized image of a fine needle aspirate (FNA) of a breast mass. They describe characteristics of the cell nuclei present in the image in the 3-dimensional space is that described in: K. P. Bennett and O. L. Mangasarian: \"Robust Linear Programming Discrimination of Two Linearly Inseparable Sets\", Optimization Methods and Software 1, 1992, 23-34.\n",
        "\n",
        "The dataset was released in November 1995 and the original source can be found [here](https://archive.ics.uci.edu/ml/datasets/Breast+Cancer+Wisconsin+%28Diagnostic%29). \n",
        "\n",
        "An example of images of cells that this data comes from of both malignant and benign tumors can be seen below."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "l4Lzlmrs9Hka"
      },
      "source": [
        "![title](img/bc_cells.png)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7SoEpj0H9Hka"
      },
      "source": [
        "Traditionally one of the main ways you could diagnose if a tumor is malignant or benign is to take a biopsy of a breast mass and then visually examine cell nuclei under a microscope. A trained specialist can then decide if there is cancer or not.\n",
        "\n",
        "We can see here in the example above that the image on the left belongs to a patient who was diagnosed with a benign tumor, and the image on the right belong to a patient who was diagnosed with a malignant tumor.\n",
        "\n",
        "### 1.1 Features\n",
        "\n",
        "The dataset we have consists of 32 features which relate to these types of images & cells and are the following:\n",
        "\n",
        "- ID number \n",
        "- Diagnosis (M = malignant, B = benign) \n",
        "\n",
        "Features 3-32 are ten real-valued features are computed for each cell nucleus: \n",
        "\n",
        "- a) radius (mean of distances from center to points on the perimeter) \n",
        "- b) texture (standard deviation of gray-scale values) \n",
        "- c) perimeter \n",
        "- d) area \n",
        "- e) smoothness (local variation in radius lengths) \n",
        "- f) compactness (perimeter^2 / area - 1.0) \n",
        "- g) concavity (severity of concave portions of the contour) \n",
        "- h) concave points (number of concave portions of the contour) \n",
        "- i) symmetry \n",
        "- j) fractal dimension (\"coastline approximation\" - 1)\n",
        "\n",
        "The mean, standard error (se) and \"worst\" or largest (mean of the three largest values) of these features were computed for each image, resulting in 30 features. For instance, field 3 is Mean Radius, field 13 is Radius SE, field 23 is Worst Radius.\n",
        "\n",
        "All feature values are recoded with four significant digits.\n",
        "\n",
        "Missing attribute values: none\n",
        "\n",
        "Class distribution: 357 benign, 212 malignant\n",
        "\n",
        "### 1.2 Further details\n",
        "\n",
        "We can see from the image examples that it is visually possible to distinguish between malignant and benign tumor cells. It is hoped that the data derived from these images of the cells is able to capture the differences between these types of cells.\n",
        "\n",
        "In this project, I will aim to develop a machine learning model that will aim to predict Malignant tumors with the highest accuracy.\n",
        "\n",
        "\n",
        "## 2 Import libraries\n",
        "\n",
        "Let us first import the libraries we will need for analysis and developing our model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "q2cqCphi9Hkb",
        "outputId": "59a5027c-0259-48f1-d8c7-013f177a448a"
      },
      "source": [
        "# Import libraries \n",
        "import numpy as np\n",
        "from numpy import array\n",
        "from numpy import mean\n",
        "from numpy import var\n",
        "from numpy import mean\n",
        "from numpy import median\n",
        "from numpy import percentile\n",
        "from numpy import std\n",
        "from numpy.random import seed\n",
        "from numpy.random import rand\n",
        "from numpy.random import randint\n",
        "from math import sqrt\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "from statsmodels.graphics.gofplots import qqplot \n",
        "from statsmodels.stats.power import TTestIndPower\n",
        "import matplotlib.patches as mpatches\n",
        "from mpl_toolkits.mplot3d import Axes3D  \n",
        "import sklearn\n",
        "from sklearn.manifold import TSNE\n",
        "from sklearn import ensemble\n",
        "from sklearn.preprocessing import MinMaxScaler\n",
        "from sklearn.metrics import fbeta_score, accuracy_score\n",
        "from sklearn import linear_model\n",
        "from sklearn.ensemble import GradientBoostingClassifier\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.model_selection import KFold\n",
        "from sklearn.model_selection import RepeatedKFold\n",
        "from sklearn.model_selection import GridSearchCV\n",
        "from sklearn.metrics import make_scorer, fbeta_score\n",
        "from xgboost import XGBClassifier\n",
        "import seaborn as sns\n",
        "import time\n",
        "from scipy.stats import shapiro\n",
        "from scipy.stats import normaltest\n",
        "from scipy.stats import anderson\n",
        "from scipy.stats import mannwhitneyu\n",
        "from scipy.stats import boxcox\n",
        "from scipy.stats import ttest_ind\n",
        "from umap import UMAP\n",
        "import plotly.express as px\n",
        "\n",
        "\n",
        "\n",
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive', force_remount=True)\n",
        "root_dir = \"/content/gdrive/My Drive/Datasets/\"\n",
        "file = root_dir + \"data.csv\"\n",
        "\n",
        "%matplotlib inline\n",
        "\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')"
      ],
      "execution_count": 131,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/gdrive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EZKk3dhd9Hkb"
      },
      "source": [
        "## 3 Initial data exploration\n",
        "\n",
        "We will now take an initial look at the dataset, its values and structure."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c_IGAJL8vD7U"
      },
      "source": [
        "# Load the Census dataset\n",
        "bcancer_df = pd.read_csv(file)"
      ],
      "execution_count": 132,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 241
        },
        "id": "otTbAVLd9Hkb",
        "outputId": "1df02e5f-1798-4bcd-dbfb-a1fe26efd9f0"
      },
      "source": [
        "# For reproducability of the results\n",
        "np.random.seed(42)\n",
        "# Observe first few rows\n",
        "bcancer_df.head()"
      ],
      "execution_count": 133,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>diagnosis</th>\n",
              "      <th>radius_mean</th>\n",
              "      <th>texture_mean</th>\n",
              "      <th>perimeter_mean</th>\n",
              "      <th>area_mean</th>\n",
              "      <th>smoothness_mean</th>\n",
              "      <th>compactness_mean</th>\n",
              "      <th>concavity_mean</th>\n",
              "      <th>concave points_mean</th>\n",
              "      <th>symmetry_mean</th>\n",
              "      <th>fractal_dimension_mean</th>\n",
              "      <th>radius_se</th>\n",
              "      <th>texture_se</th>\n",
              "      <th>perimeter_se</th>\n",
              "      <th>area_se</th>\n",
              "      <th>smoothness_se</th>\n",
              "      <th>compactness_se</th>\n",
              "      <th>concavity_se</th>\n",
              "      <th>concave points_se</th>\n",
              "      <th>symmetry_se</th>\n",
              "      <th>fractal_dimension_se</th>\n",
              "      <th>radius_worst</th>\n",
              "      <th>texture_worst</th>\n",
              "      <th>perimeter_worst</th>\n",
              "      <th>area_worst</th>\n",
              "      <th>smoothness_worst</th>\n",
              "      <th>compactness_worst</th>\n",
              "      <th>concavity_worst</th>\n",
              "      <th>concave points_worst</th>\n",
              "      <th>symmetry_worst</th>\n",
              "      <th>fractal_dimension_worst</th>\n",
              "      <th>Unnamed: 32</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>842302</td>\n",
              "      <td>M</td>\n",
              "      <td>17.99</td>\n",
              "      <td>10.38</td>\n",
              "      <td>122.80</td>\n",
              "      <td>1001.0</td>\n",
              "      <td>0.11840</td>\n",
              "      <td>0.27760</td>\n",
              "      <td>0.3001</td>\n",
              "      <td>0.14710</td>\n",
              "      <td>0.2419</td>\n",
              "      <td>0.07871</td>\n",
              "      <td>1.0950</td>\n",
              "      <td>0.9053</td>\n",
              "      <td>8.589</td>\n",
              "      <td>153.40</td>\n",
              "      <td>0.006399</td>\n",
              "      <td>0.04904</td>\n",
              "      <td>0.05373</td>\n",
              "      <td>0.01587</td>\n",
              "      <td>0.03003</td>\n",
              "      <td>0.006193</td>\n",
              "      <td>25.38</td>\n",
              "      <td>17.33</td>\n",
              "      <td>184.60</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>0.1622</td>\n",
              "      <td>0.6656</td>\n",
              "      <td>0.7119</td>\n",
              "      <td>0.2654</td>\n",
              "      <td>0.4601</td>\n",
              "      <td>0.11890</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>842517</td>\n",
              "      <td>M</td>\n",
              "      <td>20.57</td>\n",
              "      <td>17.77</td>\n",
              "      <td>132.90</td>\n",
              "      <td>1326.0</td>\n",
              "      <td>0.08474</td>\n",
              "      <td>0.07864</td>\n",
              "      <td>0.0869</td>\n",
              "      <td>0.07017</td>\n",
              "      <td>0.1812</td>\n",
              "      <td>0.05667</td>\n",
              "      <td>0.5435</td>\n",
              "      <td>0.7339</td>\n",
              "      <td>3.398</td>\n",
              "      <td>74.08</td>\n",
              "      <td>0.005225</td>\n",
              "      <td>0.01308</td>\n",
              "      <td>0.01860</td>\n",
              "      <td>0.01340</td>\n",
              "      <td>0.01389</td>\n",
              "      <td>0.003532</td>\n",
              "      <td>24.99</td>\n",
              "      <td>23.41</td>\n",
              "      <td>158.80</td>\n",
              "      <td>1956.0</td>\n",
              "      <td>0.1238</td>\n",
              "      <td>0.1866</td>\n",
              "      <td>0.2416</td>\n",
              "      <td>0.1860</td>\n",
              "      <td>0.2750</td>\n",
              "      <td>0.08902</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>84300903</td>\n",
              "      <td>M</td>\n",
              "      <td>19.69</td>\n",
              "      <td>21.25</td>\n",
              "      <td>130.00</td>\n",
              "      <td>1203.0</td>\n",
              "      <td>0.10960</td>\n",
              "      <td>0.15990</td>\n",
              "      <td>0.1974</td>\n",
              "      <td>0.12790</td>\n",
              "      <td>0.2069</td>\n",
              "      <td>0.05999</td>\n",
              "      <td>0.7456</td>\n",
              "      <td>0.7869</td>\n",
              "      <td>4.585</td>\n",
              "      <td>94.03</td>\n",
              "      <td>0.006150</td>\n",
              "      <td>0.04006</td>\n",
              "      <td>0.03832</td>\n",
              "      <td>0.02058</td>\n",
              "      <td>0.02250</td>\n",
              "      <td>0.004571</td>\n",
              "      <td>23.57</td>\n",
              "      <td>25.53</td>\n",
              "      <td>152.50</td>\n",
              "      <td>1709.0</td>\n",
              "      <td>0.1444</td>\n",
              "      <td>0.4245</td>\n",
              "      <td>0.4504</td>\n",
              "      <td>0.2430</td>\n",
              "      <td>0.3613</td>\n",
              "      <td>0.08758</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>84348301</td>\n",
              "      <td>M</td>\n",
              "      <td>11.42</td>\n",
              "      <td>20.38</td>\n",
              "      <td>77.58</td>\n",
              "      <td>386.1</td>\n",
              "      <td>0.14250</td>\n",
              "      <td>0.28390</td>\n",
              "      <td>0.2414</td>\n",
              "      <td>0.10520</td>\n",
              "      <td>0.2597</td>\n",
              "      <td>0.09744</td>\n",
              "      <td>0.4956</td>\n",
              "      <td>1.1560</td>\n",
              "      <td>3.445</td>\n",
              "      <td>27.23</td>\n",
              "      <td>0.009110</td>\n",
              "      <td>0.07458</td>\n",
              "      <td>0.05661</td>\n",
              "      <td>0.01867</td>\n",
              "      <td>0.05963</td>\n",
              "      <td>0.009208</td>\n",
              "      <td>14.91</td>\n",
              "      <td>26.50</td>\n",
              "      <td>98.87</td>\n",
              "      <td>567.7</td>\n",
              "      <td>0.2098</td>\n",
              "      <td>0.8663</td>\n",
              "      <td>0.6869</td>\n",
              "      <td>0.2575</td>\n",
              "      <td>0.6638</td>\n",
              "      <td>0.17300</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>84358402</td>\n",
              "      <td>M</td>\n",
              "      <td>20.29</td>\n",
              "      <td>14.34</td>\n",
              "      <td>135.10</td>\n",
              "      <td>1297.0</td>\n",
              "      <td>0.10030</td>\n",
              "      <td>0.13280</td>\n",
              "      <td>0.1980</td>\n",
              "      <td>0.10430</td>\n",
              "      <td>0.1809</td>\n",
              "      <td>0.05883</td>\n",
              "      <td>0.7572</td>\n",
              "      <td>0.7813</td>\n",
              "      <td>5.438</td>\n",
              "      <td>94.44</td>\n",
              "      <td>0.011490</td>\n",
              "      <td>0.02461</td>\n",
              "      <td>0.05688</td>\n",
              "      <td>0.01885</td>\n",
              "      <td>0.01756</td>\n",
              "      <td>0.005115</td>\n",
              "      <td>22.54</td>\n",
              "      <td>16.67</td>\n",
              "      <td>152.20</td>\n",
              "      <td>1575.0</td>\n",
              "      <td>0.1374</td>\n",
              "      <td>0.2050</td>\n",
              "      <td>0.4000</td>\n",
              "      <td>0.1625</td>\n",
              "      <td>0.2364</td>\n",
              "      <td>0.07678</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "         id diagnosis  ...  fractal_dimension_worst  Unnamed: 32\n",
              "0    842302         M  ...                  0.11890          NaN\n",
              "1    842517         M  ...                  0.08902          NaN\n",
              "2  84300903         M  ...                  0.08758          NaN\n",
              "3  84348301         M  ...                  0.17300          NaN\n",
              "4  84358402         M  ...                  0.07678          NaN\n",
              "\n",
              "[5 rows x 33 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 133
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PISQjJp39Hkd"
      },
      "source": [
        "Ok we can see there are quite a range of continous values for each of the features. Also note something odd, the final column is called 'Unnamed: 32' and is full of nan's - it's not mentioned in the dataset description and looks like some kind of format error in the dataset csv. Let's note that for later in our cleaning phase.\n",
        "\n",
        "Lets compute some key stats for each of the features to get a feel for them."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 334
        },
        "id": "7fV66Q2b9Hkd",
        "outputId": "87681cbc-7851-4591-c770-433198e6f029"
      },
      "source": [
        "# Compute key stats for each feature\n",
        "bcancer_df.describe()"
      ],
      "execution_count": 134,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>id</th>\n",
              "      <th>radius_mean</th>\n",
              "      <th>texture_mean</th>\n",
              "      <th>perimeter_mean</th>\n",
              "      <th>area_mean</th>\n",
              "      <th>smoothness_mean</th>\n",
              "      <th>compactness_mean</th>\n",
              "      <th>concavity_mean</th>\n",
              "      <th>concave points_mean</th>\n",
              "      <th>symmetry_mean</th>\n",
              "      <th>fractal_dimension_mean</th>\n",
              "      <th>radius_se</th>\n",
              "      <th>texture_se</th>\n",
              "      <th>perimeter_se</th>\n",
              "      <th>area_se</th>\n",
              "      <th>smoothness_se</th>\n",
              "      <th>compactness_se</th>\n",
              "      <th>concavity_se</th>\n",
              "      <th>concave points_se</th>\n",
              "      <th>symmetry_se</th>\n",
              "      <th>fractal_dimension_se</th>\n",
              "      <th>radius_worst</th>\n",
              "      <th>texture_worst</th>\n",
              "      <th>perimeter_worst</th>\n",
              "      <th>area_worst</th>\n",
              "      <th>smoothness_worst</th>\n",
              "      <th>compactness_worst</th>\n",
              "      <th>concavity_worst</th>\n",
              "      <th>concave points_worst</th>\n",
              "      <th>symmetry_worst</th>\n",
              "      <th>fractal_dimension_worst</th>\n",
              "      <th>Unnamed: 32</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>5.690000e+02</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>3.037183e+07</td>\n",
              "      <td>14.127292</td>\n",
              "      <td>19.289649</td>\n",
              "      <td>91.969033</td>\n",
              "      <td>654.889104</td>\n",
              "      <td>0.096360</td>\n",
              "      <td>0.104341</td>\n",
              "      <td>0.088799</td>\n",
              "      <td>0.048919</td>\n",
              "      <td>0.181162</td>\n",
              "      <td>0.062798</td>\n",
              "      <td>0.405172</td>\n",
              "      <td>1.216853</td>\n",
              "      <td>2.866059</td>\n",
              "      <td>40.337079</td>\n",
              "      <td>0.007041</td>\n",
              "      <td>0.025478</td>\n",
              "      <td>0.031894</td>\n",
              "      <td>0.011796</td>\n",
              "      <td>0.020542</td>\n",
              "      <td>0.003795</td>\n",
              "      <td>16.269190</td>\n",
              "      <td>25.677223</td>\n",
              "      <td>107.261213</td>\n",
              "      <td>880.583128</td>\n",
              "      <td>0.132369</td>\n",
              "      <td>0.254265</td>\n",
              "      <td>0.272188</td>\n",
              "      <td>0.114606</td>\n",
              "      <td>0.290076</td>\n",
              "      <td>0.083946</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>1.250206e+08</td>\n",
              "      <td>3.524049</td>\n",
              "      <td>4.301036</td>\n",
              "      <td>24.298981</td>\n",
              "      <td>351.914129</td>\n",
              "      <td>0.014064</td>\n",
              "      <td>0.052813</td>\n",
              "      <td>0.079720</td>\n",
              "      <td>0.038803</td>\n",
              "      <td>0.027414</td>\n",
              "      <td>0.007060</td>\n",
              "      <td>0.277313</td>\n",
              "      <td>0.551648</td>\n",
              "      <td>2.021855</td>\n",
              "      <td>45.491006</td>\n",
              "      <td>0.003003</td>\n",
              "      <td>0.017908</td>\n",
              "      <td>0.030186</td>\n",
              "      <td>0.006170</td>\n",
              "      <td>0.008266</td>\n",
              "      <td>0.002646</td>\n",
              "      <td>4.833242</td>\n",
              "      <td>6.146258</td>\n",
              "      <td>33.602542</td>\n",
              "      <td>569.356993</td>\n",
              "      <td>0.022832</td>\n",
              "      <td>0.157336</td>\n",
              "      <td>0.208624</td>\n",
              "      <td>0.065732</td>\n",
              "      <td>0.061867</td>\n",
              "      <td>0.018061</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>8.670000e+03</td>\n",
              "      <td>6.981000</td>\n",
              "      <td>9.710000</td>\n",
              "      <td>43.790000</td>\n",
              "      <td>143.500000</td>\n",
              "      <td>0.052630</td>\n",
              "      <td>0.019380</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.106000</td>\n",
              "      <td>0.049960</td>\n",
              "      <td>0.111500</td>\n",
              "      <td>0.360200</td>\n",
              "      <td>0.757000</td>\n",
              "      <td>6.802000</td>\n",
              "      <td>0.001713</td>\n",
              "      <td>0.002252</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.007882</td>\n",
              "      <td>0.000895</td>\n",
              "      <td>7.930000</td>\n",
              "      <td>12.020000</td>\n",
              "      <td>50.410000</td>\n",
              "      <td>185.200000</td>\n",
              "      <td>0.071170</td>\n",
              "      <td>0.027290</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.156500</td>\n",
              "      <td>0.055040</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>8.692180e+05</td>\n",
              "      <td>11.700000</td>\n",
              "      <td>16.170000</td>\n",
              "      <td>75.170000</td>\n",
              "      <td>420.300000</td>\n",
              "      <td>0.086370</td>\n",
              "      <td>0.064920</td>\n",
              "      <td>0.029560</td>\n",
              "      <td>0.020310</td>\n",
              "      <td>0.161900</td>\n",
              "      <td>0.057700</td>\n",
              "      <td>0.232400</td>\n",
              "      <td>0.833900</td>\n",
              "      <td>1.606000</td>\n",
              "      <td>17.850000</td>\n",
              "      <td>0.005169</td>\n",
              "      <td>0.013080</td>\n",
              "      <td>0.015090</td>\n",
              "      <td>0.007638</td>\n",
              "      <td>0.015160</td>\n",
              "      <td>0.002248</td>\n",
              "      <td>13.010000</td>\n",
              "      <td>21.080000</td>\n",
              "      <td>84.110000</td>\n",
              "      <td>515.300000</td>\n",
              "      <td>0.116600</td>\n",
              "      <td>0.147200</td>\n",
              "      <td>0.114500</td>\n",
              "      <td>0.064930</td>\n",
              "      <td>0.250400</td>\n",
              "      <td>0.071460</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>9.060240e+05</td>\n",
              "      <td>13.370000</td>\n",
              "      <td>18.840000</td>\n",
              "      <td>86.240000</td>\n",
              "      <td>551.100000</td>\n",
              "      <td>0.095870</td>\n",
              "      <td>0.092630</td>\n",
              "      <td>0.061540</td>\n",
              "      <td>0.033500</td>\n",
              "      <td>0.179200</td>\n",
              "      <td>0.061540</td>\n",
              "      <td>0.324200</td>\n",
              "      <td>1.108000</td>\n",
              "      <td>2.287000</td>\n",
              "      <td>24.530000</td>\n",
              "      <td>0.006380</td>\n",
              "      <td>0.020450</td>\n",
              "      <td>0.025890</td>\n",
              "      <td>0.010930</td>\n",
              "      <td>0.018730</td>\n",
              "      <td>0.003187</td>\n",
              "      <td>14.970000</td>\n",
              "      <td>25.410000</td>\n",
              "      <td>97.660000</td>\n",
              "      <td>686.500000</td>\n",
              "      <td>0.131300</td>\n",
              "      <td>0.211900</td>\n",
              "      <td>0.226700</td>\n",
              "      <td>0.099930</td>\n",
              "      <td>0.282200</td>\n",
              "      <td>0.080040</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>8.813129e+06</td>\n",
              "      <td>15.780000</td>\n",
              "      <td>21.800000</td>\n",
              "      <td>104.100000</td>\n",
              "      <td>782.700000</td>\n",
              "      <td>0.105300</td>\n",
              "      <td>0.130400</td>\n",
              "      <td>0.130700</td>\n",
              "      <td>0.074000</td>\n",
              "      <td>0.195700</td>\n",
              "      <td>0.066120</td>\n",
              "      <td>0.478900</td>\n",
              "      <td>1.474000</td>\n",
              "      <td>3.357000</td>\n",
              "      <td>45.190000</td>\n",
              "      <td>0.008146</td>\n",
              "      <td>0.032450</td>\n",
              "      <td>0.042050</td>\n",
              "      <td>0.014710</td>\n",
              "      <td>0.023480</td>\n",
              "      <td>0.004558</td>\n",
              "      <td>18.790000</td>\n",
              "      <td>29.720000</td>\n",
              "      <td>125.400000</td>\n",
              "      <td>1084.000000</td>\n",
              "      <td>0.146000</td>\n",
              "      <td>0.339100</td>\n",
              "      <td>0.382900</td>\n",
              "      <td>0.161400</td>\n",
              "      <td>0.317900</td>\n",
              "      <td>0.092080</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>9.113205e+08</td>\n",
              "      <td>28.110000</td>\n",
              "      <td>39.280000</td>\n",
              "      <td>188.500000</td>\n",
              "      <td>2501.000000</td>\n",
              "      <td>0.163400</td>\n",
              "      <td>0.345400</td>\n",
              "      <td>0.426800</td>\n",
              "      <td>0.201200</td>\n",
              "      <td>0.304000</td>\n",
              "      <td>0.097440</td>\n",
              "      <td>2.873000</td>\n",
              "      <td>4.885000</td>\n",
              "      <td>21.980000</td>\n",
              "      <td>542.200000</td>\n",
              "      <td>0.031130</td>\n",
              "      <td>0.135400</td>\n",
              "      <td>0.396000</td>\n",
              "      <td>0.052790</td>\n",
              "      <td>0.078950</td>\n",
              "      <td>0.029840</td>\n",
              "      <td>36.040000</td>\n",
              "      <td>49.540000</td>\n",
              "      <td>251.200000</td>\n",
              "      <td>4254.000000</td>\n",
              "      <td>0.222600</td>\n",
              "      <td>1.058000</td>\n",
              "      <td>1.252000</td>\n",
              "      <td>0.291000</td>\n",
              "      <td>0.663800</td>\n",
              "      <td>0.207500</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                 id  radius_mean  ...  fractal_dimension_worst  Unnamed: 32\n",
              "count  5.690000e+02   569.000000  ...               569.000000          0.0\n",
              "mean   3.037183e+07    14.127292  ...                 0.083946          NaN\n",
              "std    1.250206e+08     3.524049  ...                 0.018061          NaN\n",
              "min    8.670000e+03     6.981000  ...                 0.055040          NaN\n",
              "25%    8.692180e+05    11.700000  ...                 0.071460          NaN\n",
              "50%    9.060240e+05    13.370000  ...                 0.080040          NaN\n",
              "75%    8.813129e+06    15.780000  ...                 0.092080          NaN\n",
              "max    9.113205e+08    28.110000  ...                 0.207500          NaN\n",
              "\n",
              "[8 rows x 32 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 134
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tKmfn5Q19Hke"
      },
      "source": [
        "Ok quite a bit of variation in the scale and range of values (note also very variable std. dev.) we will need scaling of features later before we feed the data to our model.\n",
        "\n",
        "Lets now check for completeness of data, which according to the dataset description should not have any missing values."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "f0bsQcKl9Hke",
        "outputId": "5e57fedd-c3ad-4dc4-a308-f0d5f840e983"
      },
      "source": [
        "# Check for nans\n",
        "bcancer_df.info()"
      ],
      "execution_count": 135,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 569 entries, 0 to 568\n",
            "Data columns (total 33 columns):\n",
            " #   Column                   Non-Null Count  Dtype  \n",
            "---  ------                   --------------  -----  \n",
            " 0   id                       569 non-null    int64  \n",
            " 1   diagnosis                569 non-null    object \n",
            " 2   radius_mean              569 non-null    float64\n",
            " 3   texture_mean             569 non-null    float64\n",
            " 4   perimeter_mean           569 non-null    float64\n",
            " 5   area_mean                569 non-null    float64\n",
            " 6   smoothness_mean          569 non-null    float64\n",
            " 7   compactness_mean         569 non-null    float64\n",
            " 8   concavity_mean           569 non-null    float64\n",
            " 9   concave points_mean      569 non-null    float64\n",
            " 10  symmetry_mean            569 non-null    float64\n",
            " 11  fractal_dimension_mean   569 non-null    float64\n",
            " 12  radius_se                569 non-null    float64\n",
            " 13  texture_se               569 non-null    float64\n",
            " 14  perimeter_se             569 non-null    float64\n",
            " 15  area_se                  569 non-null    float64\n",
            " 16  smoothness_se            569 non-null    float64\n",
            " 17  compactness_se           569 non-null    float64\n",
            " 18  concavity_se             569 non-null    float64\n",
            " 19  concave points_se        569 non-null    float64\n",
            " 20  symmetry_se              569 non-null    float64\n",
            " 21  fractal_dimension_se     569 non-null    float64\n",
            " 22  radius_worst             569 non-null    float64\n",
            " 23  texture_worst            569 non-null    float64\n",
            " 24  perimeter_worst          569 non-null    float64\n",
            " 25  area_worst               569 non-null    float64\n",
            " 26  smoothness_worst         569 non-null    float64\n",
            " 27  compactness_worst        569 non-null    float64\n",
            " 28  concavity_worst          569 non-null    float64\n",
            " 29  concave points_worst     569 non-null    float64\n",
            " 30  symmetry_worst           569 non-null    float64\n",
            " 31  fractal_dimension_worst  569 non-null    float64\n",
            " 32  Unnamed: 32              0 non-null      float64\n",
            "dtypes: float64(31), int64(1), object(1)\n",
            "memory usage: 146.8+ KB\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C_yqcVX99Hke"
      },
      "source": [
        "Ok all looks complete (bar our 'Unnamed: 32' column which we will drop later in our cleaning phase).\n",
        "\n",
        "Let's now check the proportions of the dataset between benign & malignant target classes."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "O6-gsJ0a9Hke",
        "outputId": "6afec6b4-1967-493b-c35f-208b8a6fdd7d"
      },
      "source": [
        "bcancer_df['diagnosis'].value_counts()"
      ],
      "execution_count": 136,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "B    357\n",
              "M    212\n",
              "Name: diagnosis, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 136
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fXdH69iV9Hke"
      },
      "source": [
        "That's in line with our expectations from the dataset description, however it is encoded as a string/categorical value and we will need to make that numeric before feeding into our ML model."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KGy2utpP9Hkf"
      },
      "source": [
        "## 4 Clean data\n",
        "\n",
        "So we have just a couple of cleaning steps identified previously, which are to convert the diagnosis field into numeric values and to drop the id and 'Unnamed: 32' columns."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ta8LCH_G9Hkf"
      },
      "source": [
        "# Encode the 'diagnosis' feature to numerical values\n",
        "bcancer_clean_df = bcancer_df.replace(['B', 'M'], [0, 1])"
      ],
      "execution_count": 137,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 241
        },
        "id": "hpTzZJbH9Hkf",
        "outputId": "68af14bb-0289-438d-8510-9156361fada8"
      },
      "source": [
        "# Remove last column \n",
        "bcancer_clean_df = bcancer_clean_df.drop(bcancer_df.columns[len(bcancer_df.columns)-1], axis=1)\n",
        "# Remove id column\n",
        "bcancer_clean_df = bcancer_clean_df.drop(['id'], axis=1)\n",
        "# Show first few rows\n",
        "bcancer_clean_df.head()"
      ],
      "execution_count": 138,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>diagnosis</th>\n",
              "      <th>radius_mean</th>\n",
              "      <th>texture_mean</th>\n",
              "      <th>perimeter_mean</th>\n",
              "      <th>area_mean</th>\n",
              "      <th>smoothness_mean</th>\n",
              "      <th>compactness_mean</th>\n",
              "      <th>concavity_mean</th>\n",
              "      <th>concave points_mean</th>\n",
              "      <th>symmetry_mean</th>\n",
              "      <th>fractal_dimension_mean</th>\n",
              "      <th>radius_se</th>\n",
              "      <th>texture_se</th>\n",
              "      <th>perimeter_se</th>\n",
              "      <th>area_se</th>\n",
              "      <th>smoothness_se</th>\n",
              "      <th>compactness_se</th>\n",
              "      <th>concavity_se</th>\n",
              "      <th>concave points_se</th>\n",
              "      <th>symmetry_se</th>\n",
              "      <th>fractal_dimension_se</th>\n",
              "      <th>radius_worst</th>\n",
              "      <th>texture_worst</th>\n",
              "      <th>perimeter_worst</th>\n",
              "      <th>area_worst</th>\n",
              "      <th>smoothness_worst</th>\n",
              "      <th>compactness_worst</th>\n",
              "      <th>concavity_worst</th>\n",
              "      <th>concave points_worst</th>\n",
              "      <th>symmetry_worst</th>\n",
              "      <th>fractal_dimension_worst</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>1</td>\n",
              "      <td>17.99</td>\n",
              "      <td>10.38</td>\n",
              "      <td>122.80</td>\n",
              "      <td>1001.0</td>\n",
              "      <td>0.11840</td>\n",
              "      <td>0.27760</td>\n",
              "      <td>0.3001</td>\n",
              "      <td>0.14710</td>\n",
              "      <td>0.2419</td>\n",
              "      <td>0.07871</td>\n",
              "      <td>1.0950</td>\n",
              "      <td>0.9053</td>\n",
              "      <td>8.589</td>\n",
              "      <td>153.40</td>\n",
              "      <td>0.006399</td>\n",
              "      <td>0.04904</td>\n",
              "      <td>0.05373</td>\n",
              "      <td>0.01587</td>\n",
              "      <td>0.03003</td>\n",
              "      <td>0.006193</td>\n",
              "      <td>25.38</td>\n",
              "      <td>17.33</td>\n",
              "      <td>184.60</td>\n",
              "      <td>2019.0</td>\n",
              "      <td>0.1622</td>\n",
              "      <td>0.6656</td>\n",
              "      <td>0.7119</td>\n",
              "      <td>0.2654</td>\n",
              "      <td>0.4601</td>\n",
              "      <td>0.11890</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>1</td>\n",
              "      <td>20.57</td>\n",
              "      <td>17.77</td>\n",
              "      <td>132.90</td>\n",
              "      <td>1326.0</td>\n",
              "      <td>0.08474</td>\n",
              "      <td>0.07864</td>\n",
              "      <td>0.0869</td>\n",
              "      <td>0.07017</td>\n",
              "      <td>0.1812</td>\n",
              "      <td>0.05667</td>\n",
              "      <td>0.5435</td>\n",
              "      <td>0.7339</td>\n",
              "      <td>3.398</td>\n",
              "      <td>74.08</td>\n",
              "      <td>0.005225</td>\n",
              "      <td>0.01308</td>\n",
              "      <td>0.01860</td>\n",
              "      <td>0.01340</td>\n",
              "      <td>0.01389</td>\n",
              "      <td>0.003532</td>\n",
              "      <td>24.99</td>\n",
              "      <td>23.41</td>\n",
              "      <td>158.80</td>\n",
              "      <td>1956.0</td>\n",
              "      <td>0.1238</td>\n",
              "      <td>0.1866</td>\n",
              "      <td>0.2416</td>\n",
              "      <td>0.1860</td>\n",
              "      <td>0.2750</td>\n",
              "      <td>0.08902</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>1</td>\n",
              "      <td>19.69</td>\n",
              "      <td>21.25</td>\n",
              "      <td>130.00</td>\n",
              "      <td>1203.0</td>\n",
              "      <td>0.10960</td>\n",
              "      <td>0.15990</td>\n",
              "      <td>0.1974</td>\n",
              "      <td>0.12790</td>\n",
              "      <td>0.2069</td>\n",
              "      <td>0.05999</td>\n",
              "      <td>0.7456</td>\n",
              "      <td>0.7869</td>\n",
              "      <td>4.585</td>\n",
              "      <td>94.03</td>\n",
              "      <td>0.006150</td>\n",
              "      <td>0.04006</td>\n",
              "      <td>0.03832</td>\n",
              "      <td>0.02058</td>\n",
              "      <td>0.02250</td>\n",
              "      <td>0.004571</td>\n",
              "      <td>23.57</td>\n",
              "      <td>25.53</td>\n",
              "      <td>152.50</td>\n",
              "      <td>1709.0</td>\n",
              "      <td>0.1444</td>\n",
              "      <td>0.4245</td>\n",
              "      <td>0.4504</td>\n",
              "      <td>0.2430</td>\n",
              "      <td>0.3613</td>\n",
              "      <td>0.08758</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>1</td>\n",
              "      <td>11.42</td>\n",
              "      <td>20.38</td>\n",
              "      <td>77.58</td>\n",
              "      <td>386.1</td>\n",
              "      <td>0.14250</td>\n",
              "      <td>0.28390</td>\n",
              "      <td>0.2414</td>\n",
              "      <td>0.10520</td>\n",
              "      <td>0.2597</td>\n",
              "      <td>0.09744</td>\n",
              "      <td>0.4956</td>\n",
              "      <td>1.1560</td>\n",
              "      <td>3.445</td>\n",
              "      <td>27.23</td>\n",
              "      <td>0.009110</td>\n",
              "      <td>0.07458</td>\n",
              "      <td>0.05661</td>\n",
              "      <td>0.01867</td>\n",
              "      <td>0.05963</td>\n",
              "      <td>0.009208</td>\n",
              "      <td>14.91</td>\n",
              "      <td>26.50</td>\n",
              "      <td>98.87</td>\n",
              "      <td>567.7</td>\n",
              "      <td>0.2098</td>\n",
              "      <td>0.8663</td>\n",
              "      <td>0.6869</td>\n",
              "      <td>0.2575</td>\n",
              "      <td>0.6638</td>\n",
              "      <td>0.17300</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>1</td>\n",
              "      <td>20.29</td>\n",
              "      <td>14.34</td>\n",
              "      <td>135.10</td>\n",
              "      <td>1297.0</td>\n",
              "      <td>0.10030</td>\n",
              "      <td>0.13280</td>\n",
              "      <td>0.1980</td>\n",
              "      <td>0.10430</td>\n",
              "      <td>0.1809</td>\n",
              "      <td>0.05883</td>\n",
              "      <td>0.7572</td>\n",
              "      <td>0.7813</td>\n",
              "      <td>5.438</td>\n",
              "      <td>94.44</td>\n",
              "      <td>0.011490</td>\n",
              "      <td>0.02461</td>\n",
              "      <td>0.05688</td>\n",
              "      <td>0.01885</td>\n",
              "      <td>0.01756</td>\n",
              "      <td>0.005115</td>\n",
              "      <td>22.54</td>\n",
              "      <td>16.67</td>\n",
              "      <td>152.20</td>\n",
              "      <td>1575.0</td>\n",
              "      <td>0.1374</td>\n",
              "      <td>0.2050</td>\n",
              "      <td>0.4000</td>\n",
              "      <td>0.1625</td>\n",
              "      <td>0.2364</td>\n",
              "      <td>0.07678</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   diagnosis  radius_mean  ...  symmetry_worst  fractal_dimension_worst\n",
              "0          1        17.99  ...          0.4601                  0.11890\n",
              "1          1        20.57  ...          0.2750                  0.08902\n",
              "2          1        19.69  ...          0.3613                  0.08758\n",
              "3          1        11.42  ...          0.6638                  0.17300\n",
              "4          1        20.29  ...          0.2364                  0.07678\n",
              "\n",
              "[5 rows x 31 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 138
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rIVhxtrZ9Hkf"
      },
      "source": [
        "## 5 Further data exploration\n",
        "\n",
        "Ok now we have cleaned up our dataset, we are in a better position to look more closely at it.\n",
        "\n",
        "### 5.1 Compare distribution of values for each feature between benign & malignent tumor sub-groups\n",
        "\n",
        "Let's do a univariate exploration of each of the numerical features, but by dividing the dataset into sub-groups for 'diagnosis' i.e. malignant and benign sub groups, and compare the distributions for each sub groups to see what patterns there are."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9Zlib5Uh9Hkf"
      },
      "source": [
        "# Split dataset into those with malignent and those with benign tumors\n",
        "malignent_df = bcancer_clean_df[bcancer_clean_df['diagnosis']==1]\n",
        "benign_df = bcancer_clean_df[bcancer_clean_df['diagnosis']==0]"
      ],
      "execution_count": 139,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "fxlTQ7Tk9Hkf",
        "outputId": "2ee58a71-ed07-4675-b4f9-f1e3f1a23d1a"
      },
      "source": [
        "# Function for plotting feature distributions to compare benign & malignent sub-groups\n",
        "def compare_features_benign_malignent(features, malignent_df, benign_df, rows, cols, figx, figy):\n",
        "    \n",
        "    # Define constants\n",
        "    bins = 10\n",
        "    fig, axes = plt.subplots(nrows=rows, ncols=cols, figsize=(figx, figy))\n",
        "    fig.subplots_adjust(hspace=0.5)\n",
        "    # For each axes & feature\n",
        "    for index, (ax, feature) in enumerate(zip(axes.flatten(), features)):\n",
        "        \n",
        "        ax.hist(malignent_df[feature], bins, alpha=0.5, label='Malignent')\n",
        "        ax.hist(benign_df[feature], bins, alpha=0.5, label='Benign')\n",
        "        ax.legend(loc='upper right')\n",
        "        ax.set_title(feature + \": malignent v benign tumors\")        \n",
        "\n",
        "# Get feature names (remove first element diagnosis which is target feature)\n",
        "features = bcancer_clean_df.columns\n",
        "features = features[1:]\n",
        "# Plot feature distributions to compare benign & malignent sub-groups\n",
        "compare_features_benign_malignent(features, malignent_df, benign_df, 10, 3, 20, 30)"
      ],
      "execution_count": 140,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABIMAAAaMCAYAAACMwlwLAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdeZwdRbn/8c9MMiHLZJFFSOBiQOCBACFsEXUIERAICC4gEDEIiBdULiBcVIwCwsWfgFe9CLKpRKLIJrgAGm9EhIBEUAIkhEe4gEjCmn0y+/L7o2rCyeRsc5Y52/f9euWVM326q6v7dNfTXVVdXdfb24uIiIiIiIiIiNSG+lJnQEREREREREREBo8qg0REREREREREaogqg0REREREREREaogqg0REREREREREaogqg0REREREREREaogqg0REREREREREaogqg0REREREREREaogqgyqYmd1mZtPj5/klzk7ZMLPpZnZb/HymmR03iOvdL4/lXy9QPu43s2EFSOdcMxtaiDyJyMCZ2TgzOz2P5XUOZ2Bml5jZmfHzj8xs4iCt9xQz2zLHZTfEuDzzMMHMbilAOnkdpyJSfDrfi8PM5pjZEfFzQa6/s1xvzvE9xp9vFyAP+xUonYmDda8mm1JlUJkxsyG5LOfuhxY6L9XA3a9397sGaXXTgZwrgwrF3Y90944CJHUuMGg3krke+yJVbByQz0X3gM5hM6szs5q9LnD309395UFa3SlATpVBheLuy9395AIkle9xOiC1fpyKDJSZDSnl+V4r13cFvP7OxqBeoyfj7k+4+1cLkNREYNAqg2rleMyWWgzLQGyJ/BXwD2CvMMnuBcYDw4CL3f3uOO9lwPHAP4G6hDRed/dtYk+hM939xDj9QeBMYC1wJzCKUAl4vLs/lyQv04GvAV3AJODbwBbACcAy4Gh37zKzqcB/x/T+D/iMu7eY2Y3APsAI4Ifufm1f/oDbgUOBF4BPuHt3mn0yB2gG3gc0AicDFwG7AVe6+41mNibut3FxX/yHuy/ol84lwOvufr2ZvR/4EdAOLARGuPspcV1rgPcDo4GZ7r7IzBqB64BdY3L/4e6PJZsfWBH3c2dsZT7B3T3mYTLwPXc/JP59JHCcu5/Wb7PrzOwG4EDgOeDTcZ+m2tdJ96mZvQzs6u5tZvZfwCeBV+I6Lnf3BzP9Hmb2RWACsNDMngO+Atzm7gck/D63ufvvY1q/jGktBO4AvgGMJBwvL5vZTsDNwNi4bae6+/p4fD4JTAOuMrN9gaPjb/Qzd78Kkdp1ObCnmS0Cfg58j1AWfABoAL7h7r81s98DP3D3+8zse4SyupXsz+E3gLsJZc9RZvZp4GPAZjHdm5JlLpav2xPKyK2B04AzCJXiv3D3i+N8pwGfJ8Szu939m6nK7xiDLgQ6Yrpz3P3ydDsplnk/B44BXgO+CVwR8zTT3Z+I5f93geHAKuAkd3+tXzoPEuLnc2b2eeBLwJvA68C97j4nruuncf+sAT7i7mvNbGfgh8DmwFuEcvqNZPMDh8R9dI+Zve3uBybk4WxgrLtfFv++Eljs7v1b9Lcws/+N+/+2DPs66T6N1x63ufsBZjYq7sOdgUeAYxKuKTL9Hv2P07cIMeirCb/PrsABZHd9cWT8rh64y90vien0P05vIBzjfcfPg4hUqXi+3gO8SDh/HiaUVz1mdhThums48BjwBULZsOHewswO553z/RTgSGArYAfgHMI13KHAX/quT1Okm21cugR4D+HcfxQ4P8k2zSHDtX6cbzb9YpKZvZdQto4kXDOe5u5L47YdEbdtInCpu/80w77N5jr2Y4SycDPCPdhJ7t7cL52XyfH6eyDX+oT7jQ3x3d1PSMjDd4En3X1u/PsO4Bp3f6jfZr/XzBYA7wb+291vSLOvk+7TxHtOM9s65nNL4D7CfdDELH+Py4FJ8Zi6ihCHXnf36/t+nxiPTiG74/ZU3jnero73jBPZ+Hz4EFncF9cCtayUj92Bb7q7xb9Pdvd9gQ8Cl8WWsKnAh4E9gFMJF1bZmgnMd/cpwL6EgiyVycBJwP6EC+qX3H0y4ebikNgF8irCxeI+wF8JAQLgq+6+HzAFSOwGvzXwS3ffnXDcHQxgZpea2TEp8jHM3fcHricUzCfHbZ4dv28FPhrzcDShEE3nprgf9mXTFtlGd58KfB24IE77eszz/sDHgWtTze/u/4r5vNzdp/RVBAG4+9PANgn74pNAst5K7wbuc/dJhN/nCxn2ddJ92iceLwcTjpfPEIIt2SwbK/GWA+9LDDIpbE24INiVcBwf6u7vI+zvM+I8VxMqxCYDLxNusvq0xWP9fwktA5PcfS/gxgzrFal2s4FnYplyFaE19qVYJk0Dvm1mDcC/A9+JF/sHEM61gZzD7wZ+5e57EC7Ct4rr2A843czGp1l2PNAEnEe4Ufk6sCdwqpk1mtkkwoXgAcDewL4WHqdNV37vTeg5sxdwZqyo6OuCPyFFPpa4+57AesJF4LT4f98F4RKgyd33JpTVX061QWa2LaHVdV/Chec+/WZ5IZZRTxEqMiDEh9NjWfZjwo1E0vnd/R7gCeDjiRVB0S8J8abPMcBvkmTzAMI+2hM43EJ3/VT7GlLs0wRfjPncPa5v64TvMi3b/zhNJ9P1xQhCpdpRcb0fNrOmuGzicbof8EZcdi/gbxnWK1INJgOXESqDtgQ+Ea8tzwWmx2v8bkKFAWx6b5FoV2AG8FHgF/HfJMKNsqVJN9u4BPBeYJq7b1IRlCDttb6FR7CSxaTXgENiDDkH+FZCmrsT4soHCY0DxLQWpchDNtexf3b398V98Qjw2VQbNNDr74Fe62eI73cCx8Z8jCDErwVsan9CfNsPON/CY4Sp9jWk2KcJLiZU3u8B/Kvfd5mWnQ3Mi8fUz5N8nyjTcbtdTK+J0Gh/nr3z+Hfi+TCQ++Kqpp5B5WOpuy9J+PtLCZUkE4FtCLXud7t7J7DMzB4eQPqPA3PMrAe4092fTTPvo+6+CsDM1gD3xumLCS0NRghIfzIzCC2Qf4zzfMrMPgsMAf6NULv7NrA6oVb6ybhNuPtFafLRt95nCLXcfXmqi4GmDrgyXih2AzulSsjMxgF1sWKmr6b8qIRZ+i62nyTc1ECoeJsRWzcgtMQOTTN/OvcAH4utINN4J7gkanH3vnR/QWhBnUfqfZ10nyb4AHBPPF6W9zteMi07EKvc/VEAM1sCPBCnL+adADjFY+824GeEm4A+d8b/1xBaiH5kZr8itCyIyDs+TGg9OyX+PRqY4O7/NLNrCK1e+3uaXpcprHP3eQnrODq2+EHozfdewoV3Mr+LLdPPAC+7+4uwoYV0PKEXzPt552a9kRAXniZ1+f2Iu6+I6bxEaAF93t2PTLMNifHi7ZinvpgF8C7gZ2a2A+Ha55UkafTZn3CRuC7m4Xf9vk8s/99rZqMJF7m/juX0EELLbtL506wXd19mZi0WelOOBV5099VJZn3Y3ZfF/N1NuPDtJvm+fo3k+7QzIb0PEFpncfd7zawl4bukv0e67Ugjm+uLZ2MDC2Z2O2HfLmDj4/QZ4LtmdgXhRumvOeZHpJK4uy+CMG4o4bxvJ1wnPhbLnxGEG9sn2PTeItH82IPlGUKjXOJ13PbALmnSTZQ0LsXPv4rXoOlkutZPFZOWANda6H3fTei9lLhtLUCLmdWbWYO7d8Yb/2SyuY7d3szuJFTOjATSjdU60Ovv18n9Wr+/xwi/RyNwOPAHd+9JMt/97r42bvN8YCqh12WyfQ1J9mmSbb44fr4d+M+E75L+Hhm2I5VMx+2ouM2r4/T7Cb/hQjY+HwZyX1zVVBlUPjZceMWua1OBqe7eHi9oN4tf9yYsk/i5Tzcb9/jaDMDdHzKzaYSa2V+a2Vnu/scky0PoDt6nJ+HvHsJFbh3wuLsflriQme1IqOT4gLuvs/CoW1++2/ulmc3zmonr7Z+nIYRa3WGEioZuM2smtbo03yXmLzFvdcAMd1+eOGMsqJPNn85dwP8j1JY/4pmfKe6N/5Lu6355TpWPdNs80N8j6XEVZTpeMmkB8PB4wH6E4HUSoQXqlCyWF6kVdcBn+y5++tmD8DhwqnFo0p3DiTf+dYRu/rdmmadM5XQdcJ27J7ba9nXjTlV+FypeJC57KaExZU4sZ76TJq2Bxot6YFmaG41c4sWxhEfoUo151/9aoC9eJNvX00m+TxMvxishXmw4Tt39HxYeK/4IcI2ZXefuN2dYXqTSpTrvf+3u/544Y+wNkVi299cBECvOU5XdqdJNlDQuxWvldOvfKB+kjyGbxCQz+ybghOvFLdi4kipTeZcqD/3zkVguXU0YtuNBC4MdfyRNegMtT/O51t+Iu/fGCpCjCI97JX3Mm9THUrJ9vVMW+RjoNqf7PTbEEDPbrN93mY7bdBJjyEDui6uaHhMrT2OAlbEiaCqh2z6EZ24/bmYNsat8/+7lEFo7dzezoWb2HkJNM/Hza+5+HXAboWt5rp4DdjCzPWPao2JBMRpYBzTHYHFQHuvIxhjgzXgjcRyhNjip2NLQ25dnshuobD6h6zwAZrZXhvnXEfZBsvUvItRYf47UF/cjzawvuJxAaAlNta+z8SihN9LQ2M2zKdMC/SRuzxvAdnH940h+7KWzyMyOjp8/RZIuq7EVY2zsHfVlwqOGIrWsf5kyH/i8xcFzzWxK/P8gQvfng4AfWOga3n/5bM/h+cBnzWx4TNv6PufoAeDEuE7MbDsz24IBlN8FMoZ3ejedkmHexwmPLDXGcumIdDO7+xpglZkdBhBj9G7pliFNvCDEiE8QLuR/lWKeAy1062+I8z5C6n2djUcJjzD3jWs3MsvlYNNt+Sfh0a2+xyW2TrZQCg7sFrdtaMzTI/1nitdAze4+h/CIXqb4LFINdjWzyWZWRxg/dAGhJ8ghFh6Pwcy26Pucp1TpZhWXCihVTBpDGFeml8FpOBwDvBa3c1aGeQd6/Z3LtX66GHIn8GlCj5g/p5hnhpmNttCz9RBC3Msn/m+IIQn/ZytZDOk7jlINJZJKX/weE+P3DEKvoI0U+L64oqkyqDzNIzyStIQwtspTALEb9B8J3RZvJhTUG3H3f8Z5lhBaPhfHr6YDT5vZk4SbgJ/lmrnYq+VTwPVm9hTwF2And3+K0DX+OcIz/5tcwPVn6ccMyuRWwrO2TxMevXojw/xnALeb2d8JjyOtzTD/pcAEM3vazJ4l89sTfgvMMrMnLTaJ9PNr4DDC75vMm4TumUsI3TKvS7WvM+QDAHdfSAgCSwiD7D1F5m1OdBPwsJndHvPx/ZjGbfH/gTgb+HL8rXaMafU3Grgvbuev2XjMDZGaEx/NeTqWQRcANxDKuadiOXFRrPi5Bvich8EPf0YYUwJyOIfd/X5CGfW4hV6p15Fdb5ZU27CYEIv+HM//OwgVPwMtvzONGZTJdwgVZX3lf7o8v0rYp38Hfk8oQzOVnScBF8TyaxEbjxGRzBxgriV53Duuvxv4l7uvTLH8wpjGYkKX+CfS7OtsXEu40VxCGIuh/5gPKSU5ThcQGoWeJYyrMZC0WgmNML8j7Mc/er8XQ0R7Eo7RRYQxs67Jdh0iFexpwqM4SwkD4d/j7m8Szplfx/P+D4TxtfKSKt1s4lK+6+6Xj1Qx6XrgrFgGpKoU2YilHjMoG5cRyqWFwEsZ8jyg6+8cr/U3xPck3z1KGHNtfprHxv9G2J6/EcYZXJ5n/P8mcIKFx7d2ZmD3G08DI8xskZmdRHhRwC7xuNtjAOkQH5++grAPHiNs28tJZp1Oge6LK11db2+yJ41Eqo+ZjfLwBqs6wuDEj8RWxaqVsM1bEoLLPn3jYIiISHIJZecoQuXGsX3jIVWj2AtnSOyRfCBwmbtPL3G2RCSyhLf/lTovklmtXX/HHkSdsbfvpwmDcJ9S4mxJFjRmkNSSj5rZlwnjVDxOaJmudj+JvZT6XvdZtYFIRKSA/svC+H3DgRuquSIoagT+GCuF2gmPNIuISG5q7fp7IvALMxsCrCS8FU4qgHoG1ag4jkD/gbLedvdDS5EfEREpTxZeWX9Fv8m/d/evliI/IiJSOSy8MOCcfpN/4u5XlyI/IvIOVQaJiIiIiIiIiNQQDSAtIiIiIiIiIlJDSj5mkJmpa5KISAruXlfqPJSa4oSISGqKE4oTIiLppIoTJa8MAnD3UmdBRKTshLEHBRQnRESSUZx4h+KEiMim0sWJsqgMEhGRymVmDcCDwO7A6e5+V3yd6lxgNDDf3S+J834EmA30Aue4++MlybSIiIiISA3TmEEiIpKvLuA44PsJ075CeFtIE7C/mU2Krxy9DDgM+CTwvUHPqYiIiIiIqGeQiBRXT08Py5cvp7Ozs9RZKWsNDQ1MmDCB+vrKq6N3917gtX7dUJuAi+Lne4FpQA/wD3dfB6wzswYzG+7ubYOaYREpK4oT2ankOCEikg/FiewMNE6oMqgCda1fQ09bc07L1g9vZOiosQXOkUhqy5cvZ8yYMYwZM6bUWSlra9euZfny5Wy33XalzkqhjHL31vh5NbADsDmwKmGe1XHa8kHOmyTIJ6ZkQ3FHMlGcyE4VxgkZRLp/kEqmOJGdgcYJVQZVoJ62ZpqXLMhp2cbdm0CFuQyizs5OFdxZGDNmDCtWrCh1NgqpJaHXz1hgJaEiaFzCPH3TpYTyiSnZUNyRTBQnslOFcUIGke4fpJIpTmRnoHFC/UxFRKQYHgaOjJ9nxL+fB3Yxs1Fmtg3QpUfEREREREQGn3oGicigWtPczrqWjrzSGD1yGGMbN0v5/auvvsohhxzC1VdfzeGHHw7AqaeeyrBhw7jhhhs2mX/hwoXMmzePiy66iMsvv5yzzz6b0aNH55XHVJYuXUprayv77LNPUdIvFTO7A9gPaDazqcCVwC1mdh7wgLsvifNdAswnvE3sSyXKroiUMcWJ6owTIiKFUIgYAYoToMogERlk61o6eOjJZXmlMW3vbdMW3gB77LEH8+bN4/DDD+ett96ivb2dYcOGZUx79uzZeeUtk6VLl7Jq1aqqu8h39+OTTJ6RZL7fAL8pfo5EpFIpTlRnnBARKYRCxAhQnABVBolIldpyyy1pbW2lpaWF3/3udxxxxBE88sgj3HzzzfzpT39i3bp1nHjiiZxwwgkbLTdr1iz+53/+h2HDhnHuuefS3d3NTjvtxCuvvMINN9zArFmzmDRpEkuXLmXs2LH84Ac/AOCKK65g8eLFdHV1ceGFFzJ58mSOOuoompqaeOqpp9hzzz2ZPXs2t9xyC83NzTz44IP88Ic/LFqLgYiIpKc4ISIi6VR7nNCYQSJStQ4++GAeeOABHn74YQ488EAATjjhBG655RZuv/12fvrTn9LT05N02TvuuIOmpiZuvvlmJk2atNF3Bx54ILfccgudnZ24O3/+85/p6elh7ty5XHvttVx11VUAtLa2cuKJJ3LbbbexcOFCmpubOfnkk5k5cyZz587VBb5kpWv9GjpWLCvav56O9lJvokjJKE6IiEg61Rwn1DNIRKrWYYcdxhlnnMGOO+5IQ0MDAPfffz/33HMPdXV1vPnmm6xZsybpsi+//DIf/ehHAZg8eTK///3vN3zXV5hPmDCBNWvW8Pzzz7NgwQJmzZoFwPr16wEYMWIEO+ywAwDjx49PuS6RdIr9tq8RO04pWtoi5U5xQkRE0qnmOKHKIBGpWmPHjqWpqWlDLT7ADTfcwH333QfAEUccQW9vb9Jl3/Oe97B48WL23XdfnnnmmZTr6O3tZaedduLggw/m/PPPB6CjIwxqV1dXt8m8DQ0NdHd357VdIiJSGIoTIiKSTjXHiZwrg8zsGmAfwqNmXwUWAnOACcBi4Ivunry/lIjIIDnrrLOA8EYAgKamJmbOnMkuu+zCmDFjUi53/PHHc+655/KnP/2JiRMnbmgJSGb69Ok88cQTG2ryJ0+ezAUXXJB03r333puf//znLF68mG9961s0NjbmumkiIlIAihMiIpJOtcaJulS1WOmY2c7A9e5+iJn9G/Bz4HZgpLtfZWbXAve5+/1ZpNXr7gPOQy3rWLEs50cGGndvYtgW2xY4RyKpvfTSSxu6NsLgvDK4EHp7e+nu7mbo0KHcc889vPDCCykL5ELpv6/MDHevS7NITaj1OJFPmZ+NETtOofXFRUVLX3FHMlGcyJ7iRHK1HieyofsHqWSJZd9gvVq+EMo9TuTaM+gNoNXMhgLjgLeAA4Fvxu/vBQ4CMlYGiUhtGdu4WdEL3kLo6uriM5/5DHV1ddTX1/Od73yn1FkSEakJihMiIpJKpcQIKP84kWtl0Drgn4ADI4GPA5cCq+L3q4HN886diEiJNDQ0cOutt5Y6GyIiUqYUJ0REJJ1yjxO5VgZ9GNgC2BnYmtAT6AVCL6HXgbHAykJkUEREREREqo/GIBURKZ1cK4PqgZXu3mNma4FRwEPAkcBz8f95hcmiFFJvdzcdK5bltGz98EaGjhpb4ByJiIiISK2JY5Du5u4f6DcG6RMJY5AegYadEBEpilwrg/4XOMnMHgaGA5cB9wBzzOwhYCnwu8JkUQqpp6M154FCG3dvAlUGiYiIiEj+NAapiEgJ5VQZ5O7dwKwkXx2fX3ZERERERKQGaAxSEZESyrVnkIhITrrWr6GnrTmvNDI9svjqq69y7LHHsssuu9Da2sr06dM566yzsk5/6dKl/PWvf+Uzn/lMXvkUEZGBU5yoGRqDVEQGrBAxAhQnQJVBIjLIetqaaV6yIK80snlkccqUKdxwww309vby8Y9/nJNPPpkxY8Zklf5uu+3GbrvtllceRQODikhuFCdqhsYgFZEBK0SMAMUJUGWQiFS5zs5Oent7GTp0KFdccQWLFy+mq6uLCy+8kMmTJ3PUUUfR1NTEU089xZ577sns2bNZuHAh8+bN46KLLuK3v/0tP/7xj9l+++1Zt24d5513Hi0tLdx4442MGjWKF198kYsuuoipU6eWelPLigYGFZFKoThRMhqDVEQqQrXGCVUGiUhVWrRoEbNmzWLZsmV86EMf4vHHH6enp4e5c+eycuVKzjnnHObOnUtraysnnngiF154IccccwzNze90O+3u7ua6667jl7/8JUOGDOGYY47Z8F1LSws/+tGPePbZZ7n++ut1kb8pDQwqImVNcaK0NAapiJS7ao8TqgwSkarU162zp6eHs88+m8WLF7NgwQJmzQrXnevXrwdgxIgR7LDDDgCMHz+eNWvWbEhj5cqVbL311owYMQJgo66eu+66K3V1dYwfP57Vq1cP1mZVEg0MKiJlTXFCRETSqfY4ocogEalq9fX1jB49GjOjra2N888/H4COjg4A6urqNpq/t7d3w+fNN9+cN954g7a2NoYMGcJzzz234bv+y8kmNDCoiFQExQkREUmnWuOEKoNEZFDVD28MA7blmUYmfd06u7q6mDBhAgcddNCGaQCTJ0/mggsuSJvGkCFDOOOMMzjxxBPZdttt2XLLLWloaMgr7zVEA4OKSE4UJ0REJJVCxIi+dDKp9jhRl1hrVQpm1uvuJc1DpelYsSznEdRH7DiF1hcX5bRs4+5NDNti25yWldr10ksvbeg2WYk6OztpaGigo6ODY489lltvvZXRo0cXZV3995WZ4e4V2bRsZkMIbw6bSBgY9PvEgUGBbQgDg34+m7eJ1XqcyKfMz0Y+cSEbih2SieJE9qopThRSrceJbOQTS1SOS6kpTmRvIHFCPYNERNK46667uP/++2lubub4448vWsFdbTQwqIjUCsUJqXa93d10rFiW07L1wxsZmuH13SLVrlzjhCqDRETSmDlzJjNnzix1NkQqSntHF+2d3QD0tnTQ2r2u6OscPXIYYxs3K/p6RPpTnJBq19PRmteTBagySGpcucYJVQaJiIhIQbV3dvPKa6ECaNxWzSxe0V70dU7be1tVBomIiIhkqb7UGRCR6tbQ0MDatWtLnY2yt3bt2rIZTE5EZDApTmRHcUJEapXiRHYGGifUM0hEimrChAksX76cFStWlDorZa2hoYEJEyaUOhsiIoNOcSI7ihMiUqsUJ7Iz0DihyiARKar6+nq22267UmdDRETKlOKEiIikozhRHHpMTERERERERESkhqgySERERERERESkhuT8mJiZTQX+C2gA7gduBuYCo4H57n5JITIoIiIiIiIiIiKFk1PPIDPbDLgY+Ji7f8jdrwK+AvzE3ZuA/c1sUgHzKSIiIiIiIiIiBZBrz6ADgFbgLjMbCvwn0ARcFL+/F5gGPJt3DkVEREREREREpGByrQyaAEwC9ge2B24CRrl7a/x+NbBD/tkTEREREZFqpGEnRERKJ9cBpFcBj7j7endfCowFWsxsePx+LLCyEBkUEREREZHqomEnRERKK9fKoIWAmVm9mW0DtAEPA0fG72fEv0VERERERPpLHHbiD2Y2mTDsxL3x+75hJ0REpAhyekzM3VeZ2Rzgz4RunecDS4FbzOw84AF3X1KwXIqIiIiISDXRsBMiIiWU86vl3f0nwE/6TZ6RX3ZERKRaaCwIERFJY8OwE8BSMxsLrDez4e7ehoadEBEpqlwfExMREUlJY0GIiEgGGnZCRKSEVBkkIiLFoLEgREQkJXdfBcwhDDvxK8KwE1cCnzOzBcBTGnZCRKR4cn5MTHLXtX4NPW3NOS/f09FewNyIiBSFxoIQEZG0NOyEiEjpqDKoBHrammlesiDn5UfsOKWAuRERKQqNBSEiIiIiUqb0mJiIiBSDxoIQERERESlTqgwSEZGC01gQIiIiIiLlS4+JiYhIUWgsCBERERGR8qTKIMlab3c3HSuW5bx8/fBGho4aW8AciYiUVr4vBMiGXhogIiIiIoWmyiDJWk9HK60vLsp5+cbdm0CVQSJSRfJ9IUA29NIAERHpk2sjhBoWRKQ/VQaJiIiIiIhUgFwbIdSwICL9qTJIREREKl53Tw+vvrlu0Nc7euQwxjZuNujrFRGpBLkOM6HhJUSKT5VBIiIiUvFa27r4u7816Oudtve2qgwSEUkh12EmNLyESPHp1fIiIiIiIiIiIjVEPYNERESkaN7V2MAedBZ1HW31w4uavoiIiEi1UWWQiIiIFM2Q7nZWP13cN66Nm6HhIfgAACAASURBVDytqOmLiIiIVBtVBklRtXd00d7ZDUBvSwet3YUd3FMDd4qIiIiIiIgMTF6VQWbWBDwMbBUnzQVGA/Pd/ZL8sibVoL2zm1deCxVA47ZqZvGK9oKmr4E7RURERERERAYm3wGkvwQ8ET9/BfiJuzcB+5vZpDzTFhERERERERGRAsu5Z5CZfQRYAHw0TmoCLoqf7wWmAc/mlTsREREREalaetJARKQ0cqoMMrN64AvAJ3inMmiUu7fGz6uBHfLPnoiIiIiIVLFkTxrcaWb3mdkkd1fjslS8Nc3trGvpKFr6GkdVcpFrz6BPAb9x9zYz65vWYmbD3b0NGAusLEQGRUqp2AX30CH1dHX3FC19UHCQ0lKLr4iIpKInDaRWrGvp4KEnlxUtfY2jKrnItTJoT2BfM/sYMBm4lXCxfyRwNzAD+FpBcihSQsUuuPexrfi7v1W09EHBQUpOLb4iIrIJPWkg5aTYDcDtHd1FS1skVzlVBrn7V/o+m9mDhJ5CdcAtZnYe8IC7LylIDkVEpCKpxVdERNLQkwaStcGorFm45PWipb+PbZV5JpFBlter5QHcfXrCnzPyTU9ERCqfWnxFRCQDPWkgWRuM3voitSbvyiAREZEk1OIrIiIp1fKTBl3r19DT1pzTsj0d7QXOjYjUKlUG5UiFeHno7unh1TfXFS19Pd8rkjO1+IqISFZq7UmDnrZmmpcsyGnZETtOKXBuRKRWqTIoRyrEy0NrW1dRB2BWl1GR3NRyi6+IiIiISLlTZZCIiBRVrbX4lpv2ji7aO1P3chza1c3a9YXtsdrd3VvQ9ERERESksFQZJCIiUsXaO7t55bXUj9MO274z7fe5GL/lyIKmJyIiIiKFVV/qDIiIiIiIiIiIyOBRZZCIiIiIiIiISA3RY2IiIiIiIiKSs1Tj0+U6Lt3Qtg7W/eulDX8P6+5ljy06s1q2rX44L7xVW28ELvYblkePHMbYxs2Klr6UhiqDREREREREJGepxqfLdVy6Yduv5YWF77y5efyWI1n9dktWy46bPA1oGPA6K1mx37A8be9tVRlUhVQZVMMyvWGmv1xq9vVGGREREREREZHyosqgGpbpDTP95VKzrzfKiIiIiIiIiJQXDSAtIiIiIiIiIlJDVBkkIiIiIiIiIlJDVBkkIiIiIiIiIlJDVBkkIiIiIiIiIlJDchpA2szeD3wX6ACagZNiWnOB0cB8d7+kQHmUKvGuxgb2oDOnZdvqh/PCW9m/+UxEREREREREksv1bWL/BA5x9xYzOxP4IjAO+Im732lm95nZJHd/tmA5lYo3pLud1U8vyGnZcZOnAQ2FzZCIiIiIlIQal0VESiunx8Tcfbm7t8Q/O4AuoAm4N067F5iWf/ZERERERKQK9TUuHwT8ltC4/BVC43ITsL+ZTSplBkVEqlmuPYMAMLMtgC8ARwAnuXtr/Go1sEOeeRMRkQqlFl+pFd09Pbz65rpBXefokcMY27jZoK5TpNDcfXnCn4mNyxfFaX2Ny3rSQESkCHKuDDKzkcCdwNnu/raZtZjZcHdvA8YCKwuVSRERqTh6nFhqQmtbF3/3twZ1ndP23laVQVI11LgsIlIaOT0mZmZDgduAH7j7o3Hyw8CR8fOM+LeIiNQgPU4sIiKZ9G9cBlrMbHj8Wo3LIiJFlOur5WcSLuLPMbMHzewC4Ergc2a2AHjK3ZcUKpMiIlKZElp8fwyM6tfiu3nJMiYiIiWlxmURkdLK6TExd59LGPehvxn5ZUdERKqFHieWwfKuxgYaelazxxadRUm/rX44L7zVXZS0RWpYX+PyGDM7B7iP0Lh8i5mdBzygxmURkeLJawBpERGRZDK0+N5NaDz4WomyJ1VmSHc76xb/hdVvt2SeOQfjJk8DGoqStkitUuOyiEhp1XRlUNf6NfS0Nee0bE9He4FzIyJSVdTiKyIiIiJSpmq6MqinrZnmJQtyWnbEjlMKnBsRkeqhFl8RERERkfKV6wDSIiIiIiIiIiJSgVQZJCIiIiIiIiJSQ1QZJCIiIiIiIiJSQ2p6zCARERGRTN7V2MAeJH9t/bgCvNK+cfQImte1Zj3/iPa36VixdkDrqB/eyNBRYweaNREREalSqgwSERERSWNIdzurn07+wokRW47M+5X2W76viVdTpJ/MmPGjqRu12YDW0bh7E6gySKSgcn0zsd5KLJWmu6eHV99cV7T0R48cxtjGgcU1yZ8qg8pYe0cX7Z3dm0wf2tXN2vW5BZHEZbu7e/PKn4iIDEyqcj2dfMp8UFkvIlIsub6ZuBRvJe7u7i3I/UO69KV6tbZ18Xd/q2jpT9t7W1UGlYAqg8pYe2c3r7y2aQ3ssO07k07PRuKy47ccmVf+RERkYFKV6+nkU+aDynoREYGe3p6C3D+kolgjUnlUGSQiIlUr1y782VJXfxERERGpRKoMEqlyxX7Gd+iQerq6e4qWvp4hlnzk2oU/W6Xo6i8iIiIiki9VBklFSPUml2ze4tJWP5wX3hrYGB3VpNjP+O5jW+kZYhGRQdTbO/CxP3pbOmjtzr1hQBXzIlIp0r0BMpVxPavZaashNX3PILVHlUFSEVK9ySWbt7iMmzwNaChSzkRERAZXV1cPrw3wDWbjtmpm8YrcH2tUxbyIVIp0b4BMZcSWIxk+YT90zyC1pKIrg/IdC0JjPYiIiEgtyKWlPNGI9rfpWLE25ff1wxsZqlfXi4iIVIyKrgzKdywIjfUgIiIitSCXlvJEY8aPpm5U6p5Bjbs3gSqDpALl07ishmWRyrCmuZ11LR1FS79SH6UueGWQmX0OOBXoAE5z9xcLvQ4RkUIpdnCAyg0QxaI4IVJ5Mo1TlO+YRMmo7Kxdgxkn8mlcVsOySGVY19LBQ08uK1r6H9xrfEVWNhW0MsjMNgdOBz4I7A18Gzi+kOsQGah8usZrMLnSK/bb0No7ulm45PWipQ8aayNROcWJ9o4u2jvzO7eHdnUPaCDf7u7evNYnUiqZxinKd0yiZFR21qZyihNSW/K5Z6j1F9bkazCu94up2C/sKVY8LHTPoKnAg+7eBTxuZlbg9EUGLJ+u8RpMrvQG421oMqjKJk60d3bzymv5XXgM275zQGmM33JkXusTKVf5jkmUzPC2t1jxr7fDH8NG0Vo3oqDpJzN0SD1d3T1FX08i9YDaRNnECakt+dwz6IU1+dH1fmkUujJoc2BVwt/12SykMn6grq7AZUu57nzzLSIFVGVxYjDKl2Kvo9LTH4x1VHr6g7EOxVopmCqLE9WgFNfhlXbtrzJQKk+hK4NWAZMT/s7YH8vd6wqcBxERKV+KEyIiko7ihIjIICh0ZdBC4CIzGwLsBTxf4PRFRKSyKU6IiEg6ihMiIoOgoJVB7r7SzH4KPAx0Ap8tZPoiIlLZFCdERCQdxQkRkcFR19urN5uIiIiIiIiIiNSKrAZkExERERERERGR6qDKIBERERERERGRGqLKIBERERERERGRGqLKIBERERERERGRGlLoV8sLYGYNwIPA7sDp7n6XmW0JzAVGA/Pd/ZLS5bB0UuybS4BPAm8Br7v7iaXLYWmY2fuB7wIdQDNwEuH81DGTfN+cS40fM5JeNZXD1VJuVks5V01lkpltDdxDeGPTEOBM4P+AOcAEYDHwRXfvKVUes5FiO/YDvgH8C8Ddp5cqf7kwsybC27S2ipMq6jwptXLff2Y2FfgvoAG4H7iZMssjgJldA+xD6EDwVWAhJS4fBhLfzewjwGygFzjH3R8vcT5vAvYg7M9r3H1unPdzwKmEuHKau79YqjwmfPe/gLv7WeWWRzMbCnyHsC+HAse5+9tllsd9gWtjXv4JnOLu3aU8JpNRz6Di6AKOA76fMO0rwE/cvQnY38wmlSRnpZds3wBc7O7TK+UCugj+CRzi7gcBvwW+iI6ZPsn2DeiYkfSqqRyulnKzWsq5aiqT3gaa4rZ8nXCzdxrwhLsfCPQAR5Qwf9lKth0A18XfZHrJcpa7LwFPxM+VeJ6UWtnuPzPbDLgY+Ji7f8jdryq3PAKY2c7Abu7+AUJl96WUR/mQVXw3syHAZcBhhPx/rwzyeaW7vx+YBsw2s6Fmtjlwepx2AfDtEucRMzuUUInR93e55fFM4K/ufmgs498uwzyeB1zg7tMI+/KgMjgmN6HKoCJw9153f63f5Cbg3vj5XsKBWnNS7BuAr5vZw2Y2c9AzVQbcfbm7t8Q/OwiFio4ZUu4bqPFjRtKrpnK4WsrNainnqqlMcvfuhFb9ccBTwIFU3m+SbDsATjezBWZ2bomylpPYcrwAWB8nVdx5UkoVsP8OAFqBu8zsD2Y2mfLLI8AbQGvshTGO0POx5OXDAOL7zsA/3H2duy8DGsxseCnz6e7Px48dQDehd8hU4EF374q9RKyUeYzOBq5J+Lvc8vhxYA8ze9DMLi/TPC4BxplZHTCG0GhR0mMyGVUGDZ5R7t4aP68GNi9lZsrMD9x9CvAR4D/N7N9KnaFSMbMtgC8AP0bHzEb67RsdM5KLajqnKvYcqJZyrlrKpNiC/ijwA+DPhN9gVfy6Yn6TJNvxK0KX/UOAI83sgFLmL1tmVk84rq5LmFyx58lgq5D9NwGYROgZcA7wQ8ovjwDrCD0hHfgDcBXlWz4k23+JeU2cXg7OB+5w9242zWdJ78/N7DjC792aMLms8ghsBzwfe31ub2YzKL883kuIR8/GvDxDGR6Tpd5JtaQloeZvLLCylJkpJ+6+Iv6/Bvgj4fnPmmNmI4E7gbPd/W10zGzQf9/omJEcVc05VannQLWUc9VUJrn7s/ExkI8QLlxXEXoBQAX9Jv23w91Xxx5D7YSKob1Lm8OsfQr4jbu3JUyryPOkRCph/60CHnH39e6+NOap3PII8GFgC0Jvhn0I45+Ua/mQbP8l5jVxeknFypYDCI8Lwab57B70TEXxMabPATf2+6ps8hitIlRYAcwD9qT88ngdcJS770YYi+9TlOExqcqgwfMwcGT8PCP+LYCZjY3/DwXeTzhhakrc9tsIF7CPxsk6Zki+b3TMSI6q5pyqxHOgWsq5aiqT4tglfVYDLcBDvPObHEll/CabbEfCb1JHeGTk+WTLlqE9gePM7PfAZOBWKvA8KaFK2H8LATOzejPbBmij/PII4T5xZXwEcy0wivItH5Ltv+eBXcxsVNzPXf0qCQedmU0n9FyblfBo60LieDJmtg+lLau2IQy6/hvgSuDo+OhzOeURQu/PfePn/Qgxt9zyWAesiJ9XAO+iDI/Jut7e3lKuv2qZ2R2Eg7MZ+D2ha+UthFHuH3D3i0qYvZJKsm82J3SXHQLc6u7/U8LslYSZzSK0yC6Kk+4jvK2h5o+ZFPvGqPFjRjKrpnK4GsrNainnqqlMsvBmtG8TBoKtIwx4+Rzhd9kGWAp83sv/bWLJtuNjhEE6e4GH3P0rpcthbszsQcKgpHVU2HlSDsp5/5nZaYS3HjUAXyaca+WWxyGEsmAiMJwwOO49lEH5kG18N7NjgAsJ5cCX3H1hifN5LOHxu9Vxlk+6+1tmdiZwMuGNiJ919xdKlUd3/3KcPp3wlq6+t4mVTR6BKwjH4RjgZcKbw7rLLI+/A75FGB+qBTjB3deW+pjsT5VBIiIiIiIiIiI1RI+JiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGiYiIiIiIiIjUEFUGScUzszlmdkT8fL+ZDRuk9Z5rZkNzXPYUM/t2AfKwX4HSmWhmx+WbjoiUDzP7z4TP083stlLmpxok7kczO3Owys243v3yWP71AuWjIDE2n/gpIoNH14f50T2K7lHKnSqDKpSZDSl1HsqRux/p7h2DtLpzgZJezLr7E+7+1QIkNREYtIJWx6/IoPjPzLNIrtz9ene/a5BWNx3IuTKoUAoYYwc1firmiORsIoN4fVjNdI+Ss4noHqVo1CpTpszsXmA8MAy42N3vNrOJwK+AfwB7mdnuwH8DHwAagG+4+2/N7P3Ad4HhwCrgJHd/LcV65gDNwPuARuBk4CJgN+BKd78xzjcb+BiwGfADd7/JzN4L/BQYCbQDp7n7UjM7BTgC2IpwAl/q7j/NsL2vA78EDgUWAncA34hpH+3uL5vZx4ALYx7+GberuV86LwO7unubmf0X8Englfj15e7+YFzX7XFdLwCfcPduM5sa9+co4P+Az7h7S7L5gTOBCcBCM3vO3U9IyMN3gSfdfW78+w7gGnd/qN9mv9fMFgDvBv7b3W9Is6+T7lMzmw6c6e4nmtnWMZ9bAvcBJ7j7xCx/j8uBSWa2CLgK2Bl43d2v7/t93H2bmNaRMa0dgHPifjkU+Iu7nxbnPxU4P6Z9tbvfmOT4/RBwZ9zf9cDx7v4cIlXMzCbQ77gH2oB7AAf2B34BvAR8AegAZrj7ajPbH7iOUDYsAM6KZdeRwLdjene5+yVmdjmwRTyn/wT8GniXmf0W2BWY4+6Xx/PybmApsC/wJ3f/fMzrUYRyeDjwWMzPNknyv7b/tGTnciyvvgZ0AZNinrcATgCWEcr6rjRl8Y3APsAI4Ifufm1MN2mZnuY3mEOGuGdmYwjl1TigDvgPd1/QL51LiOVkjLs/IsTChcAIdz8lrmsN8H5gNDDT3ReZWWP8LXeNyf2Huz+WbH5gBSHmdJrZmYSy3WMeJgPfc/dD4t9HAsf1lcUJ6szsBuBA4Dng03GfZh334rH2MjnG2IT99kUS4ifwFeA2dz8g4fe5zd1/n+W1wU7AzcDYuG2nuvt6M3sQeBKYBlxlZvsCR8ff6GfufhUiZcTMPgecDfQAv3L3i5OV73HebM6NOUALoTwZRrhOX5jqPiGWe9cDe8Q8nE7y68NtCeXleODz7v6H2Asl2T3JiYTytRP4h7t/Mtm0FPtjDrpH0T2K7lGKSj2DytfJ7r4v8EHgMjOri9N3B77p7kYopF9y9/0JFzvfNrMGYAnQ5O57Ewr1L2dY17CYxvWEAu5k4ABgNkDs3rhVnGc/4HQzGw+8Bhzi7vsQTrhvJaS5O+Gi64PAN/smxhM5ma2BnxMujHcHDnX39wE3AWfEef7s7u9z9ynAI8BnU21QLDQPJgS0zxACSeK6funuuxPOgYNjt82rgGPi9vyVcOOTdP54E7IceF9iIRvdCRwb8zGCcPOygE3tTyi09gPON7MJafY1pNinCS4mXCjsAfyr33eZlp0NzHP3Ke7+8yTfJ9oVmAF8lHDT+gvCjd1eFmwX02siXICcFwvZvnz0Hb8zgfnx99yXEDxFql2q4343QkXJJOAU4F0xBvyFUFkC4Yb3DHffE9gcmBnLmB8CRwF7Ax82syZ3nw2siOf0l+Lye8e09wLONLNRcfok4BLC+fl+M9vZzLYktCxOj3ntJlxkJsv/QM7lycBJhPLvCkIMmwy0AodkKIu/6u77AVOAU2IeIUkZDWBml5rZMSnykTbuxfx8NObhaMJFeDo3xf2wL+FiN1Gju08Fvg5cEKd9PeZ5f+DjwLWp5nf3f8V8Xh5/T++b0d2fBrZJ2BefBJL1Vno3cJ+7TyL8Pl8YaNxLTGygMTZx2Qzxs79srg2uJlSITQZeBr6UsHxbPI/+l9CyPMnd9wJuzLBekUEVK3a/CHwwHqP/k6p8j4tkc270zbcP8On4HaS+T7iIUDkzmXAN+g+SXx++BziI0BhwcZyW6p5kNnBU3KbT47zJpqWiexTdo+gepYjUM6h8fSnhInYioTUWYKm7L4mfP0yoKT0l/j2aUBMM8DMz24HwG/fVOqdyb/z/GUJt8SoAM6uLBfmHgaNjDS+E1rf3EoLJtTGAdRNaGPrMd/cWoMXM6s2swd0740mVzCp3fzSudwnwQJy+mHcKye3N7E5CwTcSmJ9mmz4A3OPuncByM3s44bvVCTXgTxL27+uEm5Q/mRmEFpQ/ppk/nccIv0sjcDjwB3fvSTLf/e6+Nm7zfGAqodU22b6GJPs0yTb3BeXb2fgRkaS/R4btSGV+bNV4hnChnfi7bU+oRf+Du6+O0+8n/IYL2fj4fRyYY2Y9wJ3u/myO+RGpJJsc97HMedbdXwQwsxeBP8T5FxNa6MYB9e7+tzj954QLnsVx2X/FZW8nXFAlu7h7xN1XxPleIsSLTsJ5+Xyc/jShjNuVUCY+FvM3gnAx9FiS/A/kXH40Icas4Z34s5hQfhipy+JPmdlngSHAvxFaB98mRRnt7helyUemuFcHXBlvvLqBnVIlFH+bulgx09fSelTCLL9JyNt58fOHgRkWehdB6MU1NM386dwDfCy2ok9j4xvBPi3u3pfuLwgVj/PIPe4NNMbmKptrgynufnf8/DNCJWOfO+P/awg9DH5kZr8itEyLlJPphB5xawHcfaWZTSF1+Z7NuQFwu7v3Ak+a2dBYXo0l+X3CwYQKANy9C1gby4b+7vPQWyXx/E51T/IIcJOZ3co7FdXJpqWiexTdowyE7lEGSJVBZchC17SpwFR3bzezxYTueBC6e/apAz7bd6AnLP9T4G53n2NhwMnvZFhl3/OrPQmf+/4eEtfzDXe/td96vkl4rOEkQlf/JxK+bk+STroTu/96E/PU9+zm1YRH5h60MJDYR9KkV5fmu2R5qwMed/fDspw/JXfvjYXLUYSulDelmLW33+deUu/rnbLIx0C3Od3v0U3sOWhmm/X7rgPA3XvMLNnxks6G49fdHzKzaYTWgF+a2Vnu/sfUi4pUvmTHPaHLdzZlYL5SlQOpysRfu/u/908k2Xk7gHM503YmLYvNbEdCJccH3H2dhUep+8qmAZXR/fKRKu7NJFxwT4k3Pc2klq7sTcxfYt7qCI//LU+cMV7oJ5s/nbuA/0dobX3EM49JkRhvco17A4036WyIN1FizMn3vGiBcGMbr4cOJ1yzfILQS06kUmV7biS71ryUgd0n9NcOG64DE8u0ZPcknyf0wDiG8NjSnsAm02LlUzK6R9E9Sn+6RykgPSZWnsYAK2NF0FTC4wPJzAc+b2Z9J8SUhOX7xgg6pQD5mQ981syGx/VY/DyG8Mxmb4HWk8kY4LW4vbMyzPsooaV0qIUujE0Z5n8O2CEGKcxsVCzc0llHaPlI5k5Cl9z3AX9OMc8MMxttZqOBQwi10Kn2dTYeJTwiQML/2eq/Lf8kPIoBIVgPxOOExz3GxJaHGYQa942Y2XuA19z9OuA2YM8Brkek4uR63MdWrG4z2ytOmkloHXZgNwtduIcSzv1H4jw9ffEhB48RzuPtYr63MLPtkuW/wOdyqrJ4NKGcarbQpfugPNaRjTHAm7Ei6DhCa2JSsaW6ty/PZDfQ5XzCIyEAJPyuqaSMN+6+iNDi+TlSt7KPNLO+m5MTCMdOLnGvz0BjbH+J2/MGsF1c/zhC6/NALDKzo+PnT5GkV1yMRWNj76gv8058EykXDwAnWhi3BzPbnPTle7aOj+ntBXS5+xpS3yfMJ4w3Qzy3+8rdVNe6iVLdk+wQK4i+RhhLaHSKabnSPUpyukdJTvco/agyqDzNI3QZX0J49v2pFPPdQLiIeirO29cl/jvAD8zs74Ru0Xlx9/tjnh6PvZSuI9SuXg+cZeEZ26wKckv9PG42LgN+RzhpX8qQ54WEAm4JYQC5pwiDnKaav4NwEXm9mT1FGKcjU0F7E/CwhW67/T1KeL57vqceyPRvhO35G2G8g+Vp9nU2vgmcYKFr5M6k2d4kngZGmNkiMzuJMKDsLhYeGdljAOng7ssI3fQfJdxQfs/dX04y63TgaQvdjA8kdO8XqXbTyf24Pw34cTzH1xAeKWglVCr8DlgE/NHfGej4Z8AzZva9gWbS3d+M6f46lgN/IIw7kyz/+WxT//UmLYvd/SlCD6rnCGNoZLwhsvRjBmVyK2GshqcJj169kWH+M4DbE+JupvL3UmCCmT1tZs+SedyM3wKzzOxJs6TPbfwaOIwQP5J5k9C9fwmhW/91OcY9YOAxNokN8TPm4/sxjdtIfc2TytnAl+NvtWNMq7/RwH1xO39NGPxVpGy4+2LgGuDReJyenaF8z9ZbsVy6Fejr6ZnqPuEywrXfM4Sb5l3Y9PowlVT3JP8d03sa+GmsPE82LSe6R0lJ9yhJ6B5lU3W9vb2Z5xKpQGY2ysMbRbYkFJz7uPu6UuerWGLtfGdsyf40YYC7U0qcLRGRqpcQb+oIgxM/4u5zSpytoqq1GCtSaSzhzXylzotsrNbKT92jlC+NGSTV7CexBbXvFZdVW8hGE4FfWHh+eyXhjQsiIlJ8HzWzLxPGGXqc0Apf7WotxoqIFEqtlZ8T0T1KWVLPoBphZqcSXq2Y6CfufnUp8iMiItXJzLbgnTed9Hnb3Q8tRX5ERKR86R5FpHRUGSQiIiIiIiIiUkNK/piYmak2SkQkBXfP9Nrqqqc4ISKSmuKE4oSISDqp4kTJK4MA3L3UWRARKTvJXxpUmxQnREQ2VSlxwswagAeB3YHT3f2uOHjuXMLbnua7+yVx3o8As4Fe4Bx3fzybdShOiIhsKl2c0KvlRURERESkmLqA44DvJ0z7CmFsmCZgfzObFAeYvQw4DPgk8L1Bz6mISI1QZZCIiIiIiBSNu/e6+2v9JjcB98bP9wLTlR/MXgAAIABJREFUgJ2Bf7j7OndfBjTE11KLiEiBqTJIREREREQG2yh3b42fVwObx3+rEubpmy4iIgVWFmMGiUj16unpYfny5XR2dpY6K2WtoaGBCRMmUF+vOnoRqS2KE9mpwjjRYmbD3b0NGAusJFQEjUuYp2+6iNQwxYnsDDROqDIoD13r19DT1pzVvPXDGxk6amyRcyRSfpYvX86YMWMYM2ZMqbNS1tauXcvy5cvZbrvtSp0VkQEbSDwsFMXV6qE4kZ0qjBMPA0cCdwMzgK8BzwO7mNkowsDSXbGyqKYUskxVWSnVQHEiOwONE6oMykNPWzPNSxZkNW/j7k2gglhqUGdnpwruLIwZM4YVK1aUOhsiORlIPCwUxdXqoTiRnUqPE2Z2B7Af0GxmU4ErgVvM7DzgAXdfEue7BJhPeJvYl0qU3ZIqZJmqslKqgeJEdgYaJ1QZJCIiIiIiReXuxyeZPCPJfL8BflP8HImI1LaqeehYRCrDmuZ2Xn1zXV7/1jS3p13Hq6++ipkxb968DdNOPfVUzjjjjKTzL1y4kEsvvRSAyy+/nHXr1hVug/tZunQpf//734uWvohIpVOcUJwQEUmlEDFCcSJQzyARGVTrWjp46MlleaUxbe9tGdv4/9m78zi5yir/45/udIcsnXRMopiACAgcEkgTQHDrhIDKPigDQjJMQBEnKvxYh1GMEyMMjoijg4gYRxaTAdmZQbaMSIAENMIgkQ7hCMMmYQtZu9N7V//+eG4nlaK6uqq6utbv+/XKK91Vdzn3VvVz7j33uc/dKeU0+++/P0uXLuWoo45i3bp1dHR0MHz48AGXPX/+/EHFNpA1a9awceNGDjrooCFdj4hIqVKeUJ4QEelPLnIEKE+AikEiUqYmTpxIW1sbra2tPPDAAxx99NE8/vjj3HDDDSxbtozm5mZmz57NqaeeusN8c+fO5aqrrmL48OGcf/759PT0sNdee/Haa6+xaNEi5s6dy9SpU1mzZg319fVcffXVAFxxxRU0NTXR3d3NJZdcQkNDA8cddxyNjY2sWrWKadOmMX/+fBYvXkxLSwuPPPIIP/vZzxgzZkwhdo+ISMVTnhARkVTKPU/oNjERKVtHHHEEDz/8MMuXL2fGjBkAnHrqqSxevJhbb72VX/3qV8RisaTz3nbbbTQ2NnLDDTcwderUHd6bMWMGixcvpqurC3fn0UcfJRaLsWTJEq655hquvPJKANra2pg9eza33HILK1eupKWlhdNPP505c+awZMmSsjnAN7NaM3vczDaZ2cnRaxPN7AEzWxENBto37fFm9nsze8LMDilY0CIiKE+IiEhq5Zwn1DNIRMrWkUceybx589hzzz2pra0F4P777+fuu++mqqqKd955h82bNyed95VXXuFzn/scAA0NDTz44IPb3utrzCdPnszmzZt54YUXWLFiBXPnzgVg69atAIwcOZI99tgDgEmTJvW7rjLQDZwMxN9E/Q3gene/3czuM7OpgAOXATOBscCtQGO+gxUR6aM8ISIiqZRznlAxSETKVn19PY2Njduq+ACLFi3ivvvuA+Doo4+mt7c36bwf/vCHaWpq4uCDD+bZZ5/tdx29vb3stddeHHHEEVx00UUAdHZ2AlBVVfWeaWtra+np6RnUdhUbd+8F3jSz+JcbgQXRz/cSCkAx4C/u3gw0Rz2KRrh7e14DFhGJKE+IiEgq5ZwnVAwSkbwaM2o4Mw/cZdDLSNc555wDhCcCADQ2NjJnzhz22Wcfxo4d2+98p5xyCueffz7Lli1j991333YlIJlZs2bx1FNPbavkNzQ0cPHFFyed9sADD+Smm26iqamJ733ve9TV1aW9LSVmtLu3RT9vAvYAxgMb46bZFL32Rp5jE5EipjxRMXlCRCRjucgRfctJV7nmiar+qlj5Yma97l7QGLLVuX4tLatXpDVt3X6NDJ8w+C+tSKl5+eWXt3VtLCW9vb309PRQU1PD3XffzYsvvthvg5wrifvKzHD3qhSzFJVobKAmd7/DzP4AzHL3djP7ajTJo8B33H12NP22aQZYbsnmiUqRST7MFeXV8qE8kb5SzxNDpdzyRC7bVLWVUg6UJ9KXSZ5QzyARkSS6u7s544wzqKqqorq6mh/+8IeFDqnULAeOBe4CjgG+BbwA7GNmo4ExQLduERORUqU8IYm6t24m1t4y6OXEOjtyEI2IFFqx5wkVg0REkqitreXmm28udBglw8xuAz4KtJjZocAPgMVmdiHwsLuvjqZbCDwE9AIXFChcEZFBU56QRLH2lpz06Bm55/QcRCMihVbseULFIBERGTR3PyXJy8ckme4e4J6hj0hERERERPqTdTHIzH4KHARUA98EVgI3ApOBJuBsd4/lIEYRERFJQ65uUciUbmkQERERKS1ZFYPMbG9girt/0sw+BNwE3Ao85e5Xmtk1wNHA/bkLVURERFLJ1S0KmdItDSIiIiKlJdueQW8DbWZWA4wD1gEzgO9G798LHIaKQSJSAK+//jonnXQS++yzD21tbcyaNWvbIyHTsWbNGv74xz9yxhlnDGGUIiJSKMoTxUF3GohIsaqEPJFtMagZeBVwYBRwInApsDF6fxMwftDRiUjZycVtLNUj6qgZXZ9ymunTp7No0SJ6e3s58cQTOf300xk7dmxay58yZQpTpkwZVIwiIpId5YnKoDsNRCQbubolXnki+2LQZ4EJwN7AzoSeQC8Segm9BdQDG3IRoIiUl1zcxlK3XyMM0Hj36erqore3l5qaGq644gqampro7u7mkksuoaGhgeOOO47GxkZWrVrFtGnTmD9/PitXrmTp0qUsWLCA3/zmN1x33XXstttuNDc3c+GFF9La2sovfvELRo8ezUsvvcSCBQs49NBDB7VNIiISKE9UDN1pICIZy9Ut8coT2ReDqoEN7h4zsy3AaOAx4Fjg+ej/pbkJUUQkc8888wxz585l7dq1HH744Tz55JPEYjGWLFnChg0bOO+881iyZAltbW3Mnj2bSy65hBNOOIGWlu1XGnp6erj22mu58847GTZsGCeccMK291pbW/nlL3/Jc889x89//nMd5IuIlBjliYLTnQYiUtTKPU9kWwz6LXCamS0HRgCXAXcDN5rZY8Aa4IHchCgikrm+bp2xWIxzzz2XpqYmVqxYwdy5cwHYunUrACNHjmSPPfYAYNKkSWzevHnbMjZs2MDOO+/MyJEjAXbo6rnvvvtSVVXFpEmT2LRpU742S0QivT09dK5fm/f1ptOtXEqD8kTB6U4DESlq5Z4nsioGuXsPMDfJW6cMLhwRkdyqrq5mzJgxmBnt7e1cdNFFAHR2dgJQVVW1w/S9vb3bfh4/fjxvv/027e3tDBs2jOeff37be4nziUh+xTrbaHvpmbyvN5Nu5VIalCcKRncaiEhJKNc8kW3PIBGRotbXrbO7u5vJkydz2GGHbXsNoKGhgYsvvjjlMoYNG8a8efOYPXs2u+yyCxMnTqS2tjYf4YuIyBBTnig43WkgIkWt3PNEVXzVqhDMrNfdCxpDtjrXr0178Kq6/RoZPmGXIY5IpPi8/PLL27pNQv6eEpMrXV1d1NbW0tnZyUknncTNN9/MmDFjhmRdifvKzHD3ir+0XMp5It8yyUu5NHLP6XnvpVOIdYLy+VBQnkif8kRyxZInctUG57J9U5sl5SC+7cvn08RypVjzhHoGiUhe1YyuL6lbLO644w7uv/9+WlpaOOWUU4as4RYRkUB5QkRE+lNqOQKKN0+oGJQnmQx0qcEpRYrHnDlzmDNnTqHDKElm9lPgIMK4EN8EVgI3ApOBJuBsd48VLEARkRxQnhARkVSKNU+oGJQnmQx0qcEpRaTUmdnewBR3/6SZfQi4CbgVeMrdrzSza4CjgfsLGaeIiIiISCWqLnQAIlLeamtr2bJlS6HDKHpbtmwpmsHkcuRtoM3MagiPCV4HzCA8Opjo/5kFik1EiojyRHrKME+IiKRFeSI9meYJ9QwSkSE1efJk3njjDdavX1/oUIpabW0tkydPLnQYudQMvAo4MAo4EbgU2Bi9vwkYX5jQRKSYKE+kpwzzhIhIWpQn0pNpnlAxSESGVHV1Nbvuumuhw5D8+ywwAdgb2JnQE+hFQi+ht4B6YEPBohORoqE8ISIiqShPDA3dJiYiIkOhGtgQDRC9BRgNPAYcG71/LLC8QLGJiIiIiFQ0FYNERGQo/BYYY2bLgUeAy4AbgI+b2WPATsADhQtPRERERKRy6TYxERHJOXfvAeYmeeuUfMciIiIiIiI7Us8gEREREREREZEKomKQiIiIiIiIiEgFUTFIRERERERERKSCZD1mkJkdCvwLUAvcTxgYdAkwBnjI3RfmIkAREREREREREcmdrHoGmdlOwHeAz7v74e5+JfAN4Hp3bwQOMbOpOYxTRERERERERERyINueQR8H2oA7zKwG+EegEVgQvX8vMBN4btARFkD31s3E2lsGnC7W2ZGHaEREREREyo/uNChtm1s6aG7tzNv6xowaTn3dTnlbn0i5y7YYNBmYChwC7Ab8BzDa3dui9zcBeww+vMKItbfQsnrFgNON3HN6HqIRERERESkvCXcatEavXUm40+B2M7vPzKa6e0leXK4Eza2dPPantXlb38wDd1ExSCSHsh1AeiPwuLtvdfc1QD3QamYjovfrgQ25CFBERERERMpO/J0G/2NmDYQ7De6N3u+700BERIZAtj2DVgLfMrNq4ANAO7AcOBa4CzgG+FZOIhQRERERkXJT1ncaFIvenh461+em9071iDpqRtfnZFkiUnhZFYPcfaOZ3Qg8SrjH9yJgDbDYzC4EHnb31TmLUkREREREysm2Ow2ANWZWD2w1sxHu3o7uNMiJWGcbbS89k5Nl1e3XCCoGiZSNrB8t7+7XA9cnvHzM4MIREREREZEKoDsNREQKKNsxg0RERERERLLi7huBGwl3GvwX4U6DHwBfMbMVwCrdaSAiMnSy7hkkIiKSih4ZLCIiqehOAxGRwlExqEhtbumgubUzL+saM2q4HtMoIjmlRwaLiIiIiBQvFYOKVHNrJ4/9KTcj/w9k5oG7qBgkIrkW/8jgGuAfCY8MXhC93/fIYBWDRERERETyTMUgEREZCnpksIiIiIhIkdIA0iIiMhS2PTLY3dcQHhHcamYjovf1yGARERERkQJRMUhERIbCSsDMrNrMPsiOjwyGMEDo8kIFJyIiIiJSyVQMEhGRnNMjg0VEREREipfGDBIRkSGhRwaLiIiIiBQn9QwSEREREREREakgKgaJiIiIiIiIiFQQFYNERERERERERCqIxgwSemIxXn+nOS/rGjNqOPV1O+VlXSIiIiIikhu9PT10rl+77feRHZ3sP6Erq2W1V4/gxXU9uQpNRLKgYpDQ1t7N074uL+uaeeAuKgaJiIiIiJSYWGcbbS89s+335q0dbHozuwvK4xpmArU5ikxEsqFiUBHq7elhZMe7aVXaS62qrl5IIiIiIiIiIoU1qGKQmTUCy4H3Ry8tAcYAD7n7wsGFVrlinW00N61Mq9JealV19UISERERERERKazB9gy6AHgq+vkbwPXufruZ3WdmU939uUEuX0REREREypQuLouIFEbWxSAzOx5YAXwueqkRWBD9fC8wEyiaYlD31s3E2lvSmjbW2THE0YiIiIiICLq4LCJSEFkVg8ysGvg68LdsLwaNdve26OdNwB6DDy93Yu0ttKxekda0I/ecPsTRiIiIiIhUtlK7uCwiUk6y7Rn0d8A97t5uZn2vtZrZCHdvB+qBDbkIUERERCRXOjq76ejK/sELva2dtPWEMf30oAKR7JXixWURkXKSbTFoGnCwmX0eaABuJtzreyxwF3AM8K2cRCgiIiKSIx1dPbyW5aOQAca9v4Wm9eF2cj2oQGRQdHFZRKSAsioGufs3+n42s0cIjXkVsNjMLgQedvfVOYlQRERKlgYGFRGRfujismSkJxbj9XeyL+ZnQj0/pRIM9mliuPusuF+PGezyRESkrGhg0Ao12Nux4lVXVRHr7d3htZruHrZszfyBDz09vQNPJCJDTheXJVNt7d087evysi71/JRKMOhikIiISDIaGLSyDfZ2rHiTJo7izXdbd3ht+G5dWS1/0sRROYlJRHJHF5dFRPKvutABiIhI+YkbGPTauJcTBwYdn/fARERERERExSARERkS2wYGjXut1cxGRD9rYFARERERkQJRMUhERIbCNOBkM3uQ9w4MCuE2gOUFik1EREREpKJpzCAREck5DQwqIiIiIlK8VAySsqXHT4oUBw0MKiIiIiJSXFQMkrKlx0+KiIiISCXp6Oymo6tn2+813T1s2dqRk2UnLqunpzcnyxWRwlAxSEREREREpAx0dPXw2pvbe8YP361rh98HI3FZkyaOyslyRaQwNIC0iIiIiIiIiEgFUc8gERERkSzkemw6jT8nIlIcNPaoVAIVg0RERESykOux6TT+nIhIcdDYo1IJVAwSEREREZGK1L11M7H2lpwsK9aZm4GaRUTyQcWgEve+ulr2p2vA6dqrR/Diup4BpxMREZH+xefdcbFN7D9h4BycrpEd79K5fkvS96pH1FEzuj5n6xKRINbeQsvqFTlZ1sg9p+dkOSIi+aBiUIkb1tPBpj8PnMDGNcwEaoc+IBERkTIWn3dHThzFpndbc7bssZPGUDU6+a0Cdfs1gopBIiIikiNZFYPM7BPAj4BOoAU4LVrWEmAM8JC7L8xRjCIiIiIiIiIikiPZ9gx6Ffi0u7ea2VeBs4FxwPXufruZ3WdmU939uZxFWgQ6Orvp6Np+q1VNdw9btqZ3b3Cm0/b09GYVo4iIFF4ux6DIhMarEJFSoYvLIiKFlVUxyN3fiPu1E+gGGoEF0Wv3AjOB8ioGdfXw2pvbHzE4fLeuHX5PJdNpY7FYVjGKiEjh5XIMikxovAoRKSEVeXFZRKRYDGrMIDObAHwdOBo4zd3borc2AXsMMjYRESlRuuIrIiKpVOrFZQnSfQhOvGSD9ushOSLZy7oYZGajgNuBc939XTNrNbMR7t4O1AMbchWkiIiUHF3xFRGRAenicmVK9yE48ZIN2q+H5IhkrzqbmcysBrgFuNrdn4heXg4cG/18TPS7iIhUIHd/w937jtjir/jeG73Wd8VXREQqVOLFZaDVzEZEb+visojIEMqqGATMIRzEn2dmj5jZxcAPgK+Y2QpglbuvzlWQIiJSmuKu+F4HjE644ju+YIGJiEhB6eKyiEhhZTuA9BLCuA+JjhlcOCIiUi50O7FIZnp7e/t98mhvaydtPek9iAJgzKjh1NftlKvQRIZC38XlsWZ2HnAf4eLyYjO7EHi4HC4uJz6NOB2ZPIU4kZ5IXHp6YjFefyf99n2wlB+kz6AGkBYREUlmgCu+dxEuHnyrQOFJP7I5aYHkJy46Iclcd3eMNxPGw+gz7v0tNK1P/+Rw5oG76GBfilqlXFxOfBpxOjJ5CnGiSRNHZTWfFE5bezdP+7q8rU/5QfqoGCQiIkOhIq74lptsTlog+YmLTkhEREREipeKQSIiknOVcsVXRERERKQUZTuAtIiIiIiIiIiIlCAVg0REREREREREKkhJ3ybWvXUzsfaWtKaNdWY3Ir+IiIiIiIiISDkp6WJQrL2FltUr0pp25J7ThzgaEREREREREZHip9vEREREREREREQqiIpBIiIiIiIiIiIVpKRvExMRERGpBO+rq2V/utKefmTHu3Su3zLo9VaPqKNmdP2glyMiMhQybRtTaa8ewYvrenKyrGLWE4vx+jvNeVnXmFHDqa/bKS/rksypGCRSYja3dNDc2pmXdakBFxEpDsN6Otj05/TGSQQYO2kMVaMH337X7dcIKgaJSJHKtG1MZVzDTKA2J8sqZm3t3Tzt6/Kyrk8dMClv5y2gc5dMqRgkUmKaWzt57E9r87KumQfuogZVREREREQyls/CE+jcJVMqBlWIVF0ox8U2sf+E7e9VShfJXMpnd8uOTn02IiIiIiIikj0VgypEqi6UIyeOYtO7rdt+r5QukrmUz6r3Qfb+vKxHRERKV29vL1u2dgx+Oa2d1OzUoSutIiIiZUbFIBERkRLW0dlNR1foMVjT3TOoAkBPT2+uwpIC6+6O8WbchZ5sjXt/C+N36lQxSETKXvydFIl3TmRKd1pIKch5McjMvgJ8CegEznT3l3K9DhERKV3KE7nV0dXDa2+G21SH79a17edsTJo4KldhSRnJ5a3QNcOq6e6J5WRZGii0fClPSCHE30mReOdEpnJ1p8Ve7x/GiFj7oJcDKlDJe+W0GGRm44GzgE8BBwLfB07J5Tpk6GXyiEY1KuWtXB89mc8nsoFOWOLlO090b91MrL1lqBbfr1jn4G/PESkWubwV+iB7f86WpYFCy5POJ6QcZHI+laoX0rgR3bzyh8dyElMlDAVSrucuQyXXPYMOBR5x927gSTOzHC9f8iCTRzTu8clPMyLNLpTt1SMGE5YUQLk+erKjs4eVq9/Ky7pAJywJ8ponYu0ttKzOzSNnMzFyz+l5X6fIUHhfXS21g7xdIl66t14U6mJTIQrI1SPqqBldn9d1FjmdT0jJy+R8KlUvpIkfa8xlWGWvXM9dhqrwlOti0HhgY9zv1enMVLpt/E9KaNpCr19EBKi4PJFvhWqTC7HeStrWSlyvVDDliZz+3ZX7sooxplwuqxhjknKR62LQRqAh7vcBL+m4e1WOYxARkeKlPCEiIqkoT4iI5EGui0ErgQVmNgw4AHghx8sXEZHSpjwhIiKpKE+IiORBTotB7r7BzH4FLAe6gC/ncvkiIlLalCdERCQV5QkRkfyo6u3tLXQMIiIiIiIiIiKSJ2kNyCYiIiIiIiIiIuVBxSARERERERERkQqiYpCIiIiIiIiISAVRMUhEREREREREpILk+tHyEjGzrcCT0a/fBx4FbgQmA03A2e4eM7NDgKuAKuByd7+3AOEOipnVAo8A+wFnufsdZjYRWAKMAR5y94XRtMcD84Fe4Dx3f9LMqoFrgP2BN4Avuntb3jckS/1s/0LgC8A64C13nx1NW1bbb2afAH4EdAItwGmEdqXsP/t+tv18KuBzL2dm9hXgS4TP9Ux3fynuvaTttZkdB1wIDAOuc/cleQ98AJluV9Su3Qx8MHr9q+7elP/IUxtgu34KnATc6e7nRK+VxN9dFts1DzgT6AaedPfz8x91apluU9x784E57r5/PuNNVzbbZWZnAbMJ+fJyd/9tfqOWoZRFe7s78L/As9Fk57v7M/mNujTb0yxingXcBLwQTTbb3d/KZ8xRHEnjNrNxwJ3AcMJ35Fx3f7oY9nWWcc+iiPd39N6dwERgFPADd789en0h8BmgGZjr7u8Wc8zZtiPqGTR0Xnb3WdG/BwkHaU+5+wwgBhwdTfdjwsnjZ4HLzGxYYcIdlG7gZODf4177BnC9uzcCh5jZ1GjbLgOOJGzzj6NpjwFi0b55ivDFLyXJth/gO9Hn31cQKMftfxX4tLsfBvwGOJvK+eyTbTtUxudelsxsPHAWMBO4mFDIj/ee9trM3k/47I6MPvdiLARlvF3ALGBL9P28BPinvAWcpjS263vA3yW8VvR/d1lu10PAx939U8AHzexTQx5oBrLcpr75pg55gFnKZrvMbBpwiLt/JmozVAgqI1m2twC/jztvKEQhqOTa02zbFeDuuH1diMJEqrg7gNOjfXoWcEX0esFzV5ZxQ3HvbwgXGw4DjgAujebZj9BONwLXk+djoGxijmTcjqgYNHQ+ZGaPmdlNZjYBmAH09fq5F5hpZiOAGndf6+4twF+AvQsUb9bcvdfd30x4uZGE7SVs21/cvdnd1wK10T54z77JQ9g508/2A3zbzJab2Zzo97Lbfnd/w91bo187CYWxivjs+9l2qIDPvYwdCjzi7t3u/iRgfW+kaK+PJfQMe8DM7jGz3QoR+ACy2a6XCFf4AMYRersVm363C8LfKKE3XrxS+LvLeLvc/f/cve+1+PaoWGTzWQF8E/i3PMSXrWy262+BHjN7yMxuNrP35SlWyY9s2lsIF8+Wm9nPzGxk/sMuyfY023bleDNbYWaXm1lVPgJN0G/c7t4WHSvCjm15ofc1ZBc3FPH+BnD3zujHOmB19HPi/p6Rj0DjZBMzZNGOqBg0dD7i7jOB3wGXA+OBjdF7m6Lfx0c/k/B6ORgd130xfns3xk2T7PVy2QdXu/t04HjgH83sQ5Tx9kcFz68D11Fhn33CtlfU516GEj+r6oT3krXXk4EPEYpCPwGuHOIYs5HNdr0O1JnZGuBn0b9ik2q70pmnWP/ustkuYNvtqzu7+8qcRzU4GW+Tme0KTHL3p4csqsHL5rOaTMiTnwEeJhS8pHxk096+CewV9ap4FzhvqINMohTb02xifopwYj2T8Ld46hDENZAB446KJj8CfpBknkLlrmziLpX9/VtgFfBg4jzROU3dEMeYKJuYs2pHVAwaInH3Fd4CHEj4QMdFr9UDGxJei3+9HLRGV0Bg4O1Ntm9Kmruvj/7fTCgI7k+Zbr+ZjQJuJ9wf/C4V9Nknbnslfe5lKvGz6knxXvzn+HB09eYhYMqQR5m5bLbrDOAFd59CuK352qEOMguptiudeYr17y6b7cLM9iEchP/9UAQ1SNls0z8D/zo04eRMtt/B/4l+fhCYluugpKAybm/dvcPdm6PX+s4b8q0U29OMY3b3lmh/xwjHb8W6r68i9AxZlmSeQuWujOMulf3t7p8F9gEuNrP6+Hmic5qWPMQZL+OYs21HVAwaAmY2Ou4e4MMIg2Y9RrhyTPT/8qjS2G1mk8xsNKGr6It5D3hoLGf79h4T/f4CsE+0fz4IdLt7O0n2Tb6DzbWoIcHMaoBPAP9HGW5/tH23EHrEPBG9XBGffbJtr5TPvYytBA6LxgI6iO0DHpKivX4UOAjAzKYTxpIqNtlsVzXM/qSSAAAgAElEQVThyhK896CkWPS7XSmUwt9dxttlZpOA/yQMclmMt/Rl81ntCfzIzB4EPmxmxVgYyma7HgUOjn7+KCFPSPnIuL01s7Fx8x9Get+jXCvF9jSbtrI+7tei3Ndm9i3CsWL8WKSF3teQRdzFvr/NrNrCAzMAWoH26N9jhHMYCPt7RR7jhSxizrYdqertTXYrpQyGmR0M/BLYQhhQ6yzCQfWNhCezrAG+5uFpYh8jDCZXBfyru99TkKAHycxuIxzUtBCudF0JLCY8Uephd18QTXcCYTDSXuACd19pYYT8awmDRL5JGCG/9b1rKV5Jtr9v0MthwM3uflU0XVltv5nNBa4G+gYpu4/wPS/7z76fbTcq4HMvZ2b2VeB0oAv4MmEMrJfd/dH+2msz+zZhMNBqYJ67P1eQ4FPIdLvMrA74NeHveBQw34twoNsBtms+YXyWDwDPuftRpfJ3l8V2XQd8GnglWkTRPaEq021KmLfJi/dpYtl8B39C6DnaTRh09Y3CRC9DIYv29njgu4RjyI2EdmlT8qUXLO6ibE+ziHke8BWgjVCI/Yq7d+Uz5lRxE8bre4VQfOgF1rr7acWwr7OMu9j395PA/dFkw4Fr3P2maJ5LCQM0NxPa6bxeaMk05mzbERWDREREREREREQqiG4TExERERERERGpICoGiYiIiIiIiIhUEBWDREREREREREQqiIpBIiIiIiIiIiIVRMUgEREREREREZEKomKQiIiIiIiIiEgFUTFIRERERERERKSCqBgkIiIiIiIiIlJBVAwSEREREREREakgKgaJiIiIiIiIiFQQFYNERERERERERCqIikEiIiIiIiIiIhVExSARERERERERkQqiYpCIiIiIiIiISAVRMUhEREREREREpIKoGCQiIiIiIiIiUkFUDBIRERERERERqSAqBomIiIiIiIiIVBAVg0REREREREREKoiKQSIiIiIiIiIiFUTFIBERERERERGRCqJikIiIiIiIiIhIBVExSERERERERESkgqgYJCIiIiIiIiJSQVQMEhERERERERGpICoGiYiIiIiIiIhUEBWDREREREREREQqiIpBFcTM7jez4WY2zszOKnQ8hWJmj5jZvtHPD+Vxvf84iHkXmtlXcxDDCWZ2bg6WM93MPjPY5YhIaRmozazk/BLfTpvZL81s9zyt94tmNjHLeWeZ2S05iGGymS3OwXIq9vsjUioG0+aUm2hffD/6+VIz+2Se1vt5M9sry3l3N7M/5CiOnJxHDeYcSQZHxaAK4u7HunsnMA7QwRbg7vksaBS8oXP3e9z9JzlY1HQgb/vOzIbla10i0r802kzlF8Ddz3L3V/K0ui8CBT0xc/c33P30HCwqr98fM6syMx0Li2TmiyRpcyr978ndF7j7E3la3eeBrIpBuZTD86i8niPpvGK7mkIHUAnM7CvAuUAM+C93/46ZHQt8n1CQu8PdF0bTvgXcSjjRfhH4W3fviSrN/w7sBLzh7seY2eeBS6LXXgVOA4YDT7h7X8+XqcBP3f0IM3sF2Be4HJhmZs8ANwF/C/y9u/+fmdUAzwEN7t6eZFseAf4XOALoBr4GXAHsAfw/d78vWsa/AZ8EaoF/dvffmNkngB8BI4CNwGnu/qaZLQR2AaYAk4Cvufv/pNifuwN3Aw4cAvwaeBn4OtAJHOPum6IrtF+O9s//Al9291jCst5y9w9GjcIi4FPAmmh7Towmuyt67WBgmbt/LZr3OOCfo+35Q7T+3ZJNb2aXAxOifb7M3S+Ii+Eu4Mfuvjz6fSXwBXd/LWHTP2ZmZxIOli+K9ml/+zrpPjWzLwL7uvs3zWwf4OZovmXAdHefNdDnEe2rS4GdzOxo4ALgDOAWd38w+nxucfePR8vajfC92xk4E5gHfBT4tbt/J1rmt4HZhL+R+dE2zAK+Tfie7RT1aLoRGAb0Aoe7+yZEikwRtfmHEtqH0cD/AWe4e2tCrG8D/wXMAJ4n5ILWVPFGbeasKJZOwt/3je5+Oe/NL08BPwF6gK3u/ql+9tlC0msrziTkneHAXe7+XTMbG23DOKCKkItWpIgx1Wf3ShT3CcCbwHcJOW5nYI67P9VfLktYziPAV939eTP7GqGdfAd4C7jX3W+M1vUrwkH9ZuB4d99iZnsDPwPGA+sIn9vbyaYHPh3to7vN7F13nxEXw7lAvbtfFv3+A6DJ3RN78Ewws99G+/+WAfZ10n2a0O6Pjvbh3sDjwAkDfGfiJX5/1hHlrLjPZ1/g48C3CPlhKuG7OgE4FVgL/I27d6f4Hr9NyNUzgOPMbBEwme3fn0cQKRJmNhm4ndCWVwOnEAoyf3H366Np7gZ+ABhwLPB+wrHseYT88hng9+5+ZjT9W8Cd0esrgdsIx7SjCH8/r5jZzsAvCMeEbYRj6v1IaHMS/p5uN7Nh7r4gWs+PgT+6+6+TbNcX04z1Pcfb7h4zs3sJx6nDge+4+11RW5T0uD3F/n2Egc9tPkJof0cBHcCZ7r4mYTk3sv1Y+PPR57GJcF71irsvjNa1MtrGakL7+Ndk+9rd/5JsekJbdQIww8yagRnu3hzFcALwOXf/cvT714GJ7n5pwmbvZGZ3EtrP5YR8Fcvk3CZaftbnUXH7bYdzJOC/o3hmx30+XwXaSe/87xDgWsKx0grgnOi46pVonmOA86JjtQMJx2r/6u43U4EqtnqbL2bWAJwNfMrdDwCuMrORhAO94whfws+aWWM0y87Ane6+H+HzOcLMdgIWEw4IDyCcAAA86u4fc/fphAOuL7v7BuBVM5seTfMF4I6EsOYDz7r7dHe/ktC4/X303jHRct9TCIqz3t0PJPyB/TTajs8TTtwhXNV72d0PAWYC3zezWmA10BjN+3Pgn+KW+WHgMEKC6zsQnWxm9/cTwxTCgeBUQkJ8n7sfDPyecDAIcJu7H+LuDYSG9W9SbNNJhIPmKYSD/4a496YCCwkJ8BNmtreF7rHnA7Oi/d9DKKolnd7d50f7bXp8IShye7R+zOzDAEkKQX3b/CngcML3aAT972tIsk8T/DuwMPpOdSe81++87t4DLABuiLZnWZJlx5sENAIXEhrxbwPTgC+ZWV10wnoCIUkcBVxtZnXRvAcDX3T3w4F/IJzkTo/2Q8sA6xXJu2Jp881sOHAl4UDzIOCPhAOmRB8A7nP3qYQC09cHiDfegYT29wDgq1ERIDG/XAicG8V87AC7b6C2YipwNKEQcCBwsJl9lNC+fy7azr8hFMBSxdh32/TkfuJY7e7TgK3ARYS29aLoH6TOZTsws10IueLgaPsPSpjkxegzXsX23HUNcFaU064jHJgnnd7d7yYU3E6MLwRF7mT7wTiEdvaeJGF+nLCPpgFHmdlHU+xr6Gefxjk7inO/aH07x7030LyJ359UGgh/G4cQTtxejsv3nx7ge/wBQqF2f8KJ7dvRvAcQTgpFiskc4KGoHT2Y0FbfCMwFMLMJwN7u/vto+n0Jx/OfI5z8/ppwbHqAmVk0zc6Eguu+hOPVz7j7x4D/IBTiAX4MfNfdP0pol3/UT5sT//d0FXCyhV5CNYS/v/9KsW0pYx3gePv0qJ38FHCZmVVFr7/nODzaT7+Ma8cSDXRu8ybw6SjPnAd8r78NitqeHwOzotgsYZKWKO7/JBQ5IMm+7m96d19JaFvPjtrK5rhplwKHR/sekp8HQmg/LyPsq4nA32Z6bpOwvIzOo+JnHOAcKVE65383APOiPD6e8PfT55Vo21qAD7n7ftF09w2w3rKlYtDQm0WoEm8BiA7cDXjO3f/q7l2Eq8J9V0s3uftj0c9/AnaPpn+xrwIdLQNgNzN7yMyeJTTcU6PXbwdOjn4+kVCRTeWWuOlPJ5yEpHJv9P+zhCvS7UAToXIM8FlgXlThfQwYQ6hiv49wJeFZQkMxNW6Z90VFhr5t7ut23t/Jw3Pu/pK7dwAvAX09V+LjOMDMHo/W9zcJ60v0ScJVEdx9FeEKeZ817v5CFN+fo/g+QWjo/hBt52eAPVNMn8q9wLFREjuZcACfzF3u3uXua6P49qX/fQ1J9mmCA9y977O8NeG9gebNxAMeemQ9S2iE+z63Vwgnf58kXLHt8HB1/WlCwgBY7u5vRT//HrjYzP4JeL+7JxawRIrBLIqjzTdCG7Usah++RCjyJmp1974Cwa8JxZhU8cZ73N3Xe+ht9DLb2554TwA/MLNzCFfpUhmorfg0oe39X0I7MZXQ+6QqWsezhPZ0v7hlJo3Rw23Tb/QTR3yOeySKKT63pMpliQ4hnMQ1R9+JBxLe79v3fwJ2N7MxhH3939HntgDYtb/pU6yXKFe0mtleZnYw8JIn70253N3XeriN/C7Cd6C/fQ0Df+6fJMopUY6J742WzncmXU+4+0Z3X0foKdX3ufV9Vqm+x83uvjT6+VlglpldAXw04cRKpBg8Ccw1s+8Qij5t7v48MNzMdiP0rI4/jnsoOjZ/Fmh39yfcvZdQyO5rxzYmvP5w9Hp8W3cEcH3UFv0HoR1OZtvfU9TGrCL8rR1FaF/aUmzbQLGmOt6+wMxWEXq27A58MHo96XG4h9t3n+onjoHObXYCbjCzpmhfpGr3+9qeN6K2J/E8LFk7nmpfZ9LudxDy7iwz+wAwwd2fSz6pPxPt61sI7f5gzm0yPY/KVsrzPzMbB1S7e19R/yZ2PH65Pfr/pWj6q83s0+6+eRAxlTTdJlZ8OuJ+jhFuienPTwjdIh8xs5MJXcYhVOCXmdkSYHPcyXRSHrrUNVnoTr0/oSqeSmdcfJ3RMmK2/f7LKsIV6x3umzWzXxEKGjdGlfkfxr3dkWQ56cSwQxzsuM/+g9Bl8AULA5PV0b+qFO8l+0yqgP9293+InzDqnprJZ4i7N5vZc8ChhBO60/qZtDfh517639fb4k6xTwfc5gw+jx62F5cTT/je832J+32gZW87iXD3m83sSUJhb5mZHecJXXRFStCQtPnRQeCT7n5kBrH0tSuDib0rfgJ3/1cze4Dwd7vSzD7q7uv7Wd5AbUUVcK2773BF1sy+RLhNYLqHruDxvQYzao9TxBE/76X0n8sSpWpn4+PrW341sDa6cpnO9AO5g3DFdhzJrw5D/7kl2b6excCfe6b5NJX43AI75pd0jgP6E59b/hIVy44Hfmpm17r7DQPML5I37v6Ymc0ktKN3mtk57v47wsXb0wi9/v4ubpb4Y/P+jrvS+fvpBQ7yhCEWkmhN+P1GQq+lcYTbdVIZKNb+jrcPJxw3H+ruHVGRpq99yEm7n3AMfD7h9qTTCLek9ldUgszbfUi9r7Nt9z9C/72y+mv3sz23KYV2H6LvqrtvjHpyHwt8w8xmeHQbcaVRz6Ch9zAw28KYBpjZeEJjMsXCbVA1hC58j6dYxvPAR8xsStwyAMYCb1oYrG1u38Tu/i5hXIKFJO9l0kzoQRLvRkJ39NujKvFgPAR8LYqLuNsXxhK6WULo2jfURgPrLNxOdeoA0z5BdGXdzKYRet2k8gdCN/Rdo3km9P2cQsz6H1jvdsKYErXu/lI/05xoZrXRrQ37Er5H/e3rdKyKCoAQvoOZSPwOvUoYVBrCQUkmnmD7tn2Q0J1/deJEZrYHobfEj4BHeW+3W5FiUCxt/vPAHlF7hpmNtuRPHhllZn1FpVMJFwMyjTfeDm2Dme0ZXX28jNDD50NpLieZvn07Llr2rhZujxgLvBMVgk4mtP1DKZNc9iQhV9RZuP316FQTR1cnN5rZkQBRuzhlgHUky+l97iB08/88/Z8UzIg+69po2sfpf1+n4wminBLlmFFpzgfJc8sB0bIOZcdbzgaS1vc4yqkt7n4j4Ra9AzJYh8iQszCEwJvufi2hF8e06K1bgK8AXe7+8hCsejnRgO5mVt2XT0jd5gD8ltAb4wDC8dpg9He8PRbYEBWCDiXcPjSUxgJvRedIXxxg2ueBqWb2wahd/Xway+9vX/cn1WfwIKFXz6n0fxFgXzNrsHBXwimE3J/NuU2fTM+jEsWfI70G7GdmNdF3vyHFfDuIeqb1mFlfOz6HJJ0cLNwSV+XutxLGqsvk/KmsqBg0xNy9iXDv6RNRV8Zzo+6SZxO6iz8D/M7d++2NE3XdPh1YHC2j7zauy6JlrCR0t453O6EReE8xKLoq+2cz+7OZXRy9/FvCQMJLstrQHS0C3iYUG1YTurlDuHp6tZk9TRrjvVjqMYPScTmhe/syQtfKVO4gdKdfQzihcmBLfxO7+zuEz/C/zezPhG6KHxhgHf8JPGthML1EvyEki1S39DnhQHYZcEHUhbW/fZ2OCwj3WD9D6DXV7/YmsQw41Mz+FF2duY5wv/GfyPCpNu7+R8K9un8i7Mdz3T3Z92M20BTFO4aQ7ESKSrG0+dEy/g74ebSM35P8ySPvAH8TtR8fIfQGySjehNgT88uFZrY6aiefJ9w+kJVo3/4QeDRa3m2Ews/NhLGW/kwY3+ftgZZlqccMGkjauczdXyd8H54mtFmrGbitPY1wS+wqwv7/2ADT3wgsMbPl/ay/B/irb7/dMNHKaBlNwP+4+1Mp9nU6riGcaKwmjAPy1zTnS/b9WQG0WOg9++UMl5Xu93ga8GSUW/6B8HmJFJNZhL+LPxEGaf5P2FY8XkVujt2T+X/AMVFb1EQYRwdStDlRXDHCecWdg73AnOJ4eylh0OHVhOPZAXOLpR4zaCA/B86JOwZNFXMbYYy5xwjH7a8ycLvf377uzy3Ad83sGQu3F8evv51wIWIXd/9zP/P/mTAm6BrCgxDuzvLcpk9G51FJbDtHcvdXgd8R8uUPCfsjE2cC11m4lXszYV8l2pWQ354hjDP4/QzXUTaqensH2wlEykF05fEX/t4BKCuGmY12960WBjb7jUdP5ylXZjYKaHP3XgtP86qp1C6SIpXKoieBFDqOchaXW0YTihsnpegBWvKiXjjDoiv2M4DL3H1WgcMSKTsWHhTwDPCJYhvzxMyeIDwE4YVCx1IIce1+DWE8ou97mT+lsNLOo8qFxgwSzOwswtNKvljgUAptaVx1PdlTd8rNx4AfW7gf+mVCTwQREcmtf4l6UI4AFpVzIShSB/wuOgnqINzGIiI5ZGFYgP8CflZMhSAL49U9QXggQEUWgiJfN7PTCOPd3FfuhaBIpZ1HlQX1DJKkzGw+7x1H5l/cvb97T0VERFIys6MIjwGP96C7f7MQ8YiIyNCKxpBJvJXtz+6ui5AiBaZikIiIiIiIiIhIBSn4bWJmpmqUiEg/3H2gR5SWPeUJEZH+KU8oT4iIpNJfnih4MQjA3QsdgohI0TGzQodQNJQnRETeS3liO+UJEZH3SpUn9Gh5EREREREREZEKomKQiIiIiIiIiEgFGfA2MTP7BPAjoBNoAU6L5lsCjAEecveF0bTHA/OBXuA8d39yaMIWEREREREREZFspDNm0KvAp9291cy+CpwNjAOud/fbzew+M5sKOHAZMBMYC9wKNA5R3AB0b91MrL1lKFfxHtUj6qgZXZ/XdYqUspaWFtavX08sFit0KEWturqaCRMmUFdXV+hQJEPZ5CLlEpHtlCfSozwhxSAf51/KkZJIeSI9meaJAYtB7v5G3K+dQDehyLMgeu1eQgEoBvzF3ZuBZjOrNbMR7t6eyQZkItbeQsvqFUO1+KTq9msENU4iaXvnnXf40Ic+RG1tbaFDKWpdXV389a9/1UF+CcomFymXiGynPJEe5QkpBvk4/1KOlETKE+nJNE+kPWaQmU0Avg5cB4x297borU3A+OjfxrhZ+l4XkQpWVVWlhjsNtbW1VFVV/NOBRaQCKU+kR3lCRCqV8kR6Ms0TaRWDzGwUcDtwrru/C7Sa2Yjo7XpgA6EQNC5utr7XRURERERERESkSAxYDDKzGuAW4Gp3fyJ6eTlwbPTzMdHvLwD7mNloM/sg0D2Ut4iJSGna3NLB6+80D+rf5paOlOt4/fXXMTOWLl267bUvfelLzJs3L+n0K1eu5NJLLwXg8ssvp7m5OXcbnGDNmjU8/fTTQ7Z8EZFSpzyhPCEi0p9c5AjliSCdAaTnEA0KbWbnAfcBPwAWm9mFwMPuvhrAzBYCDxGeJnbBoKMTkbLT3NrJY39aO6hlzDxwF+rrdko5zf7778/SpUs56qijWLduHR0dHQwfPnzAZc+fP39QsQ1kzZo1bNy4kYMOOmhI1yMiUqqUJ5QnRET6k4scAcoTkN4A0ksIj5FPdEySae8B7hlURCIiOTBx4kTa2tpobW3lgQce4Oijj+bxxx/nhhtuYNmyZTQ3NzN79mxOPfXUHeabO3cuV111FcOHD+f888+np6eHvfbai9dee41FixYxd+5cpk6dypo1a6ivr+fqq68G4IorrqCpqYnu7m4uueQSGhoaOO6442hsbGTVqlVMmzaN+fPns3jxYlpaWnjkkUf42c9+xpgxYwqxe0REKp7yRP6YWS3wCLAfcJa732FmEwnnGGOAh9x9YTTt8cB8wsXl89z9yYIELSIVr9zzRNoDSIuIlJojjjiChx9+mOXLlzNjxgwATj31VBYvXsytt97Kr371q34fUXnbbbfR2NjIDTfcwNSpU3d4b8aMGSxevJiuri7cnUcffZRYLMaSJUu45ppruPLKKwFoa2tj9uzZ3HLLLaxcuZKWlhZOP/105syZw5IlS8riAF9EpJQpT+RNN3Ay8O9xr30DuN7dG4FDzGyqmQ0DLgOOBL4A/DjvkYqIxCnnPJHObWIiIiXpyCOPZN68eey5557bnkBw//33c/fdd1NVVcU777zD5s2bk877yiuv8LnPfQ6AhoYGHnzwwW3v9TXmkydPZvPmzbzwwgusWLGCuXPnArB161YARo4cyR577AHApEmT+l2XiIgUhvJEfrh7L/CmmcW/3AgsiH6+lzAsRQz4i7s3A81mVmtmIzQOqYgUSjnnCRWDRKRs1dfX09jYuK2KD7Bo0SLuu+8+AI4++mh6e3uTzvvhD3+YpqYmDj74YJ599tl+19Hb28tee+3FEUccwUUXXQRAZ2cnwHse7djb20ttbS09PT2D2i4REckN5YmCGu3ubdHPm4A9gPGEJxQT9/p44I08xyYiApR3nlAxSETyasyo4cw8cJdBLyNd55xzDhCeCADQ2NjInDlz2GeffRg7dmy/851yyimcf/75LFu2jN13333blYBkZs2axVNPPbWtkt/Q0MDFF1+cdNoDDzyQm266iaamJr73ve9RV1eX9raIiFQC5YmKyROtcb1+6oENhELQuLhp+l4XEQFykyP6lpOucs0TVf1VsfLFzHrdPat5O9evpWX1ihxHlFrdfo0MnzD4L59IpXj55Ze3dW0sJb29vfT09FBTU8Pdd9/Niy++2G+DnCuJ+8rMcPeqFLNUhMHkiXzIJhcpl4hspzyRvlLPE9GTh5uiAaSvBH7v7neZ2X8D3wIc+CMwgzCw9B3RmEIDLbeo80Q5ycf5l3KkJFKeSF8meUI9g0REkuju7uaMM86gqqqK6upqfvjDHxY6JBERKSLKE5kxs9uAjwItZnYo8ANgsZldCDzs7quj6RYCDxGeJnZBgcIVERm0Ys8TKgaJiCRRW1vLzTffXOgwRESkSClPZMbdT0ny8jFJprsHuGfoIxIRGVrFnidUDBIRkSFhZj8FDgKqgW8CK4EbgclAE3C2uyd/FqeIiIiIiAyZ6kIHICIi5cfM9gamuPsngS8AlwJnAk+5+wzC44OPLmCIIiIiIiIVS8UgEREZCm8DbWZWQ3gyzDrCgKD3Ru/fC8wsUGwiIiIiIhVNt4mJSF51b91MrL1lUMuoHlFHzej6ft9//fXXOemkk9hnn31oa2tj1qxZ2x4JmY41a9bwxz/+kTPOOGNQcVa4ZuBVwpNhRgEnEnoHbYze3wSML0xoIlLMlCdERKQ/ucgRoDwBKgaJSJ7F2lsG/UjSuv0aIUXjDTB9+nQWLVpEb28vJ554Iqeffjpjx45Na/lTpkxhypQpg4pR+CwwAdgb2JnQE+hFQi+ht4B6YEPBohORoqU8ISIi/clFjgDlCVAxSETKXFdXF729vdTU1HDFFVfQ1NREd3c3l1xyCQ0NDRx33HE0NjayatUqpk2bxvz581m5ciVLly5lwYIF/OY3v+G6665jt912o7m5mQsvvJDW1lZ+8YtfMHr0aF566SUWLFjAoYceWuhNLTbVwAZ3j5nZFmA08BhwLPB89P/SAsZXUL09PXSuX5vRPANdwRKR7ChPiIhIKuWaJ1QMEpGy9MwzzzB37lzWrl3L4YcfzpNPPkksFmPJkiVs2LCB8847jyVLltDW1sbs2bO55JJLOOGEE2hp2d7ttKenh2uvvZY777yTYcOGccIJJ2x7r7W1lV/+8pc899xz/PznP9dB/nv9FjjNzJYDI4DLgLuBG83sMWAN8EAB4yuoWGcbbS89k9E86VzBEpH0KU+IiEgq5Z4nVAwSkbLU160zFotx7rnn0tTUxIoVK5g7dy4AW7duBWDkyJHsscceAEyaNInNmzdvW8aGDRvYeeedGTlyJMAOXT333XdfqqqqmDRpEps2bcrXZpUMd+8B5iZ565R8xyIikozyhIiIpFLueULFIBEpa9XV1YwZMwYzo729nYsuugiAzs5OAKqqqnaYvre3d9vP48eP5+2336a9vZ1hw4bx/PPPb3svcT4RESlNyhMiIpJKueYJFYNEJK+qR9SF210GuYyB9HXr7O7uZvLkyRx22GHbXgNoaGjg4osvTrmMYcOGMW/ePGbPns0uu+zCxIkTqa2tHVTsIiKSmvKEiIj0Jxc5om85Ayn3PFEVX7UqBDPrdfes5u1cvzYnI4lnom6/RoZP2CWv6xQpZS+//PK2bpOlqKuri9raWjo7OznppJO4+eabGTNmzJCsK3FfmRnuXvGXlgeTJ/Ihm1w0cs/pWY0ZpPwj5Uh5In3KE8kVe54oJ/k4/1K+k0TKE+nLJE8M2DPIzGqBRyKOOaAAACAASURBVID9gLPc/Q4zWwh8AVgHvOXus6NpjwfmA73Aee7+5OA2RUSksO644w7uv/9+WlpaOOWUU4as4RYRkdKkPCEiIqkUa55I5zaxbuBkYF7C699x9zv6fjGzYYSnxcwExgK3AoPvvyUiUkBz5sxhzpw5hQ5DRESKlPJE9szsp8BBQDXwTWAlcCMwGWgCznb3WMECFBHJgWLNE9UDTeDuve7+ZpK3vm1my82sb6v2Bv7i7s3uvhaoNbMRuQxWRERERERKn5ntDUxx908S7ji4FDgTeMrdZwAx4OgChigiUtayHUD6andfaGb1wMNmtgIYD2yMm2ZT9Nobg4xRREpYbW0tW7ZsYezYsYUOpaht2bKlaAaTExHJJ+WJ9JRhnngbaDOzGmAcYfiJGcB3o/fvBQ4D7i9MeFIovT09dK5fO+TrqR5RR83o+iFfjwye8kR6Ms0TWRWD3H199P9mM/sdsD/wCqEh71MPbMhm+SJSPiZPnswbb7zB+vXrCx1KUautrWXy5MmFDkNEJO+UJ9JThnmiGXgVcGAUcCKhd1DfxeW+C8tSYWKdbRk/ZCEbdfs1gopBJUF5Ij2Z5omsikFmVh8VgmqATwC/BF4C9jGz0cAYoNvd27NZvoiUj+rqanbddddChyEiIkVKeaJifRaYQBhqYmdCT6AXCReX30IXlkUkojwxNNIqBpnZbcBHgRYzOxQYb2ZTgWHAze7+l2i6hcBDhKeJXTAkEYuIiIiISKmrBja4e8zMtgCjgceAY4Hno/+XFjA+EZGyllYxyP3/s3fnYZaU5cHG71mZYTbCos5gEFB5dFgEFNyGAdEooEETlUWDWzAY8VMEjVEMooZ8KvlMjCsugBARBCVBQEgIICAGMQI6MDxiwA1QYWCG6eltevn+qOqZQ9Onl9Nn6e5z/65rrjmn1udUV71v1VNvvZVHjXO6y4DLJhWRJEmSpJnuP4E3RsSNwAKKtxJfCpwbETcAa4HvtTA+SZrRau1AWpIkSZJqkpn9wHEjjBrXTWhJ0uSM+Wp5SZIkSZIkzRwmgyRJkiRJktqIySBJkiRJkqQ2Yp9BkqSGKN8++ffAPOBK4BzgfGAJcE1mnt666CRJkqT2ZcsgSVLdRcQ2wEeA12TmSzLzTOADwNmZuQo4ICJWtjRISZIkqU3ZMkiS1AgvALqASyJiLvA+YBVwWjn+cmA1cFdrwpMkSZLal8kgSVIjrABWAgcAuwBfARZlZlc5fj2wW4tiG1Xfpg0MdHeMe/qB3p4GRiNJkiTVn8kgSVIjPAr8IDM3AWsjYhmwKSIWZGY3sAx4pKURVjHQ3UHHnTeNe/qFu+/bwGja14aOHjZ29jZ8PUu2nc+yxds0fD2SJElTickgSVIj3AJ8KCJmA08CuoEbgSOA7wCHAx9qXXia6jZ29nLDbfc3fD2r99vZZJAkSWo7diAtSaq7zHwUOBf4PvBvwCnAp4C3R8RNwB2ZeWfrIpQkSZLaly2DJEkNkZlnA2cPG3x4K2KRJEmStJUtgyRJkiRJktqILYMkSZIkNV1EHAj8PTAPuBI4BzgfWAJck5mnty46SZrZbBkkSZIkqakiYhvgI8BrMvMlmXkm8AHg7MxcBRwQEStbGqQkzWC2DJIkSZLUbC8AuoBLImIu8D5gFXBaOf5yYDVwV2vCk6SZzWSQJEmSpGZbAawEDgB2Ab4CLMrMrnL8emC3FsUmSTOeySBJkiRJzfYo8IPM3ASsjYhlwKaIWJCZ3cAy4JGWRqhpp6e3j57N/eOadrCzl67+jTWtZ8m281m2eJua5pWmCpNBkiRJkprtFuBDETEbeBLQDdwIHAF8Bzgc+FDrwtN01LO5n18/OL4Ez3Y7dbBmXU9N61m9384mgzTtjZkMioh5wPXAnsDxmXlJROzICD39R8SrgFOBQeA9mXlrg+JumcH+fnrX3d/Udc5esJi5i5Y1dZ2SJElSo2TmoxFxLvB9ireJnQKsBc6LiJOBazPzzhaGOO30bdrAQHdHQ9cx0Ftb8mQq+qPF89iLzTXNu7DnYXrXPTbmdF7HaSobT8ugPuB1wAkVw4Z6+r84Iq4oe/pP4OMUHb0tBS6i6ARuRhno7aLr3tubus7Fe64CCxFJkiTNIJl5NnD2sMGHtyKWmWCgu4OOO29q6DoW7r5vQ5ffTHP6e1j/09q219LlS5i1aOyWQV7HaSob89XymTmYmQ8OG7yKood/2NrT/zOBn2fmxsy8H5gXEQvqGq0kSZIkSZImZcxkUBXDe/rfvvz3aMU0Q8MlSZIkSZI0RdSaDOqsaPUz1NP/o8B2FdP4BgBJkiRJkqQpptZk0FBP/1A813sjcA+wR0QsioinAH3layElSZIkSZI0RYzr1fIR8S3geUBHRBwIfIoRevqPiNOBayjeJvbehkQsSdI41fJmlZn0phRJkiRpJONKBmXmUSMMfkJP/5l5GXDZZIOSJM0MEbGKovXoTuWg84ElwDWZeXqj11/Lm1Vm0ptSJEmaCnp6++jZ3D+uaef29fPYptpuzPT3D9Y0n9SOxpUMkiSpRu8Fflx+/gBwdmZeHBFXRMTKzLyrhbFJkqQm6Nncz68f3Diuaefvsnnc0w63fMdta5pPake19hkkSdKoIuJVwE3ApnLQKuDy8vPlwOpWxCVJkiS1O1sGSZLqLiJmA+8E/hx4dTl4UWZ2lZ/XA7u1IjbVbkNHDxs7e5uyrp7e8T1OIEmSpIkzGSRJaoQ3AJdlZndEDA3rjIgF5ZsmlwGPtCw61WRjZy833HZ/U9a1f+w09kSSJEmqiY+JSZIaYW/gdRFxFbAPcAFFR9JHlOMPL79LkiRJajJbBkmS6i4zPzD0OSKup2gpNAs4LyJOBq7NzDtbFJ4kaYpo9VsnJaldmQySJDVUZh5S8fXwVsUhSZqSfOukJLWAySBJktS2+gcG+O0fanuF8UQt2XY+yxZv05R1SdNBxVsnh140sAo4rfw89NZJk0GS1AAmgyRJUtvq6u7jJ/lQU9a1er+dTQZJJd86KUmtZQfSkiRJkppty1snK4Z1RsSC8rNvnZSkBjIZJEmSJKnZfOukJLWQj4lJkiRJairfOqnpbHBwkMc29Yw9XWcvXf2190tnX3NqJJNBkiRJklrGt05quunrG+DBhzvHnG67nTpYs27spFE19jWnRvIxMUmSJEmSpDZiMkiSJEmSJKmNmAySJEmSJElqIyaDJEmSJEmS2ogdSEuSNA0M9vfTu+7+cU8/e8Fi5i5a1sCIJEmSNF3VnAyKiE3AreXXTwDfB84FVgBrgBMzc2CyAUqSJBjo7aLr3tvHPf3iPVeBySBJkiSNYDItg+6rfA1kRJwI/Dgzz4yIzwOHAVdOMj5JkiRJUp319PbRs7l/QvPM7evnsU0Tf1V6f//ghOeR1FiTSQb9cUTcAPwGeDdwEPDRctzlwMGYDJIkSZKkKadncz+/fnDjhOaZv8vmCc8DsHzHbSc8j6TGmkwy6OmZ+XBEvA04A9geeLQct778LkmSWmCifQyB/QxJkiS1i5qTQZn5cPnxQuAE4JfAdsDvgGXAI5MNTpO3oaOHjZ29TVvfkm3ns2zxNk1bn6SpKSJeCHwa6AU6gDdS1DnnA0uAazLz9JYF2AYm2scQ2M+QJElSu6gpGRQRi4DuzOyneBzsHuCHwBHA3eX/V9crSNVuY2cvN9w2sTvDk/Hi5yxvavIJTEBJU9SvgJdmZmdEvAM4keKGwdmZeXFEXBERKzPzrtaGOf1V6/Ohln4dBjt76eqv3vy/p3difUtIkiRpaqq1ZdCzgK9GxGNAD3A88DBwbtmP0Frge/UJUdNJV3cfP8mHmrrO1fvtbDJImmIy84GKr71AH7AKOK0cdjmwGjAZNEnV+nyopV+H7XbqYM266gmk/WOnCccnSZKkqaemZFBm/g+w3wijjppcOJKkmSQidgDeSfGGyTdmZlc5aj2wW8sCkyS1lI8TS1JrzW51AJKkmSkitgUuBt5d9jPXGRELytH2LSdJ7W3oceKDge9SPE78AYrHiVcBB0TEylYGKEkz2WTeJiZJ0ogiYi7FCwY+m5k3l4NvpOhT7jvA4cCHWhSe1BL9AwP89g8TfyVzLexPT1OdjxNLUmuZDJIkNcKxFCfxSyPiPcAVwKeA8yLiZODazLyzlQFKzdbMfvXsT0/ThY8TS1JrmAyaBgb7++ldV9sbwRb29LLXDpsnNE/37AX84iHfGCOpdpl5PkW/D8Md3uxYJElT0/DHiSOiMyIWZGY3M+hx4r5NGxjo7mj4egZ6J/YGSUntzWRQk1V7BfBo5nZ28MjaH9e0vv7+Qdb/YWKVz3b7rAbm1bS+Vmhms3uw6b0kSdJktdPjxAPdHXTceVPD17Nw930bvg5NzB8tnsdeTOzGfKWFPQ/Tu+6xUaeZvWAxcxctq3kdal8mg5qs2iuAR1PL64GHLN9x25rmm06a/Tp7m95LkiRNmo8Ta8ab09/D+p/WnghcunwJsxaNft2xeM9VYDJINTAZJEmSJKmpfJxYGtvg4CCPbRr98b/Bzl66+if3lIRPPrQnk0GSJNXZWI8Ez+3rH/Pkbrzz9PcPTji+asZqzr7dwPon9ENnP3OSJDVGX98ADz7cOeo02+3UwZp1k+svyicf2pPJIEmS6mysR4Jrefy32jz1fBx4rObsC3fclvXDTkqnWz9zkiRJMhkkSZKkSdjQ0cPGzt6mrMtHGSRJqg+TQZIkSarZxs5ebrjt/qasy0cZJEmqj9mtDkCSJEmSJEnNYzJIkiRJkiSpjfiYmDRB/QMD/PYPk3t940TYP4KkqWysN5AN59vHJEmaWpp1feN1zdTS9smgsV7/O1wtrwOuVM9XAKs1urr7+Ek+1LT12T+CpKlsrDeQDefbx5qjmTcuenpN7knSdNas6xuva6YWk0FjvP53uFpeB1ypnq8AliRJGkkzb1zsHzs1ZT2SJKl+2j4ZJEmSJEmSGsvH0aYWk0GSJKlpxtvH0HYD69lrh2I6+xlSK2zo6GFjZ2/D1+NFS2v1bdrAQHdHQ9cx0DuxLiYm2o3FkIl2Z2H3FTPDRPvuG0llnVtNPepiH0ebWkwG6QkmU6CMpyAZiSf61TW7w+q5c2bT1z/QtPV5Eiy1l/H2MbRwx21Z/3AnALu96KUsmGDdYr2iydrY2csNt93f8PV40dJaA90ddNw5/n7ParFw930nNP1Eu7EYMtHuLOy+YmaYaN99I6msc6uppS4ebqxrRevu5qp7Migi3g68FegF3paZ99Z7HWqsyRQo4ylIRmKHotU1u8Pq/WOnpq7vxc9Z3pQ7r0NMPrWe9YQmqpZ6yXplZrJj7PZgPSFNPc1IOll3N1ddk0ERsT1wPPBiYD/gE8BR9VyHpJnFt7O1F+sJSZNhx9gzn/WE1L7q8cgbwMKeh+ld91jV8bMXLGbuomWTXs90f5y43i2DDgSuz8w+4NaIiDovXzNUKx5NW7xkIR0bu5q2vlY0e3zGTnNYMNA9oXlq/X1DpvrvXND9EOt+8/DkVjh/EV2zFo57clsjPY71hCRpNNYTUpuqR+sjgKXLlzBrUfVz78V7roI6JIOm++PE9U4GbQ88WvF99nhmml5l/L+0wTrb4TdKapE2qCfGo5Yyb6LzNGMdtcwzVeOSNEVYT9SsGeXeTFlHs9YzU9bRrPVYdzdTvZNBjwL7VHwfs3lAZs6qcwySpKnLekKSNBrrCUlqgnong24BTouIOcBzgHvqvHxJ0vRmPSFJGo31hCQ1QV2TQZn5SER8HbgR2Az8ZT2XL0ma3qwnJEmjsZ6QpOaYNTg42OoYJEmSJEmS1CTj6pBNkiRJkiRJM4PJIEmSJEmSpDZiMkiSJEmSJKmN1PttYg0TEW8H3gr0Am/LzHsrxh0AfAaYBZyRmZe3JsrmG2O7fA54LfDtzHxXi0JsiWrbJSK2A74NzKfYX96dmT9pWaBNNsb+8m1gR2Bb4FOZeXFromy+0bZLOX4Z8L/AOzLzkhaEqAq11AcRcTrwMmAjcFxmPhwROwLnA0uAazLz9GbFFBGvBj4E9AD3A2/KzM0RcS6wdxnn7Zl5UhNj2hX4H+Bn5WQnZebtLd5OhwKnlZPsCPw8M/+8/Hu+HngI+F1mHtOgmJ5Qj0bEbODzwF7AA8BbMrMrIp4OfI2ifjknM79Sa0w1xnUC8DagD7h1aN+JiOuBeRQd8V6VmZ9oYkyHAN9g69ugjsnM39VzW9UQ0xuAvyoneSpwWWae3Ixjr9o5SLP2qZnCY7ahMR1Cg4/ZGuNq+HE7WlytPHZriKnh+1WNcR1CC/etUeJqWZ1Qjhvx+qvR561DpkXLoIjYHjgeWA28Hxi+8/4TxUnhnwAfL19FOeONY7v8A/CGZsfVamNslx6Ki66Dymk+2fwIW2Mc+8uxmXkwcCjwsSaH1zLj2C4A76N41a1arJb6ICL2BA7IzFXA2cDflNN+ADi7HH5ARKxsVkzA7cCqzFwN/Ao4qmL6EzLzkEmeeNRab/6wXPchmXl7Oaxl2ykzrx2KB/gexYnckI+U4yaTCKqlHj0cGCjrkR9TnOBRzvt+4CDg+HLZzYzrGuAFmfli4CkR8eKKca8ut9VkLiprPee4tGKf+l05rC7bqpaYMvOCin3qNh6/TzX62Kt2DtLwfWqm8JhteEzQwGO21rgafdyOI66WHLs1xtTQ/WoScUFr960R42pxnQAjXH81+ry10rRIBgEHAtdnZl9m3grE0IiIWADMzcz7M7MD+DnwzBbF2WxVtwtAZj4AtOPr4qpul8zsysz7y6+9FFnzdjHW/tJbflwM3Nns4Fpo1O0SEU8GdgdubUVweoJa6oODgKEWo5eX3wFWDRu+ulkxZeavMnNzOdnwsujzEXF9RLy0xnhqiqkcfUBE3BgRX4iIheWwlm2nivGzgD8F/q1ieR8uYz22xnhGjQmq1qPD96eh7bFHZt6amf3A9eWymxZXZv5vZg4Nq9ynBoFLI+LqiNi3mTGVXhURN0XEGeXfEeq3rWo+D4qIJRR382+uGNzQY2+Uc5Bm7FMzhcdsA2MqNfKYnUxcjTxuR42rhcfuhGNqwn5VU1yllu1bY10DtqJOKOMa6fqr0eetW0yXZND2wKMV32cPG7e+4vv6clg7GG27tLMxt0tZAH0a+FSzgpoCxrNd/hO4A7iqWUFNAWNtl1NpoxZk00At9cGWeTKzi6LCBVhUfq+ctlkxAVA2j34FcGk56H2Z+XyKu6KfrUjINCOmB4FnlHfNHgbeU45v+XaiOOG5PTM3ld8/m5n7Aq8C3hcRf9yAmMYzT2Wcsyqmmey5SM31e0S8EHhyZg61Znx9+Tc9GTinyTH9mOLEdzWwAji6HF6vbTWZ86A/B/694qKpGcceMOI5SDP2qZnCY7axMTX6mK01riGNOm7HFVcLjt2ar2kauF/VGteU2LdGuQZsZZ0w/Pqr0eet1YOZoh4Ftqv43j/KuGXAI80IagoYbbu0s/Fsl89QZGmva05IU8KY2yUz/wTYA3h/FP3ktIOq2yUidgO2y8yfNj0qVVNLfbBleNkCpaMc31l+r5y2WTEREU8C/hX4i6E7Q5n5cPn/A8AaYNdmxZSZPZm5sRx2IbBf+bml26l0LPDNoS+Zua78fwPwXxR38+od03jmqYyz8k72ZM9FaqrfI2IPipPcvxgaVrFP3QlsnsQJ7YRjysyOcr8aAC5m6z5Vr201mfOg4ftUM469IcPPQZqxT80UHrMNjKkJx2xNcVVo1HE73riafezWdE3T4P2qprim0L5V7RqwZXXCCNdfjT5v3WK6JINuAQ4u+37Yn60dTw1ly/oiYnlELKJoVv6LFsXZbFW3S5sbdbtExIeAvsz855ZE1zpVt0tEzI6IeeXXTqC7/NcORttf9gOeHhFXUVSop5bP8ap1aqkPbqB4rh/gCOCm8vON5XfK8Tc2K6aIWEzxXPp7M7PyWFxW/r8IWAn8tokxLa2Y/+CKeVq2nQDKsumlFH0GUQ4b2k5zgRdSdPBe15hGcQNbt8cRbN0eP4+I/aPof+lg4Ec1xlRTXBGxnCK5eFxmPlQxfGn5/1N4/F3FZsRUeVOhcp+q17aq6TyoTMQuz8w7hsfayGOvXP5I5yDN2KdmCo/ZxsbU6GO2prjK2Bp53I4ZV4uO3QnH1IT9qta4Wr5vVbsGbFWdMMr1V6PPW7eYNTg4PbqUiYh3AG+i6AH9LymembsvM78fEc+n6HhyFvB/M/Oy1kXaXGNsl1Mpmrw9CbgrM1/Rukibq9p2Ae4FfklxUA0C92fmG1sUZtONsl1uBa4sJ5sPfD4zv9GSIFtgtOOoYprTgTXp28Rarpb6ICI+RtE530aKDgQfioidgPMo3spwbWae9sS1NSamiPg74ETg7nIRX8vM8yPiSoq7PXOBT2fmRU2M6VXARynuQD1K8VaU9a3cTuU8rwT+LDOPr1jOVylOzuYAF2TmZxoU0xPq0SjeHvPFcv0PUmynzoh4BsUbUeYBX8/Ms2qNqca4vkaRNPtluYgzKFpN3UpxkjkXODUzr21iTCcAbwe6KBJ2b8/irXl121a1nAdFxInAH2Xm31csp+HHHlXOQZq1T80UHrMNjanhx2wtcZXzNPS4HS0uWnjs1hBTw/erGuNq6b5VLa5ynlbVCVWvvxp93jpk2iSDJEmSJEmSNHnT5TExSZIkSZIk1YHJIEmSJEmSpDZiMkiSJEmSJKmNmAySJEmSJElqIyaDJEmSJEmS2ojJIEmSJEmSpDZiMkiSJEmSJKmNmAySJEmSJElqIyaDJEmSJEmS2ojJIEmSJEmSpDZiMkiSJEmSJKmNmAySJEmSJElqIyaDJEmSJEmS2ojJIEmSJEmSpDZiMkiSJEmSJKmNmAySJEmSJElqIyaDJEmSJEmS2ojJIEmSJEmSpDZiMkiSJEmSJKmNmAySJEmSJElqIyaDJEmSJEmS2ojJIEmSJEmSpDZiMkiSJEmSJKmNmAySJEmSJElqIyaDJEmSJEmS2ojJIEmSJEmSpDZiMkiSJEmSJKmNmAyqIiI+HRFrIuL9k1zOIRHxvHFMd31EPGucy7wwIg4pP18zmfhqERFXRsT8Zq93qir/xheWn98REa9r4nrH3LdGmf93dYqjLvtDRJwUEXPrEZOk8Wl1fdIolXVqM39XRLxvEvOeHhHvqEMMR0bEu+uwnH0j4mWTXY6k1omI7SLi+EnMP+PPzYbOhyPieRHxiSatc9fJXC9M5LpxjOV8LCJeVIflvCYinjHZ5aj5TAZV9wZg78w8s3JgRMyZ4HIOAWq+YB9LZjb9RC0zj8jM3mavdzrIzC9l5iVNWt0hNHDfGq867g8nAU074ajhWJamhVr37VbUJ83Q5N9VczKoXjLzssz8lzosal+gadvOMllqiO2AmpNBTPDcLCJmRcS0vL7MzB9n5t82aXW7Ak25eTyazDwtM2+uw6JeAzQtGWR9UT8zOtNbq4i4BNgBuC0iPgB8ELgNWA2cGRHbAX8JbAP8D/CXmTlQZmjPoih4u4GjgHcAm8u7fUcDLxlp3nHE9PFyeb8CZlUM/11mPqW8s/shoA9YCXyi/A1HA/cDf5qZfRFxIPD/gEXA/wJvzszOMit+EcWJ3y+APy9X8XVgP2AA+L+ZeUFE/BJ4VmZ2R8SHgWPK8adm5nfLWD4I9ALPAs7NzDOq/K7Jxv1lYH9gIfCFzPz80HYZ/nsys3+U7Xsu0AE8H1gMvAk4DXg28KnM/HJELAX+jeLvOwv4P5l507DlnA78LjO/FBEvBL4K9AC3AAsz8y3lujYALwSWAMdm5u0RsRj4YrnNKJf/3yNND6xj2L6VmVnGsA/wT5n50vL7EcDrMvNtw372rIg4CzgIuBv4i3Kbjnsfycz+YfvD3wOvB35druOMzLx+rL9HRJwIrABuiYi7gQ8AF2bmCyr+Phdm5lXlsr5dLusW4FvA3wHbUuwvvyzvTpwDLCt/21szc1NEXM/jj+XnAn9a/o3+dXjyV5ouImJXivLp58BzikFxObAcmA98JDO/U047nvrkHZl5TDn8eory5jHgYoqyYTZwVGbePUIsewPnAnOAQeAlmbk+Ik6lOGHcBvhsZn5ljN9zKZDAAcA3gfuAd1LULYeXy3wHY9SpFb9rDkUd/WJgLbAb8GflZN8phz0XuC4z/7qc95UU5csC4L/L9e8y0vQRcQawQ0TcXg57b0UM36Eol28sv98CvD4zh8rKIc+PiLdR1DOnlHXqXIoy+UXAPODvyuGnAztT1FPLgb/OzP+IiLdQlMl/GxF7ABeU810H7JuZh1SbtyLeOcDHgG0i4jDgvcCb2VoO71p+fkG5rF0o6q4nA28DTqC4WfHNzPxIucxq5wsfpjgH2KZs0XQuw/YdJNXqDGDvslz6BvBPjFyeXEVRLl8REf9EcQ7exfjPzX5PUS4eBLwyIv6CcZT3EXEMxfn2ZuDnmfn6aufD1X5gWeYdAexEUa6/h+Ic8WXAD4fOf6tdM1Qs5xDKui8inkxx3rojcAXFefau5boOK9e1K/CxzPx6Of8T6rhRpj8DWFn+Xc7MzG+Uy5gD3APsmZld5TXnDzPz2SP89BMi4lCKOvHYzPxFGfeXKcr3Loo68edlPX5LuU1mA0dm5m+G/Q1fA3wKWA/cBfwyM0+vNm/Fdns+cCRwUERspNgHfsbWa4O3sLVOup6irj6Uotz/a+CT5d/t/5T737bA14C9KM473lr+htOBp1HsFzdHxK0M23dG2EYaw7TM3DZaZr4OWJeZ+2bm1eXg7sx8bmZeCHwrMw/IzH0oDrQ/Lac5Hzg9M58DvBz4DfAligvifcuL9WrzVlVenP8JxUHxVuAFVSbdB3gjxYnzJ4H7Ktbz0ige5TmT4iDeH/gRxYktFCdw387MPSn2i0Mp7gr+cWbumZl7UxSGw+M6kuJk+BXAZ8sCHIoE0lsoLkreERGLRvmJA2I1ygAAIABJREFUk4n7bzPzeWWsb4mIHUf5PUPNIY+sEsf8zDyA4m/2LYqE0AuAU8vxXcCryxj+lKIyHc1XKBI3z6WoTCotzswDKU6Chx5F/HAZ8wEUFyifrzZ9WQgP37cAyMyfAk+p2BavB0ZqrfQk4IrMXElxUfjOGvaRLcr94VCK/fTNFIk1xjNvWSE/ADw/M48eIdZKT6Y4oXkWsCfwssx8PsX2PqGc5l8oLrz2AX5JcSEzpDsznwv8J8VdmZXlMfvlMdYrTXV7Ah/NzCi/v6nc118MfLy8Yzve+mQkxwLXZOa+FOXar6pM91fA58rpXgx0lAmFncry7XnA8RGxPCJWRMSVVZbzbIqbBSsp6pM/Kn/PDyluGMDE6tTXAsvKk+qPUtQ9Q1YCp1NswxdGxDPLMvQk4JDyt/Sz9UbJE6bPzFPZeu5QWeZAkUR7LUBEPA1ghETQ0G9+McWNo89ExAKKu/r3ldtuNfCJiJhXTv804GCK5N5HRljeP7P1vKRv2Liq85bJ+tOAc8rfc90Iy660HFgFnEyRxPswsDfw1ohYPMb5wnOBt2TmSxhh3xljvZJGdyrws/I4PpPq5clfAf8YEa+gqBf+aYLnZk8C/i0z96Iox55Q3o8S3yvLMmqoBdOI58NRPAL7sSrLeRZwOPBqipsH36Qop58TEUN1YrVrhpF8BLik/D2/GTZuT4q65sUUdQnV6rhq05e/++ry7/KNoQWXZe9/UpSTlL/nu1ViHCy3299TlPVQJPs+Wv7Ok4FPV0zfUdah/0pxg2eLiFhYzntIGWfweFXnzcxbgMuAE8vfs7FKvEPWZeZ+wE3A54BXUiTRPlyOfxfwUHnteQbwhYp5nw6szsxTGHnf0QSZDBq/iys+PycifhARP6M4uFeWrUaWDJ0wZeaGHLnFzxPmHce6XwR8JzM3Z+b9wI1Vprs5Mx/NzIcoWpJcXg5fQ3HXLihOfq8rM9FvpTgZBFifmTeUn2+jyF7fC+wSEZ+NiJdm5oYR4rokM3sy80HgJxQFHsAPMnNdZnZS3M1dMcrvm0zcb4iI2yiyzM8AnjnK7xlqDnlZlTiG1vsz4LYypocpWtDMo7iD/qnyb3d5xW99gjKTPyszf5qZgxTJpUpDMWyJjeIC7aPlb7wceFJsfU57pOlHcynwmnL+1cBIfWZ0VmyLb1KcyE90H6n0IuDScj99gMfvp2PNOxGPZubN5Xa9E7i2HD60v0Bx9/s75ed/pajYhgwdyxsoLjS+WiYIvejQdLc2M++s+P7eiLiD4ljcFXgK469PRnIrcFxEfAR4ZmZ2VZnuh8D7I+JvKE6O+yjKtz8ty5UfUVw4PD0zH8jMI6os567MvDczeyjqo6GWK5XH+kTq1BdRlsWZeQdFq8EhazPznvJE/KcU2+uFFOXhf5dxvwzYfZTpR3M5cEREzKJIQn+7ynSVf5u7KS5w/oTiDvDtwA0ULUSH6tQryhiqlavPycyhuu2iYePGmncivlee8/yM4m7y0N/tlxSJotHOF27MzKE+7EbadyTVz4jlSZmc/hxFC9O35yit6avYmFtvoI9Y3leZ7wfAV8oWkYMV8z/hfDiLR2BPq7KcazKzm6IM6h52njhUX1S7ZhjJi9haZg4vO6/JzM6y3JpdXiOM9ptHmn40F7P1EbLXMfINXYALATLz3yluwkNxs/XsMo6vUJS/Q0a7lgiKOveBzNxM0cqr0kSvQ0ZTeb11c/l3q6zXX0Rx7k5mXkmRXBzyb2V8MPK+ownyMbHx66z4/BWKJur3RNFZ5OIq84yk1nkHq3yuVNlvy0DF9wGKJtezgFsz8+UjzNszbN45mfloFI8cHQF8ICIOyszTxxnvE5Y3yrQ1xR0Ru1O0BHlRZm6M4pGIbWpY//A4KmOonP9Yisct9s3i8ajRkgezRhlXGV9lbLMo9o0HKicsb2iMNP1oLgH+L8XdjB/k2H36DJb/JrSPDBs/2m+e6N+jn8cnq7ep+DzW/jKWToAsHj98HsXdlzdS3PF/yzjml6aqLfVURLwEOBA4MDN7ImINW4+jseqTEY+/zLwhIlZTJF2+HRHvysz/Gj5zFo8T31pOd10Uj1rNongc4YIJ/J7xHOsTqVMnWkbNAv49M/+qcsIoHpGaUJlW1lF3UfxNXkdR5oxk+N9mqFz+yxzWr0Nl3ZDFo+ojxTDmbx5l3uEq94ttho0bq/4czZb9dqR9JzPXjiM2SeMzYnlSGnosp1qLmdHOzSqvkyZS3v81ReL9SIrH0famyvnwGHphS3n2hDJojGuGkdRSXzzhN0fRbcFEz4GvB74UETsBe2Tmj6pMN1JdPgjsX6VBwmjXErVct4xmQvXFBOqhyv3sCfuONxAmzpZBtVkEPFQ23z4aIDMfAzaUJ+BExLIoOlDbSJF1rzrvONwM/FlEzIuIFRTPYtbibmC3sqAlIhbFKD2/l80nZ2XmRRTN9PYdJa6nUGSl76T+qsW9hGL7dpQn5wc3YN2VlgJ/KBNBr6P4W44oMx8FBodiZnydxF0DnDj0JSKeM8b0w/etyvXfTpFhfzvV7yhsGxGvKj8fTdFcc0L7yDA3U7ZGKpvGrhrnfEMqf8/vgaeW69+Oie/zt0fE0OMib6D4bY9TPqKwrGwd9Tc8cf+WprOlwCNlIuhAtt5ZG0998mtgz/JYfhrl41Tl5wcz84sUdyT3HmFeImI34BeZ+Wng+xR3HK8B/rKs+4jCgjr8zonUqTdTlsVlGTfWm1j+m+JR5aeW8+ww9HkUA1G989SLKR5ZnZeZ91aZpvJv8yyKPpOuAf56aLkRMZGy6o4o+o2D4pHhiRhex/yKreVktcetqxnX+UKVfUdS7YYfxyOWJxFxMEVrvYMpHuNcOML84z03m0h5v1uZmPoQRR9GS5j4+fB4TPSa4Wa2lpnjKTsnWseNdg7fR9HH278wrIuOYY4u1/UqihY7ULT2Pb4cPrviOmQsd1M86fKUsuXSa8Y535AR64tyP3vVyLNUdTNbf9thFP0XjWSkfUcTZDKoNmdQNHG+jq0HHxR9zHw0imb5V1HsmN+laFZ/WxS38arNW1WZEf4viiZ051CcoE5Y2TrkDRTZ5jsommOPdqH/VOD7ZVPD/0fRufPwuK6g+B3/Abw7M+v+qE21uMtm/v9LUYB9gaK54Khi9D6DxnIBcGhE/JTi0avfjzH9CcBFEfETikeQHhtj+o8BKyLip+Ud5LGefx2+bw337xR9V109wjiAP1A0ab2ToinrF2vYR7bI4pnh71Oc4H8duIOxf3OlrwA3RsRFZRz/XC7jwvL/iXg38Dfl32p3tj5LXWkJcEX5O/+dopNYaaa4mqIz4zspEhB3wPjqk8z8VTnNncA/ltNC0ZfAT6NoZn8QZTPuERwDrCnrjiXAVWVT76uBW6NopfRFiru1o/UZNB4TqVMvATojYi1Ffz/JKGVUZv6B4oLk38uy5D8omv6P5l+Bn0XRAetw36U4wR7e/P5xq6Woy64D3ls2nz+Lor65o/x7VntMYiTvpegv6naKVlMTKZOvAw4s65iXUHTo+efl33+0vjaeYALnC0/YdyayHkmPl5nrKMrtn0bE+xmhPCkTP5+jeDzsbopy7OPlIiZ8blatvK8S4v+L4jHfnwJfL2+mjng+HKP3GTTWdpjoNcNHgaPL2J7JGGXnBH8zFL93YUTcHhEjtRS9mKI8HO0NxXPKc9iPsLVvzP8DHF4OX0PRH8+Yyse+T6F4dPAHFMmcidQXF1I+2hcRSyj6MTqPIrEzkRZeUOyLK8p69++oSAwOM9K+owmaNTjoI3ZSI0TEoizeYDWLonPiH2TmuS0Oq6EqfvOOFImk/XPsjuQkqSkqyqhnAt/NzLFaB01rUbyVpSszB6N4m9fcCTzuLUltqWzVs7l8GuAvKF5W8pYWh9VQFfXjXIp+fT6Rmde3OCw1mH0GSY3z6ig6wZxP0fHqRPrKmK7OLlspDb2u1ESQpKnk6vKuJWx9U+JM9nzgn6Loi+E+ihbMkqTR7Qp8syw7H6E9ys53lq2UtqF4wcD1LY5HTWDLoCkmIm7hiR1tHTiODoCntIjYgeKxg0oPZ+bLWhGPJGlyLNclSeMRxSvrPzls8FWZ+betiEdSwWSQJEmSJElSG/ExMUnSpJRvnrie4k0kx2fmJRHxFYrX1M4GPpeZ55fTbqJ4bBKK59HtIFaSJElqspYngyLCpkmSVEVmzmp1DOPQR/HK7hMqhn0qM++JiG0o3lryzfJ1qfdl5iETWbj1hCRVN03qiYaynpCk6qrVEy1PBgFkZqtDkKQpp+iLe+rLzEHgwcp4M/Oe8mMv0A8Mnaj/cUTcAPyG4vXS68a5jvoFLEkzxHSpJ5rRgtR6QpKeaLR6YnYT45AktZ9TgG9lZn/5/emZuZqi4+EzWheWJKmJhlqQ/nPFsE9l5guB1cCp5SutoWxBWv7zUWJJapAp0TJIkjTzRMTrgBcARw0Ny8yHy48X8vjHyiRJM1QzWpBKkibGZJCkhhoYGOCBBx5g8+bNrQ5lSps3bx4rVqxg9uyZ0WAzIg4B3gm8MjMHymGLgO6yldDBwD3VlyCpXVhPjM9MqycqjNSC9OGIeBtFC9J3tC40SVOB9cT4TLSeMBnUQH2bNjDQ3dGUdc1esJi5i5Y1ZV3SRDzwwAMsXbqUpUuXtjqUKe2xxx7jgQce4KlPfWqrQ6lJRHwLeB7QEREHAq8FNgLfK+8Evx7YBfhqRDwG9ADHtyjcLcZTTlu+So1lPTE+072eGEk7tyBt5nVCJes0TUfWE+Mz0XrCZFADDXR30HHnTU1Z1+I9V4EFu6agzZs3W3CPw9KlS1m3bvq2hM/Mo4YN+psRJnsI2K8J4YzbeMppy1epsawnxme61xPDtXsL0mZeJ1SyTtN0ZD0xPhOtJ0wGSZIkSWqo6dqCVJJmqhn30LEk/fa3vyUiuPrqq7cMe+tb38oJJ4zc2vyWW27hYx/7GABnnHEGGzdubFhsa9eu5Sc/+UnDli9JGpv1RPNl5lGZuXtm7pOZf5OZT8/MfSveHPZQZv5PZu6XmQdn5ssz89etjltSe2qHesKWQZKaakNHDxs7eye1jCXbzmfZ4m1GnWavvfbi6quv5hWveAUPPfQQPT09zJ8/f8xln3rqqZOKbSxr167l0UcfZf/992/oeiRpurKesJ6QpGrqUUeA9QSYDJLUZBs7e7nhtvsntYzV++08ZuG944470tXVRWdnJ9/73vc47LDD+MEPfsA555zDddddx8aNGznmmGM4+uijHzffcccdx2c+8xnmz5/PSSedRH9/P894xjP49a9/zVlnncVxxx3HypUrWbt2LcuWLeOzn/0sAJ/85CdZs2YNfX19fPCDH2Sfffbhla98JatWreKOO+5g77335tRTT+W8886jo6OD66+/ni984QssWbJkUttCkmYa6wnrCUmqph51BFhPgI+JSZrBDj30UK699lpuvPFGDjroIACOPvpozjvvPC666CK+/vWvMzAwMOK83/rWt1i1ahXnnHMOK1eufNy4gw46iPPOO4/NmzeTmXz/+99nYGCA888/n89//vOceeaZAHR1dXHMMcdw4YUXcsstt9DR0cGb3vQmjj32WM4//3xP8CWpxawnJEmjmcn1hC2DJM1YL3/5yznhhBPYfffdmTdvHgBXXnkll156KbNmzeIPf/gDGzZsGHHeX/7yl7z61a8GYJ999uGqq67aMm6oMF+xYgUbNmzgnnvu4aabbuK4444DYNOmTQAsXLiQ3XbbDYDly5dXXZckqTWsJyRJo5nJ9YTJoBlisL+f3nWTby43XrMXLGaur6XUFLds2TJWrVq1JYsPcNZZZ3HFFVcAcNhhhzE4ODjivE972tNYs2YNz33uc/nZz35WdR2Dg4M84xnP4NBDD+WUU04BoLe3eI551qxZT5h23rx59Pf3T+p3SZLqw3pCkjSamVxPjJkMioh5wPXAnsDxmXlJROwInA8sAa7JzNPLaV8FnAoMAu/JzFsnHaHGZaC3i657b2/a+hbvuQpMBmkaeNe73gUUbwQAWLVqFcceeyx77LEHS5curTrfUUcdxUknncR1113HrrvuuuVOwEgOOeQQfvzjH2/J5O+zzz68//3vH3Ha/fbbj2984xusWbOGf/iHf2Dx4sW1/jRJUh1YT0iSRjNT64lZ1bJYQyJiFvAU4ARgTZkMOhP4UWZeHBFXAO8HEvgxsBpYClyUmavGCiAiBjOzpuCnut5199Nx501NWdfC3fdtejJo/g47N219mr7uu+++LU0boXlviZmswcFB+vv7mTt3Lpdeeim/+MUvqhbI9TJ8W0UEmTlrlFnaQiPrifGU05Z3UmNZT4yf9cTIpuP1RDOvEypZp2k6qiz7mvk2scma6vXEmC2DMnMQeDAiKgevAk4rP19OkQAaAH6emRuBjRExLyIWZGZ3bT9D0ky0bPE2DS9466Gvr483v/nNzJo1i9mzZ/OP//iPrQ5JktqC9YQkqZrpUkfA1K8nau0zaFFmdpWf1wO7AdsDj1ZMs74c9kDt4UlSa8ybN48LLrig1WFIkqYo6wlJ0mimej1R66vlOyNiQfl5GfAIRSJou4pphoZLkiRJkiRpiqg1GXQjcET5+fDy+z3AHhGxKCKeAvT5iJgkSZIkSdLUMq7HxCLiW8DzgI6IOBD4FHBeRJwMXJuZd5bTnQ5cQ/E2sfc2JGJJkiRJkiTVbFzJoMw8aoTBh48w3WXAZZMNSpI0fUTEPOB6YE/g+PKtkzsC5wNLgGsy8/Ry2lcBp1LcNHhPZt7akqAlSZKkNlZrB9KSVJO+TRsY6O6Y1DJmL1jM3EXLqo7/7W9/y2tf+1r22GMPurq6OOSQQ3jXu9417uWvXbuWH/3oR7z5zW+eVJxtpA94HXBCxbAPAGdn5sURcUVErAQS+DjFGyiXAhdRvJ1SkrawnpAkVVOPOgKsJ8BkkKQmG+juoOPOmya1jMV7roJRCm+Afffdl7POOovBwUH+7M/+jDe96U0sXbp0XMt/9rOfzbOf/exJxdhOMnMQeDAiKgevAk4rP19OkQAaAH6emRuBjRExLyIW2L+cpErWEzOPLUgl1Us96giwngCTQZJmuM2bNzM4OMjcuXP55Cc/yZo1a+jr6+ODH/wg++yzD6985StZtWoVd9xxB3vvvTennnoqt9xyC1dffTWnnXYa3/3ud/na177GLrvswsaNGzn55JPp7Ozky1/+MosWLeLee+/ltNNO48ADD2z1T51qFmVmV/l5PbAbsD3FmyepGL498ECTY5OkLawnmsIWpJKmrZlaT5gMkjQj3X777Rx33HHcf//9vOQlL+HWW29lYGCA888/n0ceeYT3vOc9nH/++XR1dXHMMcfwwQ9+kCOPPJKOjq3NTvv7+/niF7/It7/9bebMmcORRx65ZVxnZydf/epXueuuu/jSl77U7if5I+msaPWzDHiEIhG0XcU0Q8MlqemsJ5rHFqSSpqOZXk+YDJI0Iw016xwYGODd7343a9as4aabbuK4444DYNOmTQAsXLiQ3XbbDYDly5ezYcOGLct45JFHePKTn8zChQsBHtfU81nPehazZs1i+fLlrF+/vlk/azq5ETgC+A7FCwc+BNwD7BERiygeC+jzBF9Sq1hPtJwtSCVNaTO9njAZJGlGmz17NkuWLCEi6O7u5pRTTgGgt7cXgFmzZj1u+sHBwS2ft99+e37/+9/T3d3NnDlzuPvuu7eMGz5fu4uIbwHPAzoi4kDgU8B5EXEycG1m3llOdzpwDUVfEO9tUbiStIX1RMvYglTStDBT6wmTQZKaavaCxUWHbZNcxliGmnX29fWxYsUKDj744C3DAPbZZx/e//73j7qMOXPmcMIJJ3DMMcew8847s+OOOzJv3rxJxT5TZeZRIww+fITpLgMua3xEkqYr64m2YQtSSRNWjzpiaDljmen1xKzKrFUrRMRgZrY0hkbpXXd/XXo6H4+Fu+9L1723N2VdUPS+Pn+HnZu2Pk1f991335Zmk9PR5s2bmTdvHr29vbz2ta/lggsuYMmSJQ1Z1/BtFRFkZtvfWm5kPTGectryTmos64nxm871RGULUuAq4EzgPIqkz7WZeVo53ZHABylbkGbmLeNY9rS7nmjmdUIl6zRNR9YT4zeResKWQZI0iksuuYQrr7ySjo4OjjrqqIYV3JKk6cl6YnxsQSqpXU3VesJkkCSN4thjj+XYY49tdRiSpCnKekKSNJqpWk+YDJIkSZLUlvo2bWCgu2PsCetsoLen6euUpEo1J4Mi4nPA/sBs4G+BW4BzgRXAGuDEzByoQ4ySprF58+bx2GOPsXTp0laHMqU99thjU6YzOUlqJuuJ8bGeaIyB7o6W9N2zcPd9m75OabqynhifidYTNSWDIuKZwLMz80UR8cfAN4CLgB9n5pkR8XngMODKWpYvaeZYsWIFDzzwAOvWrWt1KFPavHnzWLFiRavDkKSms54YH+sJSe3KemJ8JlpP1Noy6PdAV0TMBbYDHgIOAj5ajr8cOBiTQVLbmz17Nk996lNbHYYkaYqynpAkjcZ6ojFm1zjfRuBXQAL/QfFqyO2BR8vx68vvkiRJkiRJmkJqbRn0J8AOwDOBJ1O0BPoFRSuh3wHLgEfqEaAkSZIkSZLqp9Zk0GzgkcwciIjHgEXADcARwN3l/1fXJ0RJkhqnp7ePns39VccPdvbS1b+xrutcsu18li3epq7LlCRJksar1mTQfwJvjIgbgQXAx4FLgXMj4gZgLfC9+oQoSVLj9Gzu59cPVk/2bLdTB2vW1fcVwKv329lkkCRJklqmpmRQZvYDx40w6qjJhSNJkiRJkqRGqrVlkCRJbeGPFs9jLzZXHd89ewG/eKj6Y2aSJE0Vg/399K67v6nrnL1gMXMXLWvqOiWNzWSQJEmjmNPfw/qf3lR1/Hb7rAbmNS8gSZJqNNDbRde9tzd1nYv3XAUmg6Qpx2SQJKnuImIl8IXy62KKFw/8C/B3wG8AMvOQlgQnSZIktTmTQZKkusvMu4BDACLiRGC7ctQXM/MfWxWXJGlq8KaBJLWWySBJUqMdC7wJWA0cHxGvAS7JzH9ubViSpFbxpoEktdbsVgcgSZq5ImJXYE5m3gv8G7An8FLgiIh4QStjkyRNGccC3yw/Hx8RN0XESa0MSJJmOpNBkqRGOga4ECAz12dmf2b2UCSG9mtpZJKklvOmgSS1hskgSVIjHQ1cBBARy8r/Z1E8MnZPC+OSJE0N3jSQpBYwGSRJaoiI2BNYl5m/KwedEhH/DdwM/Cozr2lddJKkKcKbBpLUAnYgLUlqiMy8E3hZxffTgNNaF5EkaSqpctPg5cAgcIM3DSSpcUwGSZLUZP0DA/z2Dxubus4l285n2eJtmrpOSRqNNw0kqXVqTgZFxIHA3wPzgCuBc4DzgSXANZl5ej0ClCRppunq7uMn+VBT17l6v51NBkmSJAmoMRkUEdsAHwFek5md5bAzgbMz8+KIuCIiVmbmXXWMVZKkCenbtIGB7o6q4wd6e5oYjSRJkjQ11Noy6AVAF3BJRMwF3gesYmuzzsspOn0zGSRJapmB7g467ryp6viFu+/bxGgkSZKkqaHWZNAKYCVwALAL8BVgUWZ2lePXA7tNPjxV6unto2dz/4jj5vb189imxtzh3mbeHLaZb/dSkiRJkiTNBLVe4T8K/CAzNwFry9dAboqIBZnZDSwDHqlXkCr0bO7n1w+O3OHo/F02Vx03WbssX2IySJIkSZKkGaLWK/xbgA9FxGzgSUA3cCNwBPAd4HDgQ3WJUFPSYH8/vevub8q6Zi9YzNxFy5qyLkmSJEmSZrqakkGZ+WhEnAt8n+JtYqcAa4HzIuJk4NryVZGaoQZ6u+i69/amrGvxnqvAZJAkSZIkSXVR87M/mXk2cPawwYdPLhxJkiRJmr5G6+dzyET7+5w9axYDg4OTDe1x67VfUKm9efRLkiRJUp2M1s/nkIn297l8x2158OHOyYb2uPXaL6jU3ma3OgBJkiRJkiQ1j8kgSZIkSZKkNmK7wBps6OhhY2fvmNMt7Oll4wSeBR5Lf//knxOWJEmSJEntzWRQDTZ29nLDbWO/Vn2vHTazfgLPAo9l+Y7b1m1ZkiRJkiSpPZkMkiQ1RERsAm4tv34C+D5wLrACWAOcmJkDrYlOkiRJal8mgyRJjXJfZh4y9CUiTgR+nJlnRsTngcOAK1sVnCSptbxpIEmtYzJIktQofxwRNwC/Ad4NHAR8tBx3OXAwMyAZ9EeL57EXm6uO7569gF881N/EiCRp2vCmgSS1iMkgSVKjPD0zH46ItwFnANsDj5bj1pffp705/T2s/+lNVcdvt89qYF7zApKk6aMtbhpI0lTkq+UlSQ2RmQ+XHy8E9qNIBG1XDlsGPNKKuCRJU8bTM3M18F/M4JsGkjQVmQySJNVdRCyKiDnl14OBe4AbgCPKYUcAN7YiNknS1OBNA0lqnUk9JhYRqyhO5ncqB50PLAGuyczTJxeaJGkaexbw1Yh4DOgBjgceBs4tHwlYC3yvhfFJklooIhYB3ZnZz9abBj+kuFlwd/n/1a2LUJJmtsn2GfRe4Mfl5w8AZ2fmxRFxRUSszMy7Jrl8SdI0lJn/Q3GXd7ijmh2LJGlK8qaBJLVQzcmgiHgVcBPw6nLQKuC08vPlwGrAZJAkSZKkx/GmgSS1Vk19BkXEbOCdwBcrBi/KzK7ysx2+SZIkSZIkTUG1diD9BuCyzOyuGNYZEQvKz3b4JkmSJEmSNAXVmgzaG3hdRFwF7ANcQNGR9NBbYg7Ht8RIkiRJkiRNOTX1GZSZHxj6HBHXU7QUmgWcFxEnA9dm5p11ibCO+jZtYKC7Y9LLWdjTy147bB5zuu22ncX6Sa9NkiRJkiSpfib7NjEy85CKr4dPdnmNNNDdQcedN016ORs39bD+wY1jTrfj81fxKWbsAAAgAElEQVRNel2SJEmSVG+Dg4M8tqmn7svdZt4ctpk/6ctMSQ3mUSpJkiRJbaavb4AHH+6s+3J3Wb7EZJA0DXiUSpLUBvoHBvjtH8Zu1VpPS7adz7LF2zR1nZIkSRqbySCNaaQmpHP7+hvSrLSSTUwlqX66uvv4ST7U1HWu3m9nk0GSJElTkFfaGtNITUjn77KZX4+j36TJsImpJEmSNL0N9vfTu+7+pq939oLFzF20rOnrlaYLr7QlSZIkSQ0x0NtF1723N329i/dcBSaDpKpmtzoASZIkSZIkNY/JIEmSJEmSpDZiMkiSJEmSJKmN2GeQJKnuIuKFwKeBXqADeCNwEvB64CHgd5l5TOsilCRJktrXtE4GbejoYWNn77inX9jTy8Y6vA69v39w0suQpBnuV8BLM7MzIt4BnFgO/0hmXtLCuCRJU4A3DSSptaZ1MmhjZy833Db+1xTutcNm1tfhdejLd9x20suQpJksMx+o+NoL9AHzgA9HxHuAL2TmN1sSnCRpKvCmgSS1UE3JoCqZ/LnA+cAS4JrMPL1OMarNDfb307tu/Em/yZq9YDFzfQ2lVBcRsQPwTuAwYDAzT4+IZcC1EXFTZv6mtRFKklrBmwaS1Fq1tgwaKZO/HXB2Zl4cEVdExMrMvKtukaptDfR20XXv7U1b3+I9V4HJIGnSImJb4GLg3Zn58NDwzNwQEf8F7AWYDNL/Z+/e4+Sq68P/v3LZkJCrgCigFCzwhgARRPCrXSLa1gJa1IoI9Qui0oLaH9daC1hMtfaLtdUqbdVWkZJKvaC0CAgtAhK0RSgCBuIbKChykxCSkMtuNnv5/XHOhmWz19nZmZ2Z1/PxyCM7Z858zvszZ+bzOfM+n/M5klqYJw0kqT4quptYZj6RmZvLh/2Z/HbgmnLZNcDSiYcnSWpEETET+DpwSWb+qFy2cMBzrwX+t34RSpLqbfBJg8xcA8VJA6D/pIEkaRJMaM6gQZn8d2dmR/nUOmDvCcamFtfX18dzm7Yws7uH56ow8fdY7NA2g3k12ZLU9E6iOCmwoBzufy0QEbEYmAFckZkP1DNASVL9DHfSoBw92n/S4Mv1jFGSmlnFyaDBmfyI2BwRszOzE1gIPFutINWaurt7efKZzczacyuPVmHi77HYc7f5NdmO1OwycznFPHKSJA3FkwaSVEeVTiC9XSYfWAEcC3wHOAa4oCoRSpIkSWoqnjSQpPqqdGTQUJn8vwIuj4hzgZsy874qxShJkiRJVbelq/rTEfT09FW1PEmaDBUlg0bI5B8zsXAkSWouL5rXxkFsfcGyRb3rOGjn55d1Tp/NQ6t7ah2aJLW8ru6eqk9HsNsuO1a1PEmaDBOaQFpqNn19fWzY3EVHT23mKOo3f8dZLJy3Q023Kak2ZvRsYd29t71g2ZxddmTdM5u3PV60ZCnQVuPIJl9Pby+PPW17KkmtpP8mMP2qcTOYHdpmsMMsf7pK1eQ3Shqgu7uXx3+1kZVranP3sn5LD93DHy+Smk5HZzd35eqabtP2VJLqq/8mMP2qcTOYPXebbzJIqjK/UZIk1dlQl5IN5GVkkiRJqiaTQZIk1dlQl5IN1KyXkUmSNFn6enroWvN4zbc7ffY8Zs5dWPPtSuNlMkiSJEmS1FR6uzroePjumm933oHtYDJIDWB6vQOQJEmSJElS7TgySBpktLk7qq1z+uyabUuSJEmSJJNB0iCjzd1RbcVcIJIkSZIk1YbJIEmSJEnSlNXX18dzm7aM6zUzu3tGfM0ObTMm5Xb1TlytRmEySJIkSZI0ZXV39/LkM5vH9ZpZe27l0Sc3DPv8nrvNn5RkkBNXq1GYDJLq7EXz2pjduZo1v3xm8jc2ay4d0+Zsezh/x1ksnLfD5G9Xkmqkp7eXx54e/uB/MtiWSpKkRlP1ZFBE/AHwXqALeF9mPlztbUjNZEbPFtbe/V/jPttRiUVLlrJyTdu2x0sP3aNpfsB0b1pPb+fGmm3PobiVs5/QZOro7OauXF3TbTZTWypNBfYTkjT5qpoMioidgNOA3wAOBS4GTqjmNiRVbvCd0uZseYauNc9N3ganz4Te7skrf4Deri1sfvCOmmwLHIpbKfsJSdJI7CckqTaqPTLoCOCWzOwG7oiIqHL5kiZg8J3SFuw2n2lzJ+9s9pxXHFKza6bnvOKQUdfZ0tXNlq09Vdle3+YuOnrGdimKl5C8gP1EBQYncgebN38OGzd0jFjGrm2bOWjnysvonD6bh1ZX5/sjjdX6jVvYsLmrZtubOWM63T29Ndse2EcMwX5CNVHJpNSjmT5t2qgTV1dSZm9f36jrjefYtJpt3Zy+DujaxKyZM9hh1oyqlDkW9RilX+srEfpNVl2rnQzaCVg74PH0sbyoedv4zzfptmq9vVrXrdZ8L9VSGrCfGO17M9Hnq1WGJDWFBuwnKlGvdr0e222lukqNo9rJoLXAkgGPRz2FmJnTqhyDJGnqsp+QJI3EfkKSaqDayaDbgYsiYgbwSuDBKpcvSWps9hOSpJHYT0hSDVQ1GZSZz0bEPwMrgK3A+6tZviSpsdlPSJJGYj8hSbUxrW8Mk1FJkiRJkiSpOYxpQjZJkiRJkiQ1B5NBkiRJkiRJLcRkkCRJkiRJUgsxGSRJkiRJktRCqn1r+aqKiD8A3gt0Ae/LzIcHPHc48DlgGvDJzLwmIvYC/gf4abna2Zl5d22jro7x1r1c/mbgXGAG8JXMXF7zwKuggv3+RuCicpVdgAcy8/dqHHZVVFD3NuAK4KXl8jMyc2XtI5+4Cuo+F1gOvAh4HDgtMztrH7kmSyXtYCMYpV63AG0Ud9C5PjMvrkuQ41S2RbcAB1J8F68c9PxbgAuBPuCszLyj5kFWYAz1+l/gl+XDyzLzspoGWKGIeC3wGYrP4Ebg3Zm5bsDzjbq/RqtXo+6vlwBXUbQLMxjU1zdye6jRjdYOtYrRvt+tYrT2oNVERDvFnQdfnJnP1DueeoiITUB/P31xZl4/1tdO2WRQROwEnAb8BnAocDFwwoBVPgu8E1gPrIiI75XL/ysz31LLWKutwrrvRPHj4k2Z2VPbiKunkrpn5k3ATeXrPw00agKwkv1+FPBcZr4zIo4E/gQ4pZZxV0OFdT8duCUzPx8RZwKnAl+sZdyaPBW2BVO+7RtDvQDe2oAHNN3A8RTfyxeIiBnAJ4ClwALgG0B7TaOr3LD1KnVk5lG1C6dqfgH8ZmZujogzgA8Bn4SG31/D1qvUqPvrGaA9M3sj4ijgT4H/O+D5hmwPNWajtUOtYrTvd6sYrT1oNecAd9Y7iDp7pNK+bcomg4AjKH7odQN3RET0PxERs4GZmfl4+fgBYF+gEzg8IlZQjA46LzM7ah/6hFVS99dQZMm/FxGdwB9l5qO1D33CKqn7z8rH04DfBZbVOugqqaTuDwOzytUWAatrG3LVVFL3fYF/KVe7i+KgwGRQ86i4LZjihq1XqQ+4KiI2Ax9plNGtmdkHPLl9dYBi3zyQmRuADRHRFhGzG2Ek3yj1AphVjuZaTzEa+ZFaxTYRmfnEgIddFD82+zXy/hqpXtC4+2tgYmcRcE//gwZvDzUGY2iHWsIYvt8tYaT2oNWUo1hvA95a71jq7OURcSvFyNczM3PNWF84lecM2glYO+Dx9EHPDRwWuK5c9iSwT2YeSZE1PWuyg5wkldR9d+DlwLHA54FPT3KMk6WSuvdbCtydmZsmL7xJVUndHwPmRcQq4B/Kf42okrr/FHhTuey3KS4XU/OYSFswlY1UL4B3ln3YucBXaxbV5Bpc50baX6N5XXk27vM0YPsbETsDHwS+MmBxw++vYeoFDby/ImJxRPwIuAT4wYCnGrk9lMZthO93yxihPWgZETGd4nPwhXrHMgX8emYuBb7POEfLTeVk0FqKbGe/nhGeWwg8m5lbyjNZAF+nGILfiMZd93L5TZnZnZk3AgdMepSTo5K69zsJ+NfJC23SVVL39wAPZuYBwNE0boNYSd2/AuwaETcBcymSwWoeE2kLprKR6kX/5WGZeR+wNSLm1DC2ydLI+2tEA/bX9ylOyDSMiNgR+BbFWcSBlyU29P4aoV4Nvb8y8/7MfB3wFoofgP0aen9J4zHS97uVjNAetJLfB65uhFGrk23Ad2Hc+Y+pfJnY7cBF5bXrrwQe7H8iMzsiojsidgOeoxgO+1BELMjM58rVXj/wNQ1m3HWnmDTwLwAi4hCK62obUSV1759c7zeBM2sfctVUUvejKEbBwfYHhI1k3HUvL7P5AEBEXEiLnhlpYhW1BQ1g2HoB9PdjEfFSYG6DXuo82IPAflFM+j4f6G6Gg7eI2AGYlpmdEXEQMOZh2fUWETMpDhovycwfDXq6YffXSPVq8P21Q2ZuKR+uAzb3P9fg7aE0ZqO0Wy1jpPagxRwMHBYRbwOWUNxQ500jv6T5lH11Z3n54LjzH9P6+vomJbBqKCcHO4VitvT3U0xg+Ehm/iAiXkMxYd404P9l5tXldYN/TjF3zlrg1EadZX68dS9f81GKy2WmA6dn5v11CX6CKqz7m4G3Z+ZpdQq7Kir4zM+jGA01H9gRuDAz/7M+0U9MBXU/mOJsSDewIjP/vE6ha5JU0hY0guHqRXE3jDsoDuxmUnyfb6pXnOMVEd8EXk3RB19PMTrhqszMiDgOOJ9iTqRzMvP2+kU6PsPVC9gAXFv+D8VcfffWJchxioiTKdrP/jmprqW4K01D76+R6kVj76/XUkw230vR5p1L8eOn4dtDjc3gdigz/6TOIdXcUN/vzGzUaTEqNlR7kJl31Teq+irngju+FUeLRcRhwJcpTgZsobjj4JjnDZ7SySBJkiRJkiRV11SeM0iSJEmSJElVZjJIkiRJkiSphZgMkiRJkiRJaiEmgyRJkiRJklqIySBJkiRJkqQWYjJIkiRJkiSphZgMkiRJkiRJaiEmgyRJkiRJklqIySBJkiRJkqQWYjJIkiRJkiSphZgMkiRJkiRJaiEmgyRJkiRJklqIySBJkiRJkqQWYjJIkiRJkiSphZgMkiRJkiRJaiEmgyRJkiRJklqIySBJkiRJkqQWYjJIkiRJkiSphZgMkiRJkiRJaiEmgyRJkiRJklqIySBJkiRJkqQWYjJIkiRJkiSphZgMkiRJkiRJaiEmgyRJkiRJklqIySBJkiRJkqQWYjJIkiRJkiSphZgMkiRJkiRJaiEmgxpcROweEZdXoZxFEXFaNWKqt4jYKyL+u/z7uIg4s0bbPSQifmsCr/95RMyuQhxfjoi9qlDOqRGxy0TLkaTJZJtvmy9pckTEHw/4+6iI+Ho945kKIuKWiNi//PvGGm73j0dfa9jXLouIM6oQQ1X62In2n6oek0ENLCJmZOYTmXlKFYpbBIwrGRQRM6qw3UmVmVdn5udrtLlDgLo3bJl5Wmb+vApFnQrU7IdBI3yepGbSjN852/wJORXbfEkvVHECohVkZi37gLrviyr2sTXtP+1vhjetr6+v3jG0tPJs3lXAw8BiYAVwRmb2RsSbgT8DZgP/DXwQ2BP4N+AB4JXA7wBfz8z/ExGnAscCLwb2Bs6i+KL9FvBfmfm+cptDlbsceBvwIPA14LPA3wCvA9qAP8vM70bEMuDXgP2BH2XmeUPU6WzgDGALcHNmnh0RLwH+EdgD6ADen5kPjPC+LCvruj/wEuB9wOnAq4F/zcyPletdA+wGzAI+lpnfKd/Tge/J/pn5pxGxH3BFWZ+bgUMy86hyW3sAB5RlfSAz/yMiZo7wHrxgfeD7wCPADsCTwDmZeXMZ407le9V/FmEx8HeZ+cZBdf458A3gzcDTwLsyc3VE7Av8A7ATsBp4T2b+qlz/n8v9th54S2Y+FxG3UHyGfhYRHwDOKct7CrgmMy8b7rUDYnk7xWfil8AzmXlkRDyVmS8dsH+eyswvlmV9DTiurPufA58q99tJmXlnRLwYuBx4GfAEcEpZh8uAzcDhwNeBHgZ9dpBUkRHax4F9yIEM3c69FvgMRT+xFnh3Zj45zHZs823zbfOlGomI3YFvAXMpTuyfAHRS/J5Iiu/Xv1K0UR8EuoBjMnNdRBwOfIGi7boN+KPM7ImIY4GLy/KuzMxlEfFJ4E+A+yja0H8Hzi/L2x+4LDM/WbbB3wFWAYdRfJc/UMY61G+Olw4R/3ODl2Xmz4ao+8HAZcAMoA94Q1mvCynatx2ASzLzn0Z4//Ya43t1BvD+ssz/oejHege1uU9l5kvLZMOXgN8o34e9gbeXmxzPe7PnUOsP3heZec6A+nwH+Gxmrigf3w68MzMfHbDOMorfbwdSDAA4r+zfxtzvlf3kqVTYxw6IZQaD+k/gPRT9+PWD+vRljO3Y4KPAiUAvcGFZh6OAjwLd5bbOZIjPDi3OkUFTwxLgExTJoF2A3yuHap8NHJWZh1AcMP1euf6BwJ9nZgxR1v7AMcBbKRq3fy3LfWUUhiv3QuCnmXlIZn6aYpTQI5l5OLAUuDgi2spt/DqwdKhEUOlC4NDMfCXwsXLZZ8uYXw2cS/Ejg4g4Y4Rhi7sB7eX6V1F8oQ8G3hsR88p1TsnMwyga309ExLRhygL4W2BZGVf3oOd+DXg9RYfUH/NI78EL1s/MHuAi4Kvle3hzf8GZ+Szwi4g4pFz0TuDKYWJ8PDMPouhw++P4e+C0sp5foeg4+j1U1uce4F0DC4qIPSj29WEUScJXDdrWsK/NzKuAO4G3Z+aRw8Q60H2ZeTCwCTiP4v06r/wHxY+F/yzX+XfgkwNeuwg4IjP/hqE/O5IqM1z7OLAPGa6duw9oz8xDgS9SHIQOxzbfNt82X6qdk4Aby+P4w4BflMsPAC6gOO4/FXhR2Y78F89/578KnF5+N3cCToqIORQJ6DcDhwK/HRHtmXkhsKZs4/qTD4eWZb8SOCMi5pbLFwPLKPqX10bEviP85hgq/uHqNNgfUiTXD6HoBzZGxNHAi8u2+9XAaRGxWxRTaVw3TDljea++mZmHZ+YSipMavztMWQDvABZm5gEU7d+SAc+N570Zcv1h9kW/b5XbJyJ+DWBgImhQnX8DeAPwuSguUx5zvzdEeePtYyljG7b/HMaIxwYRcQTFyYnDKAZJXDLgmOEw4NTMfANDfHZG2W5LMBk0NWRm3p2ZfRRnytqB11I0JP8dEXdTjO55Rbn+qsy8b5iybszMTuCnQGdm/qgs9z6KzOpI5Q7028Dp5Tq3AvOB3cvn/i0zt45Qn/8B/iUiTqQ42wfwRuDSsrx/ovhik5lfzMwvDlPO9zKzt6zLzzPz4czcAvy8//XAORFxD8WIqr0ozjYM55WZeU359zcGPXdt2Tj9pCwHRn4Phlp/JN8Cji//fjtF1n8o/ddi/yvQHhHzKRqsfy/juIjiTGu/q8v/h4rjcIrPw4YszgB/b9DzI712vPrf158Ct5T7bSXFZw6KMw7/Uv79LxR16ndl+RmFoT87kiozXPs4sA8Zrp17EXBVRPyU4sB28Qjbsc3fnm2+bb40We4ATo6IjwH7ZmZHufz+Ae3mw0D/aIyVwJ4RsQiYnpn/Uy7/GsV3M8rX/rI8vv8GL/zODvTDzFyTmZspRnf0t5GrMvPBsp28l6KNGe43x1DxD1enwf4L+HBE/AlFAqibou3+3XIbPwZ2BX49i6k0jh2mnBHfq/LvV0bED8t+8HcZuR98HfBNgMy8Bxg4qmk8781w64/kGuDY8uTI8cC3h1nvO5m5NTMfL+Pbn4n1e+PtYys12rHB6yj6lS1ZjGC+iyKRBrAiM58q/x7qs9PyZtY7AAHFULWBf/cB04B/z8w/HLhiOXRu8whldQFkMYyxa8DyXophcSOVO9A0iuGQPxq0HqNsH4ozC2+gyFKfARxV1ulV5Zd5rPrj7x3wd//jGRHxBuAIijOMWyJiJcUwwOGMdAZ5C2x732YMWH+492Co9Ufyb8DNEbEcWD+gYRqsb8D/fRQJ28fLLPawcfP8/h1opPqO9tqRYoPt3+eh9tVYyx34eRrqsyNpnEZpHwd+54Zr5/6Z4sDxsoh4NfDXI2zONn97tvnDs82XJiAzb42IpRQJim9HxB8B/8v27eZ4v5tjMTBp21/u1mGWD/mbA2Bw/Jn5/aGWDX5dZl4REXeU691cXmo1jeLypivGUY+xvFf/RHHJ2INRTN48j+GN2t8MKn+k32NDrT+szNwQEfdT9I/HA+8eZtXhfm9W2u+Nt48dSQ/PD1IZS3/T/3i0srf1N0N9djJz1Rhia2qODJoa9o+IJWVG9wSKa3j/G/jNiHgZQETs3P/3BA1X7gaKbHC/G4EPRMT0cr3hDkxfoFz/ZZl5I8Vw8f3Kp1ZQTlAdEdOjuOZ3ohYAz5Y/Co6gGP44knuiuCYaimH7oxnvezD4PdwmM5+hmL9hGcNn7OH5oanvAm7LzPXA2oh4UxlDW0SMVs9+d1Ds63nlcMmjx/i6foPrsykiXhYROwBvGmdZP+L5uv0+xWf8BUb47Egav7G2j8O1cwsoruWHYgj9kGzzbfOHYZsvTZLyUqAnM/MLFKMLx9S+ZjE/Sk9EvLJcdBLFdzOBA6K4rGomRXv5w3Kd3v42sQJD/uYYKv6x1iki9qa45PUzwA8oRjXdCLw/yrszRmHCd2qkmL9odVnWu0ZZ90eUo0HL/m7/Udav5HfeSPviWxRz77Rl5sPDrPP2sk/ZvYwvqfD3Xmm8fexAg/ubX1BMKg3FJV/j8SOer9tLKS5l3O4KmmE+Oy3PZNDUcC/F9ZSrKCbqvCoznwY+RDFU/F6K4Yu7TnRDw5WbmWuAeyPi3oj4MMUkaL+i+KLfRzFUfSxmAFeUZd8+4HX/H3BMFMP7V1KcDRxt/ojR3ADsXMZ3DsU8CCM5h2KOibspsvvPjbL+eN+Dm4EjIuIn5Rnswb5F0ZmM9MPgZVEMR30H8PFy2bsphjXeA9wNvGaUOADIzMeAv6MYLnk9RcM4Wp0HugxYHhEryscfo2g8/4OiAxmPZRRDWO+luCb6o0OsM9xnR9L4jbV9HK6d+2uK6+7vYuTr6m3zbfOHsgzbfGmyHEVxzP4T4EievyRzLN4HfKVsd9ZTTNTbQfHb4HsUbc73M7M/gfsvwE8j4rPjDXKE3zJDxT/WOp0IrCzb9fnA9Zl5HUX/cEcUI0a/QDGadKQ5g8bikxTt6c0UlzuN5Epgc0Ssomj/khHa3wp/5420L75LMYH2cJckU8b0Q4r6nFNOK1Lp7z0Yfx870OD+8ysUc+b+hHHe1TIzfwxcS7GP/gM4MzOHOm7Z7rMznu00K+8mVmcxYMb0esfS7CJiR6AjM/uimHV+ZmYuq3NYkyoi5mbmpigm+LsNeMcIZwwkqWnY5tvmS1KtDGh/9wW+m+UdJZtVK/axzcg5g9RKXgN8Np6/peEpdY6nFv6izLjPBr7kjwJJLcQ23zZfkmrlhihuAgDFbeKbXSv2sU3HkUGqWES8Fzhr0OJLM/Pz9YhHUn1EcRvSWyju3nBaZl4ZEcsoriFfDTyVmSeW676F4nbSfcBZmXlHXYLWuNnmS5JqISJ2BgZPIP1MZv5WPeKRmpXJIEnShEQx+f1LgdOBlQOSQSsz88oB680A7gSWUkwG/I3MbK9DyJIkSVJLcwJpSdKEZGZfZj45xFMfjYgVEXFS+Xhf4IHM3JCZjwNtVbrjhyRJkqRxGHXOoGGG/+8CLKeYifvG/smiKhn+HxEOTZKkYWTmtHrHUKFLMnNZRCwEboqI24CdKO6Y2G9dueyJkQqyn5Ck4TVwP1E19hOSNLzh+omxTCDdDRxPMfy/30co5gn4VkRcGxGLKW5X9wkGDP8HxjT8P3O8dyyVpOYXEfUOoWKZuab8f31EfB84CPg5sGjAaguBZ8dYXrVDlKSG18j9RLXZT0jS9kbqJ0a9TGyY4f/twDXl39dQJIAc/i9JAqAcEUREzAReC/wv8CCwX0TMjYiXAt2Z2VnHMCVJkqSWVOmt5edmZkf59zpgbyoc/i9JanwR8U3g1cDGiDgC2KkcNToDuCIzHyjXWwbcSHE58Tl1CleSJElqaZUmgzZHxOzyjG7/MP+1VDj8X1Lz6u3t5YknnmDr1q31DmVKa2trY/fdd2f69Mac1z8zTxjjelcDV09yOJIaiP3E2DR6PyFJlbKfGJvx9hOVJoNWAMcC3wGOAS5gwPB/ioml6zb8v3vTeno7N263fPrsecycu7AOEUmt64knnmDBggUsWLCg3qFMac899xxPPPEEL3vZy+odSlMb3D/YL0j1Zz8xNvYTrWu43zaVsu9To7GfGJvx9hNjSgYNMfz/r4DLI+Jc4KbMvK9cbxlTYPh/b+dGNt5323bL5x3YDjZ8Uk1t3brVhnsMFixYwJo1a+odRtMb3D/YL0j1Zz8xNvYTrWu43zaVsu9To7GfGJvx9hNjSgYNM/z/mCHWc/i/JEmSJEnSFOZFx5Jqav3GLTz29IYJ/Vu/ccuI23jssceICG644YZty9773vdy+umnD7n+7bffzsc//nEAPvnJT7Jhw4bqVXiQVatWcdddd01a+ZLU6Own7CckaTjV6CPsJwqVzhkkSRXZsLmLW3/y+ITKWHroHiyct8OI6xx00EHccMMN/M7v/A6rV69my5YtzJo1a9SyL7zwwgnFNppVq1axdu1aXvWqV03qdiSpUdlP2E9I0nCq0UeA/QQ0eTJoS1c3W7b2bHvct7mLjp7Jy9ABzN9x1qgfKkmTb5dddqGjo4PNmzfzve99j6OPPpof/vCHfPWrX+Xmm29mw4YNnHjiibzrXe96wetOPvlkPve5zzFr1izOPvtsenp62GeffXj00Uf50pe+xMknn8zixYtZtWoVCxcu5JJLLgHgU5/6FCtXrqS7u5vzzz+fJUuW8OY3v5n29nbuueceDj74YC688EIuv/xyNm7cyC233MI//MM/MH/+/Hq8PZLU8uwnJEkjafZ+ormTQVt7ePTJ55M/i168kZVrRh4ONo1Rst0AACAASURBVFFjyTBKqo03vvGN3HTTTaxYsYILLriAH/7wh7zrXe/ive99L11dXbztbW/jne9855Cv/eY3v0l7ezunnnoqV111FY8++ui254488kjOP/98zjjjDDKTp556it7eXpYvX86zzz7LWWedxfLly+no6ODEE0/k/PPP57jjjmPjxo2ccsoprF27lve///21ehskScOwn1C1VfvOXwC9XZP7+0XS8Jq5n2jqZJCk1vamN72J008/nVe84hW0tbUBcN1113HVVVcxbdo0nn76adavXz/ka3/+85/z1re+FYAlS5Zw/fXXb3tu8eLFAOy+++6sX7+eBx98kNtuu42TTz4ZgE2bNgEwZ84c9t57bwB22223YbclSaoP+wlVW7Xv/AUw5xWHVLU8SWPXzP2EySBJTWvhwoW0t7dz5JFHblv2pS99iWuvvRaAo48+mr6+viFf+2u/9musXLmSww47jJ/+9KfDbqOvr4999tmHN77xjZx33nkAdHV1ATBt2rTt1m1ra6Onp2e7ciRJtWc/IUkaSTP3EyaDJNXU/B1nsfTQPSZcxlj90R/9EVDcEQCgvb2dk046if32248FCxYM+7oTTjiBs88+m5tvvpm99tpr25mAoRx11FHceeed2zL5S5Ys4cMf/vCQ6x566KF87WtfY+XKlfzlX/4l8+bNG3NdJKkV2E/YT+iFc5/O7O7huU3Vu1RsuHlUnftUjaAafUR/OWPVrP3EtOGyWLUSEX2ZWdUyu9Y8zsb7buO5TVteOGfQkqWsXDP8DqiGpYfuwct2daI/qd8jjzyybWhjI+nr66Onp4eZM2dy1VVX8dBDDw3bIFfL4PcqIsjMaSO8pCVUs5/o7x/6zTuwnVk7T/yAQlLl7CfGrlH7iYhoA24BDgROy8wrI2IXYDkwH7gxM5eV674FuBDoA87KzDvGUH7Vf09UanA/U6mBv2P2eU07D91evUvPhvtN5O8YTVX2E2M3nn7CkUGSNITu7m7e8573MG3aNKZPn85f//Vf1zskSdIUYj8xLt3A8cDpA5Z9BLg0M78VEddGxGIggU8AS4EFwDeA9loHK0nVMNX7CZNBkjSEtrY2rrjiinqHIUmaouwnxi4z+4AnI2Lg4nbgovLvaygSQL3AA5m5AdgQEW0RMTszO2sasCRVwVTvJ6bXOwBJkiRJLWduZnaUf68Ddir/rR2wTv9ySVKVmQySJEmSVGubI2J2+fdC4FmKRNCiAev0L5ckVZnJIEmSJEm1tgI4tvz7mPLxg8B+ETE3Il4KdHuJmCRNDucMktR0HnvsMd7xjnew33770dHRwVFHHbXtlpBjsWrVKn784x/znve8ZxKjlCTVi/1E7UXEN4FXAxsj4gjgr4DLI+Jc4KbMvK9cbxlwI8XdxM6pU7iSWlwr9BMVJ4Mi4u+AV1GMLvpT4HbgMmB3YCXwoczsrUKMkppI96b19HZunFAZ02fPY+bchSOuc8ghh/ClL32Jvr4+3v72t3PKKaewYMGCMZV/wAEHcMABB0woRklSZewnmlNmnjDE4mOGWO9q4OrJj0hSI6pGHwH2E1BhMigi9gUOyMzXRcTLga9R3Prxzsz8dET8PXA0cF31QpXUDHo7N7LxvtsmVMa8A9thlMa739atW+nr62PmzJl86lOfYuXKlXR3d3P++eezZMkS3vzmN9Pe3s4999zDwQcfzIUXXsjtt9/ODTfcwEUXXcR3v/tdvvKVr7DnnnuyYcMGzj33XDZv3sw//uM/MnfuXB5++GEuuugijjjiiAnVSZJUsJ+QJA2nGn0E2E9A5SODfgV0RMRMikneVgNHAn9ePn8N8HpMBkmqk7vvvpuTTz6Zxx9/nDe84Q3ccccd9Pb2snz5cp599lnOOussli9fTkdHByeeeCLnn38+xx13HBs3Pn+moaenhy984Qt8+9vfZsaMGRx33HHbntu8eTNf/vKXuf/++/niF7/oQb4kNRj7CUnSSJq9n6g0GbQB+AWQwI7A24GP8/ytIL0NpKS66h/W2dvby5lnnsnKlSu57bbbOPnkkwHYtGkTAHPmzGHvvfcGYLfddmP9+vXbynj22Wd5yUtewpw5cwBeMNRz//33Z9q0aey2226sW7euVtWSJFWJ/YQkaSTN3k9Umgz6bWBnYF/gJRQjgR6iGCX0FN4GUtIUMX36dObPn09E0NnZyXnnnQdAV1cXANOmTXvB+n19fdv+3mmnnfjVr35FZ2cnM2bM4Gc/+9m25wa/To1hS1cPz23asu1x3+YuOno2TOo25+84i4XzdpjUbUiqnP2EJGkkzdpPVJoMmg48m5m9EfEcMBe4leL2kD8r/7+hOiFK0vj1D+vs7u5m99135/Wvf/22ZQBLlizhwx/+8IhlzJgxg9NPP50TTzyRPfbYg1122YW2trZahK9J0tXdw6NPPp/8WfTijaxcs2WEV0zc0kP3MBkkTUH2E5KkkTR7PzFtYNZqrCJiBsWdw/YCZgN/C1xVLnspsAr4wFjuJhYRfZk57hhG0rXmcTbedxvPbdrywoP+JUtZuWZy3/ilh+7By3adP6nbkBrJI488sm3YJNTuLjHVsnXrVtra2ujq6uId73gHV1xxBfPnT853fPB7FRFkZsufWq5mP7Hml49w338+P52d/YJUf/YTY2c/MbTJ+D1Rqf7fIRM18HfMPq9p56HbJ15mv+H6PvsrTVUD275a3k2sWqZqP1HRyKDM7AFOHuKpoW4ZKUnbzJy7cMwz908FV155Jddddx0bN27khBNOmLSGW5JUsJ+QJA2n0foImLr9RKWXiTWkF81r4yC2Dvlc5/TZPLS6p8YRSZrqTjrpJE466aR6hyFJmqLsJyRJI5mq/URLJYNm9Gxh3b1DD7FctGQpMDWu3ZMkSZIkSZos0+sdgKTm1tfXx9atQ4/I0/O2bt1KJXO4SVKjs58YG/sJSa3KfmJsxttPtNTIoFro6e3lsacn9zbFg3nbYk1lu+66K0888QS9vaPOJ9/Spk+fzq677lrvMCSp5uwnxsZ+QlKrsp8Ym/H2EyaDqqyjs5u7cnVNt+ltizWVzZs3j3nz5tU7DEnSFGU/0boi4u+AV1FcrfCnwO0UdyfeHVgJfGgsdyeW1NzsJyaHySBJ0oRERBtwC3AgcFpmXhkRuwDLgfnAjZm5rFz3LcCFQB9wVmbeUZegJUl1FRH7Agdk5usi4uXA14BvAHdm5qcj4u+Bo4Hr6hmnJDUr5wySJE1UN3A88LcDln0EuDQz24HDI2JxRMwAPgG8CXgn8NmaRypJmip+BXRExExgEbAaOBK4pnz+GmBpnWKTpKZnMkiSNCGZ2ZeZTw5a3M72B/T7Ag9k5obMfBxoi4jZNQxVkjR1bAB+ASTwH8CngZ2AteXz68rHkqRJYDJIkjQZ5mZmR/l3/wH9wIP8gcslSa3nt4GdKU4UvAr4e4o+YlH5/ELg2fqEJknNz2RQ6UXz2jho563b/dvnxTPqHZokNaLNA0b99B/QDzzIH7hcktR6pgPPlhNEPwfMBW4Fji2fPxZYUafYJKnpOYF0aUbPFtbde9t2yxctWQq01T4gSWpsKygO5L8DHANcADwI7BcRcykmlu7OzM76hShJqqP/BN4dESuA2RRzyl0FXBYRtwKrgO/VMT5JamomgyRJExYR3wReDWyMiCOAvwIuj4hzgZsy875yvWXAjRR3EzunTuFKkuosM3uAk4d46oRaxyJJrchkkCRpwjJzqIP3Y4ZY72rg6smPSJIkSdJwKk4GlWd+/4LiGqrrgK8CyymG/t+YmcuqEaAkSZIkSZKqp6IJpCNiB+BjwNsy8w2Z+WngI8ClmdkOHB4Ri6sYpyRJkiRJkqqg0pFB/wfoAK6MiJnAHwPtwEXl89cAS4H7JxyhJEmSJEmSqqbSZNDuwGLgcGBP4J+AuZnZUT6/Dth74uFJkiRJkiSpmipNBq0FfpiZm4BVEbEQ2BQRs8vbBC8Enq1WkJIkSZJUL+s3bmHD5q5R15uzpYsNm7ZMeHs9PX0TLkOSRlJpMuh24IKImA7sCnQCK4Bjge9Q3EHmgqpEKEmSJEl1tGFzF7f+5PFR1zto562se3LDhLe32y47TrgMSRpJRRNIZ+Za4DLgB8C/AecBfwX8QUTcBtyTmfdVK0hJkiRJkiRVR8W3ls/MS4FLBy0+ZmLhSJJUOy+a18ZBbN32uHP6bB5a3VPHiCRJkqTJV3EySJKkRjejZwvr7r1t2+NFS5YCbfULSJIkSaqBii4TkyRJkiRJUmNyZJAkSZKkmouII4C/oBiSeR3wVWA5MB+4MTOX1S86SWpujgySJEmSVFMRsQPwMeBtmfmGzPw08BHg0sxsBw6PiMV1DVKSmpgjgyRJkiTV2v8BOoArI2Im8MdAO3BR+fw1wFLg/vqEJ0nNzWSQJEmSpFrbHVgMHA7sCfwTMDczO8rn1wF71yk2SWp6XiYmSZIkqdbWAj/MzE2ZuQpYCGyOiNnl8wuBZ+sWnSQ1OUcGSZI0iXp6e3ns6Q013+78HWexcN4ONd+uJI3R7cAFETEd2BXoBFYAxwLfAY4BLqhfeJLU3EwGSZI0iTo6u7krV9d8u0sP3cNkkKQpKzPXRsRlwA8o7iZ2HrAKuDwizgVuysz76hiiJDU1k0GSJEmSai4zLwUuHbT4mHrE0kpeNK+Ng9i63fI5W56ha81z4y5v+ux5zJy7sBqhSaohk0GSJEmS1CJm9Gxh3b23bbd8wW7zmTZ3/CNK5x3YDiaDpIbjBNKSJEmSJEktxJFBkqSm1b1pPb2dG7c9buvrrmM0kiRJ0tRgMkiS1LR6Ozey8b7nh8LP3PPgOkYjSZIkTQ0TSgZFRDvFLSBfXC5aDswHbszMZRMLTZIkSZIkSdU20TmDzgHuLP/+CHBpZrYDh0fE4gmWLUmSJEmSpCqrOBkUEW8BbgM2lYvagWvKv68Blk4sNEmSJEmSJFVbRZeJRcR04IPA7wFvLRfPzcyO8u91wN4TD29kgycG7dfbtWWyNy1JkiRJktSQKp0z6PeBqzOzMyL6l22OiNmZ2QksBJ6tRoAjGTwxaL85rzhksjctSZIkSZLUkCpNBh0MHBYRbwOWAFdQTCR9LPAd4BjggqpEWGcvmtfGQWzdbnnn9Nk8tLqnDhFJkiRJkiRVrqJkUGZ+pP/viLiFYqTQNODyiDgXuCkz76tKhHU2o2cL6+7dfvTRoiVLgbbaByRJkiQ1Ce9OLEn1MaFbywNk5lEDHh4z0fIkSc0hIjYBd5QPLwZ+AFwG7A6sBD6Umb31iU6SNEUMdXfib0XEtRGxODPvr2NsktS0JnpreUmShvNIZh5V/rseeB9wZ2YeCfQCR9c3PElSPXl3YkmqnwmPDJIkaRgvj4hbgV8CZwJHAn9ePncN8HrgujrFJkmqo6lyd2JJalWODJIkTZZfz8ylwPeBTwI7AWvL59aVjyVJrWnb3YkHLNscEbPLv2tyd2JJalUmgyRJkyIznyn//DpwKEUiaFG5zIN8SWptBwPHR8T1bH93YijmIl1Rp9gkqel5mZgkqeoiYi7QmZk9FJeDPQj8F8VB/s/K/2+oX4SSpHpqpbsTS9JUZDJIkjQZ9ge+HBHPAVuA04BngMvKeYRWAd+rY3ySpCnCuxNLUu2ZDJIkVV1m/g/FpWGDnVDrWCRJkiS9kHMGSZIkSZIktRBHBkmSVHrRvDYOYuu2x53TZ/PQ6p46RiRJkiRVn8mgJtDT28tjT2+o6Tbn7ziLhfN2qOk2JWmyzejZwrp7b9v2eNGSpUBb/QKSJEmSJoHJoCbQ0dnNXbm6pttceugeJoMkSZKkJtHX18dzm7aM/3Wbu+joqezEtCeYpfoxGSRJkiRJLa67u5cnn9k87tctevFGVq4ZfxIJPMEs1ZMTSEuSJEmSJLUQRwZJkiRJkioy+OYL4zFnyzN0rXluu+XTZ89j5tyFEw1N0ggqSgZFxGuBzwBdwEbg3WVZy4H5wI2ZuaxKMUqSJEmSpqDBN18YjwW7zWfa3O0vE5t3YDuYDJImVaWXif0C+M3MfD3wXeBDwEeASzOzHTg8IhZXKUZJkiRJkiRVSUUjgzLziQEPu4BuoB24qFx2DbAUuH9C0UmSJEmSJKmqJjRnUETsDHwQOBp4d2Z2lE+tA/aeYGySJEmSmpDTTkhSfVWcDIqIHYFvAWdm5jMRsTkiZmdmJ7AQeLZaQU5Fw02UtmBGVx2ikSRJkhpK/7QTmyPiDIppJxZRTDvxrYi4NiIWZ+akXmnQvWk9vZ0bR11vzpYuDtp59EmSF+04jXXVCEySJlmlE0jPBL4OXJKZPyoXrwCOBb4DHANcUJUIp6jhJkrb9aij6xCNJEmS1DimyrQTvZ0b2Xjf6JMfb9i0hXVPbhh1vV1e016NsCRp0lU6MugkisZ5QUScBVwL/BVweUScC9yUmfdVKUZJkiRJTchpJySpPiqdQHo5xfW8gx0zsXAkSZIktYJWn3ZCkuqp0lvLS5IkSVJFRpl2AoqTzCvqEZsktYIJ3U1MkiRJkirgtBOSVEcmgyRJkiTVlNNOSFJ9mQxSRXp6e3ns6dHvqFBN83ecxcJ5O9R0m5IkSZIkNRuTQapIR2c3d+Xqmm5z6aF7mAySpDEyaS9JalR9PT10rXm8auVNnz2PmXMXVq08qRmYDJIkqQmZtJckNarerg46Hr67auXNO7AdTAZJL+DdxCRJkiRJklqIySBJkiRJkqQW4mViVbbDzGkctPPW7ZZ3Tp/NQ6t76hCRJEmSJLUu5yCStmcyqNq2drDu3lu3W7xoyVKgrfbxSJIkSVILcw4iaXsmg2rkRfPaOAhHDElSIxncdttmS5IkqRmYDKqRGT1bWHfvbdstd8SQJE1dg9tu22xJkqqnr6+P5zZt2W75zO6eIZdXamB5O7TNYIdZ/gyW/BaoYfT09vLY0xtqus35O87yNsmStnGkkCRJ1dPd3cuTz2zebvmsPbfy6JPVO+4fWN6eu803GSRhMkgNpKOzm7tydU23ufTQPUwGSdrGkUKSJKnaE1KDk1Kr9qqeDIqIPwDeC3QB78vMh6u9jWYy3FxC8+bPYeOGju2WexZaUqNrpn5iYBtu+1yfEZzgKE6p2TRTP6HmVO0JqcFJqVV7VU0GRcROwGnAbwCHAhcDJ1RzG81muLmEdnlNO485x5CkJtNs/cTANtz2uT4jOMFRnKqe9Ru3sGFzV023aTLzhZqtn5DqpXvTeno7N1atPEcuNZ9qjww6ArglM7uBOyIiqly+pCZSj4PumTOm093TW9NteqD/Ai3bT+zz4hnM7u3c9tiRRNXjnHKqlg2bu7j1J9W99GM0JjO307L9hFpbtS896+3awuYH76haeWMZudQKCfVmqmO1k0E7AWsHPJ4+lhfVro3/fAXPjXP5p6pUzriXS1JDmGL9xOA2dSKPbZ8lqQqmWD8xXpPRF1S7zKle3mSUaR8tDVbtZNBaYMmAx6Oe8szMaVWOQZI0ddlPSJJGYj8hSTVQ7WTQ7cBFETEDeCXwYJXLlyQ1NvsJSdJI7CckqQaqmgzKzGcj4p+BFcBW4P3VLF+S1NjsJyRJI7GfkKTamNbX11fvGCRJkiRJklQjY5qQTZIkSZIkSc3BZJAkSZIkSVILMRkkSZIkSZLUQkwGSZIkSZIktZBq31p+QiLiD4D3Al3A+zLz4TqHNGVExCbgjvLhxZl5fT3jqaeIaANuAQ4ETsvMKyNiF2A5MB+4MTOX1S/C+hjmfVkGvBNYDTyVmSfWL8L6iIjXAp+haFc2Au+maPta+vPSjJqlDxlPGxcRbwEuBPqAszLzjiELnSLG831stLoBRMRLgKso7oA0AzgD+F/gMmB3YCXwoczsjYjDgc8B04BPZuY1dQl6nCKineIuTy8uFzXFvoPtj7WAH9BE+061NVKfNNxnqDxu+y1gA3ByZj4zFY7lxluXiNgd+C4QwFGZeWe57q8DXwFmAV/NzH+qbU2qWpfLgIMp9tXdmXl2TStCRXV5K3ABsAV4HDglM7fWe79UsR6X0Xj75JXAF8r1e4H/m5lP1Oq37ZQZGRQROwGnAUuBD1N0wnreI5l5VPmvZRNBpW7geOBvByz7CHBpZrYDh0fE4rpEVl9DvS8AHys/Ny2XCCr9AvjNzHw9RWf+Ify8NJ0m60PG1MZFxAzgE8CbKH4ofLbmkY7fmL6PDVo3gGeA9rJ+HwX+FHgfcGdmHklxoHd0ue5nKer228Anyjo3gnOAO8u/m2nfwfbHWs2271QjY+iTtvsMRcSBwOHl9+lS4E8GrF+3Y7lK6gI8S9EGXDlo3YvLMo4ETivLrpkq1wXg9HK/1CPpUEld7qboo5ZS9McnlOvWbb9UuR7QePvk/sx8XWYeBVwOfLBctya/VaZMMgg4ArglM7vLM0hR74CmmJdHxK0R8bWI2LnewdRTZvZl5pODFrcD/WfmrqH4EraUYd4XgI9GxIqIOKnmQU0BmflEZm4uH3ZR/NBu+c9LE2qaPmQcbdy+wAOZuSEzHwfaImJ2DUMdt3F8HxuubgCZ2ZOZveXDRcA9FAfXL6hfWZeZmfl4Zm4EHqCo85RWjvi5DdhULmqafVcafKzVNPtONTdsnzTCZ2jw5+3IAeXV81hu3HXJzM7MXDNEWftl5h2Z2UMxAvaIGsQ/UDXrAvD3EXFLRPzm5Ie+nUrq8ovM3Fqu1t8HQ333SzXrAY23T7YOeP1C4N7y75r8VplKyaCdgLUDHk+l2KaCXy+zn98HPlnvYKaguZnZUf69juLzJLgkMw8B3gL8cUS8vN4B1Ut5YP9BimGwfl6aT7P3IUN9ZgfXuWE+y2P4PjZy3RZHxI+ASyguMxpYl4H1WzfgZVO+fhExnWKffWHA4qbad2x/rNUU+051MVKfNNxnaNtryu/VvPL5eh/LVVKX4Uwbx7qToZp1+ePMfA3w+8AlETGnalGOTcV1KS8L+x2Ky5qhvvulmvVoyH0SEW+MiB9T9LF3lc/X5LfKVDpYXktxFq1fT70CmYoy85nyz68Dh9Yzlilq84AzjwsphnS2vP4zGZm5nuLg9qD6RlQfEbEj8C3gzPK75Oel+TR7HzLUZ3ZwnRviszzG72ND1g0gM+/PzNdR/HC7hBfWpZHr9/vA1ZnZOWBZs+27wcdazbLvVHsj9UnDfYa2LS+/VxthShzLVVKX4fSNY93JULW69LcXmfkExZxie1UtyrGpqC4RsSvwLxRz03SVz9dzv1StHo26TzLzpsw8guLSsr8sn6/Jb5WplAy6HXh9ec3sq4AH6x3QVBERcwdcj/56fG+GsgI4tvz7mPJxy4uIheX/M4HXUkxk2lLKun+d4szaj8rFfl6aT7P3IUN9Zh8E9iv7iJcC3YN+qE854/g+NlzdACJihwEP1wGbgVt5vn7HAivKs33dEbFbRMyluETkoZoGO34HA8dHxPXAEuAKmmvfDXWs1Sz7TrU3bJ80wmfoVorvERSft9tgShzLVVKX4TwQEa8qv2uvB348mYEPoWp1GbBf5gKLgccmNfLtjbsuETEP+DZwTmYOPE6q536pWj0adJ8MPm7oHw1Uk98q0/r6+kZfq0Yi4gzgFIq7cLw/M+1cgYg4DPgy8BzFrOmnZeaj9Y2qviLim8CrKc6aXA98mmLSrfnATZl5UR3Dq5sh3pedKBrDGcAVmfm5OoZXFxFxMsXZ+bvLRddS3B2m5T8vzaaZ+pCxtnERcRxwPsVZvXMy8/b6RDw24/k+NlrdYNvd0i6mmGx4GnAu8DOKOr4UWAV8IIs7Ur2GYjLJacD/y8yr6xJ0BSLiFopJzqfRPPtuu2MtignBL6OJ9p1qZ3CfRDEHyCOZ+YPhPkMR8XHgjRR3QzolM1dHxJep87HceOsSxV0xv1fG/UvgG5n5mYjYh+Ly4DbgnzPzSw1cl+soRmzMBD6Tmd9ogLr8GcWNG35WFvGVzFxe7/1SxXo04j75PeAsiuOGrcAfZubPI+LF1OC3ypRKBkmSJEmSJGlyTaXLxCRJkiRJkjTJTAZJkiRJkiS1EJNBkiRJkiRJLcRkkCRJkiRJUgsxGSRJkiRJktRCTAZJkiRJkiS1EJNBkiRJkiRJLcRkkCRJkiRJUgsxGSRJkiRJktRCTAZJkiRJkiS1EJNBkiRJkiRJLcRkkCRJkiRJUgsxGSRJkiRJktRCTAZJkiRJkiS1EJNBkiRJkiRJLcRkkCRJkiRJUgsxGSRJkiRJktRCTAZJkiRJkiS1EJNBkiRJkiRJLcRkkCRJkiRJUgsxGSRJkiRJktRCTAZJkiRJkiS1EJNBkiRJkiRJLcRkkCRJkiRJUgsxGSRJkiRJktRCTAZJkiRJkiS1EJNBkiRJkiRJLcRkUIuKiL0i4vh6x1EPZd3/u/z7uIg4s0bbPSQifmsCr/95RMyuQhxfjoi9qlDOqRGxy0TLkdRYIuK6iJgVEYsi4rR6x1OJiDgqIr5e/n1GrfrDcruvnsDrn6pSHNdFxKwqlHN2RMysRkySWkNE3DjK843ctzxV/v/qiLi4Rtuc0G+6iLglIvavQhwfj4jXVaGct0XEPhMtR2NjB9669gKOB66scxx1lZlX13BzhwD7AyN2gpMtM6vVwZ4K/DfwTJXKG1FEzMjMnlpsS9LwMvNYgIjYHTgN+HJ9I5qYzPxiDTd3FPAUcGcNt7md/n1YBWcDXwS6q1TeiOwHpMaXmaOdGF1Eg/ctmXkntWvn92IK/KbLzIuqVNTbgE7goSqVN6JW71em9fX11TuGphcRfwCcCfQC/5aZH4uIY4GLKUZnXZmZy8p1nwK+DfwWcDvwTeDPgB2B383Mn0fEZcBm4LXALOB9mXl7RLwW+AwwG1gLvDszn4yIBRQHaweVMZwGfA5YDPwC+DSwL7AHcACwG/CBzPyP8ozf3wCvA9qAP8vM70bEicBFwFbggcx851DLhnk/zgbOgP+fvTuPQ8MRLQAAIABJREFUl7Iu/z/+OsBhkYMQagouqamXgiK4YOoBycrcskVTqHDLvlj2df1aGaWYWS6VqbmlGUoqblmG2y8ScivCzAXFS03JPWU/h+Xsvz8+n4HhMDNnzpzZzpn38/HgwZl7Pvc9133PPZ/rvq97owGY4+5nmtmWwK9jDGuAr7v7KxmW6TRgO0JxZUvgZGAKsA9wh7tfENvNivPTF7jA3X8fz4qZ6e6fMLMTgV3d/Xtmtgtwe5zPOcBod58QP6szy2aj9sBfgDeAfsB7wFnuPifGOBR4yt13ja9HAL9y94PbzfMi4E7gCOAD4Dh3/9DMdgauBYYCHwInuPt/Y/tbCJ3qCuBId19pZnOBU939ZTP7JnBWnN77wCx3n55u3KRYvgjMAN4CFrv7ODN73923Svp+3nf36+O0bgOOivN+IXBp/N4mufvTZrYFcCuwDfAucHych+mEdX1fYCbQQrt1B5EKkUMuuZOQS14DvuTuLfGo3S8JfdG77n6YmX0BOC8O+w/wVUKfmbJfir/pXYHfEPqIVwm/8S8BX3P3f8f+8SVglLuvTTEvlwOfI/yWf+ful6fryzIsj+lAPbAfUAMcT8hBuwGXufuvY/77A2Hnogr4X3d/wswmEPrBie36q/0JOyANhBw8wN1PjJ+1gpB3BxH6rmfNrAa4Li4P4vT/nqo9sASYT8iRSwh9uMd5GQVc4e6fiq8PB45x95PbzfN/4/yMA16Oy3u1mY0l5KOBwL/jsludYT1YRMh9a83sx8CXgTfjx1zs7nPTjZsUy2nAFcDCGMt3ibk16fuZ6e4PZ7ltsxPwW2BwnN5J7r4q5qx/AeMJ2yt7027dQaSHKaP+PmXf0i7WdP1S2njdfavYD58HNBL60OnufrGZ3caGueVp4CrCNuAqdz8wzTJLtW+Sso/OsNxPBA4HtgB2AM6Iy/XTwN8SfbKZ/RrYCxgAXOvu16SYt0SO2TJ+P5sDDxD6/u3jZx0aP2t74EfufkucztS4DPoBV7v7jenam9mTJO3TufttcRq94zIc6e5rzGxInIfd2s3zXEIfe3D8Lia5+2vp9s1i+3lxmfQCjnL3t9r1+V8ALgOWE7YFFrn7tHTjJsWyH/AQYT+2jrBOvcD6fHUi6/fb5gL/jHE3E/a3Lo3f2/+6+wNmtglhW2V3YCUhr7wS8/7HCOvFU4Tc3OE+bE+ky8QKLG7gnQYc6O57Alea2QDCBu8RwBjgM2ZWG0fZktDx7QqMBD7t7vsBNxKKHSS12wv4WnwP4EWg1t3HEIo/34nDzyes2KMIxZJXgKnAI+4+OtFpEH4UBwHHAhfEYacAb7j7voQNsUvMrDqOf0Scp8SZJqmGpTIVGBPbJT7nCuBCd98HOJtQ1Eqcvn9qmukMA2pj+/uAHwB7ACfFzh9CUWFv4EDgIjOryhDXL4FpMa72Rzk7s2w2ah83oM8HfhuX+ZzEhN19KfAfMxsdB32Z9NX9d9x9d+CPSXFcA5wS5/M3hA3shNfi/DwHHJc8ITPbmnBUd29C4tur3WelHdfd7yMk5y+6+7g0sSZ70d33AFYB5xCW1znxH4QC0Z9jmz8CFyeNOwQY6+4/J/W6I9Lj5ZhL7nX3kYRcf7CZ9SMUXU+I0/hqbPtXd9/P3UcDTxI2+LLpl6YCL8Q+7XJCAflr8b3D4nRTFYI2IxzFHBHj+HV8K2VfFk89PyrNoukb++DrCQWG44FPxNggbMB+3t33IhQQfp5mOgk3Ego3exM23JPVuPtYQq45Nw77AWE57wt8Mc5DyvZxg/d6QrFldKIQBODuzwNb2fpLb9PlgY8CD7h7YsP/WxYu97qcsFG9F/AP4Fux/UbrQfLE4o7ewYQN5RMIhTWyGTfu/LwL7OfuG+SIFLLZtrmKUBAbBSwiHKhIWBvXiz+Tet0R6THKpb/voG9JlqpfyhRvsjGEM833BE41s4FsnFvOBk6PMWc6qzHVfkjKPtrCbSJ+lGY6uxJy2OeBO+K/EcCeZmaxzffiPstooKPbJlxAKIbtTjiImmwkITcdSNgWxswOBbaIMe8DnGJmw9K1J/U+HXHf48/AZ+OgzwN/ShNjW1xuPybsD0GafbOoPvbJvyMcpF0nfvdXEM6EPRAwNpR2XHefB9wPnBbnpy5NvAlL4n7vE8CvCOvbFwjfO8C3gQ/j/sXFhHUy4ePAeHc/h+z3YXscFYMKbwKhSroS1u34G/CSu7/l7k2EanGiyr3M3Z9y9zZCcefROHwB4UyYhDvdvc3d/wX0idXejwD3mdkLhA5iRGx7MGEDFHdv9qQzPNp5IHYc/yJUnAE+A0wxs2eBxwhHOIcTEsiNZnYykDi9LNWwVP4J/C5W8BuSYrw5fs6NhEIP7n69pz+F/yF3byVUjBe5++vu3kDYiEx0mmeZ2XPA43GetsoQ157uPiv+fWe79zqzbNK1z+RuwgYuhGT1+zTtZsb/7wBqzWwQYd35Y4zjfMLZNQmJy+BSxbEvMNvd6+I68VC79zON21mJ5foCMDd+b8nr9AGEpED8P/mozz3x9wCp1x2RSjCBzuWS5e7+WPw78Rs2QpF3YdI0ALYzs9kxd0xhfe7Itl9KmJnU/njCjkgqKwhn9NwUizz1mfoydz/f01/Sm9y3/Mvdl7n7YqAqFuergMvivM0ibEinFPNolbs/H/ucu9o1SdUnfga4MMY8C/iorb+HTmf70PuAL8Txx5P6kuLVScviDsIBEQNGAXNiHCcRDkhA6vUg2QHAfe7e5O7vEnIlWY7bGdls24x298Q61j4P3B3/32jd6UJMIuVqAuXR32fqW5Kl65fSxZvsSXdf4uFsozdYvx2d7ClCP/5twpky6aTaD0nZR7v7/Z7+sqbZ8UDGC4RCdHLfleivvmJm/yJsl+5EuMIinQNYv1/Rfv9itruvdvf3gV4xb30G+FyM+R+EYtvHM7TPJPl7zXQp2UwAd/8joUAHafbNokz5LfHdvxu/+/bbDoXav3gqfm8p9y/c/UHCmcMJf4jxQfb7sD2O7hlUfhqT/m5Net0K9E56r63d323Aj4Dfe7jMZx/gZ5387AYAd2+NpxZC2JD+urs/ldzQwuVF+xMu/ZlnZnsQTs/bYJi7p7qPwBHAJ4GjCRXhCTH+vWKRIFvJy6b9cuttZp8ExhLOKmkwswVkTiKZzhrqzLJJ1z6TPxCS7QxgRezkU2lL+r+NUNB9Jx4tSRs3G68/ifgzyTRupthg4+Wc6rvKdrrJpyOnWndEZGPJxdKOfmtXEc5gnGvhJpRHxuHZ9ksAuPtyM1sQLw3YnXCkLlW75pijPks4Wv0lwqn4mfqydDLmAcJZPn0JhYYWM8tUPMilT6wCDouFlHWS8wDZ93X3AD8lHDl+0t0bO2ifyANVwHx3PyRDzOni6DDvZRi3vRY2PMiYnAey3bZJZzWkXXdOzGJ8kZ6sIP29mX2U9H1LOol+qSuxNyU3cPefmtlDhDNi5pnZPu6+JMW0Uu2bpOyjO9AYP7fVzFLtX+xIKKQd4O51Fm5J0aX9i+Tpx/Y/dPfbkxtauJS2s/3yXOB6C7dj2MXd/5GmXfv9ysT/6fbNirl/kZxbOty/6MS+V/L+Rbb7sD2OzgwqvEeBiRbuW5C4P4wDu5nZ8HgE8MuEimRnHBuntyfQ7O4rgE0J92SBDTeOZhNPwzOzPvEIbB3hTJaOzAa+aWa94viJDfUdYhHk+4T75QxKM2wDcTrbuPtswiVCu8S3HieelmdmvWIH3lWbAktjIWgsG1aDU3ku7sRA+E46km7ZpJN2mccj2e8D0wj3VUjnuKT/n4jf+zIzOyTGUG1mHc1nwnzgU2ZWEy+rOzTL8RLaz88qM9smnprcmQ0HCEd9EvP2FVLsRGZYd0QqQT5yycvAxxN9RJwGxNwRf2OTE42z6JdS9WnTCZd43Z10Rt8GYn8zOB5J/g6hUNOVviyTTYEPYiHoGMJ9L1Jy92VAW1L+yebpLLMJl3MA63JyJpnywLOEo5nfIP3R203MLLHzdhyhr3wZ2CERt5kNtOyfxPIU8WykeBlCqss4Mkmen/8C28TPH0K410NnPGtmn4t/p8sDG607nfwMke6gXPr7bPuWVP1SV/Z1NugnzWxHd3/W3S8inP2/bZrxUu2HdLaPzkZiP6rewn1ID+qg/VOs36/Idv/i6xafIGxBpqcJZ8orzYT7oF5FuF9ROsfFzzqScMYO5L5v9jIwwsy2imcufSHL8RLaz89/gNFxnT0y9Shprdu/sHD53Utp2nW4D9tTqRhUYO6+gHAN41MWLlc63d3XEDqmh4Bngb+4e8ojqBl8aGbPEG54/D9x2M+Aq+Pw5KOfFwG7xFNC5xN2op8HBpjZs2b2VdK7gbCB95yZvUg4dR/g53F6zwO3xI3oVMPa6w3cbmbPE24glpje/wKHxWW0gHAGSEf3DOrII8BmMe6zCPe+yeQswn2FniXcjDTd5XQJ6ZZNOnOAsWb2r3jWUnt3EzqsTMWgbeIyPppwJhiEo6PnxmX3LBve8yEtd3+bsG4+AzxMOP21o3lONh2YYWaJywouAP4K/D/CRkBnTAMOj+vFl1h/rW+ydOuOSI+Xj1wSzzQ5Hrg1TiNxGddFcRrzCKfpJ0vbL8Ujs8+b2fNmlriHzp8JG1IzMszOIOCBGMMfWX+fs5R9mWW+Z1BHbifcP+N5wqVXaW9IHU0B7kzKox31iT8Chsdl8BId32vgT8DkmAfa30cBwvI4hJC/UvmAcPnAi4TLBq6L3+tXCEd/nwP+RrhsoUMe7s/wV0L/fwshT3YmD9wIPG5md8Y4fhmnMZOOc257pwPfid/Vjqy/b0WydOuOSI9RLv19J/qWVP1Szvs6KXLL2Wb2YuwbXiZ935JqPyRlH22Z7xnUUXzPEW6m/TLhHjQdFbkuBI6Lse1MB31svJzpEWC+hasariPzGTQd7dPdDUwk89PGesfv+ALW368t5b5ZR+J3fw7hFhpPEoo5nckrM4mX9lk4geHHhPX3KcJ96jrjV8Tvn5AvTkvTLpt92B5JTxPrhizpbu2ljqUnsXDH+TXu3mZmPwD6eHzyQU9lZgM9PK1lIOFIztHu/nqp4xKR7ikehf61Z3dj+bKT1CdWEW5O/KS7Ty9xWAWVNM+bE3b29vKOb9opIgKsf4JWqeMoV/GsnqZ4hurXCDfQP7HEYRVUUl7pQ7ivzyXuPrfEYUkKumeQyHr7AVdYuM70DcIRlZ7ux/Espf7ADSoEiUiuzOwUwpG3E0scSld83sy+Q7jP0HzCmUU93c3xLKVqwn0qVAgSEcmf7YE74v7FUipj/+Jb8SylfoSH6swtcTyShs4MkoIxs5MINwVNdrO7X1WKeEREpLjMbCob3yPhx+6e6XR1ERGRlMzss8Cl7QY/7O7fK0U8It2ZikEiIiIiIiIiIhWk5JeJmZmqUSIiabh7R4/oLDkz2xK4j/A42N6Epxf+m3CT8eGEGw+eFh/3uS9wJeHRoxe7+6wspq88ISKSRnfIE4WmPCEikl66PFHyYhCAe2cfPCQi0vOlfthQWVoM1MZizwTge4Qb0T7t7peb2TXAocCDwBWEy4ZWEJ5C9JC7t3T0AcoTIiIb60Z5ouCUJ0RENpYpT+jR8iIi0iXu3uLurfHlEMJjX8cRniBB/H98fKJGH3d/x93rgVcIj1kVEREREZEiKoszg0REpHszsxHATcC2wNHAZ4Bl8e3lwND4b3nSaInhIiIiIiJSRDozSEREuszdX3L3A4AjgasJhaAh8e3BhMepJg9LHi4iIiIiIkWkM4NEpKDq6+tZsmQJra2tHTeuYL169WKzzTajpqam1KF0mpn1c/eG+HI5sBp4DDgceDn+/4i7rzGzZjMbBqwkXCL2WiliFpHyoTyRne6cJ0REukJ5IjudzRMqBmXQvGoFrWvrCzLtXv1r6DNwcEGmLVJOPvjgA7bddluqq6tLHUpZa2pq4q233uquG/l7mdklQCvhKWFnE4pA083sMWAh8FBsew5wb2x3obs3FzKwQvbjmaiPF8me8kR2unmekHaUn0SypzyRnc7mCRWDMmhdW0/9i08UZNo1I2tBHbFUgKqqKnXcWaiurqaqqns+Hdjd/wYclOKtY1O0nQccUPCgokL245mojxfJnvJEdrpznpCNKT+JZE95IjudzRO6Z5CIiIiIiIiISAVRMUhEimpFfQNvf1DXpX8r6hsyfsbbb7+NmfHII4+sG3bSSScxZcqUlO3nzZvHj370IwAuvvhi6urq8jfD7SxcuJBnnnmmYNMXEenulCeUJ0RE0slHjlCeCDq8TMzMqoG5wEjgFHe/x8xuBHYnFJN+5e4zYttVwPw46iXu/nCXIxSRHqVudSOP/eudLk1j/JitGVzTL2Ob3XffnUceeYTPfvazfPjhhzQ0NNC3b98Opz116tQuxdaRhQsXsmzZMvbaa6+Cfo6ISHelPKE8ISKSTj5yBChPQHb3DGoGjgGSS2CXufurZtYPeM7M7og3AX3D3Sd0KSIRkTzYfPPNWbNmDatXr+ahhx7i0EMP5cknn+S3v/0tc+bMoa6ujokTJ3LcccdtMN7kyZO58sor6du3L2eeeSYtLS3stNNOvPnmm9xwww1MnjyZESNGsHDhQgYPHszVV18NwKWXXsqCBQtobm7mvPPOY9SoURxxxBHU1tby3HPPscceezB16lRuvfVW6uvrmTt3Ltdeey2DBg0qxeIREal4yhMiIpJJT88THV4m5u5t7v5eu2Gvxj8bgRagLb7e1sweM7PbzGyznCISEcmTgw8+mEcffZTHH3+ccePGAXDcccdx6623cuedd3LLLbekfUTlXXfdRW1tLb/97W8ZMWLEBu+NGzeOW2+9laamJtydv/71r7S2tjJjxgyuueYaLr/8cgDWrFnDxIkTmTlzJvPmzaO+vp7jjz+eSZMmMWPGDG3gi4iUmPKEiIhk0pPzRFefJnYOcJe7t8TXH3f3xWZ2MnAxcGoXpy8ikrNDDjmEKVOmsOOOO657AsGDDz7IfffdR1VVFR988AErVqxIOe6iRYv4/Oc/D8CoUaN4+OH1V70mOvPhw4ezYsUKXn31VZ544gkmT54MwKpVqwAYMGAAO+ywAwDDhg1L+1kiIlIayhMiIpJJT84TOReDzOwY4BMkPTrY3RfHP2ey4WVlIiJFN3jwYGpra9dV8QFuuOEGHnjgAQAOPfRQ2traUo77sY99jAULFrD33nvzwgsvpP2MtrY2dtppJw4++GDOOeccABobGwE2erRjW1sb1dXVtLS0bDQdEREpPuUJERHJpCfniZyKQWY2AfgWcIS7t8ZhA4G18Syhg4BX009BRCrVoE36Mn7M1l2eRra+/e1vA+GJAAC1tbVMmjSJXXbZhU033TTteMceeyxnnnkmc+bMYfvtt193JCCVCRMm8PTTT6+r5I8aNYpzzz03ZdsxY8Zw2223sWDBAn7yk59QU1OT9byIiFQC5QnlCRGRdPKRIxLTyVZPzRNV6apYyczsLmAfoB54GDgaqAOWxyZfBrYDbgJWAg2EJ4+9mcW029w9p+ALrXHJO9S/+ERBpl0zspa+m3V9JRYpd2+88ca6Uxu7k7a2NlpaWujTpw/33Xcfr732WtoOOV/aLyszw92rMoxSEbqSJwrZj2eiPl4ke8oT2VOeSK2c9yfSUX4SyZ7yRPY6kyeyOjPI3Y9tN+g7KZp9CIzJNkgRkXLW3NzMCSecQFVVFb169eJnP/tZqUMSEZEyojwhIiKZlHue6OoNpEuqedUKWtfWF2z6rY0NBZu2iJS36upqbr/99lKHISIiZUp5QkREMin3PNGti0Gta+sLenrlgB1HF2zaIiIiIiIiIiKl0K2LQSIiIiIiUt7MrBqYC4wk3Ff0HjObRrjv6IfA++4+MbY9EpgKtAFnuPv8kgQtItLDqRgkIiIiIiKF1AwcA0xpN/wCd78n8cLMegMXAeOBTYE7gdpiBSkiUklUDBKRosrHvb569a+hz8DBad9/++23Ofroo9lll11Ys2YNEyZMWPdIyGwsXLiQf/zjH5xwwgldilNERDpPeaLncfc24D0za//WD8zsDOBad78D2Bl4xd3rgDozqzaz/u6+tsghi0iZytd9g5UnVAwSkSLLx72+akbWQobOG2D06NHccMMNtLW18cUvfpHjjz+eTTfdNKvp77bbbuy2225dilFERHKjPFExrnb3aWY2GHjUzJ4AhgLLktosj8PeLUWAIlJ+8nXfYOUJFYNEpIdramqira2NPn36cOmll7JgwQKam5s577zzGDVqFEcccQS1tbU899xz7LHHHkydOpV58+bxyCOPcP755/OnP/2J3/zmN2y33XbU1dVx9tlns3r1an79618zcOBAXn/9dc4//3zGjh1b6lkVEZEcKE+Uhrsvif+vMLO/ALsDi4AhSc0GA0uLH52IyHo9NU+oGCQiPdKzzz7L5MmTeeedd/jkJz/J/PnzaW1tZcaMGSxdupQzzjiDGTNmsGbNGiZOnMh5553HUUcdRX39+tNOW1pauO6667j33nvp3bs3Rx111Lr3Vq9ezU033cRLL73E9ddfr418EZFuRnmitMxscCwE9QH2B24CXgd2MbOBwCCgWZeIiUip9PQ8oWKQiPRIidM6W1tbOf3001mwYAFPPPEEkydPBmDVqlUADBgwgB122AGAYcOGsWLFinXTWLp0KVtuuSUDBgwA2OBUz1133ZWqqiqGDRvG8uXLizVbIiKSJ8oTxWVmdwH7APVmNhYYamYjgN7A7e7+Smw3DZhNeJrYWSUKV0Skx+cJFYNEpEfr1asXgwYNwsxYu3Yt55xzDgCNjY0AVFVVbdC+ra1t3d9Dhw7lv//9L2vXrqV37968/PLL695rP56IiHRPyhPF4e7HZtnufuD+AocjIpK1nponVAwSkR4pcVpnc3Mzw4cP56CDDlo3DGDUqFGce+65GafRu3dvpkyZwsSJE9l6663ZfPPNqa6uLkb4IiJSYMoTIiKSSU/PE1XJVatSMLM2d89p3MYl7+TlTuLpDNhxNGtef7Yg064ZWUvfzbYuyLRFyskbb7yx7rRJKM4jg/OpqamJ6upqGhsbOfroo7n99tsZNGhQQT6r/bIyM9y94g8tl3OeSEd9vEj2lCeypzyRWlfyRKkoP4lkL7nvK9aj5fOpXPNEh2cGmVk1MBcYCZzi7veY2ebADMKN3Wa7+7TY9khgKuEa3zPcfX7XZkVEepo+Awd3+BjHcnLPPffw4IMPUl9fz7HHHluwjltERALlCRERSae75Qgo3zyRzWVizcAxwJSkYd8Fbnb3u83sgXjzNwcuAsYDmwJ3ArV5jldEpKgmTZrEpEmTSh2GiIiUKeUJERHJpFzzRK+OGrh7m7u/125wLTAr/j2LUADaGXjF3evc/R2g2sz65zVaERERERERERHpkg6LQWkMdPc18e/lwND4b1lSm8RwEalg1dXVrFy5stRhlL2VK1eWzc3kRESKSXkiO8oTIlKplCey09k8kevTxFabWX93XwsMBpYSCkFDktokhotIBRs+fDjvvvsuS5YsKXUoZa26uprhw4eXOgwRkaJTnsiO8oSIVCrliex0Nk/kWgx6HDgc+D1wGPB94FVgFzMbSLixdHMsFolIBevVqxfbbLNNqcMQEZEypTwhIiKZKE8URlbFIDO7C9gHqDezscBlwK1mdjbwqLu/GNtNA2YTniZ2VkEiFhERERERERGRnGVVDHL3Y1MMPixFu/uB+7salIiIdB9mtj/wC6ARqAe+SsgvMwhnis5292mx7ZHAVMJBgzPcfX4pYhYRERERqWS53kBaREQk4T/Ap9z9IOBPwGnAd4Gb3b0W2NfMRphZb+Ai4BDgy8AVpQpYRERERKSS5XrPIBEREQDc/d2kl41AM1ALnB+HzQLGA63AK+5eB9SZWXXSwwhERERERKRIdGaQiIjkhZltBnwL+A0w0N3XxLeWA0Pjv2VJoySGi4iIiIhIEakYJCIiXWZmmwB3A6e7+2JgtZn1j28PBpYSCkFDkkZLDBcRERERkSJSMUhERLrEzPoAM4Gr3f2pOPhx4PD492Hx9avALmY20My2App1iZiIiIiISPHpnkEiItJVkwj3BNrUzM4AHgAuA241s7OBR939RQAzmwbMJjxN7KzShCsiIiIiUtlUDBIRkS5x9xmEx8i3d1iKtvcD9xc8KBERERERSUuXiYmIiIiIiIiIVBAVg0REREREREREKoiKQSIiIiIiIiIiFUTFIBERERERERGRCqJikIiIiIiIiIhIBcnpaWJmNgK4Nr6sIRSVrgJ+CLwF4O4T8hCfiIiIiIh0Y2ZWDcwFRgKnuPs9ZrY54UmUg4DZ7j4ttj0SmAq0AWe4+/ySBC0i0sPlVAxy95eACQBmdhowJL51nbv/LD+h9RwNjc00NLVsMKxtdSNrWuoK9pmDNunL4Jp+BZu+iIiIiEiWmoFjgClJw74L3Ozud5vZA/FgswMXAeOBTYE7gdpiBysiUglyKga1Mwk4ntBpn2JmXwDucfdf5mHaPUJDUwtvvrdh4WfIFvUsWNJQsM8cP2ZrFYNEREREpOTcvQ14z8ySB9cC58e/ZxH2JVqBV9y9Dqgzs2oz6+/ua4sasIhIBejSPYPMbHugt7u/DvyBcOrnp4DDzewTXQ9PRERERER6oIHuvib+vRwYGv8tS2qTGC4iInnW1RtITwRmArj7cndvcfcGQmFoTFeDExERERGRHmm1mfWPfw8GlhIKQUOS2iSGi4hInnW1GHQc4VpezGxw/L+KcJrnq12ctoiIiIiI9EyPA4fHvw+Lr18FdjGzgWa2FdCsS8RERAoj53sGmdlIYIm7vx8HnWNmhxDu/P+Yu8/OR4AiIiIiItK9mdldwD5AvZmNBS4DbjWzs4FH3f3F2G4aMJuwT3FWicIVEenxci4GxQ7700mvz2f9TeCKGwLMAAAgAElEQVREREREREQAcPdjUww+LEW7+4H7Cx+RiEhly8fTxERERERERKTMtbW00LjknaJ+Zq/+NfQZOLionykiHVMxSEREREREpAK0Nq5hzevPFvUza0bWgopBImWnqzeQFhERERERERGRbkTFIBERERERERGRCqJikIiIiIiIiIhIBVExSERERERERESkglTkDaQbGptpaGrpsF2f5hZWrmro8ue1tLR1eRoiIiIiIiIiIvlQmcWgphbefK+uw3Z9t2vKql1Hhm2+SZenISIiIiIiIiKSD7pMTERERERERESkgqgYJCIiIiIiIiJSQVQMEhERERERERGpICoGiYiIiIiIiIhUkJxvIG1mq4D58eUlwF+B6cBwYAFwmru3djVAERERERERERHJn66cGfSGu0+I/x4GTgaedvdxQCtwaF4iFBERERERERGRvOnKo+W3NbPHgLeA04FxwIXxvVnAQcCDXQtPREREREREiqmhsZmGppa8TKttdSN9+jUwuKZfXqYnIvnRlWLQx919sZmdDFwMDAWWxfeWx9ciItLDmVk1MBcYCZzi7veY2ebADGAQMNvdp8W2RwJTgTbgDHefn3KiIiIiUjINTS28+V5dXqY1ZIt6hvZrVDFIpMzkXAxy98Xxz5nAFGARMAR4HxgMLO1qcD3ZR2qq2Z2mgkx7ba/+BZmuiEgazcAxhFyQ8F3gZne/28weMLMRgAMXAeOBTYE7gdpiBysiIiIiUulyKgaZ2UBgrbu3EC4HexX4G3A48HL8/5F8BdkT9W5pYPnzTxRk2kNGjS/IdEVEUnH3NuA9M0seXAucH/+eRSgAtQKvuHsdUGdm1WbW393XFjVgEREREZEKl+uZQbsCN5nZSqABOAVYDEyP9xFaCDyUnxBFRKQbGujua+Lfy4Ed2PBy4sTwocC7RY5NRERERKSi5VQMcvd/AmNSvHVs18IREZEeYnXSWT+JS4eXES4nTtAlxSIiIiIiJdCVR8uLiIik8zjhkmGAw+LrV4FdzGygmW0FNOsSMRERERGR4uvK08REREQAMLO7gH2AejMbC1wG3GpmZwOPuvuLsd00YDbhaWJnlShcEREREZGKpmKQiIh0mbunukz4sBTt7gfuL3xEIiJS7sxsFTA/vrwE+CswHRgOLABOc/fW0kTXfTQ0NtPQ1JJV2z7NLaxc1dBhu5aWtq6GJSJlTsUgEREREREphTfcfULihZmdBjzt7peb2TXAocCDpQquu2hoauHN9+qyatt3u6as2g7bfJOuhiUiZU7FIBERERERKYVt45OI3wJOB8YBF8b3ZgEHoWKQiEhBqBjUQ7W0tvL2B9kdIciHQZv0ZXBNv6J9noiIiIh0ex9398VmdjJwMTCU8ORJgOXxdUE1r1pB69r6Qn/MRlobO75US0SkkFQM6qHWrG3mGf+waJ83fszWKgaJiIiISNbcfXH8cyYwBVgEDAHeBwYDSwsdQ+vaeupffKLQH7ORATuOLvpniogk06PlRURERESkqMxsoJn1ji8PAl4FHgMOj8MOBx4vRWwiIpVAZwaJiIiIiEix7QrcZGYrgQbgFGAxMD3eR2gh8FAJ45M8yvctLHSLCpGuUzFIRESkA9k+trdfdW/69VVqFRHpiLv/ExiT4q1jix2LFF6+b2GhW1SIdJ22WEVERDqQ7WN7txs2SMUgERERESl7umeQiIiIiIiIiEgFyenwpZntD/wCaATqga8CZwJfBj4E3nf3ifkKUkRERERERERE8iPXc9n/A3zK3Veb2anAaXH4Be5+T35CExERERERERGRfMvpMjF3f9fdV8eXjUBz/PsHZva4mU3KS3QiIiIiIiIiIpJXXbpnkJltBnwL+A1wtbuPBo4E/s/Mts1DfCIiIiIiIiIikkc5P/LEzDYB7gZOd/fFieHuvsLM/gLsDrzV9RClsz5SU01163J236ypINNf26s/r33Y8SOWRUQqTVtbGytXNdC2upE1LemfPjZok756JK6IiIiIlEyuN5DuA8wknA30VBw2OBaC+gD7AzflL0zpjN4tDdQt+BvLF6/uuHEOhowaD1QXZNoiIt1Zc3Mr7y1ezQ47NLDsv4vSthuwZQ2NDX3z/vm9+tfQZ+DgvE9XRERERHqWXM8MmgSMBzY1szOABwAzsxFAb+B2d38lTzGKiIh0K71bGlj+/BNp39902CCqBub/zKCakbWgYpCIiIiIdCCnYpC7zwBm5DkW6cZaWlt5+4P0l0QUgi6zEBEREREREem8nO8ZJJJszdpmnvEPi/qZ48dsrWKQiHRJQ2MzDU0d3wOtpaWtCNGIiIiIiBSHikEiIlKxGppaePO9js9qHLb5JkWIRkRERESkOFQMEhERERERkYIoxJOOBzQspnHJyoxt9FAFkcxUDJJO+0hNNbuzYWc+JE8dvB5bLyIiIiLScxTiScfZPIhBD1UQyUzFIOm0VE/JGbD5Jnnp4PXYehEREREREZHCUjFIRESkyNra2li5qqHDdv2qe9Ovr1K1iIiIiOSXtjBFRESKrLm5lfeyOJtyu2GDVAwSEelhGhqb6dPcktVBgWxU4hMvszmo0ra6kTUtHT8kAmDQJn31lGKpONrClG6rpbWVtz/IroPPByUJEREREemqhqYW2lY1ZfU0y2xU4hMvszmoMmSLehYsya7gNn7M1trOl4qjYpB0W2vWNvOMf1i0z1OSEBERERERkZ5AxSARERERERGpWPm+4kBXFEh3oGKQVIydtuhN/9a1OY8/oGExjUtWpn2/V/8a+ujxlSIiIiIiJfeRmmp2pymrtv1WL+aDpV1/MjLA2l79Gb7NVioGSdlTMUjKSmc67SGty9l9s+zaAgzp38yivz+Wa2hsOmwQVQPTd+o1I2uhQMWg5lUraF1bX5Bpq4glUr46+9SxtpYWGpe8U4TINqR+REREyk3vlgaWP/9EVm0HbL4Jy7N4sEM2howan5fpiBRa3otBZvYN4CSgETjZ3V/P92dIz1XITnvz/WpzDQvoeKesM08syFbiFNPWtfXUv5jdcumsQhaxRFJRnshetk8d23arGhqaWuizup6lC59O265Qj6pXPyIi+aQ8Id3ZR2qqO7yioBB0YEY6K69bhGY2FDgFOBAYA1wCHJvPzxAplY52ynbYoYFl/12U18/s/9GBLFlWRXVbc94eP9pechGrT+9eNLe0FuRz0tE11ZVFeaIwEv1T3+0yP51Gj6oXkXKnPCHdXe+WBla+MI+6qqq8TrejAzoDd92/YFcSZJIoQq2ob6BudWPepqt9hMLL9xbhWGCuuzcD883M8jx9kbLVmbOasjVg8014b/FqdtqvNm+PH20vuYj10aGbsDRP10snrO3Vn9c+bEn7frGf0pbvRJUNJbMNKE+UULaXnfWqqqK1rS3rdn3WNlL31htdnl7fPr3p17d3h+1AR0CLpZCXKWdSqu+3FPOrdXkjyhPS7WV7Zm9nJM4CTqejs4PbyzY3d2TQ7uNYs6oXDY0tzHvx/S5PL6Ecn+Tc03JivotBQ4FlSa97ZTNS+fbxV2naRZ9+d512oaffXactspEelifSKdXvSr9nEen2KiRPlEop8kSl5UTlYuke8l0MWgaMSnqdvnQZuXt+z58TEZFypjwhIiKZKE+IiBRBvotB84Dzzaw3sCfwap6nLyIi3ZvyhIiIZKI8ISJSBHktBrn7UjO7BXgcaAK+ns/pi4hI96Y8ISIimShPiIgUR1VbHm4aJSIiIiIiIiIi3UNWN2QTEREREREREZGeQcUgEREREREREZEKomKQiIiIiIiIiEgFyffTxHJmZt8ATgIagZPd/fWk9/YFrgSqgIvdfVYcPg34NFAHTHb3xeUcs5l9Hvg+0AC8Axzv7k3lHHPSe18DrnP3QcWMN352LuvGEcDZQG/gN+4+o1zjNbNq4HZgqzj8VHdfUKx4s4j5V8DRwL3u/u04rBdwDbA78C5woruvKfOYpwAnA83AfHc/s5jx5hJz0ntTgUnuvnsx4+0OyjV35NAPbA/8E3ghNjvT3Z8tYkwl+53nENcE4DbWP2Fooru/X6y4zGwIcC/Ql/Adnu7uzxRjeeUQ0wRKuKzie/cCmwObAJe5+91x+DRK+zvcKK5S/w7j+4OBfxO2Be4ph3zb3ZVrnsiXHPLNcOBPgAET3P3pEoTdKd1xn6szcpi/PYHrYvtW4Gvu/m7xI89ed9wH7Yxy3O7LRlmcGWRmQ4FTgPHAucAl7ZpcAXwZ+AxwkZn1NrORwL7uXgvcDHyniCHnFDPwLFDr7uOB/wDHFi/inGMmFiu+DLxVvGiDHNeNLQg/xkPcfUKRC0G5LOMJwEp3HwecR/mtyz8BvtJu2GFAa4z5acLyLpocY54NfMLdDwS2MrMDCx5okhxjTow3ouABdkPlmjty7WuBv8U+a0IBdkDL8nee6+8CuC9pWRWiuJEprgbCjsW42ObSOLygyyvHmKC0ywpCIfsg4GDgR3GccvgdbhRXVMrfIcD/ER6tnlDSfNvdlWueyJcc881S4BDgniKGmrPuuM/VGTnO30vufoC7TwBuBb5VvIg7rzvug3ZGOW73ZassikHAWGCuuze7+3xCpRoAM+sP9HH3d9y9HngF2BkYBySqhrPi67KO2d3/k1SVbiScoVDWMce3pwDTCZXnYssl5sOBeuAhM7vfzLYr83hfJxzNBRgCfFjEeDPGDBCPNLR/7GD739/4gke5oU7H7O7/dvfEsLL6/cX4Ui1ngO8BPy9CfN1RueaOXPvafc3scTO71swGFCsmKOnvPNffxZFm9oSZXWxmVcWMy93XuPs78WVyX1Lo5ZVLTFDCZRVja4x/1gAvxr9L+jvMEBeU8HdoZlsCOwLzkwaXOt92d+WaJ/Ill/2Rte6+pETx5qI77nN1Ri7zl3yW02Dg+aJG3HndcR+0M8pxuy8r5VIMGgosS3rdq917y5NeL4/D1o3j4XTZmgLH2F4uMQNgZh8HPgvcV8gAU+h0zGY2EDjU3Ysda0Iuy3k4sC2hKHQVcHmBY0yWS7xvAzVmthC4Nv4rpkwxZzPOBut3keQSMwBmtj+wpbvP67BxfnU6ZjPbBhjm7s8ULKrurVxzRy5xvQfsFI/+LwbOKGJM2YxTqN95LnE9TdjQGk/o748rRVyxsPIL4LIU4xRieeUSU7ksqz8DzwEPtx+nRL/DdHGV+nc4lQ3P6mo/TinybXdXrnkiX3LeH+lGuuM+V2fkNH9mdrCZ/YNwVlC5byd2x33QzijH7b6slEsxaBnhjIiElgzvDSac3rhueKy41Rc4xvZyiRkz+yjwO8K1nY0UVy4xnwX8qvChpZXruvForM7OBnYreJTpY8om3hOAV919N+BQwjXAxZQp5mzGWbd+F1EuMWNmuxB2lL5WiKA6kEvMPwR+WphweoRyzR2djsvdG9y9Lg6bCYwpYkzZjFOo33mn43L3+ri8WoG7yf+yyjauKwlHAuekGKcQy6vTMZXLsnL3zwC7AOfGe+KU+neYMq5S/g7NbAdgiLu3P8Jf6nzb3ZVrnsiXnPZHupnuuM/VGTnNn7s/6u5jCZcl/aTQQXZRd9wH7Yxy3O7LSrkUg+YBB8XrdPdi/Y0OExX5ZjMbFiuEOwOvAY8RrqOGcAbIE+Ues5nVEG7weJa7v5pyqmUWM7ArcKaZPQx8zMxu7gYx/xXYC8DMRhOuFS7neHsRKsKwcYdR0pgzeIzwuyP+/3ihgkuj0zGb2TDCRsFkdy/2pXiQ23LeEfhF0u9PhaENlWvuyCU/bJo0/kFkt37kJaYMivE7z+W3PDjpZSGWVYdxmdn3gWZ3/2XS4EIvr07HVOplZWa94j0fAFYDa+O/kv4O08VV4t/hGODjsb//GjDVwr1rSp1vu7tyzRP5ksv8dTfdcZ+rM3KZv35J4y8Hyv2m8t1xH7QzynG7LytVbW2pLsUvPjM7FTgeaAK+DtQCb7j7X81sP8KNl6qAn7r7/XGcHxFu/FdHuHFiUXfwOhuzmf0QOA14OU6iqE+5yiXmduMu8BI8zSjHdeMHhJt09QKmuPtL5RpvTFh3AIMITzWZ6u5/Lla8WcQ8FfgS8FHCDes+a+HpJtcRbmz8HuHpJqvLPObfAJ8CFsVJXFzuy7nduCX5/ZW7cs0dOfQDRwIXEo5ALyP8ppannnpBYirZ7zyHuKYA3yBs/P4b+IYX4Ckx6eIi3OdtEWEHsQ14x92/WozllUNMpV5W84EHY7O+wDXuflscp2S/w3Rxlfp3mNRmGrDA1z9NrKT5trsr1zyRLznkm2rgIcI69RZwp7v/ojTRZ6c77nN1Rg7z9yXCZUWtcZz/cfdFJQk+S91xH7QzynG7LxtlUwwSEREREREREZHCK5fLxEREREREREREpAhUDBIRERERERERqSAqBomIiIiIiIiIVBAVg0REREREREREKoiKQSIiIiIiIiIiFUTFIBERERERERGRCqJikIiIiIiIiIhIBVExSERERERERESkgqgYJCIiIiIiIiJSQVQMEhERERERERGpICoGiYiIiIiIiIhUEBWDREREREREREQqiIpBIiIiIiIiIiIVRMUgEREREREREZEKomKQiIiIiIiIiEgFUTFIRERERERERKSCqBgkIiIiIiIiIlJBVAwSEREREREREakgKgaJiIiIiIiIiFQQFYNERERERERERCqIikEiIiIiIiIiIhVExSARERERERERkQqiYpCIiIiIiIiISAVRMUhEREREREREpIKoGCQiIiIiIiIiUkFUDBIRERERERERqSAqBomIiIiIiIiIVBAVg2QdMzvRzDYvdRz5YGaLzKy/mQ03s1uL9JlDzOyULow/3cwOzUMcp5rZMXmYzgQz26er0xGpVGb2CzNbYGbndnE6Wf0WzWyume2a5TRnmtmE+PfsrsSXCzN70Mz6FvtzCyV52RdzeZrZ/3Vh3GlmdmoeYjjKzE7Pw3RGm9mnuzodEemcUueDzirG9qmZbW9mf49/56WPy/Jzu9QPJvZ/8hDHTWa2fR6m02P2LXsqFYMk2YnARj9YM6sys265rrj7u+5+fJE+bgiQczEoX9z9ene/Jw+TmgAUrRhkZr2L9VkiRfIVYA93vzx5YA7r+gQK+Ft096LvgLv74e7eWOzPLYYiL8+ci0H54u73u/tVeZjUaKBoy045R3qyXNfvUuSDdDLMwwSKuH2axz4uG0XtB9Nx91PcfVEeJnUiKfYtC0X9eudVtbW1lTqGimVmw4G7gYGEwtyxhB/NK+5+c2xzH3AZYMDhwBbADsAZhM7i08Df3P3k2P594N44fB5wF/BDYBPgc+6+yMy2BH4NbA2sAb4OjARmAG8Bi919nJn9F/g9MC7G2dvdz4+fcwXwD3e/I8V8fRK4CmgBVrn7gWbWB/g5cABQDfzQ3f+UYdlMAL4PNAMjgEuAzYDjgHfivDSb2YVxuQwA7nf378fxFwG7AlsBM939E2Y2ELgN2Bl4EjjK3beKn3Ue0BjHme7uF8fpnAx8E+gL/N7dL0zX3sxuA74AvArclrwDaGb/BL7o7m/GwtorhJ3ENUltpgOrgf3j553s7vPMrAa4Ln4WwP+6+99j+xWx/SBgkrs/a2bTgPfd/Xoz2x+4CWggrA8D3P3EdOMmxbItMB9oApbE5X4DcKq7vxyXwanuPjFOqx7YD6gBjgfOB3YDLnP3X8d5voqQwBuA0+I8nAgcSViv3wFupN26g0g3ZGb3AJ8HXgS+S+gz/gWMBy4nFI+/DvQD/gl83d1b49klN8T31xLywjw2/C1+Ms24c4m/0TQxXRSn9x+gCrjY3eea2ftJfWE2/e5YQn8+EPg3cIK7r475505C/nkN+FL86FuAMUAr8FN3vz3RR7v7WjP7ATAxvj/V3f+UqV9OMV97ANOB3kAb8El3X25mUwl9cj/gane/MdX4cRrbA/cBDuwL3AG8AXwrxnBYnOapHS37pOXZm/BdHggsJOTuL8aP/H0ctjcwx92/GeM4gpCz+wN/j5+/Xar2ZnYx8B3COjbH3c9Kmp/fA1e4++Px9Tzgy+7+ZlKbacDHCPl/CHBOXPYp83VsvzWhbx8GfNPd/1/sx3d19++Z2S7A7XG8OcBod5+QbtykWHrH5d0PeA84CziBkL8fjt9PIpdPi8tkV2BL4GRgCmHn8A53vyBOM9169QPCOt4POJ0U6w4i3VD8nfyBsI25p7ubmc0i/Ob6Ahe4++9j22zywanuPjG2nwucCqyk3b5LqpxjZjcCt7v7HDN7jtCHX2FmtxD6xWeB3wC7x2me5O6vJPVLuwJPEbZFzyfkwFeAs2m3ferunmZ5TCO7vmKjZdSuzzmRHPu4zvSnwF9o1w+6+5wY41DgKXdPnIE6AviVux/cbp4XEfLwEcAHcfl8aGY7A9cCQ4EPCXn7v7H9LYRcuQI40t1Xtstp3yT0yR8A7wOz3H16unGTYvkiG+9bvu/uWyV9P4n9lUWEfbSj4rxfCFwav7dJ7v60mW0B3ApsA7wLHB/nYTph/2lfYCZhH+JUwv7GHHc/E0mrW57t0YNMAma7+2jCBt5/CBslkwHMbDNgZ3f/W2y/K3AYYQfjjvhvBLCnmVlssyXhx7QrYQPv0+6+H2Ene0pscwVwobvvQ+hUf+Hu9wFPEwoW42K7jwJ/cPfdgSuBY+JZQn0Incwf0szX2cDpcb4Oj8NOAd5w930JO0OXmFm1me1jZjelmc4o4KuEH/elcfxRhALWp2KbK+M0RwGjzGzPNNMCOA14zd1HAvfHZZUwhlCI2xM41cwGxo72UOAT8f29k05L3ag9MBV4wd1Htz8TALiH9TtG44BnkgtBSbYE9gK+RvjOIGy43hvn84vANUnta9x9bGyT6lKUGwnr2d5sXJlPO667vwVcT9g4GJ0u0SbpG+O7nlCAPJ6w3KbG948hdN57EJbb9KRxRxESyFdIve6IdDvufgywJP5+HomD17r73u4+E7jL3fdN6tM+F9vMAKa5+57AIYSNqPa/xXTjphULOJ8hbHifRPh9ppKx37VwadflhGL6XsA/CAULCP3XvbGP7QUcTDjKua27j3T3PYAHUsR1FKGP+ixwdSyAQ+p+NpX/IWwUjyYUXuotXHK7ReyX9gFOMbNhFi4dfjDNdHYjFMNGxM/9iLvvDfyNUBCDzi37o4HB7r4bYcN2VNJ7I4BphDy9v5ntHE+lPxOYEOelhfV5Y6P27j6V9evYWWzo7vj5mNnHAJILQe3m+UBCgfFKC5cXpMzXsf3HgIMIO5EXpJjeL1m//ja3ey/tuO7eQtjh+22cnzkppp1sGFBLyBn3EfLYHsBJZlbTwXq1N3Ciu3+SFOtOB58rUu5GErbxE/sFx8d+7EDgorgdn20+SCXVvksqTwIHmNmmhIL6AXH43oRC+reBD2NeuJhQqEj4ODDe3c8hbEceEfuUU1Jtn1q4PUK6S14z9hWxzUbLKMP8d7aPy7o/zdQPuvtS4D9mNjoO+jJh3yKVd+K+2x+T4riGsPz2JhThfpjU/rU4P8+xPtcBYGZbE/LS3oTt8r3afVbacdPsW2byYlwfVgHnEJbXOfEfhDz659jmj4T1JmEIMNbdf05YZ8bEuFLlKUmiYlBpzQcmm9kFhKLPmlhd72tm2xGOaN2Z1H62u68FXiDsVDzl7m2Eo4LbxTbL2g1/NA5fkNTmYOBmM3uWUCwYlia+usROjIcjZc8ROsrPAo+nKWZAqORfZmbfJlS3ISSdKfEzHyOcjTLc3Z9293SXVj3l7svc/UNCxXlWinn5lJnNJxxh2JewwZzOAcTl6e6zCFXkhCfdfYm7ryZU5YcTCk77E5LWM3HaO2don8m6DXNCYSRdB36nu7e5+7+APmY2hLDsLozLbhbw0ViQg1DUgnDGwfbJE4rjVrn783F9uKvdZ6UdNweJ7+YF4F/xe1sMVMWkdwDhCFGbu78ArLZwhhrAw+5eF/9Ote6I9BR3J/29p5k9aWYvEAoKI+JG86DEBqC7r3D31hTT2WjcLD77AMLZjU3u/g7weJp2HfW7RihqzIl90kmEDVqA5e7+WPw70a+8DmxnZleb2afcfUWKuO5x9wZ3f4/Q146M72Xbz/4NONfMvkMoADUT+s3PxRj/QTi48XEPlw6nKzS/5O6vu3tDjDtx5kpyzunMsj+A2O+6+3NA8tHzhe7+atz4f56wrPYnLNu/x7g/DeyYoX0ms4DD407NMYQzhlNJXideJhxISpmvY/sHYgzp8saeMb/Chtsv2YzbGQ/F38YLwKKk720RYZsm03r1uLu/H/9Ote6IdGcL3f3FpNdnxTNzHif87rYi+3yQykb7LmnaPUnYZ9ifcPB4y7jdtzz+Vg8Afgfg7g8SCtMJf3D3pqTp3GjhTP2Ul7N4uD3C9Wni6KivgNTLKJ3O9nFd6U/bu5vQn0M4OPz7NO1mxv/vAGrNbBDhu/hjjON8wgHahEz7A/sS9j/rPJz181C79wu1LzE3fm/J+XfdOhP/T7564J64rwNhv+13ZjaRcHaQZNCn4yZSKO7+mJmNJ2xQ3mtm33b3vxBOgfsq4ajWV5JGaYzjtZpZ8r0WWgmnOK9rkzS8MUWbNmCvNDsZyVa3ez2dcNbSEMJlS+nm66dm9lCcr3nxbJoqwun0T3Xwmckyzks8gvkLYG93/8DMfkXmAkKmSn9yZ5FYVlXAde7+k+SG8dTZVO2bSMPdX4tnG21NOLvre2matrX7uy3GcZi7v9sujuS4k7/fhEzz29G4qbSwvoDcfjknfzfp1s101q1nqdYdd1+SRWwi3UFyn3oj4Xf9qoWbANekGSeVXMdt37+k0lEOqQLmu/shKcbdqF9092VmNopwRPG7ZjbO3adlGW+qfnYjHi47m0/oN+ZYuNSqinA6/u1ZfhZklz87s+xzyTl/dPf/SW4YL1fIalkkuHudmb0EjCXsPHw1TdN0OWejfJ2ccyHwBv4AACAASURBVOJ2SKoYOpznDOO2V4ycs9G64+4Ls4hNpFytW78t3LZhLOGMiQYzW8D631JH+SD590divAz7LhuIfeROhLPh/z97dx4nV1Xmf/zTnXRISCeNgKMEREHggQQCYVOwAxEUAjiCIpAMExBkBhWGdRjFOBBlmJ8Io6MIGEcFEkWWCIhsUWRLQCMMBugQHsKwJ6whWyed3n9/nNNJpVNVXVVde33fr1de6aq699zn1q06z61zzzn3j4SLqdMIF/0y3gfC8KmDCL+JFlgYFpyNtHXFAO9RMtnWcYOpT/u7k1BPzQZWJTRq99eb8H8v4Tgu9dCbK2XcFOe3ROJnLZN6PdNyEz8zxxB6ux5PGC42KYP1a5Z6BpVQ7Lr9prtfR2jF7avgbgb+Ceh095cLsOl5xImOzaw+oWJdQ2ixTuWPhFbYvYFHUi1kZju7+0J3v4zQ8v4R4AHgaxYnok7o5jgYwwmVxAoLY2kHGirxOKFbJWZ2NGEepXQeBKbEHjaY2Q4Whu6lMtD791vgKuAZd1+bYpkT47b2BrriVfQHCEPcSHhtQO6+AuhNOL7Z3mGs//68ShjyASEpZ+Nx4ITYPXkcYe6it/svlOKzI1KNRgLvxkbtkwDiVbdV8eQUM2uKdWb/7+Jm62bgceALFobnjiGcoOfieWCnvnolNnLvkmrhOPSpzt1vIXTp7l/3J8b1YcLQsEVkwcx2InRV/wEhNxmh3vxKfI+wYNB3WCG79/5xYr0b36+B7vT2F0Jv1x3iOtv0/Z1Gj6W+wcNthHkeGtz9pRTLJH4mdifMmTSYfP10zK8Q820W8p1zBvxcpfjsiFSL0cD7sZHjQDb2vskkH7wGjDOzofH3ynhI+9slmecIF7X/Sujhc178vy+Gk2KZk+OyyewUG1K+RZhzZxQDn29nI9V7lEq2dVy29WnKfYs97t8iDBlO1dsTNuamk4D58bfECjM7IsbQYGYD7WefJwh5qdHCsLps73rcf3/Wxt9TWxCGwmdjw2eG8Lma33+B+D7v4O4PEIaX7ZblNmqOGoNKaxLwjJn9jVAR93WXXEUYkjW7QNv9F+Co2CWyhdCCCqHnz2wzS9pdNPYk+iNhToh0M49fYGaLzOwZwg+HpwmTxb1NqEQXEbooYunnDErLw9C13xAm1byd0N07nWuA3eP2jyXMxZGu/BZC480jcV9uJfwQSLX8csLxfMaS30r6NsLQv3R3+nrXzJ4iTE7Xd3X4u8CYWO5zZHfHsjOBW2KZrYRJ+jL1e0JX4L9ZuITxQ+DbZvYkm4+THsgcwvF/lvC5Pi3Fcsk+OyLV6HLC0JWHCN2r+5xCGBb6NHA/4eS3/3cx1bopuftfCZNTtgDXExoesubhDmD/APw0xvhnIGVjEKEr+iMWuqb/F2FS6v5x3UPYjz8Q5gzLdt6WKUBL3MYowtDTe4G5wBPxSu91hKvA6eYMykQ27/0cwpDYxYSTdydNHezu7xAa/n8X68A/EIa3pfMr4FkLN3Xo7/eEiT1TDSUgxvQYYX/O9zAUPWm+ztD5hDk3FhJ6TWWTcx4CDoyf808T5rX4YjxHyupuNFl8rjb77GSzHZEyNxfYJn6PzyeeU2WSD9z91bjMIsK5cEt8aRJJfruk8BhhXqC2+Pf2bOwZ9BPiuS1h/pqzkhfBf1kYlvsMcGO80LlJTrT0cwYNJOl7lEa2dVy29Wn/erC/2wgNIukag3aI79nxhN8QEHqHXhTz9kLCTV8G5O5vEI7VU4T6cRHZ1es3sOlvy0sJDe9/IOSfbMwgDH9+hjCf3reTLDMEuCkus4Ds8ldN0t3EypCFCToXAgclmV+hpMzsccIM9EtKHUu2LMyzMyS2/k8ELnP3SSUOq6DMbKS7r7Uwb8TPCHNw3FDisEREakJCHbwr8HuPd4KpVma2JdDm7r0W7uY1NIthgSIiZa0W67iEPDaS0Bvn+DS9TaXCaM6gMhO7D94JXFtODUFm9neE1vz7KrEhKGoE/hQbhdoJQ/Gq3bEWJsYcRujqmc38GSIiMjhzLUzeCRvvulbNPgH80DbeKv6UEscjIpJPtVjH/UfspTQcmKmGoOqinkGSszgHQv+hbM+4ey1UjCIiZcnMFrD5xIwHxiFeFcvCnG39Jyp9z90/U4p4RERqneplkcqmxiARERERERERkRqiCaRFRERERERERGpIyecMMjN1TRIRScHd60odQ6kpT4iIpKY8oTwhIpJOqjxR8sYgAPds7ywnIlL9wl3EBZQnRESSUZ7YSHlCRGRz6fJEWTQGiYhI5TKzg4AfAB1AK3AycB5wAvAu8Ja7T4nLfg6YDvQC57r7EyUJWkRERESkhqkxSEREButV4HB3X2dmXwXOis9f6u5z+haKt2K9DDgEGA3cAjQXO1gRERERkVqnxiARKaienh6WLVtGZ2dnqUMpaw0NDYwZM4b6+sqb19/dlyU87AC6gAbg22Z2LnCtu/8G2BV4wd3XAGvMrMHMhrv7+uJHLSLlQnkiM5WcJ0REBkN5IjPZ5gk1BmWpa+0qeta35r3c+uGNDB3ZlPdyRUpt2bJljB49mtGjR5c6lLK2evVqli1bxg477FDqUHJmZtsAXwcmA73uPsPMmoAHzWw+sDWwImGVlfG5ZZsVlieFqrMHojpdJHPKE5mphjwhhZNLvlOukkqhPJGZbPOEGoOy1LO+ldZF8/NebuO4ZlBlLFWos7NTFXcGRo8ezfLly0sdRs7MbEvgNuAcd3+v73l3X2VmfwL2BF4BtkpYrQl4v5BxFarOHojqdJHMKU9kptLzhBRWLvlOuUoqhfJEZrLNE+pnKiIig2JmQ4Gbgavd/fH4XFPCawcB/wcsAXYzs5Fm9mGgS0PERERERESKT41BIlJ13njjDcyMuXPnbnjutNNO48wzz0y6/IIFC/jud78LwOWXX86aNWsKFtvixYt56qmnClZ+iUwlTAp9rpk9bGYXAf9lZo8DjwFz3P0Fd+8CZgAPALcDF5YqYBGpbcoTIiKSTi3kCQ0TE5GiWtXazpp1HYMqY9SWw2hq3CLtMnvuuSdz587lyCOP5N1336W9vZ1hw4YNWPb06dMHFdtAFi9ezIoVK9h3330Lup1icvfZwOwMl70LuKuwEYlIJVOeqL48ISKSL/nIEaA8AWoMEpEiW7Oug0f/tnRQZRwyYfsBK+9tt92WtrY21q1bx3333cfkyZN57LHHuP7663nooYdYs2YNU6ZM4aSTTtpkvWnTpvGjH/2IYcOGcd5559Hd3c0uu+zCa6+9xsyZM5k2bRpjx45l8eLFNDU1cfXVVwNwxRVX0NLSQldXFxdffDHjx4/nmGOOobm5maeffpq99tqL6dOnM2vWLFpbW3n44Ye59tprGTVq1KDeCxGRaqM8oTwhIpJKPnIEKE9ABsPE4q1/HzOzlWb2pfjc/5jZn81sgZlNS1h2bRwi8LCZTc4pIhGRPDnssMN48MEHmTdvHhMnTgTgpJNOYtasWdxyyy3ceOON9PT0JF331ltvpbm5meuvv56xY8du8trEiROZNWsWnZ2duDuPPPIIPT09zJ49m2uuuYYrr7wSgLa2NqZMmcLNN9/MggULaG1t5ZRTTmHq1KnMnj1bJ/giIiWmPCEiIulUc57IpGdQF/AlIHFw3PfdfYmZbQE8bWa/iXNBvOzuk3KORkQkj4444gjOPPNMdt55ZxoaGgC49957ueOOO6irq+Odd95h1apVSdd95ZVXOPbYYwEYP348999//4bX+irzMWPGsGrVKpYsWcL8+fOZNi20ja9duxaAESNGsNNOOwGw3XbbpdyWiIiUhvKEiIikU815YsDGIHfvBd40s8TnlsQ/O4BuoDc+/oiZPQq8Tri9sO5/KSIl09TURHNz84ZWfICZM2dyzz33ADB58mR6e3uTrvvRj36UlpYW9ttvP5599tmU2+jt7WWXXXbhsMMO48ILw3zIHR1hHHNdXd1myzY0NNDd3T2o/RIRkfxQnhARkXSqOU8M9m5iFwK3untfJB9390OAPwGXD7JsEZFBO/vss9l77703PG5ubmbq1KlceumljB49OuV6J554Io8++ihf/vKXWbhw4YYrAclMmjSJ3t5epk2btmGMcCoTJkzgwQcf5JxzzqG1tTW3nRIRkbxRnhARkXSqNU/UpWrF6s/MZgAt7j4nPv4SMAU40d17+i27JfCQu38ig3J73T3buEumY/lSWhfNz3u5jeOaGbbN9nkvV6TUXn755Q1dG6F4d4kZrN7eXrq7uxk6dCh33HEHL774IhdddFFBt9n/vTIz3L0uzSo1YTB5olB19kBUp4tkTnkic8oTyVXa74lCyCXfKVdJpUis+4p5N7HBKvc8kdPdxMxsEvB14Ji+hiAzGwmsj72EDgWWpC5BRGpVU+MWBa9486Grq4tTTz2Vuro66uvrueqqq0odkohITVCeEBGRVColR0D554mMGoPM7FZgf6DVzA4EjgfWAPfFuYROAHYEfm5mq4F24IyCRCwiUgQNDQ3cdNNNpQ5DRETKlPKEiIikU+55IqPGIHc/sd9T/5ZksXeBCYOOSERERERERERECianYWIiIiIiIiKDYWY/AfYl3NTmm8AC4AZgDNACnNV/blIREcmPwd5NTEREREREJCtmtiuwh7sfTJhy4rvA6cCT7j4R6AEmlzBEEZGqpsYgEREREREptreBNjMbCmxFmHJiInB3fP1u4JASxSYiUvU0TExEiqpr7Sp61rcOqoz64Y0MHdmU8vU33niD448/nt122422tjYmTZrE2WefnXH5ixcv5q9//SunnnrqoOIUEZHsKU/UjDXAq4ADWwJfIPQOWhFfXwlsXZrQRKRc5SNHgPIEqDFIRIqsZ30rrYvmD6qMxnHNkKbyBthnn32YOXMmvb29fOELX+CUU05h9OjRGZW/xx57sMceewwqRhERyY3yRM34LLANsCvwIUJPoBcJvYTeApqA90sWnYiUpXzkCFCeADUGiUiV6+zspLe3l6FDh3LFFVfQ0tJCV1cXF198MePHj+eYY46hubmZp59+mr322ovp06ezYMEC5s6dyyWXXMLvf/97fvGLX7DjjjuyZs0aLrjgAtatW8fPfvYzRo4cyUsvvcQll1zCgQceWOpdLRkzOwj4AdABtAInE/LLbGAU8IC7z4jLfg6YDvQC57r7E6WIWUSkj/JEydQD77t7j5mtBkYCjwJHA8/H/+eWMD4REaB684Qag0SkKi1cuJBp06axdOlSPv3pT/PEE0/Q09PD7Nmzef/99zn33HOZPXs2bW1tTJkyhYsvvpjPf/7ztLZu7Hba3d3Nddddx29/+1uGDBnC5z//+Q2vrVu3jp///Oc899xz/PSnP631k/xXgcPdfZ2ZfRU4i3Bl95fufpuZ3WNmYwlDAS4jzAExGrgFaC5V0CJS25QnSu6PwMlmNg8YTsgPdwA3mNmjwGLgvhLGJyI1rtrzhBqDRKQq9XXr7Onp4ZxzzqGlpYX58+czbdo0ANauXQvAiBEj2GmnnQDYbrvtWLVq1YYy3n//fT70oQ8xYsQIgE26eu6+++7U1dWx3XbbsXLlymLtVlly92UJDzuALkIjzyXxub5JQHuAF9x9DbDGzBrMbLi7ry9qwCIiKE+Umrt3A9OSvHRisWMREUmm2vOEGoNEpKrV19czatQozIz169dz4YUXAtDR0QFAXV3dJsv39vZu+Hvrrbfm7bffZv369QwZMoTnn39+w2v91xMws22ArxNuBXyyu7fFl1YCOxEmAl2RsErf5KCJjUkiIkWlPCEiIulUa55QY5CIFFX98MYwYdsgyxhIX7fOrq4uxowZw6GHHrrhOYDx48dz0UUXpS1jyJAhnHnmmUyZMoXtt9+ebbfdloaGhkHFXq3MbEvgNuAcd3/PzNYl9PrpmwR0BWH4WB9NDioim1GeEBGRVPKRI/rKGUi154m6xFarUjCzXncvaQzZ6Fi+NC+zl/fXOK6ZYdtsn/dyRUrt5Zdf3tBtshJ1dnbS0NBAR0cHxx9/PDfddBOjRo0qyLb6v1dmhruX/aVlMxsK3A5c7+53xOeuBP7s7reb2e+AbxHmDPorMJEwsfQcdx8wmw8mTxSqzh6I6nSRzClPZK5S80ShVdrviULIJd8pV0mlUJ7IXDZ5Qj2DRETSmDNnDvfeey+tra2ceOKJBau4K9xU4qTQZnYucA/wfWCWmV0APOjuiwDMbAbwAOFuYueXJlwRkfxRnhARkXTKNU8M2BhkZg3Aw8A44Ax3n2Nm26JbBotIDZg6dSpTp04tdRhlzd1nE3JCf0clWfYu4K6CByUiUiTKEyIikk655on6DJbpAr4E/HfCc98g3DK4GTjAzMaa2RDCLSGPAE4AfpjvYEVEREREREREZHAGbAxy9153f7Pf082EWwXDxlsG70q8ZbC7LwUazGx4XqMVkYrT0NDA6tWrSx1G2Vu9enXZTCYnIlJMyhOZUZ4QkVqlPJGZbPNErnMGjdQtg0UkE2PGjGHZsmUsX7681KGUtYaGBsaMGVPqMEREik55IjPKEyJSq5QnMpNtnsi1MUi3DBaRjNTX17PDDjuUOgwRESlTyhMiIpKO8kRh5NoYNA84mnAr4aMItwxeAuxmZiMJE0t3xcYiERERERERKSO93d10LF+a1Tr1wxsZOrKpQBGJSDFl1BhkZrcC+wOtZnYgNX7L4PaOLto7u/NaZu+6Dtq616RdZtSWw2hq3CKv2xURERERkdrT09FG20sLs1qncVwzqDFIpCpk1Bjk7icmebpmbxnc3tnNa2+mb7jJ1lYfbKVleXvaZQ6ZsL0ag0RERERERERkUDK5tbyIiIiIiIiIiFQJNQaJiIiIiIiIiNQQNQaJiIiIiIiIiNSQXO8mJiIiIiIikrN4Y5r/ABqAe4HrgdmEOxM/4O4zShediEh1U88gEREREREpKjPbArgUOM7dP+3uVwLfAH7p7s3AAWY2tqRBiohUMfUMEhERERGRYvsk0AbMMbOhwL8CzcAl8fW7gUOA50oTnohIdVNjkIiIDIqZNQAPA+OAM9x9jpnNAE4A3gXecvcpcdnPAdOBXuBcd3+iJEGLiEipjQHGAgcAOwL/A4x097b4+kpgpxLFJiJS9dQYJCIig9UFfAk4s9/zl7r7nL4HZjYEuIxwpXc0cAvhKrCIiNSeFcBj7r4WWGxmTcBaMxvu7uuBJuD9kkYoIlLFNGeQiIgMirv3uvubSV76tpnNM7Op8fGuwAvuvsbdlwINZja8eJGKiEgZWQCYmdWb2YeB9cA84Oj4+lHxsYiIFIB6BomISCFc7e4z4pXeB81sPrA14Upwn5XxuWWlCFBERErH3VeY2Q3AI4S7iV0ILAZmmdkFwIPuvqiEIZZU19pV9KxvzXj5no72AkYjItVIjUEiIpJ37r48/r/KzP4E7Am8AmyVsJiGAIiI1DB3/yXwy35PH1WKWMpNz/pWWhfNz3j5ETvvU8BoRKQa5dQYFG/zeG182EgYbvZj4N+B1wHcfVIe4hMRkQpkZk2xIWgocBDwc+AlYDczGwmMArrivBAiIiIiIlJEOTUGuftzwCQAMzuLjVd6r3P3q/ITmoiIVAozuxXYH2g1swOBreOFgyHATe7+QlxuBvAA4W5i55coXBERERGRmpaPYWJTgVMId4c5w8yOA+a4+3/noWwREakA7n5ihsvdBdxV4HBERERERCSNQd1NzMw+Bgxx95eAO4FxwOHA0Wb2ycGHJyIiIiIiIiIi+TTYnkFTgJsB3H1lfK7bzO4EJgB/GWT5IiIiIiIiIiKSR4PqGQScBNwCYbLQ+H8dYcjYkkGWLSIiIiIiIiIieZZzzyAzGwcsd/e34lMXmtkRhElBH3X3B/IRoIiIiIiIiIiI5E/OjUHuvgj4TMLjS4BL8hFULfpAYwN70pl2mRHt79GxfHVW5dYPb2ToyKbBhCYiIiIiIiIiVSQfdxOTPBjS3c7KZ+anXWb0dqOoG7lFVuU2jmsGNQaJiIiIiIiISDTYOYNERERERERERKSCVG3PoK61q+hZ35r3cns62vNepoiIiIiIiIhIsVRtY1DP+lZaF6UfdpWLETvvk/cyRURERERERESKRcPERERERERERERqSEX3DFrV2s6adR1JXxvR3sGatfkf0jW0q5vu7t68lysiIiIiIiIiUgwV3Ri0Zl0Hj/5tadLX9tymk5Vvrsn7Noft2ElPT0/eyxURERERqTVm1gzMAz4Yn5oNjAIecPcZpYpLkuvt7qZjefLfX8nUD29kaBne2Thdp4J8GbXlMJoas7sTtEgxVXRjkIiIiIiIVLTzgSfj398Afunut5nZPWY21t2fK2FseZHLjW3K9aY1PR1ttL20MOPlG8c1Qxk2BqXrVJAvn9p7u4I2OKmxSQZLjUEiIiIiIlJ0ZvY5YD5wbHyqGbgk/n03cAhQ8Y1BudzYRjetqXxt67t4yt8tWPmHTNhejUEyKGoMEhGRQTGzBuBhYBxwhrvPMbNtSdLVP574Twd6gXPd/YmSBC0iIiVlZvXA14EvsrExaKS7t8W/VwI7lSI2EZFaoMYgEREZrC7gS8CZCc9t1tUfcOAywpXe0cAthKvAIiJSe/4BuMvd15tZ33PrzGy4u68HmoD3SxadpNTe0UV7Z3dGy/au66CtO7t5XDX8SaQ41BgkIiKD4u69wJsJJ/OQvKt/D/CCu68B1phZQ8JJv4iI1Ja9gP3M7DhgPHATYSLpo4HbgaOAb5UuPEmlvbOb1zK8Uc9WH2ylZXl28x9p+JNIceTcGGRma4G+7v3fAx4BbgDGAC3AWe6u226JiNSmZF39twZWJCyzMj63rMixiYhIibn7N/r+NrOHCT2F6oBZZnYB8KC7LypReCIiVW8wPYNedvdJfQ/M7CzgSXe/0syuASYD9w4yPhERqUzJuvqvALZKWEZDAEREhMTfFIQeQVLDunt6eOOd7IaWZau9I7NhbiLVbDCNQR8xs0eB14FzgInAd+JrdwOHosYgEZFalayr/xJgNzMbSZhYuktDxERERCRRoe/CBbCvfbCg5YtUgsE0Bn3c3d8zs9OBy9m0+39f138REakBZnYrsD/QamYHAt8nSVd/M5sBPEC4m9j5JQpXRERERKSm5dwY5O7vxT9vJtxB5hVC9/+3UNf/stHb3U3H8qV5LbN+eCNDRzbltUwRqWzufmKSpzfr6u/udwF3FT4iERERERFJJafGoNjFf727dxOGgy0B/kwYEvB8/H9uvoKU3PV0tNH20sK8ltk4rhnUGCQiIiIiIiJSkXLtGbQ78HMzWw20A2cA7wE3xHmEFgP35SdEERERERERERHJl5wag9z9f4EJSV5KNkxARERERERERETKRH2pAxARERERERERkeJRY5CIiIiIiIiISA1RY5CIiIiIiIiISA3J+dbyIiIiklxvdzcdy5cWfbv1wxsZqrs9ioiIiMgA1BgkIiKSZz0dbbS9tLDo220c1wxqDBIRERGRAWiYmIiIiIiIiIhIDVFjkIiIiIiIiIhIDdEwMREREREREZEK0t3TwxvvrCnoNkZtOYymxi0Kug0pHTUGiYiI9NPe0UV7Z3fGy2/RMIQthimlioiIJPpAYwN70pnVOqOHdBQomurStr6Lp/zdgm7jkAnbqzGoiunMVUREpJ/2zm5eezPzq207bjdKjUEiIlkws4OAHwAdQCtwMuG3yWxgFPCAu88oWYCSF0O621n5zPys1vm7SZMLFI2IJNKZq4iIFISZrQWeiA+/BzwC3ACMAVqAs9y9pzTR5Vdvby+r17ZveDy0q3uTx/2pJ5GICK8Ch7v7OjP7KnAWsBXwS3e/zczuMbOx7v5cacMUEalOOhMVEZFCedndJ/U9MLOzgCfd/UozuwaYDNxbquDyqaurhzffW7fh8bAdO9P2LFJPIhGpde6+LOFhB9AFNAOXxOfuBg4B1BgkIlIAOZ2JpujWeR5wAvAu8Ja7T8lXkBL0v/KciYGuTmdCV7BFJEcfMbNHgdeBc4CJwHfia3cDh1IljUEiIpIbM9sG+DrhAsHJ7t4WX1oJ7FSywEREqlyuv/CTdesEuNTd5+QnNOmv/5XnTAx0dToTuoItIjn6uLu/Z2anA5cDWwMr4msr42MREalRZrYlcBtwTswX68xsuLuvB5qA90sboYhI9crpF36Kbp0NwLfN7FzgWnf/TR7iExGRCuXu78U/bwbOBF4hzAfxFjrJFxGpaWY2lJAfrnb3x+PT84CjgduBo4BvlSi8tLrWrqJnfWvGy/d0DK6XvohIIQyqu0e/bp297j7DzJqAB81svru/no8gpbz0dnfTsXxp3sutH97I0JFNeS9XRIrPzEYC6929mzAcbAnwZ8JJ/vPx/7mli1BEREpsKmFOoNHxYvI9wPeBWWZ2AfCguy8qZYCp9KxvpXVR5nfIGrHzPgWMRkQkNzk3BvXv1tn3vLuvMrM/AXsS5omQKtPT0UbbSwvzXm7juGZQY5BItdgd+LmZrQbagTOA94Ab4jxCi4H7ShifiIiUkLvPJtxGvr+jih2LiEgtynUC6c26dZpZU2wIGgocBPw8f2GKiEglcff/BSYkeenEYsciIiIiIiKbyrVnULJunWZmY4EhwE3u/kKeYhQRERERERERkTzJdQLpVN06RURERERERESkjOl+4SIiUvXaO7po7+zOePnu7t4CRiMiIiIiUlpqDBIRkarX3tnNa2+uyXj57bbdsoDRiIiIiIiUVn2pAxARERERERERkeJRY5CIiIiIiIiISA3RMDEREZEi6+3tZfXa9oyX36JhCFsMU8oWEZH0spkjb2hXd1a5qG8dzasnUh10ZikiIlJkXV09vPneuoyX33G7URk1BvV2d9OxfOlgQstJ/fBGho5sKvp2RURkU9nMkTdsx86s5tPrW6enpyeX0ESkzKgxSEREpEr0dLTR3qVRHAAAIABJREFU9tLCom+3cVwzqDFIRESkqnT39PDGO9k1GGZj6JB6uroL27g4asthNDVuUdBtVCo1BomIiJS5TIeVJXb519AyERERGYy29V085e8WrPx97YMFLR/gkAnbqzEoBZ0lyoD6/wjJZXxxRttZ10Fb98aWZ7XiiogEmQ4rS+zyn+nQMhGRWta1dhU961uzWqenI//nwSIixaazRBlQ/x8huYwvzsRWH2ylZfnG5KpWXBEREREppJ71rbQump/VOiN23qdA0YiIFI8ag6RsfKCxgT3p3PB4RPt7dCxfPfiC64dCT1dGi7Z3dNPRldkdGBg2kra6EYMIbKNijJdNRr2vRERERKScbDG0jj236Rx4wQTr64fz4rsZnsOLCFCAxiAz+yfgNKADON3dX8r3NqQ6DeluZ+UzG6/MjPpwI2vq6gZd7tZ77M/7i5/MaNnu7l6WvpNZV+Gtxh9Cy/KGwYS2QTHGyyaj3ldSCsoTxaHb14tIpVKeqHGdbax85tGsVtlq/CFAfs7LpboUehJsqNwL7Hk96zOzrYEzgE8BE4DvASfmcxtSO7K99XIq2Qxr227bLQe9PRFJTXmieAp1+/pkSnFLe93OXqQ6KU+ISD4VehJsqNwL7Pm+BHgg8LC7dwFPmJnluXwREalsyhNVor2ji/bO0CV/6LrWAXtgZtvzKLH8ZEbtOZG2tfUbH1foVTkR2YzyhGSt/3QTAxk9pKOA0UitKXTvo0Kd4+S7MWhrYEXC4/pUCyaqrDr+xyq3omKtxHJFqloN5AkoXf2geklEKl6N5IlykEvOyHadHLZxRTHiEpF8NwatAMYnPB5wFi93H/ykMCIiUimUJ0REJB3lCRGRIsh3Y9AC4BIzGwLsDSzJc/kiIlLZlCdERCQd5QkRkSLIa2OQu79vZjcC84BO4Cv5LF9ERCqb8oSIiKSjPCEiUhx1vb29pY5BRERERERERESKJKMJ2UREREREREREpDqoMUhEREREREREpIaoMUhEREREREREpIaoMUhEREREREREpIbk+9byRWdm/wScBnQAp7v7SwmvHQD8CKgDLnf3u81sDPB7wIBJ7v5kmcZ5LPAtoB1YCpzi7p1lGOfewHVx+R7gH919WbnFmfDaPwLXufuoQseYS5xm9jHgf4Fn42LnufvCcoszPn8McAEwBPiFu88utzjN7DDgkrjItsAL7v7FMoyzAbgJ+HB8/qvu3lLoOMtRjp/FGcBngDXANHd/z8y2BWYDo4AH3H1GMfcjxpWvfZkBnAC8C7zl7lOKuR8JMecl35rZx4FfAMOA6939f4q7Jxtiztf+3ADsRThmC939vKLuCPk7x6jgY5Nqf26g8o5N0vOqcqjTJLWYxx8GxgFnuPucSj9mKfZpBmWQj3JlZgcBPyB8v1qBkwm/hyv5OCXbp/Oo7OP0IeAOwt0EhwBfBf4PuAEYA7QAZ7l7T6lizFaKfdof+HfgdQB3n1TsuCq6Z5CZbQ2cARwCXAR8r98iPyR8ET4LXGZmQ4D3gSOAOWUe50Kg2d0PAV4FTizTOJ9z94Pjh3cW8PUyjbMvqZ1A/MKVa5zAn919UvxXjIagrOM0sw8STm6PiHEWoyEo6zjd/cG+9xK4D/htOcYJTAJWu/tE4GLg3wodZznK8bM4DjjA3ZuBX7LxvfsG8Mv4/AFmNrYY+9Anz/sCcGn8LJeqISif+fZ7sYyJwBmx7KIqwPnDmfH4lKKxIZ/nGJV6bNKdM1XasUl1XlXSOk0G1AV8CfjvhOcq/Zgl2ycocT4apFeBw939UELj/llU/nFKtk9Q2cfpPUKdfijwbeCbwOnAk/FcuQeYXML4cpFsnyB0Uuj7rVJ0Fd0YBBwIPOzuXe7+BOFqHQBmNhwY6u5L3b0VeAHY1d3Xu/vyCojz1YSeQB2ECrkc40zsrdQEPFOOccaXzyS0KBerFTnXOA8ws3lmdq2ZjSjTOI8mXH24z8zuMrMdyzTOvtfrgL8H7izTOF8iXIkH2IpwJacW5fLeTQT6ev/dHR8DNPd7/pAixJ8on/sC8O1YL0wtTvibyWe+3c3dn3D3bsJV5wOLEH9/+T5/uMbMHjazwwsf+mbyeY5Rqccm3TlTpR2bVOdVpa7TJA1373X3N/s9XdHHLMU+QenzUc7cfZm7r4sP++qKSj9OyfYJKvs4dSf0+tkKeJrNz5Eq7Tgl2ycIF17mm1nRL1hA5TcGbQ2sSHhc3++1lQmPV8bnSiHnOGOX7SMJ3coKLac4zewwM/sr4erVU4UOkhziNLORwGR3L8b7mBhLtu/nm8AusdX7PeDcQgdJbnGOAT5CaBT6MXBlgWPsiyXX7/shhCECawsX3iaxZBvnG0CjmS0Gro3/alEu792Gddy9DWiMr4+MjxOXLaZ87svV7r4P8DngX83sI4UKOo185tu6LJYtlHzuz7+6+yeAfwCuLlIjfqJ8nmNU9LFJsj8VeWxSnFeVuk6T7FXjMSuHfDRoZrYN4fv1C6rkOPXbp4o/TmY21sweB64GHmHT+rQij1OSfbqTMAzzcOBoM/tksWOq9MagFYSWtT7daV5rInTxLoWc4jSzvwN+RRgv3lHoIJPEklGcHobiHEjoAv2fhQ4ySSyZxHk+8JPCh7aJrON093Z3XxOfuxmYUNgQk8aSyfu5AngwXu18ANij4FEO7vs+FfhN4ULbRC5xngoscfc9CN1eryt0kGUq18/iVrDhSntrfH1dfJy4bDHlbV/6eqO4+yrgT8CeBYs6tXzm294sli2UvO2Pu78X/19GmMfgY3mLMjP5PMeo2GOTbH8q9dikOK8qdZ0m2au6Y1Ym+WhQzGxL4DbgnFhHVPxx6r9P1XCc3P05dz+Y0KB1NZvWmRV5nPrvk7uvjD2G2gkNQ8X43beJSm8MWgAcGuda2BdY0vdCbOHtMrPtYq+QXYEXKyVOM2skzG9yvrsvSVpqecS5RcL6K4E2Ci+X4747cJ6Z3Q981Mx+WY5xmtnohPUPTVynnOIktGbvC2Bm+xDGK5djnH1zRR1OmDOoGHKJs57QEww2/4FQS3J57x4FjoqLHQ3Mj3/Pi4+Jr88rQvyJ8rYvZtYU/x8KHESYRLHY8plvXzCzfeP8KIcCfy1k4CnkbX8Sjs9IYCyhp18x5fMcoyKPTar9qdBjk+q8qtR1mmSv6o5ZmeSjnMW4byb8EH88Pl3RxynZPlXBcepfD64jnCP1HaejqbzjtNk+JRynOsIohmL95t+grre3d+ClypiZfRU4hTAz91cI4z5fdvdHzOwThMn56oD/5+53xR+G9xFOCl4HbnH3H5RhnP9OmADs+VhEse7WlG2cXyQMZeqJ6/yzu79SbnH2W7fF3YvSQp7D+/k54DuEHgErgC+7+8rkpZcuzrjOtwmTXtYTJuh8rkzjPAb4grufUej4co0z/pD5DeFOFlsC0939j8WKt5zkeIy/CxxGuGPQKe7+roVJzmcR3tMH3f2SzbdWMfvyc0LOGgLc5O4/Kva+xNjykm/NbBdCN/YG4EZ3n1mC3cnn/txLuEo5FPiBu99SAfuS9Byjgo9Nqv2pxGOT9LyqHOo0Sc/MbiXcHagVuJ8whL6ij1mSfdqaMshHuTKzaYReJn03aLmHMJ9oxR6nFPtkVPZxOogw4X4PoX68gFC/30C48+5i4GteWXcTS7ZPxxFuTNELPOru3yh2XBXfGCQiIiIiIiIiIpmr9GFiIiIiIiIiIiKSBTUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGiYiIiIiIiIjUEDUGlZCZbWVmZwxi/fPMbGg+YyoWM/uymW1bhG18L/79XTM7uJDbS9jucWa2S47rfszM/pKnOB7IUzn/mo9yRKQwzGyMmc3KQzmDyknFUozcZ2YzzOyr8e+fm9nHCrm9hO3mnBvNbJKZ3ZyHGGrq8yQi5U11fsp1VefLoKkxqLS2AgbzpTkPyLhyNLM6MyvqMTezISle+jJQ0MagRO5+ibs/XqTNHQfk1BiUT+7+mTwVVdTGoDSfGRHpx8yGuPsydz8lD8VlnZMK+X1NU3ZWuW+w3P0Md3+lSJv7MkXMjcmU8vM0GKU4xxGpdOV0zqU6vzRU59e2ut7e3lLHULPM7NeEhoMlwK+BHwL/BRwMNAD/7u6/N7P7gavd/R4z+yGwFGiLyy8Gnge+Adzs7p+MZd8QH99vZm8DtwMTgWOAf4zb3SKW+z8p4msB9gNGAu8B+7n738xsMbAX8DHgeqApxnCau681s4eBvwGHAFfGMv4eaAd+BbwIzAZeB95z94lp3qOHgf8FDgO6gK8BVwA7Af8S35OPAzcCW8ZtnO7ui83sy8Du7v7Nfu/HccD3gZXAc8Ar7j4jbmsB8BlCQ+nn3f11M/sQ8DNg+/i+f8XdX0i2PDAGuA9YAawBJrr7mrgvnweOdfevxMdfB7Z19+8m7O/HgDuAl4CxwDzgq+7eY2bHAP8ODAf+Anwd2JFwbBfH9/khd/9aLOstd/9wTK4zgU/F5XYCvhA3mXTdhHguB/4NWAQ8BPwuxjMl4fh8FVgf43bgAOA3wMsxxg7gKHdfaWYHANcRPnvzgbPdvdvMXonrHAWcC/wTMAHoAf6fu9+ESJXK8Xt/J/ACsDdwJLH+j/Xe0cAHCd/1cwl11GeAP7v76XGbycqdTWY5aQbwUWB34HF3v7Df/gwFnnX3PcxsAqEO3xZYBzzh7nvlWhcAHyAh97n7SWne11fifnweeBP4DiF/fAiY6u5PmtlBwA/i+7ACONnd34z7+Ja7/7SvnnP3583sa8D5wDvAW8Dd7n5D3NaN8f1bBXzO3Veb2a7AtcDWwLvAqe7+drLlgcNJkRvN7Bygyd0vi4+/D7S4+6yEZSYB0+PDHQmfiUvja6cT8ucw4HZ3/05c/mJCHb07cIO7Xx4/j32fp5HxPdwVeIyQFz+cat1+73//c5x3iTk54fjsDnwS+BYhx48FvgdsA5xEON/5e3fvMrOj42v1wBx3nxHL6X+OM5OQi+sI5wkPI1KjzOxuYDvCd/9Sd789fscTc8g4ktf1SevHJNtQna86X3V+hVJrWmlNJ1Se+7j7lYTW1Jfd/QBCQ8r3zKwB+GfgKjM7kvAF+qG7XwMsAz6RrmKM/g640933BPYAPhi3sT9whpltl2K9J+MyBxMadw42sx0IlWUX8OMYy3jgFUJl2We9u+8H/BH4EjDW3fcGfubud8Syv9BX8cUumfuniGO5u08gJI+fEL74xwHfjq+/CRzu7vsSEsl/pnojzGwEIalMIjSOWL9FWmPcvyI0chCX/4677w9cQEgiSZd39wXAXcBZ8biuSVh2LvDphK6uJwBzkoQ5HriMUEFuC3wxdiE9D5jk7vsA3cAX4/JjgRmEZH5QTESJjicklD0IiXF8wmtp13X36YT3fx93Tzy+yexBqNzHEq50fCC+N38mVPAQGg/PdPe9CIlyasL6r8R9awU+4u7j4nL3DLBdkWqQ7fd+HKFe6l+HQTjZOgo4lnCS/ZtY7t4WpCo305wE8HHgkP4NQQAxP7xtZtsT6tm/AQcR8smTcbGc6oJkuc/M7jWzMSne10Vx3bXAhXE/Loz/IDR0N8cc81NC43dScX/OIzSeHw3s22+RF2Oee5qNdd41wBmxLvwFoQEu6fLJcmOC37KxER/Cj527koT5SUL9uxdwpJntb2ZjgcnxtQnAfgn5dkJcfm/gq/GHQKKzYpzj4vY+lPDaQOv2/zylMx44mXAx4QrC52484QLM4TF3X0vI/xOAz5pZc1w38Rxnf+DtuO7ehB+lIrXslFj/fAq4zMzq4vOJOSRVXZ9R/ag6X3V+pDq/AlXkfDNV7LPA2HhlF2AUMMbdXzWznxBa8Q9w9+4sy13j7nMTtvH3sYUXQq+ejxMaVPp7jFCpf4DQk+ZYQg+hvuFW+7j77fHvXxG+zH1ui/+vIlTuPzezO0nxw97d03UrvDv+/yzQ6+7rY6+lHePzWwDXmNl4wo+a4WnKMuA5d18GYGa3E3o+9emraP8GfCX+fRjhuPQt0z3A8km5e7uZPQ5MMrNngG3c/bnki/rCGN/NQDOhx9N44C8xjhHAq4Qkstjdl8TlnyH02FqSUN7BwK2x4KfN7PmE1wZaNxvPuftLsayXgD/E51uAj5vZVkC9u/dV1L8m/GD9VXzc95l5CdjRzK4mVPZ/yjEekUqSy/d+UYqyHoj15LOEhvnHY7mLCPXmbmnKTZQ0J8W/73T3zjT705c/PkXIH58i5IPH810XuPvRaeJIzB/veehtlZg/PgD8ysx2IpwTvZamrAMI721fb8/7+r2emA8+bmaj4n7/Lr7PQ4D/S7V8mu3i7kvNbJ2F+eiagJfcfWWSRee5+9IY3+2Ez1E34YdZ3/vdSLjq+ybwmLsvj8u/TDi+icf1YODyGMPdZrYu4bVk6+aaPx539xWxrFVsPG59x6ovd78el7mF8N7OZ9NznGeBH5jZFcBv3f2vOcYjUi3Ot9AzHcI53ofj34k5JF1dn2n9qDpfdX42VOeXCTUGlZc6whCkZHPb7AmsJvW40m427em1RcLfiV/kOkL3z0yG3TxO6GUzgtB75CzCF/H+DNZdB+FqQWyNPpLQAvxFQqtyNjri/z19f8fKvW9s8XmE4UknE7oZ9v9Bk6guzWsQfnz1bauv/F5gX3fvyXD5dOYQeup8nNC4l0xvv797CXH/zt3/OXHB2LWzPeGpZHGk2+eB1u0v3eesI+HvHjY9bpm8N32fmRWxYe9o4BtmNtFj11CRKpbt9z6xXu8vsZ7s/70cMkC5iZLmpHiSm277EPLHEYThtXcSGsvXA98cYL0NZeepLtgsf7BpnfRdQhf6G2KuuipNWdnmj3pgabzincnyA+nLH1uRvFcppP4cXefum/SajReFkuWAxB8GlZA/NnwWPQzh3o8wBOMnZnadu18/wPoiVcnMPg0cCBwYL0i2sPF71/+3QbK6/kYyrx9V56vO7091fgXQMLHSWkNofe/zAPA1i5Nhmdk+8f9DCd05DwWujl3n+q//NrCDmY2MLfCp5uF5APiKmQ2PZVvf30k8RxhaUOfu7YQxrccRhv0ALDSzv49//wOhtXYTZtZIGKJ0F6ErZl8F2X/fB2M0YehaLwM3ND1PuPrx4dgF9rgMyp9HnBDNzOrNbK8Blk+3b/cT5u44idQV++5mNj525T2R8L7+hdBtcocYxzZ9f2fgccJQPWLsu2e4Xp8e2zhB22vAODMbamYfZdMhZ2nFKxrdZrZ3fGoqyT8z2xI+c7cQrk6kSqoi1STf3/t0UpWbUU7K0OOELu6vxfxRTxhKujgPdUG+80dfz9gvD7DsE4T3rTHmtsnpFnb3VcAKMzsCwMwazGyPAbaRbt/mEC6oHEfqiwkTLdwZpiEu+xjwIDAlnhtgZjuY2TYDxNHnccKQZizM37BlhuvB5vvyKqEbP2Z2IJsOPxiIA3vEfRsaY3qs/0IWho60uvsNhOEae/dfRqSGjAbejw1BBxLq4GRS1fXZ1I+q81Xnq86vQGoMKqHY1e4ZM3vGzC4iTID1NvC0he78l8SGn58A/+TuzxO6VF4Wi/gfYJ6Z3eLuHcB/E8ah3hz/T7bNewlz1zwRrxBcR4oW2Ni48gLwVHzqMULXvBXx8TnAv1kYXrRz3H5/o4B7zOxpwuTDfWNnbwBmm9k8GHDOoIH8FDjbzBYyQLJw9zbCuOFH4/68Suhxlc6/AEfFfWghjF9N52bgO2a2MHYZTdz+ekJy2d7dn0mx/jPApYTJ8lYAd7j7O4SeWb+L7/cfCGNmMzEHWGdh4u8ZhAp2oH1O9CvgWTP7obu/CvyJMOb6KsL7kY3TgV9YGL6yivBe9bcD8Eg8nv9FmDxOpNrl+3ufUqpyM8lJWWyjbxL9vpO3p4AXYl6BwdUFG3IfDDh/xECuIlxkeYowpDndPr1ByMdPERr2FzFwXXoycFHMHwuBTwyw/A0k5MYk2+8GXnf391OsvyCW0QL8wd2fdPcWwn4+Eo/3rWw6PDqdawgNlYsIQ8Vfz3C9ZOc484FWM3uO0Gsgm7LaCJ/Z+wjv45/cfbMfk4R5M56In5l/JhwvkVo1F9gmfn/PJ8VvA1LX9dnUj6rzA9X5qvMriu4mJjXHzEZ6uOvZUMIY1e95lc88n7DPuwK/d/dseweJSIFYwp08Sh2LpJdQl44knOge3zdXWjWKeXJI7FkwEbjM3SeVOCwRkaJQna86v9ppziCpRV83s5MJY1fvqfaGoGhuQi+lr5c0EhGRyvUfFubhGA7MrOYfBVEj8Kf4A6GdcNtnEZFaoTpfqpp6BgkWbll/Rb+n73f3TCZ9ExGRGmVmpwHn9nv6l+7+41LEIyIihaM6X6S6qDFIRERERERERKSGaAJpEREREREREZEaUvI5g8xMXZNERFJw97pSx1BqyhMiIqkpTyhPiIikkypP5NwYZGY/AfYl9C76JhtvbTeGcHu7s9y9J8Pgcg1DRKRqmVmpQygbyhMiIptTnthIeUJEZHPp8kROw8Ti7an3cPeDgROA7wKnA0+6+0SgB5icS9kiIiIiIiIiIlI4uc4Z9DbQFm87txXwLjARuDu+fjdwyODDExERERERERGRfMp1mNga4FXAgS2BLxB6B62Ir68Eth50dCJS8Xp6eli2bBmdnZ2lDqWsNTQ0MGbMGOrrNa+/iNQW5YnMKE+ISK1SnshMtnki18agzwLbALsCHyL0BHqR0EvoLaAJeD/HsiUPutauomd9a8HKrx/eyNCRTQUrX6rHsmXLGD16NKNHjy51KGVt9erVLFu2jB122KHUoUgZyKYOV30slU55IjPKE5VloHpcdbdI5pQnMpNtnsi1MageeN/de8xsNTASeBQ4Gng+/j83x7IlD3rWt9K6aH7Bym8c1wxKYJKBzs5OVdwZGD16NMuXLy91GFImsqnDVR9LpVOeyIzyRGUZqB5X3S2SOeWJzGSbJ3LtZ/pHYJSZzQMeBi4Drgc+aWaPAlsA9+VYtoiIiIiIiIiIFEhOPYPcvRuYluSlEwcXjohUu1Wt7axZ1zGoMkZtOYymxi1Svv7GG29w+OGH8+Mf/5gjjzwSgNNOO41hw4Yxc+bMzZZfsGABc+fO5ZJLLuHyyy/nnHPOYdSoUYOKMZXFixfT1tbGvvvuW5DyRUQqnfKE8oSISCr5yBGgPAG5DxMTEcnJmnUdPPq3pYMq45AJ26etvAH23HNP5s6dy5FHHsm7775Le3s7w4YNG7Ds6dOnDyq2gSxevJgVK1boJF9EJAXlCeUJEZFU8pEjQHkC1BgkIlVq2223pa2tjXXr1nHfffcxefJkHnvsMa6//noeeugh1qxZw5QpUzjppJM2WW/atGn86Ec/YtiwYZx33nl0d3ezyy678NprrzFz5kymTZvG2LFjWbx4MU1NTVx99dUAXHHFFbS0tNDV1cXFF1/M+PHjOeaYY2hububpp59mr732Yvr06cyaNYvW1lYefvhhrr322oJdMRARkfSUJ0REJJ1qzxO6N6WIVK3DDjuMBx98kHnz5jFx4kQATjrpJGbNmsUtt9zCjTfeSE9PT9J1b731Vpqbm7n++usZO3bsJq9NnDiRWbNm0dnZibvzyCOP0NPTw+zZs7nmmmu48sorAWhra2PKlCncfPPNLFiwgNbWVk455RSmTp3K7NmzdYIvIlJiyhMiIpJONecJ9QwSkap1xBFHcOaZZ7LzzjvT0NAAwL333ssdd9xBXV0d77zzDqtWrUq67iuvvMKxxx4LwPjx47n//vs3vNZXmY8ZM4ZVq1axZMkS5s+fz7RpYSq1tWvXAjBixAh22mknALbbbruU2xIRkdJQnhARkXSqOU+oMUhEqlZTUxPNzc0bWvEBZs6cyT333APA5MmT6e3tTbruRz/6UVpaWthvv/149tlnU26jt7eXXXbZhcMOO4wLL7wQgI6OMKldXV3dZss2NDTQ3d09qP0SEZH8UJ4QEZF0qjlPqDFIRKra2WefDYQ7AgA0NzczdepUdtttN0aPHp1yvRNPPJHzzjuPhx56iI997GMbrgQkM2nSJJ588skNLfnjx4/noosuSrrshAkT+PWvf01LSwv/+Z//SWNjY667JiIieaA8ISIi6VRrnqhL1YpVLGbW6+4ljaEadSxfSuui+QUrv3FcM8O22b5g5Uv1ePnllzd0bYTi3DI4H3p7e+nu7mbo0KHccccdvPjiiykr5Hzp/16ZGe5el2aVmlCLeSKbOlz1sVQ65YnMKU8kV455YqB6XHW3SOYS675i3Vo+H8o9T6hnUAl1rV1Fz/rWgpTd09FekHJFBqupcYuCV7z50NXVxamnnkpdXR319fVcddVVpQ6p4pjZT4B9CTcr+CawALgBGAO0AGe5e/IZ90SkZilPiIhIKpWSI6D884Qag0qoZ31rwXrvjNh5n4KUK1IrGhoauOmmm0odRsUys12BPdz9YDP7CPBr4BbgSXe/0syuASYD95YyThGRXClPiIhIOuWeJ9QYJCIihfA20GZmQ4GtgHeBicB34ut3A4eixiARkZqlHqQiIqVTX+oARESkKq0BXgUc+ANwJbA1sCLwBTRbAAAgAElEQVS+vjI+FhGRGpTYgxQ4AfgucDqhB+lEoIfQg1RERApAjUEiIlIInwW2AXYlXPW9htAQtFV8vQl4vzShiYhIGUjVg/T/s3fn8XKV9eHHP1lJSEIoqwSlQpUHAsSAgtsF4lLLYlGrRdBfcCn9Yas/UKilSAXc+nNp689dXBBJxQWUVgHFIiIEagrKYiB8xYIiiAIJCbm5+/L745xJJjdz586dO9ud+bxfr7wy98yZc75z5pznOed7nvM8V+fvXw0c06TYJKntmQyS1HYefvhhnv/857Ny5Upe97rX8elPf3pSn1+3bh1f/epX6xRdx5gJbMib9z8FLABuAk7I3z8BuLlJsUnqcNYTLcEWpJJaVifUE/YZJKmhajGK3sx5C5m9YHHZeZYvX87FF1/M6Ogor3nNazjttNPYZZddKlr+wQcfzMEHHzylGMV/Am9MKd0MzAM+AFwFXJpSuglYB3y/ifFJalHWEx2juAXp3mQtgX5F1kro99iCVFIJtRqR23rCZJCqNDo8zMD6R+q2/EoOTk1PtRhFb+EhXVDh/jE4OMjo6CizZ8/mIx/5CGvXrmVoaIjzzjuPZcuWceKJJ9LV1cVdd93FYYcdxvnnn8+aNWu47rrruOCCC/je977Hl7/8Zfbbbz82b97M2WefTU9PD1/4whdYsGABDzzwABdccAFHHXXUlL5Tu4mIYWBlibdObnQskqYX64mOsbUFaUppbAvS+/L/r2tifJJaUK1G5LaeMBmkKo0M9NL7wJ11W/5kDk6plDvvvJOVK1fyyCOP8JKXvITbbruNkZERVq1axYYNGzjrrLNYtWoVvb29nHLKKZx33nmcdNJJdHdvu9MwPDzM5z73Ob797W8za9YsTjrppK3v9fT08KUvfYl7772Xz3/+857kS9I0Yz3RdLYgldTS2r2eMBkkqS0VmnWOjIxw5plnsnbtWlavXs3KlVljlS1btgAwf/589t9/fwD22WcfNm3atHUZGzZsYO+992b+/PkA2zX1POigg5gxYwb77LMPGzdubNTXkiTViPVEc9mCVFKra/d6wmSQpLY2c+ZMFi1aREqJvr4+zjnnHAAGBgYAmDFjxnbzj46Obn2922678Yc//IG+vj5mzZrFfffdt/W9sZ+TJE1P1hOSpHLatZ4wGSSpLRWadQ4NDbFkyRKOPfbYrdMAli1bxrvf/e6yy5g1axZnnHEGp5xyCvvuuy977LEHc+bMaUT4kqQ6s56QJJXT7vXEjOKs1WSklI4CPgjMAa4FvgKsAhYB10fERRUuZzQiqophuhtY/0hNOr8qZf4By+vap0+9l7/wkC7m7r5v3ZavxnnwwQe3NpuExo0SUyuDg4PMmTOHgYEBXvva13L55ZezaNGiuqxr7LZKKRERHX9ruRPricnUD5aXmu6sJypnPVFaK9YTE5Xjlt1S5YrLvkaOJlYrrVpPVNUyKKW0E3Ah8OqI6MmnfQy4JCKuSCldk1JaGhH3VrN8Se1r9oLF06pz8CuvvJJrr72W7u5uTj755LoV3JKkjPWEJGk8062OgNatJ6p9TOwFQC9wZUppNvB3QBdwQf7+1cAxgMkgSdPaqaeeyqmnntrsMCRJLcp6QpJUTqvWE9Umg5YAS4Ejgf2ALwILIqI3f38jsP84n5UkSZIkSVKTzKzyc08Ct0TElohYBywGelJK8/L3FwMbahGgpOltdHSUwcHBZofR8gYHB6m2DzdJms6sJypjPSGpU1lPVGay9US1LYPWAO9JKc0E9gL6gJuBE4DvAMcD76ly2ZLayF577cXvfvc7RkZGmh1KS5s5cyZ77bVXs8OQpIaznqiM9YSkTmU9UZnJ1hNVJYMi4smU0qXAT8hGEzsHWAdcllI6G7ghIu6pZtmS2svChQtZuHBhs8OQJLUo6wlJUjnWE/VRbcsgIuIS4JIxk4+fWjiSJEmSJEmqp2r7DJIkSZIkSdI0VHXLIEmSJEnqVENbNjHS113yvZGB/rKfHR0eZmD9I+O+P3PeQmYvWDyl+CSpHJNBkiRJkjRJI33ddN+zuuR78w9YXv6zA730PnDnuO8vPKQLTAZJqiOTQZKkukgpHQV8kGyggWuBrwCrgEXA9RFxUfOiax8T3V0u5p1mSZIkgckgSVIdpJR2Ai4EXh0RPfm0jwGXRMQVKaVrUkpLI+LepgbaBia6u1zMO82SJEkCk0GSpPp4AdALXJlSmg38HdAFXJC/fzVwDGAySJI6lC1IJal5HE1MklQPS4ClwF8CZwGfBRZERG/+/kZgtybFJklqsjEtSF8SER8DziVrQdoFHJlSWtrUICWpjdkyqIxyIwTUwkSjDEjSNPYkcEtEbAHWpZQWA1tSSvMiog9YDGxoaoSSpGayBakkNZHJoDLKjRBQCxONMqD62NTdz+aegbqvZ9HOc1m8cKe6r0dqUWuA96SUZgJ7AX3AzcAJwHeA44H3NC88SVKTFVqQHgnsB3yRHVuQ7t+k2CSp7ZkMUsfZ3DPATXdUNvLOVBxz+L4mg9SxIuLJlNKlwE/I+oI4B1gHXJZSOhu4ISLuaWKIkqTmsgWpJDWRySBJUl1ExCXAJWMmH9+MWCRJLccWpJLURCaDJEmSJDWULUhrr1FdIRTYJYI0vZkMUln9A0P0Dw7vMH320DBPbaltB9g7zZnFTnPdJSVJkjqBLUhrq1FdIRTYJYI0vXnlrbL6B4d56NHNO0yfu99gyelTsd8+i0wGSZIkSZJUZ155S5LUIUaHhxlYX/ld45nzFjJ7weI6RiRJkqRmMBkkSVKHGBnopfeBOyuef+EhXWAySJIkqe2YDFLLGB0d3doP0WjPAL3DtX0MraB/YMc+kCRJkiRJ6hQmg9QyhoZGePSJHgB23bObtetr20F1wRFpz7osV5IkSeoUwyMjPPxYfW7ejuXIZVLtmQySJEmSJE1Kb98QP4/HG7IuRy6Tam9KyaCUUhdwM1BoarEKWARcHxEXTS00SZLUbjZ197O5Z6Cu6/AOsiRJUnlTbRn0LuD2/PW5wCURcUVK6ZqU0tKIuHeKy5ckadob2rKJkb7uiucfGajPY7KtYHPPADfdUfmIZtXwDrIkSVJ5VSeDUkqvBFYDr8ondQEX5K+vBo4BTAZJkjreSF833fesrnj++Qcsr2M0kiRJ6nRVJYNSSjOBvwX+gm3JoAUR0Zu/3gjsP/Xw1Kn+aOEcDmWwLsveZVZ9H0+QJEmSJKmVVdsy6A3AdyOiL6VUmNaTUpoXEX3AYmBDLQJUZ5o13M/Guyu/iz4Ze604ri7LlSRJkiRpOphZ5ecOA16XUvoBsAy4nKwj6RPy94/P/5YkSZIkSVILqaplUEScW3idUrqRrKXQDOCylNLZwA0RcU9NIpQkSZIkSVLNTHU0MSJiRdGfx091eZIkSZLUbBONBNnOIz9Kan9TTgZJkjSelFIX2WPDe+aTVgGLgOsj4qJmxaXS+geG6B8c3vr3aM8AvcOba7yO4YlnkqQWMNFIkI78KGk6MxkkSaqndwG356/PBS6JiCtSSteklJZGxL1NjE1j9A8O89Cj25I/u+7Zzdr1tb3zfUTac+KZJHUMbxpIUnNU24G0JEllpZReCawGtuSTuoCr89dXA8c0Iy5JUkspddOgCzgypbS0eWFJUnuzZZAkqeZSSjOBvwX+AnhVPnlBRPTmrzcC+zcjtlqZqC+JYvYr0V42dfezuWegrutYtPNcFi/cqa7rkJqt6KZBoZ7oAi7IXxduGtiCVJLqwGSQJKke3gB8NyL6UkqFaT0ppXkR0QcsBjY0LboamKgviWL2K9FeNvcMcNMdj9R1Hcccvq/JILW1TrhpIEmtzMfEJEn1cBjwupTSD4BlwOVkfUKckL9/fP63JKkzbb1pUDStJ6U0L3897W8aTMXo8DAD6x8p+W9oy6ZmhyepDdgySJJUcxFxbuF1SulGspP+GcBlKaWzgRsi4p4mhac2NzwywsOP1XYUtGKOiCbVxGHAc1NKr2bHmwbfIbtp8J7mhddcIwO99D5wZ8n3Fh7SBQsWNzgiSe3GZJBUJ/W+GCmwXwm1uohYUfTn8c2KQ52jt2+In8fjdVu+I6JJU9cpNw36B4boH9wxgTx7aJintozfn1y590d7Bugd3vEc00S1pMkwGSTVSb0vRgrsV0KSJE1n7XzToH9wmIce3TFxM3e/wZLTK3l/1z27Wbt+x0RROyeqG3WTtcCbreoEJoMkSZIkSS2rUTdZC7zZqk5gB9KSJEmSJEkdxGSQJEmSJElSBzEZJEmSJEmS1EFMBkmSJEmSJHUQk0GSJEmSJEkdxGSQJEmSJElSBzEZJEmSJEmS1EFMBkmSJEmSJHWQ2c0OQGq0nWbP4NDdB+u2/L6Z8/jV48N1W74kSZIkSVNRVTIopfRC4F+BAaAbeGO+rFXAIuD6iLioRjFKtTXYy8a7b6rb4ndddgwwp27LlyRJkiRpKqp9TOw3wMsi4ljge8DbgXOBSyKiCzgypbS0RjFKkiRJkiSpRqpqGRQRvyv6cwAYArqAC/JpVwPHAPdOKTpJkiRJ0lZ/tHAOh7Jjlwe7jmzk0N0H7bJAUkWm1GdQSml34G+B44A3RkRv/tZGYP8pxiZJkppovAuOUrz4kKTGmDXcz8a7V+8wff4eO7PxiR67LJBUkaqTQSmlnYErgDMj4omUUk9KaV5E9AGLgQ21ClKSNL3Yt1x7GO+CoxQvPiRNhvWEJDVXVX0GpZRmA98APhURt+aTbwZOyF8fn/8tSepM9i0nSSrHekKSmqjalkGnkvUJtEtK6SzgGuCjwGUppbOBGyLinhrFKEmaZuxbTpJUjvWEJDVXtR1IryJrwjnW8VMLR5LUTuxbTpJUjvWEJDXHlDqQliRpPPYtJ1VveGSEhx/bXNd1LNp5LosX7lTXdUjlWE9IUvOYDJIk1dwEfct9h6wl6XuaFF5JQ1s2MdLXXfH8IwP9dYxGna63b4ifx+N1Xccxh+9rMkhNMx3rCUlqJyaDpqn+gSH6B8cfwnf20DBPbZn6hcrw8OiUlyGpI027vuVG+rrpvqeykbMA5h+wvI7RSFLbm3b1hCS1E5NB01T/4DAPPTp+8/G5+w2Wfb9S++yx85SXIanz2LecJKkc6wlJaq6qhpaXJEmSJEnS9GTLIEmSJEkdqVx/cfYNJ6mdmQySJEmS1JHK9Rdn33CdqxEjOhY4sqOaxWSQJEmasj9aOIdDGZxwvl1HNnLo7oP0zZzHrx4ffyAESZKapREjOhY4sqOaxWSQJEmaslnD/Wy8e+LR2ObvsTMbn+hh12XHAHPqH5gkSZJ2YDJImuYa1YzVJqySJEmS1B5MBkk1VnhUovAoRK2NfbSiUc1YbcIqSZIkSe3BZFCN9Q8M0T9YWR8Is4eGeWpLdaMUDA+PVvU51V/hUYnCoxC1tv+LXsa8oiRTLZNO9uEhSZIkSe3PZFCN9Q8O89CjlT2yM3e/wYrnHWufPXau6nOa/sb2y1HLpJN9eEitrdQNh8ncWJhoXm80SJIkdQaTQZIkTROlbjhM5sbCRPN6o0GSpMZyGHs1y7ROBg1t2cRIX3fdlj8yUN0jXJIkSZ1uU3c/m3sG6roOL2wkTXcOY69mmdbJoJG+brrvmXgY22rNP2B53ZYtSZLUTPW+G90/MMyae35ft+WDFzaSJFVrWieDJEkqZzItSG0Nqk5T77vRR6Q967ZsSeMrjGxbioOFSCowGSRpq3InD/P7n2Bg/VNTWv7MeQuZvWDxlJYhTcZkWpDaGlSS1A7GDjZSzMFCJBWYDJK0VbmTh0VPW8jmGTOmtPxFhx5N75aZ5eex/wdJkiRp2mtE33EFXkNMXs2TQSmlvwbeAgwAb42IB2q9DkmNNzQ0wqNTHMJ+1z27Wbu+/KM4L37OPg2pNGbPmsnQ8Ejd1wNWTmNZT0iSyrGekOqjkSOXQWP6jiuwD7nJq2kyKKW0G3A68GLgcODDwMm1XIek9taoERWOSHs6ckMTWE+ooNxjqWPZx4XG04gLGxP6jWU9UV8Tlb2Wt+2tkSOXgX3Htbpatww6CrgxIoaA21JKqcbLlyRNb9YTAso/ljqWfVxoPI24sKl3i1WTTTuwnqijicre/V/0MubtbufT0kTa4RG4WieDdgOeLPq7fOcguc4t4z85TZc9zZf/kWkce92XP51j1zRhPVFzkzmuJnsMTtdlS5rGOqSemKhcK/d+sz4rqZ3UOhn0JLCs6O8JU8cRMbUeaSVJ04n1hCSpHOsJSWqAWieD1gAXpJRmAc8B7q/x8iVJ05v1hCSpHOsJSWqAmiaDImJDSumrwM3AIPBXtVy+JGl6s56QJJVjPSFJjTFjdHS02TFIkiRJkiSpQSrqkE2SJEmSJEntwWSQJEmSJElSBzEZJEmSJEmS1EFMBkmSJEmSJHWQWg8tP22llOYANwKHAKdHxJUppYuAvwQeB34fEac0Mb4XAv8KDADdwBvJfr9VwCLg+oi4qIVieyets+32Bq4iG5FiFvA24H+AS4ElwFrg7REx0kLxPQ94L/BbgIhY0YzYiqWUushG9tgzn9T0fa9gTGzvoEX2vTy2LcBt+Z8fBn5Ci+x77ajVtndKaSnw2fzPhWQ3YT5Jg4/vceq4PShxHKeUXgmcD4wCZ0XEbSUXWr+4vggcSratPh0Rq/J5t/ttI+IHDYzpIkqUK43aVuPE9FHgqHyWQ4G3RsR3U0r/Q75vAZdGxKV1iqni85IGbqdSMX2M5u5PFZ8jNfLYazcppb8G3kK2nd8aEQ80cN0V1TsppSOBTwAzgA9FxNU1jGFKZXxKaSbwGbJj5XfAmyOitw4xXUSF+36tY5pqmVWnbTSl8qFOMVV83TTePp3/zi8HNgMrI+KJGsdT8jqpEfGMiW3Ca6NG/W7VsmXQNkPA64D/N2b6hRGxotkXlMBvgJdFxLHA94C3A+cCl0REF3BkftHRKrFB62y7J4CuPL5/BP4BeCtwe0QcDYwAx7VYfACfy7ffiqZFtr13Abfnr1tl3ysojg1aZ98DeLDwO+YXG62077WjltreEXFv0XH8FeDb+VuNPr5L1XE7HMcppVnAB4BXkJ2MfrwJcX00Il4IHAOcn1Iq3Lga+9s2MiYYU640eFvtEFNE/H2+/7wM2AD8MH+rt2g7XVrHmCo6L2nwdioVU7P3p4rOkZpw7LWNlNJuwOlkv/G7yRIyjVRpvfNxst/2T4EP5L95rUy1jD8eGMljvp0ssVaPmKDyfb/WMU21zKrHNppq+VCPmCZz3bTDPp1SOgQ4Mt+mlwB/X4d4oPR5VCPiKVb22qjBv1tVTAblImI0Ih4t8dY/ppRuTimd2vCgikTE7yKiJ/9zgKyA7QIKdxWuJqsEWyU2aJ1tN1zUEmBX4C7gaFpg28G48QGcnlJanVJ6Z5NC2yrPaq8GtuSTWmLfg5KxQYvse7lnpJRuSil9LaW0Oy2077WpVt7epwJfz1839Pgep44rdRw/G/hlRGyOiEeAOSmleY2MKyLuz18OAMNkd9Ngx9+2YTHlxpYrDdtWZWICeCnwXxHRl/89N6V0Y0rpP1JK+9cjnjymSs9LGrmddoipBfanSs+RGnrstZmjgBsjYiiy1lSpweufsN7Jf8vZEfFIRHQDvyT7zWuiBmV8zevKGpSlNY2pBmVWPbbRVMuHesRU0XVTmX167LxH1yEeGHMe1ah4Ciq8NmrY71Ytk0HlfSoilgOvBP4upfSMZgeUVzJ/C3wZWBDbmpRtBHZrWmDsEFtLbbs8O3sr8Cmy5ru7AU/mb7fCthsb37+TNat9GXBCSukFTYxtJtnv+rmiyS2x740TW0vte8CfRMQxwI+AD9Fi+14basntnVJ6JjArskcXWuX4LnUcF2+v4unNcA7wrYgYzv8e+9s2UqlypVW21RvYlmQEeFF+p/STbHtEsW4qOC9p+HYaE1NBU/enCs6RWmV/mo7GbrtGX99UUu/slr9mzPR6mszx2Ki6cjL7fl1imkKZVbdtNIXyoV7bqJLrpvH26a3z5tt2YR3iKXUe1ch4Kr02avbxNiGTQWVExPr8/01kBfyhzYwnpbQzcAVwZmTPOvYU3TVaTNZMvCVia7VtF9mjGi8iK1g/RXYA7pq/3dRtBzvGFxEb80x4P1mBd3gTw3sD8N2iu87QOvveDrG14L5XeC75G2S/Y0vte+2mhbf3KXlMtNDxXeo4Lt5exdMbKqX0OuAFZM2rgZK/bcOMU640fVullHYiuxt5fWFaYTtFxI+AuibDKzwvaeh2KhFT0/enCs+Rmr4/TWNjt93weDPWQ4X1TjN+38kcjw2pKye579c8pimWWXXZRlMsH+oSU4XXTRPGlG/b7lrHM855VMPiofJro6Yeb5UwGVRGSmlx/v9s4IVknWc1K5bZZJXMpyLi1nzyzcAJ+evj879bIrYW23Y7Ff25EegBbmLbtjuBJm07KB1f0fabQdZ08P5Sn22Qw4DXpZR+ACwDLqdF9r1SsbXYvrcgbesT4Fiy37Fl9r120+Lb+/XAN2G78rHZx3ep4/h+4MB8Wz6N7DGbvvEWUA8ppRVkd9xWFpqGj/PbNjKmUuVK07cVcCLww4gYyuPbqXAymlI6FFhfrxVP4rykYdtpnPORFTRxf5rEOVIr7E/T1Rrg2JT1D3IEDSwfKq138hYDQymlfVJKC8geH/lVncObzPHYkLpykvt+TWOqQZlV821Ug/KhHjFVdN1UZp++iWxbFuZdXet4Sp1HNSqeXKXXRk093ioxY3R0dOK5OkRK6VtkvZN3Az8ga7K1lKzn8ssj4hNNjG0lWWb2znzSNWS9ul9G1mv5DRFxQQvFlmidbfdCss4ER8h6lz8buI9s+z0NWAf8TTRvNLFS8b2arLOxUeCmiDi3GbGNlVK6kawzwBm0wL5XrCi2D9M6+95zgS8BTwH9ZB1cPkGL7HvtplW3d8o6L/xERLw8//v9NOH4LlHHfYwSx3FK6STgvDy+d0XEmgbH9VqyET8Kzb3/EtiPMb9tRDzUwJhKng80cluNjSki/j6ldAXZvrU6n2cJWR28Of/YOyLi7jrFU/F5SaO20zgxvY0m7k+TOUdq9LHXTlJKbwNOIxtx6K8iot6JlsJ6K653UkrPJ+tAdgbwfyPiuzWOpeoyPn/s5XNk++WjZKMb9ey4linHVHFZWuuYplpm1WMbTbV8qFNMFV83jbdP5+c5LyUre0+LiMdrHE/J66RGxFMivhspc23UyOOtGiaDJEmSJEmSOoiPiUmSJEmSJHUQk0GSJEmSJEkdxGSQJEmSJElSBzEZJEmSJEmS1EFMBkmSJEmSJHUQk0GSJEmSJEkdxGSQJEmSJElSBzEZJEmSJEmS1EFMBkmSJEmSJHUQk0GSJEmSJEkdxGSQJEmSJElSBzEZJEmSJEmS1EFMBkmSJEmSJHUQk0GSJEmSJEkdxGSQJEmSJElSBzEZJEmSJEmS1EFMBkmSJEmSJHUQk0GSJEmSJEkdxGSQJEmSJElSBzEZJEmSJEmS1EFMBkmSJEmSJHUQk0GSJEmSJEkdxGSQJEmSJElSBzEZJEmSJEmS1EFMBkmSJEmSJHUQk0GSJEmSJEkdxGSQJEmSJElSBzEZNE2llP6u6PWKlNI3mhlPPeXf73l1XsczU0o/zV+flFI6s57rK1rv8pTSy6fw+V+nlObVII4vpZSeWYPlvDmltMdUlyOp/vJy73XNjqPWUkrvTCnNrvM6LkopvS1/XZPys8L1Vl3G1upcIaW0JKV0WQ2Ws2tK6fSpLkdS46WUrk0pzW3V47j4OqmO67g0pXRc/vralNLceq8zX1fVdVxeh3y4BjE8r0bLacvzkOnEZND0VfdCrtFSSrPGeWsFUNdkULGI+G5EfLJBq1sOVJ0MqpWIOD0ifl2DRb0ZaFgyqMw+I2lizwSm7UlYmeP/nUBdk0HFalh+VuLNNLCMLSUifhcRp9VgUbsCDbuITCnNSCl53ivVQEScEBEDNPg4LjbBMd3Q66Si7dEIDa3jSomI2yPiH2qwqGfSwPMQrxt2NGN0dLTZMbSNlNIS4ApgAVmi7WSgD7gKCOBI4OvAg8DfAgPA8RGxMaV0JPA5YCdgNfCOiBhOKZ0AfDhf3pURcVFK6UPA3wP3AD8G/gM4L1/eQcClEfGh/E7ld4B1wHOBH0fE3+Sxngi8F5gH/DSP52kl4n9q7LSIuK/Ed/8icHlE/DildFcew8dTSl8FLgbuBL4MHJov8y0R8cuU0kXAH+dx3wrcBlwADAK/BM7Opw0C64HXR0SMs/0vAvbLl7U38FbgDLJE0tcj4sJ8vquBfYC5wIUR8Z18W30jIl6QUnozcFBE/ENK6UDgcmBOvq2XR8SKfF37Agfny/qbiPhhnqn/F+BF+WfeGxHfKzU/8COyfWEn4FHgXRHx4zzG3YBbI+Kg/O+lwKcj4qVjvvOvgW8CJwKP5dvn8ZTSs4HPArsBjwNviog/5PN/FXg1sAl4ZUQ8lVK6EXhbRNyXUvob4F358n4PXB0Rl4732aJYXgOsAn4LPBERR6eUfh8RTyv6fX4fEZ/Pl/U14KT8u78P+Ej+u50aEbenlPYELgOeDvwOOC3/DpcCPWTH0zeAYeBtQD/ZPv5OpBaTUvpr4ExgBPj3iLiwVPmez/t74NtkieI1wLfIyuudgT+PiF8XHQcvJCvL3hoRa1JKLwT+laxsfxJ4Y0Q8mlLaBfg8WRk8Qnby/glgKfAb4GPAs5lcuXYKReV1RPxlqWkltsVs4BcRcXBK6XDgZ2QJjh7gtog4rEyd+GuyevR44Czgr4HD8+/0f4E/Aj5OVu/dFxGvL/Ob/JqJy6HxtudFbCvPbqTK8nMyZTXwMsaUsUXf5UxgcUR8IP/7o8DaiLisaJ4VwPn5n/uR1XmFevGtZPXSXOA7EfG+fP7xzi0K9eWCfBs+G7gFOCkinjbeZ8ds/6/l3+/+fBmPk+I3Jo4AACAASURBVNe9Rb/PQcALgPcAQ2T764eB3YHXA4+QHRNDZY6nP5CdCx1NVldeDCwBZgD/JyJuRJomqqhLvklWl/wK+Iu8HH0R8P/IytffRcTxKaVXkx2zO5HVCW8kKw9KnosWHZ9fZvvj+C+A/xUR/5OX9fcCyyKib8z32Bf4Wn5O/RrgErJy8JnAxRHximqPaeBPKbpOioh3ldmeldS3O2ybiOjO6+FvRMQPCtsjIvpSSh8E/hJ4KF/NhyLixjK/x1FkdewC4H/I6oGeUvOTne+WrONSSv8K3BERq/K/v5X/XjcVzfPmfJvtA+wF/EtEXJy/d37+W+4EfCoivpjPfxywZ/7bvD8ivpqX8W+LiFNSSnvnce4BXEN2LfLM8T47Zvvfwo7nIb+PiM8Xfp+8TnkzcEK+rP3J6v+X5//+KyLems//FuCcfPGfjIgv5HXWv5NdVz4HeAkVXNt2Cu+Q1NapwPURsZws+fKbfPrBZCcyS8nu6v1RRDwX+C+ykxmArwBnRMRhZIXhqSml+WQniSeSnez+aUqpKyLOB9ZHxPKiAu7wfNnPAd6Wn6CRr/Mi4BDghSmlZ6esifk7gRV5rMNkBUyp+Mf7TmPdArwov+AYILtoIP/Mz4B3AI/n3+9D+fcq+BPgmIg4h+xE9cSIeA5wekT8luwC5kP5942U0ttS3jy/hH2ALrIk0lXAPwKHAW9JKS3M5zkt3/4vBj6QUpoxzrIgqywvyuMZGvPeHwPHkiXNLsynnQ48GBFHAscAH04pzSk1f0QMk100fSX/bj8uLDgiNgC/SSktzyf9JXDlODE+EhGHkiUFC3F8hmz7PZeson5v0fy/yr/PXWzb/4CtlfM7yX63E4Ajxqxr3M9GxFXA7cBrii9Syrgn3x+2kBXcx+T/Fwrx9wH/mc/zH2T7TcGuwFER8S9k+8zheVwXIrWYlNIy4O3Ai/P99BPjle/5R/YmO6k+iKzsfnlEPB/4IlmCm6L5jgD+V/4eZCe/XRFxOFnZ+ff59AvIkjPLyBLkvyQ7dq7Ly5+v5fNNplzbrrzO5y01bTsRMQT8IS9vXgzcQZbUeh5ZGQIl6sSiRfw6r5O6gWdExCH5fNdExGfIksfPL5wkp6z5/pJSsTBxOTTe9txBleVnxWX1BGXst4HXFP19EvDdEmG+gOxc4TDgz1LW1H8p2Qn7C8j2xeembY9mj3duUfD2PM5D8vXtXfTeRJ89nywpuDwiPlYi1mLLyC5OjyRL2D2Y78u9wMsmOJ72IrtoPpRsH/tD/tnnkJ2fSNNClXXJt/Pjcybw0pTSTmQ32t6UL+ON+bw/iYjn52XrLcBfVXguOvY4/ipZnQRZ0v4nYxNBABHxCLBXXpe8GHiA7JrlxcAtUzmmS10npZTuHGezVlLf7rBtxlkWeWLnpWQ3Xt4EPH/Musb+HnPJkiAnRcQRwH+T3aAvOX+pOq7IFcBr8zjmk9VBq0uEeSRZHfU84JyUPfp7HLBnXs8/Dzg9pbRPPv8hwJ+T/TbvK7G8C8mSdYeS3awoNtFnS52HjOcgsn3qVWQ3hb5Ots88J2Weni+vi+yc4uy07RHuQ4D3RUSi8mvbjmAyqLZuA1amlC4Enh0Rvfn0eyPigYjoJyvsfphPXwvsl1LaFZgZEYWTkq+RHTQp/+xvI2KQLOv64nHWfUtErI+IHrLWJoUT33URcX+eeLibLDP7QrITq5/mhePLgQPGiX+877TD+vPYXkiWfd07zxRvzL/3i4B/A4iIa8kSZAX/nn+/wnK+mN+lLNlsLSI+X8gYl/D9iBgBfkF2wVDY7r8mSxQBvCtlrZduzrfH08ZZFsBzIuLq/PU3x7x3Tb5d78iXA9ndiDPy7XoTsIhtv0Wp+cu5gm1NJ19DdheklEIfEF8HulJKi8h+i//I47iArHVNQeECoVQcR5IVkJsja/Xz/THvl/vsZBW26y+AG/PfbS3ZHWso2mfy/4v3/SsjorB//Az4t7xFQv8UY5LqYQXZ3cOnYGuyt1z5/mRE3Jrv4/cAN+TTi48PgG9GxGhE3AHMzuuSPwKuSin9guzEa2k+70vJkhlExFAUteobYzLlWqnyesIyvGi+F+f/Plr0+tYydWLBFfn/D5DVoZ9KKb0sIjaVWlFkzfd/N04cE5VD423PUiZVfk6xrN5OfmHVk1J6VkrpucADEbGxxKw3R8QjkT3O8B2yk+aXkdXdPwN+nn/HZ+fzj3duUfAi8roxryt7it6b6LOTcWtEPBkRj5O1lCr8boXfqtzxtDkirstf/wJYkVL6CPC8iNg8hZikRlvB5OqSjbGtVUihHElkCdx1RcuArCy9Pi/rzmBbWVfpuWjBN4rmP40s8TSeO8gSPYcDn6aoHpjge036mM4v/EuppL4db9uU8iLgqogYzOudm4veG+/3WAb8OK8H3kJ2U2a8+cv5KbA0v/l9AvDDvE4b69qIeCrfj64HjiKr5/88j+G/yRJuf5LPf31E9ETE74GZRTe5i79z4Rpp7LXSRJ+djOvzxOIvgL4xv9t+ZEmsH0bExnw/uJZtybh1EXFP/rrSa9uOYDKohvID9hiy5uHfTim9LH+r+BnSkaK/R4BaPbtYfBFcvNxS02cA/5FnYZdHRIqIj5aKv8x32k5E3A88i6zJ5mqyk/SVZAX6RIpPHv+G7IT4QGBNmnwHacXbdux2n5VSeglZoXdUfkek8JjWeMq1GuoHyAvawvaeQXY3pbBt94uI35SZv5x/B16VUkrAprwgLWW06P9RsuP6kaIYDouIV4+Nm9L7X7nvO9Fny8UGO27nUr9Vpcst3mdOJHuc5FjgBxV8Vmp1ldYZo2NejwLvJ3vM5zCyMrhc+VbKZMq1UuV1pWX4rWQn9vuSlXXPy/++pYIYe/IYnyQ7iV4NnJuyR7cma6JyaDLbc7Ll51TK6lKuJLsr/DrGb0laap+ZAXyuKI5nRcTXx8QwXhwT1pFlPjvWMNuflxZv66mcR22tLyLil2R3gtcBn84fKZDa1WSOwU8CH8zLunPZdvxVei4KQJ6EXps/4nUopVumFNxKdkN6iKwrhheTtVD8abl1UNtjupKyZbxtU8pky8QZZI9HF8rfpRHxf8rMP648MXIt2XlxNfXAe4vi2D8iCr9dU+qBvBVbsQHYen6ywzXeBMst3mcqurbtFCaDaiil9MfAoxHxObLM+GGVfC4vOIdTSs/JJ51KVngGcHDefG82WfPMwonySKq+I8SfkjWrfnoe9+4ppaeXin+S3+le4A1kGeVbyJrLF+K9lbxZfN4U8d5xlrF/RNxK9ljdHLI70Jvz/2thF2BDRPTnTTkPnmD+u/IKDbLtP5Hrgb8p/DZFTWvHM+53i4gnyPqcuIjsEYDxvL7o/9X53fEnU0qvyGOYk1Ka6HsW3Ea2byzM7ywcV+HnCsZ+ny35vrUT8IpJLmvrPkO2X+1wQpFv56dHxPVkj3UcOMl1SI1wA3BK/hhtoU+wcuV7pU7Ol/ccYCg/9nch6/8Gskd0Cq4n62uAlNLsvFVKpWXreOVaqfK61LRSbiW7y/xQZK03Z5KVx+vK1InbSdkjzzMi4ptkj5EW4qp1nVFqe5YyqfKzyrK63He7kuyR71eTXcCVcnS+z83J572FbfvnrnkcT08p7T5BHAW3kteNeV25c4Wfgx2/y2/IHvMoPGqxd6kPjaOi4ylljwt2R8SlZI/oPWfsPFILq0Vdch/wJ4WyJl8G5GVdXs6vLMxcwbloqTLpUrLHXq+Iba24S7mF7JGo2yLrhH8pMJK36pjqMT2V66SxSm6bcdwKvDqvZwtdV5RzH7B/SukwgJTSgpTSsyb4TLl64Aqyx/SeD/xknHmOTyktys8DXkZWd10P/FXKRyhOmUpHK95aD1DZtVKxUvVAoS4/aZLLKtTBu+R18PFk/UBtp9rr9XZlMqi2VgB3p5TuIGsh82/lZ9/OW4Ev500QN5E1A+0lezb4+2QdMP+oKEv7b8AvUkofn2yQEfFYvtz/SCndTfbY2l7jxD+Z73QLWb9AvfnrfdnWMujTwJJ8fe/N11/Kv+Tb4G7gq/md3++RNee7Iy+cyvUZNJHrgN1TSveQdfJ51wTzv4usX6E7gYVknV+XczHwB7Ik0j1kd8jL+TFwVP7dXlLi/SvIEiLlkkFPz7fZa8nuYkP2DPi7U/Y43J1s/8zyuCLiYbLf6udkrWzuYeLvXOxSYFVKqdAs9kKyyuiHZBX7ZFwEnJDvM39B1v/TWLOAy/N51jDx9pYaLiLWkh1Xt+bH5JkTlO+Vejyl9HOyTu7/dz7tn4FP5dO7i+b9AHBgXlbcRpY4vRuYn1K6M6X0RsY3XrlWqrwuNW0H+fTNbDu5/zlZn0aFC4cd6sQSi3k68JO8fP4Xso5GIevr4eaU0jdhwj6DJjLe9iz1naopPydbVl/K9mXs2PUPA78tevRjrDX5MtaSNae/Pd8//5lsW95N1oHq2P59xvMZ4KB8v3gVO/YXMa6IWE92fnF3SundZAm/7pTSvWR9ckxmWZUeT4cBt+X7zP8m+72kaaEWdUlkj4ieBlyWL6PwGNcH8mWsIWs1X2zcc9ESxzHAf5LdDFg1wVf6BVmipVAPPEjeKqgGx/R210lp/D6DKlFu22wnItaQnffeQ9Z/0l2UqQfy3+MNwOfz3+O/yJ60KGe7Om6MW8keu7s+ske+S/kZ2ff5GfDxyEaIvJbsGum2lNJashb3lT698j7g9Xl9/Wwmd90w9jzkO2TnKneTtSyrWGSPS3+EbBv8lOy7/brErCuo/nq97TiamFRGSmlnoDciRlNK/wjMjnw0g3aVUloQEVtS1tnnauC1EfFAs+OStE0qGsWk2bFom04rP/M79rPy1rZHAx+IiBVNDktSE+Wtjr4QlQ0m0naK6oE9yJI7R0Qb90+WtyAajGxktP9F1gn3m5sclio02f5YpE7zfODjKaVZZHcDTmtyPI3wwbyV0jyy4T3b9kJGkmqs08rPhcCP8qRQP/DXTY5HUhOllE4newLgzU0OpZkuSSklstZR723nRFDumcDX82ulDXTGtVLbsGWQJiXvR+BHYyY/EREvb0Y8kqTWlbLOPM8aM/mSiPhkM+KRJDVWSul8duxL5oMRMV4Hx5IaxGSQJGlK8s5obwQOAU6PiCvz5tGryDoGvL7weGVK6ZXA+WSjV5wVEbc1JWhJkiSpgzX9MbGUktkoSRpHREw0XHUrGCIbxvSMomnnkrUAuSKldE1KaSlZJ+IfIBvScxfgm0w80ob1hCSVMU3qibqynpCk8Y1XTzQ9GQQQMdlBhiSp/WWPnLe+fASoR8fE28W2UaeuJksAjZCNGLUZ2JyyobTnRURfBeuocdSSNP1Nl3qiEawnJGlH5eoJh5aXJNXDgnxoWICNwG75v+KhxgvTJUmSJDWQySBJUj305MONAiwmG2HiSWDXonkK0yVJkiQ1kMkgSVI93AyckL8+Pv/7fuDAlNKClNLTgKFKHhGTJEmSVFst0WeQpPbV3d3N+vXrGRkZaXYoLW3mzJnsvvvuLFy4sNmhVCWl9C3geUB3Suko4KPAZSmls4EbIuKefL6LgOvJRhN7V5PCldRCrCcqM53riXFGnbyIbMjxx4HfR8Qp+byOOilpO9YTlZlsPWEySDU3tGUTI33ddV/PzHkLmb1gcd3Xo6l57LHHeMYznsGcOXOaHUpLGxwc5Le//e20PMkHiIiTS0w+vsR83wW+W/+IWsNUy0PLOXUC64nKTPN6otSokwAXRsSVhT9SSrOoYtTJKQXWoPPWsSzfpcpZT1RmsvWEySDV3EhfN933rK77ehYe0gVWoi1vxowZFtwVmDNnDjNmdPzowG1nquWh5Zw6gfVEZaZzPTHOqJMA/5hSOgv4bER8HXg2VY46Wa1GnbeOZfkuVc56ojKTrSfsM0iSJElSo30qIpYDrwT+LqX0DBx1UpIaxmSQpIba1N3Pw49tntK/Td39Zdfx8MMPk1Liuuuu2zrtLW95C2ecMbZ1embNmjW8//3vB+BDH/oQmzdvrt0XHmPdunX8/Oc/r9vyJWm6s57ojHoiItbn/28CfgQciqNOSppALeoI64mMj4lJaqjNPQPcdMcjU1rGMYfvy+KFO5Wd59BDD+W6667jz/7sz3j88cfp7+9n7ty5Ey77/PPPn1JsE1m3bh1PPvkkRxxxRF3XI0nTlfVEZ9QTKaXFEbEppTQbeCHwJeAB8lEngUU46qSkMWpRR4D1BJgMktSm9thjD3p7e+np6eH73/8+xx13HLfccgtf+cpX+PGPf8zmzZs55ZRTeP3rX7/d51auXMknPvEJ5s6dyzvf+U6Gh4d51rOexUMPPcTFF1/MypUrWbp0KevWrWPx4sV86lOfAuAjH/kIa9euZWhoiPPOO49ly5Zx4okn0tXVxV133cVhhx3G+eefz2WXXUZ3dzc33ngjn/3sZ1m0aFEzNo8kdTzricYqMerkbimlpcAs4PKI+GU+30U46qSkFtDu9YSPiUlqWy996Uu54YYbuPnmmzn66KMBeP3rX89ll13GN7/5Tb761a+OO0Tlt771Lbq6uvjKV77C0qVLt3vv6KOP5rLLLmNwcJCI4Cc/+QkjIyOsWrWKz3zmM3zsYx8DoLe3l1NOOYVvfOMbrFmzhu7ubk477TROPfVUVq1a1TYn+JI0XVlPNE5EnBwRB0TEsoj4+4g4PSJeFBHPj4hPFM333Yh4Yf7emmbGLEntXE/YMkhS23rFK17BGWecwQEHHLB1BIJrr72Wq666ihkzZvDYY4+xadOmkp/99a9/zate9SoAli1bxg9+8IOt7xUK8yVLlrBp0ybuv/9+Vq9ezcqVKwHYsmULAPPnz2f//fcHYJ999hl3XZKk5rCekCSV0871hMkgSW1r8eLFdHV1bc3iA1x88cVcc801ABx33HGMjo6W/Owf//Efs3btWp773Ofyi1/8Ytx1jI6O8qxnPYuXvvSlnHPOOQAMDAwA7DC04+joKHPmzGF4eHhK30uSVBvWE5Kkctq5njAZJKmhFu08l2MO33fKy6jUO97xDiAbEQCgq6uLU089lQMPPJBddtll3M+dfPLJvPOd7+THP/4xz3zmM7feCShlxYoV3H777Vsz+cuWLePd7353yXkPP/xwvva1r7F27Vr+6Z/+iYULF1b8XSSpE1hPWE9I0nhqUUcUllOpdq0nZoyXxZpISunTwBFk/Q79A7AGuBRYAqwF3h4RpR+e2345oxFRVQyavKEtmxjp667rOkYG+um5/7a6rgNg4SFdzN196gWB6uvBBx/c2rRxOhkdHWV4eJjZs2dz1VVX8atf/WrcArlWxm6rlBIRMaPMRzrCdK4nBtY/Qvc9q6v+vOWcOoH1ROWsJ0qbSj0x1XK6WpbvUuWsJyo3mXqiqpZBKaVnAwdHxItSSs8AvgZ8E7g9Ij6WUvoMcBxwbTXLV/2M9HXXvcKbf8Dyui5faoShoSHe9KY3MWPGDGbOnMk///M/NzskSVILsZ6QJJXT6vVEtY+J/QHoTSnNBnYFHgeOBt6Xv381cCwmgyRNU3PmzOHyyy9vdhhqsqm2phwZ6K9hNJJaifWEJKmcVq8nqk0GbQZ+AwSwM/Aa4P3Ak/n7G4HdphydJElNNNXWlLaUlCRJUiuqNhn0p8DuwLOBvclaAv2KrJXQ74HFwIZaBChJkiRJ7WR0eJiB9Y80fL0z5y1k9oLFDV+vpNZTbTJoJrAhIkZSSk8BC4CbgBOA+/L/r6tNiJKk6ahWAw1IktRuRgZ66X3gzoavd+EhXWAySBLVJ4P+E3hjSulmYB7wAeAq4NKU0k3AOuD7tQlRUjupxYh2E93Vevjhh3nta1/LgQceSG9vLytWrNg6JGQl1q1bx3//93/zpje9aUpxdjIHGpBULesJSdJ4ajU6tvVElcmgiBgGVpZ46+SphSOp3dViRLtK7motX76ciy++mNHRUV7zmtdw2mmnscsuu1S0/IMPPpiDDz54SjHKgQYkVcd6QpI0nlqNjm09UX3LIEmaFgYHBxkdHWX27Nl85CMfYe3atQwNDXHeeeexbNkyTjzxRLq6urjrrrs47LDDOP/881mzZg3XXXcdF1xwAd/73vf48pe/zH777cfmzZs5++yz6enp4Qtf+AILFizggQce4IILLuCoo45q9ldtNQ40IGlasJ6QJJXTrvWEySBJbenOO+9k5cqVPPLII7zkJS/htttuY2RkhFWrVrFhwwbOOussVq1aRW9vL6eccgrnnXceJ510Et3d25qdDg8P87nPfY5vf/vbzJo1i5NOOmnrez09PXzpS1/i3nvv5fOf/7wn+TtyoAFJLc16QpJUTrvXEyaDJLWlQrPOkZERzjzzTNauXcvq1atZuTJ7wnXLli0AzJ8/n/333x+AffbZh02bNm1dxoYNG9h7772ZP38+wHZNPQ866CBmzJjBPvvsw8aNGxv1taYTBxqQ1NKsJyRJ5bR7PWEySFJbmzlzJosWLSKlRF9fH+eccw4AAwMDAMyYMWO7+UdHR7e+3m233fjDH/5AX18fs2bN4r777tv63tjPaQcONCBpWrCekCSV0671hMkgSW2p0KxzaGiIJUuWcOyxx26dBrBs2TLe/e53l13GrFmzOOOMMzjllFPYd9992WOPPZgzZ04jwp/2HGhAUquznpAkldPu9cSM4qxVM6SURiOiqTF0koH1j9Sk9/Vy5h+wnN4H7qzrOiDrAX7u7vvWfT2amgcffHBrs0lozJDBtTQ4OMicOXMYGBjgta99LZdffjmLFi2qy7rGbquUEhHR8beWm1lPTLXMnGp5aDmnTmA9UTnridKmUk804ty4lEadL49lvaLpqLjsa9TQ8rXUqvWELYMkNdTsBYsnHMaxlVx55ZVce+21dHd3c/LJJ9et4JYkZawnJEnjmW51BLRuPWEySJLKOPXUUzn11FObHYYkqUVZT0iSymnVemJmswOQJEmSJElS45gMklRXc+bM4amnnmp2GC3vqaeeapnO5CSpkawnKmM9IalTWU9UZrL1hI+JSaqrJUuW8Lvf/Y7169c3O5SWNmfOHJYsWdLsMCSp4awnKmM9IalTWU9UZrL1hMkgSXU1c+ZMnv70pzc7DElSi7KekCSVYz1RHz4mJkmSJEmS1EFMBkmSJEmSJHUQk0GSJEmSJEkdpOo+g1JKRwEfBOYA1wJfAVYBi4DrI+KiWgQoSZIkSZKk2qkqGZRS2gm4EHh1RPTk0z4GXBIRV6SUrkkpLY2Ie2sYqyRJkqRpJqU0B7gROAQ4PSKuTCntQYkbySmlVwLnA6PAWRFxW1OClqQ2V23LoBcAvcCVKaXZwN8BXcAF+ftXA8cAJoNUN6PDwwysf6Su65g5byGzFyyu6zokSZLa3BDwOuCMomnnMuZGMhDAB8iuI3YBvkl2jSFJqrFqk0FLgKXAkcB+wBeBBRHRm7+/Edh/6uFJ4xsZ6KX3gTvruo6Fh3SBySBJkqSqRcQo8GhKqXhyqRvJI8AvI2IzsDmlNCelNC8i+hoasCR1gGo7kH4SuCUitkTEOmAx0JNSmpe/vxjYUIsAJUmSJLWdsTeSd8v/PVk0T2G6JKnGqm0ZtAZ4T0ppJrAX0AfcDJwAfAc4HnhPTSKUJE1LDjQgSSqjp6jVT+FG8pPArkXzeINZkuqkqmRQRDyZUroU+AnZSf45wDrgspTS2cANEXFPzaKUJE0rDjRQG7XoG60ZfZ9t6u5nc89AQ9cJsGjnuSxeuFPD1yupKqVuJN8PHJhSWkB242DIR8QkqT6qHlo+Ii4BLhkz+fiphdO5hrZsYqSvu+7rGRnor/s6JAkHGqiJWvSN1oy+zzb3DHDTHfXt4L+UYw7f12SQ1KJSSt8Cngd05y1HP0qJG8kppYuA68lGE3tXk8KVpLZXdTJItTXS1033Pavrvp75Byyv+zokCQcakCQViYiTS0ze4UZyRHwX+G79I5KkzlZtB9KSJJXjQAOSJElSi7JlkDpG/8AQ/YPDk/rMaM8AvcObJ/UZ+6yQAAcakCRJklqWySB1jP7BYR56dHKJnV337Gbt+sn1s2SfFZIDDUiSJEmtzGSQJKkuHGhAkiRJak0mg9RU1Ty6VTB7aJintlTeamd4eLSq9UiSJEmS1E5MBqmpqnl0q2DufoOT+uw+e+xc1XokSZIkSWonjiYmSZIkSZLUQUwGSZIkSZIkdRCTQZIkSZIkSR3EZJAkSZIkSVIHMRkkSZIkSZLUQUwGSZIkSZIkdRCHlpckqY31Dwzz2GObG75OSZIktS6TQZIktbGBoWFuuvuRhq7ziLRnQ9cnSZKkyTEZJElSg/QPDNE/WHmrmdlDwzy1pX9K69x5eHRKn5ckSVL7mVIyKKXUBdwMFG4BrgIWAddHxEVTC02SpPbSPzjMQ49W/sjW3P0GJzV/KQcmk0GSJEna3lQ7kH4XcHv++lzgkojoAo5MKS2d4rIlSZIkSZJUY1Ung1JKrwRWA1vySV3A1fnrq4FjphaaJEmSJEmSaq2qx8RSSjOBvwX+AnhVPnlBRPTmrzcC+089PEnSdObjxJIkSVLrqbZl0BuA70ZEX9G0npTSvPz1YmDDlCKTJLUDHyeWJEmSWky1HUgfBjw3pfRqYBlwOdmd3xOA7wDHA++pSYSSpGmp6HHiQgvSLuCC/HXhceJ76xnD0JZNjPR1V/35kYGpjeQlSZIktaKqkkERcW7hdUrpRrKWQjOAy1JKZwM3RMQ9NYlQkjTttMrjxCN93XTfs7rqz88/YHkNo5EkSZJaw5SGlgeIiBVFfx4/1eVJktrC1seJU0qFaT0ppXn5I8Y+TixJkiQ1yZSTQZIkleDjxJIkdbhN3f1s7hloyLoW7TyXxQt3asi6pHZgMkiSVHM+TixJkjb3DHDTHY80ZF3HHL6vySBpEkwGSZLqyseJm2un2TM4dPfBqj/fN3Mev3p8ccANvQAAIABJREFUuIYRSZIkqdlMBqmk/oEh+gerO/mfPTTMU1sqG4FneHi0qnVIkio02MvGu2+q+uO7LjsGmFO7eCRJktR0JoNUUv/gMA89urmqz87db7Diz+6zx85VrUOSJEnTW0ppC3Bb/ueHgZ8AlwJLgLXA2yNipDnRtaf+gWEee6y6c/xq1yepNZkMkiRJktQMDxY/SpxSejtwe0R8LKX0GeA44NpmBdeOBoaGuenuxvThA3BE2rNh65I0OSaDJEmSJDXDM1JKNwG/Bc4Ejgbel793NXAsJoNqas5MptSP3GTtOrKRZ+05y77npBZkMkiSJElSM/xJRDyRUnor8CFgN+DJ/L2N+d+qpSn2IzdZ8/fYmXlLnod9z0mtZ2azA5AkSZLUeSLiifzlN4DDyRJBu+bTFgMbmhGXJHUCWwZJNTY8MsLDDeqYb9HOc1m8cKeGrEuSJKlWUkoLgL6IGCZ7HOx+4L+AE4D78v+va16EktTeTAZJNdbbN8TP4/GGrOuYw/c1GSRJkqajg4AvpZSeAvqB04EngEvzfoTWAd9vYnyS1NZMBlVgaMsmRvq667qOkYH+ui5fkiRJahUR8TOyR8PGOrnRsUhSJzIZVIGRvm6671ld13XMP2B5XZcvSZIkSZIEdiAtSZIkSZLUUUwGSZIkSZIkdZCqHhNLKf1/9u483q753v/465zkRIaTATEkXC1VHyIiKG2JNNXe1tBLWy2iF6X6o+Uar7a4N9y2Ormt24nqYEopNXVAKZeQUCklSMSnXFQJJYkMJ2cefn98vzvZOdl7nz2P7+fjkUf22XsNn7X3Wt/vd33Wd33X+4HvAd1AG/CZuKy5wFjgPne/uEgxilTM5q0tTKUnp3km9K9i6pa5zdPZPJIX3urLaR4RERERERGRfOQ7ZtDfgA+5e7uZnQqcBkwArnL3m83sTjOb4u7PFi1SkQoY1tfFqqdzGy9q1MTRrFrentM8E6bNBFpymkdERERERIK+/n5efXNtWdY1dvQIPdFXal5eySB3X5b0ZzfQC8wA5sT37gBmAkoGiYg0IPUgFRERkXLq6OzlCX+rLOuaudd2SgZJzStozCAz2xL4IvALYIy7d8SPVgFbFBibiIjUrkQP0g8Avyf0IP0yoQfpDGBfM5tSyQBFRERERBpV3skgMxsN3Ayc4e7LgXYzGxk/Hg+sLEJ8IiJSg9x9mbsn7pdM7kF6R3wv0YNURERERETKLK9kkJkNB24Efujuj8S35wOHxteHxL9FRKSBqQepiIiIiEj1ybdn0GzCFd0zzWyemZ0HfAf4vJktAJ5y9yXFClJERGqPepCKiIiIiFSnfAeQnksYBHSwQwoLR0RE6sEQPUhvI9QXF1QoPMnB5q0tTKUnp3km9K9i6pZhns7mkbzwVl8pQttEOZ8kk6AnyoiIiEgtyvfR8iJSBcp14qOTHclDogfpODM7E7iT0IP0OjM7B7hfPUhrw7C+LlY9vSCneUZNHM2q5WHIqAnTZgItJYhsU+V8kkyCnigjIiIitUjJoBrS1d1LV09hV1eH9/axZl3XkNP19Q0UtB4pj3Kd+OhkR3KlHqQiIiJSr8rdE1UXZqUUlAyqIV09fbzyemGFzogderJaxqSJowtaj4iISCPQrWkiIo2n3D1RdWFWSkHJIBEREZE86dY0ERERqUX5Pk1MRERERERERERqkJJBIiIiIiIiIiINRMkgEREREREREZEGojGDREREREREKqQYTwzO1oT+sqxGRGqAkkEiVWDz1ham0pPzfBP6VzF1y+zm62weyQtvlaehISIiIiLZKcYTg7M1fpeBsqxHRKqfkkEiVWBYXxernl6Q83yjJo5m1fL2rKadMG0m0JLzOqC8j07WI5NF6ku+ye4EJbJFREREik/JIBEZUjkfnaxHJku5ZNMtf3hvH2vWdeW9jsHz9/U13hXZfJPdCYUksutVORP0CUrUi4iI1Bclg0REpCFl0y1/xA49BXXdHzz/pImj816WSEI5E/QJStSLiIjUFyWDREREREREpCQKvV04W4PH0tRtxiKZKRkkIiIiVSuXk4hUg+rrZEBEpLIKvV04W4PH0tRtxiKZKRkkIiIiVSuXk4hUg+rrZKB2rW7rYm17d1nXqbGRRESkUdR0Mqh33Wr6O9tKvp7+7syDh2YzCOlQshmktBEHHpXiKaSLbj0+wr6cJxk6uRCRWleJQau7uvtYuOSNsq5TYyOJSDUqZxk8fFgzvX39ZVmX2siVVfRkkJl9HjgR6AZOcvcXi72OhP7ONtqWlKHL4U7TM36ezSCkQ8lmkFINPCqFKKSLbi6PsN9x/w8xMsvEUcp1dS2ne8WajNM0j2xl+Jjxea8DYG17Nw89+VpBy8iWTi42Vs56QkSKoxKDVu9tW5V1fVI9VE+IbKycZfDetpWeItwgipoMMrMtgJOBA4C9gG8BRxVzHSJS3Qq9L3zcpLE0jclcKbTuPgMKTAZJZaiekHIrdODS1rGjaFvbkfbzoXpO1kpvSZFqUS31RDF6/meSfFeAev+XRikHrs5U9g9VbxR7fcVeby53BEDt1HPlvvW4Fno9Fbtn0H7APHfvBR4zMyvy8kVEpLapnpCyKjRBPfG9M3g1w/xD9ZzUmEW1pRK3w9XCCUOZVUU9UYye/5kk3xWg3v+lUcqBqzOV/UPVG8VeX7HXm8sdAVBYPVfOMrcYtx7vvNUwRvZ3ZjXtqG1a6e4aUdD6EopxV0QqTQMDxctEm9mxwA7u/q349zPuvscQ8ygVLiKShrs3VTqGYlI9ISJSXKonVE+IiGSSrp4ods+gt4FpSX8P2V+s3iowERHJSPWEiIhkonpCRKQMip0MWgjMMbNhwJ7A80VevoiI1DbVEyIikonqCRGRMihqMsjdV5rZtcB8oAf4XDGXLyIitU31hIiIZKJ6QkSkPIo6ZpCIiIiIiIiIiFS35koHICIiIiIiIiIi5aNkkIiIiIiIiIhIA1EySERERERERESkgSgZJCIiIiIiIiLSQIr9aPmqY2afB04EuoGT3P3FpM/2Bb4PNAGXuPsdZtYC3ABsG98/1d0XVyi+HwFHAre6++nxvWbgx8BUYBnwWXfvqJLYTgFOAnqBx9z9rFLElW98SZ9dCMx296nVFJuZnQwcQzguL3H3e6slPjObAPwaGAH0A//q7svKGVuM4dYYQxNwhrs/UQ3HRIbYynpM1Ltcy/PKRFm4XPezigVaoEy/Z/x8PPB/hHr4lgqEWBRD7LdbE8qvLYE33P3YykRZmCG28QTgdELdcYO7f78yURYuthHnAbsDJw/eL83sY8CFwABwprs/VvYgJaNarkdqvW6o9TK/lsvyWi6ja73czRS/me0MXE2IvQf4nLu/XM746rpnkJltAZwMzATOA741aJLLgE8D/wx8zcyGAbOANe5+IHA+8KUKxvcNYHBhcgjQH+N7nHBgV0ts9wHvc/cDgG3N7IBSxFZAfIn5ppQqrnxjM7M9gH3d/cPuPqvEiaB8vrtPA/PdfRah0Dq1ArF1AcfHff9k4Nvx/Wo4JtLFVrZjot7lWZ7XnDz3s5qTxe8J8O/AwnLGVWxZbOd3ga+4+0HVdvKQrSy28XxC2+r9wClmNqKsARZXL/Ap4H8GfxDLnK8BHyGURZeVNzQZSi3XI7VeN9R6mV/LZXkdlNG1Xu6mjR94GzjC3WcClxKSWmVV18kgYD9gnrv3xiyhJT4ws5HAcHd/zd3bgL8C7wZeJGTWASYAb1UiPoDY82Jg0DwHAokrFXcQDuyqiM3d/8/dE+91E3b+UsnnuwP4CqHALqV8Yvsk0Gdm95nZDWa2eZXFtxQYG1+X8rhIG5u7d7j7a/HP5P2r4sdEutjKfEzUu3zK81qUzzFQizKWQ2a2DbATUFVX+PKQab8dBuwKXGxmD5rZ0ZUKskAZf0vgOaAVGAV0AH1ljq9o3H3A3V9P8/G7gb+6+9p4nLbEskmqRy3XI7VeN9R6mV/LZXlNl9G1Xu5mit/dV7j7yvhnRY7dek8GbUHIuCU0D/psVdLfq+J7rwKtZrYUuDz+q0R82cyTiLkU8okNADN7P7CNu5cyu59zfGa2PTCpDF1n8/nuJgNj3P3DwP2EpFWp5BPfYmCGmT1D6Er6y1IERhaxmVkT8D3gOynmqegxkSK2xPvlOCbqXT7leS3Kez+rMUNt54VU4dXtPGTazq2BPYGvA4cCX4lXcGvNUL/lLcCThBOOa929qk40imjw91DL5VC9quV6pNbrhlov82u5LK/nMrouyt2YwLqYcJtqWdV7MuhtQi+GhL4Mn40HVgInAM+7+27AwcAVFYovm3kSMZdCPrFhZrsQKqJ/LUVQSfKJ7z+Bb5YmnI3k+7v+Mb6+G9ij2EENWleu8X0J+KW77wGcTekaGtnE9n3CFY4HUsxT6WNicGzlPCbqXT7leS3Kaz+rQWm308x2BCa4+9Nlj6r4htpv/+bBOuAvwM7lDK5IMv2WY4ELCFei3wUcaWY7lDe8sqmncqhe1XI9Uut1Q62X+bVcltdzGV3tx+2QYs+yXwLfd/fnyr3+eh9AeiEwJ37JewLPJz5w9w4z6zWzScAaQjezFwj3TC6Pkw3ewcoWXwYPEbLOd8f/51dLbPG7/CVwrLuX8va6vOIjdD/9npkBvMPMvunu51dJbA8S7lG/HngPYQC9UsknvmY2HBcrgFLdxpYxNjO7AOh19+T7bqvimEgVW5mPiXqXT3lei/I5BmpRpu3cC3iXmd1NaFCvNbOl7r6kAnEWKtN+22lmr8XbI5YTBsH/W2XCLEim37Kf0PV9nbv3m1k7MK4CMZbD88AuZjaGcFt1r7t3Vjgm2Vgt1yO1XjfUeplfy2V5PZfR9VDuXg485O63VmLlTQMDqYZVqR9mdipwPHGEbmAG8JK7P2hm7yUMNNUEfNPdf2dmrcCvCDvUaOBCL+1gvpniu5AwlszWwLPu/lELT066gjAI8uuEJye1V0lsvwA+BLwcF1HqJ2LlFN+geRd7aZ8mls/v+gNCBdJLGAiwJE/ryjO+7YDrgGGEMbW+6O6LyhkbYTyvl4EFhDGNXnP3z1TDMZEhtrIeE/Uu1/K8cpEWJtf9rEJhFizT75k0zcXAYq/CJ8tka4j99j2EnoMjgOvd/YeVizR/Q2zjWcBswj77J3c/u3KRFs7Mfk24aNNGuAixErjd3d3MDicMxjoAnK1bg6tPLdcjtV431HqZX8tlea2X0bVe7qaLn3CudS/waJz0L+5+bjljq/tkkIiIiIiIiIiIbFDvYwaJiIiIiIiIiEgSJYNERERERERERBqIkkEiIiIiIiIiIg1EySARERERERERkQaiZJCIiIiIiIiISANRMkhEREREREREpIEoGSQiIiIiIiIi0kCUDBIRERERERERaSBKBomIiIiIiIiINBAlg0REREREREREGoiSQSIiIiIiIiIiDUTJIBERERERERGRBqJkkIiIiIiIiIhIA1EySERERERERESkgSgZJCIiIiIiIiLSQJQMEhERERERERFpIEoGiYiIiIiIiIg0ECWDREREREREREQaiJJBIiIiIiIiIiINRMkgEREREREREZEGomSQiIiIiIiIiEgDUTJIRERERERERKSBKBkkIiIiIiIiItJAlAwSEREREREREWkgSgaJiIiIiIiIiDQQJYNERERERERERBqIkkGyETO7b4jPJ5jZyeWKJxtm9lkzm1iGdXwrvv6qme1fyvUlrffjZrZznvO+08weLVIcGfeLHJbz78VYjoiUTjnK1EIVUjbmsI5ZZnZjfH2qmX2qlOsbtN73FDD/G0WK4y4zG1GE5ZxlZsOLEZNIrTGz75nZYjM7r8DlZFUumNk8M9s1y2XeaGaz4uuitPNyUawyptQKLZOzXMf6NruZHW5mZ5RyfUnrnW5mHy5g/pfNbGQR4vi5mb2zCMup+vZLtVEySDbi7kMVCBOAiiSDzGxYmo8+C5TtwHf3Oe7+SJlW93GgpCc82chiv8hWWZNBGfYZEUnvs6QoU82syczK2m7IcAyXtWx095+4+y1lWt0soKQnHtlw90PdvbsIizoLKFsySOW+VJljgT3c/dLkN/PYT2dRwnKhiO28XNZZrDKmKDL8JrMoY5ns7r9z9x+UaXXTgbL/9oO5+8nu/nIRFvVZynhOWA/1TdPAwEClY6g5ZvZ54AygH/iNu19kZocC3yIk2G5x94vjtG8ANxEOtBeAT7p7X+xZ8j/AZsAydz/EzD4OnB/f+xvwGWAE8Ii77xqXNwX4kbsfZGb7Ad8FxgD/B5zg7u2DYv0H8BvgQOA54F/dvT1TvO6+bbxScD7QDewKXOPul5jZ9YRG+PPA9cDjwA+APmCdux+Q5jtbDOwTY10O7OPuT5rZUmAP4J3A1cD4GOeJ7r7OzOYBTwIzgUvjMv4F6AJ+Gb/TucDfgeXufmCG320e8BfgIKAX+ALwbWBH4N/c/U4zexdwLTA6ruMkd19qZp8FdnX3r5jZNcCN7n53/M2+A6wCngVedveL47oWEn73ZuBwd/+7mW0D/BTYDugAPufuf001PTAZ+APwNrAWONDd18ZtORw4wt0/F//+IjDR3b+atL3vBG4HXgSmAPOBU92938wOA/4TGAk8CnwR2AG4DVgav+cH3P0LcVmJ/WIYcCVwQJxuR+ATcZUp502K5xLgS8AS4AHgtzGeY5J+n1OBzhi3A/sCvwJeijF2A4e4+yoz2xe4gnC8LABOj8fWy3GeQ4Azgc8DexGO12+6+w2IVBEzmwzcTCgfm4GjCA2av7r7VXGa2wlljQGHAlsRjr8zCeXGh4E/uftJcfo3gFvj+wuBXxOO+dHAv7j7y6nKI2B3BpWpsR65jVCP3AwMc/c5cT2XAX9291+l2K57gf8HvAysJJSnt5vZg8BsoAe4DtgeWAYc7+7/iGVsO+H4v5FQv5xKKJMfIBzfKcvGFDFcA7QB7wVageOBOcBuwHfc/admNo5QT04Amgj1wYJYD57q7seY2cXAG+7+EzN7P/DzGM9CYJS7fzauazXwfmAsMNvdF5lZK6GsSlyt/zd3fzTV9MAK4LH43awAjnZ3j9syDbjM3T8U/z4U+FTiN0/a5nT1fso2Q4Z2ysuEeq/TzL4OfBp4Ja7mEnefl27epFhOAy4j1A3PAV8m1J/vS/p9EvVpNvvszuTRVhh8Ii5SDmZ2C3AEod3zZUK7Onk/nUAodzcjtE8/F9touxLaWhMIbaKjCMfE+nIB+GCaeecRyq3n0sT0tbi8vxHKu/XHclL7/wJCO3kK4Vxhy7jO1wjHYm8u5Ulc9bUMaosNKmP+Azgmfn6hu/8+3blImu36GXCDuz9gZk/FaS8zs2vjd7kI+AUwFVhDKDv+Gsv2d8TlP0Iof+fE7/qvwDmkKZNTxHAxoS29K7ANcBJwCiGR9Ct3vyhOdwcwiXCOd5G73xbb7De6+/sGnXPsAtwAtBDqv+nuPiuuaztCXTYJ+IK7/zH2wvwusH+c5z/jd7nJ9MD/EtrXmwGvA2e7+wMxxi1Ic/45aJtfJvzehwFvxu/nLTN7N3A5sAXwFmH/+Eec/lrCueRq4GPuviZ5vzWzLwBnx+W9Adzh7tekmzcplk+wafvlDXffNun3SdTjLxPOYw+P2/5fhHPCbQh19+NmthV5tFHc/SxqiHoG5Sg2xk4DDnD3PYHvm9kowg5/GKGg+2czmxFn2Qa41d13J3zfB5nZZoSd64S4jM/EaR909/e6+3TgYULBvhL4m5lNj9N8GrjFQrfKSwlJhr2BPxNOlgfbGrjT3acQCv4vDhFvsr0IJyR7Aqea2RjgQuAZd58eG1fnAGfEmA/N8NU9TigM9ydUhPub2faEg7KXkFC6zN2nEU4czk6at9Pd9wHuBT4FTInf20/d/fa47E8kEkGxq2G6DP4Kd9+LkDz4UfwOPg78R/z8deBD8Ts9E/hGug2K3+NlhCsGBxBO0pK1xbh/SSgkiNP/l7u/h/DdfS/d9O6+EPgdcFr8vpNPdu4BPmgbut5/Gkh11Xoa8DVChT4R+GTsPnkWMCv+bn1sqKynABcTTgjfHwvzZEcC4919N0LBOS3ps4zzuvuFhO9/ursn/76p7EZojEwh7IObx+/mT4QGCYQTglPcfQ9CZTM7af6X47a1Af/k7rvH6e4cYr0ilTAbuC/us/sQyuprgOMAzGxL4N3u/qc4/a6EZOcRhMTIrwjHyp5mliiHtiE0dHYlHJMfdvf3Aj8jNE4hRXmUqkwl1CO/cfepwPeBT8VeQsMJZehv0mzXw4Qyf3dCUnp/M2sBtnL3ZYQy5N54bP4WSG7kTwD2c/fvEuqdvWK5f1GqstHC7buHp4ljhLvvC/yEkGA4HnhfXC6ERNgRsdz/F0JDOpOfEX6zfdj0CmSru+9HqFMSt4X8B6EdsC8hef7jdNO7+99jnJfEbVt/0uHuTwPb2oYu8OnK/VT1fqY2wybtlOSFxZO+gwgnUScQEmtkM6+7/5jQiH6vux9NZtnss3m1FYZYr0hJuPun2NDuuSe+3enu+7j7jcCv3X3fuD93EMofCCe0F8f99yOEk9vB5UK6edOKx/I/E47lEwnlYCrTCOcm+xJOkF9KWs+H8ihPppOhLRbjOpxQpn4U+GFMokPqc5FUHibUMeMIyaPEcA77EJJlpwNvxfVfQjgPSngXMNPdzyXUC4fF7/7kVGWyhduGTyW1ScAMQp16O6Fs3wM4MWmbjo9l1QHA18ysKc2yIHQcSOwLvYM+ewfwAUJy76L43smE32tfQtLxW7He3WR6D4n7OcDVcdseSCw43flnmhhfi+2D3ybF8WPC97cPIQn3n0nTvxC35yk2tOkBMLPtCOco+xDOK/cetK6086Zpv2SyJO4P64BzCd/XufEf5NlGyWK9VUXJoNzNImRu18D6g8WAZ9397+7eQ8iQJnrIrHL3h+LrJwk9YIywMy9NWgbADmZ2n5k9Q2j4TInv30xo2EBoSN4WlzENeMDMFhEK9XekiLfd3X8XX/+KUEBlijfZw+6+wkNvo5cIPVUGewT4jpmdTsgsp/NwXMcBhKvbideJ262mu/tt8fUvB8Vzc/x/NeHk/uex0d+WakUeuho+niaOO+L/zxAy3p3AYkImn7gNV1voyfQzNvwGqSS+x2Xxe7xt0OeJ7z3xu0OoFK+Kv9nPCJVGpulTcvcuwnc3y8y2BrZ092dTT+qL3H2AkL2eQbgKPQ14NMbxYWCnOP1Sd38+VhBPp4hjf8LJFO7+FOHKLFnOm4tn3f3FuJ0vAn+M7y8mHCcTgGZ3/0t8/3pS7zMvxul/aGYfcvfVBcQkUiqPAceZ2UWEpE+Hh6u6I8xsB8LV0puSpr8vll3PEE4sHonH+BI2lGVvD3r//vh+cnmXqTxKtjZxIuPuqwiNsAMIjfb57t6RZr7kcv+HhEb93oQyDkJ58sv4enC5f0uMHUJD/pdmdgzhytsmPNy++7tUn7Fxuf+ku7/t7suBpthIbiLUY8/EaXdPsxxi2dPk7k/H+H49aJJU5fg/A/8Vv+c7gK2TEvlZl/vR7cDH4/wzgVTjfKSr99O1GVK1U5LtD9zu7j0xiTc/6bOh5s1FNvts0doKIhVyc9LrPc3s4Vj2/AswJSYzxiZOzN19tbv3p1jOJvNmse79gdvisfwaGx/LyR6J5eRbhOMpUYYmjsVcy5Oh2mL7E8r8Lnd/HXiCDeVwNucisKG+eT/hAsU2Fnq/roptyfX1jbvfRbjomPCb2I5PLOdnZnYSkPL2GQ+3Df8kTRx/iL/XM4QLk4m27MtsqGPPjr2X5sfvZ9s0ywLY090T3/9Ngz67M7a5B9c3p8Tf5SFCr9PJGabPJNX5Zyo3xv9/Bcwws7GE3+K3MY45hN41CZnqvX0JbZy18Vz7D4M+z7XOzCS5bTAv/m7J9U3R2ijVTAP6lV7yTtEPZLq38AeETO08CwNVfiy+/xtCgTsXWO3ub8QEwGPu/pEcYhkgTcGWQ+w9yRO4+zfN7A+Eimihmb3H3VekWNYjhF42owi9R04jHFR3ZxFHe1xXb+zx81HCFYtPEq4W5CJxb3J/4rWHbrWJ3+Uswu1JnyF0i02XVIJwApFJ4vtL/t0HgL3TVOypps/kFkJPnXeR/sr8wKDXA4S4f+vu/y95wthFdaj9NdM257KvQ+iRlJyQTk4mJt9D3s/Gv1s2301in3k79uY7FPiymR3o8ZZIkWrh7g+Z2UxCOXqrmZ3u7v9L6EH6GcIV02OTZkkuuwYfK8OSp0l6P9UxlKk8StY+6O9rCL2WJhBuf0rnUcJV5VbCVcHZwIcIje2hJK/zMMItEUcSelnOymL+ZJuU+0l/D4txjSAkGvrMLFPyIJ9yv4lwe+uy5AljJ658yv1vEnoKPOxDj7eRXO6nazPUW7lfjLaCSKkkl20/I5QNz1t4wEZrmnlSyXfewe3CVIY6FnMqTwpsi2VVxsTvYWfC7bH3Au8m1FPZjPGZ/Jt8gZBQOpxwXrNHlnEmZKxvzOyDwH6EXiVd8eJzpovpQ5a/g85jmgh3lmy03cn1zaDpM9nk/DPNdANJ/w8QyvjXPPR2Ths3udc3Q82bKTbY9HtO9VvlVN9EhbZRKko9g3J3P3BMzNwn7ql0YDczmxyv1n2azI3d54B3mdluScsAGAe8bmGAzuMSE8crmG8Qkii3Ji1jx0QhZWZjLPWTVUabWSKpdDTh9qhc4022lpBlJq53p9jz5GuErPc/pZnvWcJVi6aYIf874fasxG0Pi8ws0cX12BjnRmL3yvHxiueXCN1ON4mpQOMIt64NMHTj8TnCVZxt49Xlj2ex/PnEAbjNrDmLSibTtt1N6NVzNOm7bu5qZtNiF9SjCN/ro4RuvtvHOLZMvM7CI8SrBDH2rJ5YkaTfNgxA+wqwu5kNN7N3sPEtZxnF3gl9ZrZnfGs2qfeZiYR97iZC984gia95AAAgAElEQVR0FZNIxcT9/3V3v4JwhS1RLtxIGPOqx91fKsGq05VHQ5Wp9xKS+XsCD6abyDfc2rqbu/+NkFz/Ihvqm0fY0M07XbnfDGzv7vcRum7vkmWMuRgHvBkTQZ8ijIGRkru/DQwkfVfZPGHsPsIFEACSyq100m6buy8iXLX8POnL/VT1frZthlQeIfZGMrPELRC5SN6efwDbx/VPIJy45aKQtoJItRkDvGXhaUxHA8TeEKtj0gAzGx/LwcHlwibzZuER4BNm1mJhrLpcj7+EnMqTLNpiyXFtS+hFuiSPuJ4llAt/JtQzZ5GivjGzg+O0qewYEykXEMbcGUvx65uVMRG0Hxv3UErlKQvjw0E4XxvKfcAXEm1t23CbVzqZ6ptU55+pHJ30/4LY8+ttM/tIjKElcc6bhccI5yitsSw/OMv5EgZvzzoz297CEC25dKCAwtooNUPJoBy5+2LCWDOPxC5+Z8Qu8qcRurItAv7X3TfZYZKW0U0Ys+C6uIzr4kdfi8tYSOgKmexmwg55a9IyjgV+EpfxJ1I/WeVN4F/MbAmhB8kVucY7KPYVwNNm9rSFx2SeY2ZLzOxpQuXwVJr5BggDsT0R33qYcOvB2/HvM4AvxeXsRLhHdrCxwJ1xe3/LhvtPrwHmmtl8GHLMoKH8BDg9dmvMWPDH7/FcQjfMhwljM6zJNA/wb8AhcRsWE7LJmdxIvL0gdrtMXn8nodDczsNYEqk8Tbh/dSlhsNXb3f1Nwu//2/h9/5EwxkQ2bgHaLQz8fTEhsTjUNif7JfCMmV0WTw7/l1Dh/zfh+8jFScAvLHSRXs2GbqrJtgcejL/ndwkDIYpUm1mEcvVJQuM80ZV9NaFMnVui9aYrj64hqUwdLPYkupcwLsRQvU3/RLhFAEI5OY7QJRtCGXJoLIc+yYax25INA26I0ywkdDeHQWWjZR4zaCg3EMbze5pw69U/hpj+FOAmM3uCcAvSUGXgV4HJsd58lqGfyPl7wm2DT5qtHwMq2W8Jjdp7UnwGqev9bNsMm/AwRtODhLL6WsI+mUu5/zNgvpndFOP4n7iMG0nTZsigkLaCSLW5hNAufoANt89COEf4r7gP301ISgwuF9LNm5a7/5nQ7lpMGHfx0XyCzqM8ydgWi3HdSdiOPxLOrfK5vfNhwrhAHfH1dmzoGfQjYjlMKBNOS70IvhvblU8D18bzlI2+e8s8ZtBQ7gG2jOXz2QxdBp5NGFdoEaH311Bl75WEOuypuI45Q0z/ALBf3LYPpvh8o/PPNLaP39mRhPoOQq/M8+L+sYiNx5pLy91fJfxWTxD2/SXkVt9cw8btl4sI9dcfCecsubiY/NsoNUNPE6tzljSKuhSfmY3x8CST4YR7T7/l7vMqHFZJJW3zu4Hfe3zSgIgUl4VBOhcB708xxkJFmdkjhIcgPF/pWMotqQxsIgxO/LC7X1PhsEoqaZsnEk789vY0T3ATEZHiMLPRQIe7D1h44trweh/uIKm+GUPojXOku7841HySH40ZJFKYL5rZZwj3od5Z74mg6J6kXkqpnmAnIgWKXbt/A1xeTYkgC+PVPUIYJLPhEkHREWb2JcI4Q48RehbVu6tib4TEo4qVCBIRKb33ApdZGOPnJUKvsXr39dhLaSRwpRJBpaWeQVJUZvZRwoChye52969UIh4RESmtOG7E4FvZnnb3Rmi0iohUHTNbyKYD5u6XxaDzVc3MtiTc6pZsubt/uBLxiNQ6JYNERERERERERBqIBpAWEREREREREWkgFR8zyMzUNUlEJA13b6p0DJWmekJEJD3VE6onREQySVdPVDwZBOCe65PeRETqX+qnSjcm1RMiIptSPbGB6gkRkU1lqid0m5iIiIiIiIiISANRMkhEREREREREpIFUxW1iIlK/+vv7WbZsGT09PZUOpaq1tLQwefJkmpuVoxeRxqJ6IjuqJ0SkUameyE6u9YSSQVWod91q+jvbCl9Q83Do7y18MSNbGT5mfOHxSENatmwZ48aNY9y4cZUOpaqtWbOGZcuWsf3221c6FKkThdQlKvelnFRPZEf1ROMo1rmAynKpF6onspNrPaFkUBXq72yjbcmCgpczaqfpdLy4qODltO4+A1SRSJ56enpUcGdh3LhxrFixotJhSB0ppC5RuS/lpHoiO6onGkexzgVUlku9UD2RnVzrCfUzFRERERERERFpIEoGiUjdefXVVzEz7rnnnvXvnXjiiZxyyikpp1+4cCFf/epXAbjkkktYu3ZtyWJbunQpTzzxRMmWL/Wjd91qule8lve//u6uSm+CSNVSPSEiIpk0Qj2h28REpKxWt3Wxtr27oGWMHT2C8a2bZZxm6tSp3HPPPXz0ox/lrbfeoqurixEjRgy57AsvvLCg2IaydOlS3n77bfbee++SrkdqX6G3CYzaaXoRoxEpH9UTqidERNIpRh0BqidAySARKbO17d089ORrBS1j5l7bDVl4T5w4kY6ODtrb2/nDH/7AwQcfzMMPP8zVV1/NAw88wNq1aznmmGM4+uijN5rvuOOO4/vf/z4jRozgrLPOoq+vj5133plXXnmFK6+8kuOOO44pU6awdOlSxo8fzw9/+EMAvv3tb7N48WJ6e3s5//zzmTZtGocddhgzZszgqaeeYo899uDCCy/kuuuuo62tjXnz5nH55ZczduzYgr4LEZF6o3pC9YSISDrFqCNA9QToNjERqWMHHXQQ999/P/Pnz+fAAw8E4Oijj+a6667jpptu4tprr6W/vz/lvL/+9a+ZMWMGV199NVOmTNnoswMPPJDrrruOnp4e3J0HH3yQ/v5+5s6dy49//GMuvfRSADo6OjjmmGO48cYbWbhwIW1tbRx//PHMnj2buXPn1n0D38x+ZGaPmNmjZjbLzEaZ2U1mNt/MrjAz1UEiUlGqJ0REJJN6rifUM0hE6tZHPvIRTjnlFHbaaSdaWloAuOuuu7j99ttpamrizTffZPXq1SnnffnllzniiCMAmDZtGnfffff6zxKF+eTJk1m9ejXPP/88CxYs4LjjjgNg3bp1AIwaNYodd9wRgEmTJqVdVz0ys3cDu7n7/mb2T8D1wE3A4+5+qZn9GDgYuKuScYpIY1M9IcVUrEfCa8w3kepRz/WEkkEiUrfGjx/PjBkz1mfxAa688kruvPNOAA4++GAGBgZSzvuOd7yDxYsXs88++/DMM8+kXcfAwAA777wzBx10EOeeey4A3d3hPuampqZNpm1paaGvr6+g7aoR/wA6zGw4MAF4CzgQ+K/4+R3AB1AySEQqSPWEFFOxHgmvMd9Eqkc91xNKBolIXTv99NOB8EQAgBkzZjB79mx22WUXxo0bl3a+o446irPOOosHHniAd77zneuvBKQya9YsHn/88fWZ/GnTpnHeeeelnHavvfbi+uuvZ/HixXzjG9+gtbU1302rdmuBvwEOjAY+AXwVeDt+vgrYojKhiYhsoHpCREQyqdd6oildFqtczGzA3SsaQ7XpXvFa0a4qdLy4qODltO4+gxFbblfwcqQxvfTSS+u7NkL5nhJTqIGBAfr6+hg+fDi33347L7zwQtoCuVgGf1dmhrs3ZZilapnZR4CTgGOBbQg9gV4ALnL358zsYGCWu38li2U1ZD1RaF1QSB2gcl/KSfVE9uqpniimaqkn1IYXKY3ksq+cTxMrVLXXE+oZJCJlNb51s5IXvMXQ29vLCSecQFNTE83Nzfz3f/93pUOqNc3ASnfvN7M1wBjgIeBQ4Ln4/z0VjE9EqpTqifpjZi3APGB34GR3v8XMJgJzgbHAfe5+cZz2Y8CFwABwprs/VpGgRaQq1UodAdVfTygZJCKSQktLCzfccEOlw6hl9wKfMbP5wEjga8DtwDVm9hCwFPhDBeMTESmI6omc9AKfAk5Jeu/LwFXufrOZ3WlmUwi3Fn8NmAmMIzx4YEa5gxURKYZqryeUDBIRkaJz9z7guBQfHVXuWEREpLLcfQB43cyS354BzImv7yAkgPqBv7r7WmCtmbWY2Uh37yxrwCIiDaC50gGIiIiIiEjDGePuHfF14qECW7DhQQPJ74uISJEpGSQiIiIiIuXWbmYj4+vxwEpCImhC0jSJ90VEpMiUDBIRERERkXKbT3iYAMAh8e/ngV3MbIyZbQv06hYxEZHS0JhBIlJWvetW09/ZVtAymke2MnzM+LSfv/rqqxx55JHssssudHR0MGvWLE4//fSsl7906VL+/Oc/c8IJJxQUp4iI5E71RH0ys18D7wHazGw/4DvAdWZ2DnC/uy+J010M3Ed4mtjZFQpXRKpUMeoIUD0BBSSDYiH+daAFuAu4mhSPhxQRSdbf2UbbkgUFLaN19xmQofAGmD59OldeeSUDAwN84hOf4Pjjj2fcuHFZLX+33XZjt912KyhGERHJj+qJ+uTuqR4gcEiK6X4H/K70EYlILSpGHQGqJyDPZJCZbQZcBHzc3dvje5cy6PGQ7v5sEWMVEclZT08PAwMDDB8+nG9/+9ssXryY3t5ezj//fKZNm8Zhhx3GjBkzeOqpp9hjjz248MILWbhwIffccw9z5szh97//Pb/4xS/YYYcdWLt2Leeccw7t7e389Kc/ZcyYMbz44ovMmTOH/fbbr9KbKkJXdy9dPX0ADO/tY826rryWM9DeTUff2qynHzt6BONbN8trXSKVpnpCREQyqdd6It+eQe8DOoBbzGw48O+kfjykkkEiUhGLFi3iuOOO47XXXuODH/wgjz32GP39/cydO5eVK1dy5plnMnfuXDo6OjjmmGM4//zzOfzww2lr29DttK+vjyuuuIJbb72VYcOGcfjhh6//rL29nZ///Oc8++yz/OQnP1EjX6pCV08fr7wekjgjduhZ/zpXE7ZqY/GK7BNJM/faTskgqTmqJ0REJJN6ryfyTQZNBqYA+wI7AD9j08dD7lh4eCIi+Ul06+zv7+eMM85g8eLFLFiwgOOOOw6AdevWATBq1Ch23DEUV5MmTWL16tXrl7Fy5Uq22WYbRo0aBbBRV89dd92VpqYmJk2axKpVq8q1WSIiUiSqJ0REJJN6ryfyTQa9DTzs7uuApWY2HlhnZiPjiP96DKSIVIXm5mbGjh2LmdHZ2cm5554LQHd3NwBNTU0bTT8wMLD+9RZbbME//vEPOjs7GTZsGM8999z6zwbPJyIitUn1hIiIZFKv9US+yaCFwAVm1gxsDXSy4fGQtxEGg7ugKBGKSF1pHtkaBmwrcBlDSXTr7O3tZfLkyXzgAx9Y/x7AtGnTOO+88zIuY9iwYZxyyikcc8wxbLfddkycOJGWlpaCYhcRkcxUT4iISDrFqCMSyxlKvdcTTclZq1yY2UnAiYSniX0JWApcR3ia2P3uPifD7MnLGXD3vGKoV90rXivKCOmjdppOx4uLCl5O6+4zGLHldgUvRxrTSy+9tL7bZC3q6emhpaWF7u5ujjzySG644QbGjh1bknUN/q7MDHdv+EvLjVpP5FMXrFnXtX6coJ3fO4MXFuZXl0yYNpPFK7JvqMzcazu237o0x4XUP9UT2VM9kVq11BNqw4uUhuqJ7OVST+T9aHl3vwq4atDbmzweUkSklt1yyy3cddddtLW1cdRRR5Ws4BYRkdqkekJERDKp1noi72SQiEgjmD17NrNnz650GCIiUqVUT4iISCbVWk80VzoAEREREREREREpHyWDRKSkWlpaWLNmTaXDqHpr1qypmsHkRETKSfVEdlRPiEijUj2RnVzrCd0mJiIlNXnyZJYtW8aKFSsqHUpVa2lpYfLkyZUOQ0Sk7FRPZEf1hIg0KtUT2cm1nlAySERKqrm5me23377SYYiISJVSPSEiIpmonigNJYNERKQkzGw/4OtAC3AXcDUwFxgL3OfuF1cuOhERERGRxqVkkEiRrG7rYm17d1nWNXb0CMa3blaWdYnkw8w2Ay4CPu7u7fG9S4Gr3P1mM7vTzKa4+7MVDVREREREpAEpGSRSJGvbu3noydfKsq6Ze22nZJBUu/cBHcAtZjYc+HdgBjAnfn4HMBNQMkhEREREpMyUDBIRkVKYDEwB9gV2AH4GjHH3jvj5KmDHCsUmQ9i8tYWp9GQ9/aiu5XSv2PCUj+aRrQwfM74UoYmIiIhIESgZJCIipfA28LC7rwOWmtl4YJ2ZjXT3TmA8sLKiEUpaw/q6WPX0gqynHzdpLE1jNvRWbN19BigZJCIZmNkU4PL4ZyvQDPwA+E/g7wDuPqsiwYmINAAlg0REpBQWAheYWTOwNdAJzAcOBW4DDgEuqFx4IiJSSXHMuFkAZnYaMCF+dIW7/3el4hIRaRRKBsmQBvr66F5R+Fg4um1ApHG4+9tmdg3wIOFpYucCS4HrzOwc4H53X1LBEEVEpHrMBo4njCV3spl9HLjF3f+nsmGJiNQvJYNkSP3dHXS8uKjg5ei2AZHG4u5XAVcNevuQSsQiIiLVyczeCQxz9xfNbCUwl3CO8nsze9TdH61ogCIidUrJIBERkRLoXbea/s62vOfv7+4qYjQiIlXrGOBGAHdfFd/rM7PfAHsBSgaJiJSAkkEiIiIl0N/ZRtuS7AdhHmzUTtOLGI2ISNU6mthr1MzGu/tqM2si3DL284pGJiJSx5QMEhERkYIMDAywZt2GnkwD7d109K0t6TrHjh7B+NbNhp5QRKqWme0OrHD3N+Jb55rZR4AB4CF3v69y0YmI1Dclg0RERKQgvb39vL68ff3fE7ZqY/GK0t7mNnOv7ZQMEqlx8UECH076ew4wp3IRlU9Xdy9dPX0pPxve27dRgj1fqRLzSqSLSIKSQSIiIiIiImXU1dPHK6+n7kE5YoeetJ/lIlViXol0EUlQMkikBvX19/Pqm6W9BSNBV5BERERERETqi5JBIjWoo7OXJ/ytsqxLV5BERERERETqS3OlAxARERERERERkfIpqGeQmc0A5gNbxbfmAmOB+9z94sJCqz2961bT39lW8HL6u0s76KaIiIiIiIiINK5CbxM7G3g8vv4ycJW732xmd5rZFHd/tsDl15T+zjbaliwoeDmjdppehGhERERERERERDaV921iZvYxYAGwLr41A7gjvr4DmFlYaCIiIiIiIiIiUmx5JYPMrBn4InBF0ttj3L0jvl4FbFFgbCIiIiIiIiIiUmT53iZ2LPA7d+80s8R77WY20t07gfHAymIEKFKI1W1drG3vLsu6urr7yrIeERERERERkULkmwzaA9jHzD4OTANuIAwkfShwG3AIcEFRIhQpwNr2bh568rWyrGtv22roiUREREREREQqLK9kkLt/OfHazOYRego1AdeZ2TnA/e6+pCgRioiIiIiIiIhI0RT6NDHcfVbSn4cUujwRERERERERESmdgpNBIiIi6ZjZDMJtxIn7KOcCY4H73P3iSsUlIiJS7zZvbWEqPRu9N6prOd0r1uS0nOaRrQwfM76YoYlIFVAySERESuls4PH4+svAVe5+s5ndaWZT3P3ZCsYmIiJSt4b1dbHq6QUbvTdu0liaxmyW03Jad58BSgaJ1J28Hi0vIiIyFDP7GLAAWBffmgHcEV/fAcysRFwiIiIiIo1OPYOkbAb6+uheUfiTvdRVVaT6mVkz8EXgk8AR8e0x7t4RX68CdqxEbCIiUh3MbB3wWPzzW8CDwDXAZGAxcJq791cmuvo0MDDAmnVduc3T3k1H39q81zl29AjGt+bWG0lESk/JICmb/u4OOl5cVPBy1FVVpCYcC/zO3TvNLPFeu5mNdPdOYDywsmLRiYhINXgp+WE0ZnYa8Li7X2pmPwYOBu6qVHD1qLe3n9eXt+c0z4St2li8IrcEUrKZe22nZJBIFdJtYiIiUgp7AJ8ys7uBacANhIGkD42fHxL/FhGRxvVPZvaQmV1vZlsCB6LbiUVEykI9g0REpOjc/cuJ12Y2j9BTqAm4zszOAe539yUVCk9ERKrDu9x9uZmdBFwCbAG8HT9bFf8WEZESUDJIRERKKvkWAEKPIBEREdx9eXx5I3AK8DIwAXgD3U4sIlJSSgaJiIhIUW3e2sJUevKev7N5JC+81VfEiESk2pjZGKDT3fuADwDPA38i3E78XPz/nspFKCJS35QMEhERkaIa1tfFqqcX5D3/hGkzgZbiBSQi1WhX4OdmtgboAk4GlgPXmNlDwFLgDxWMT0SkrikZJCIiIiIiZeXufwH2SvHRUeWORUSkESkZBPSuW01/Z1vBy+nvzv+RiyIiIiIiIiIi5aBkENDf2Ubbkvy7syeM2ml6EaIRERERERERESmd5koHICIiIiIiIiIi5aNkkIiIiIiIiIhIA1EySERERERERESkgWjMIBEREak5ff39vPrm2rKvd+zoEYxv3azs6xUREREpJiWDREREpOZ0dPbyhL9V9vXO3Gs7JYNERESk5ikZJCIiUmSr27robe9m7bquvJcxvLePNTnO39c3kPf6RERERKRxKBkkIiJSZGvbu1n5jzZWvZ7/bUwjdujhlRznnzRxdN7rExEREZHGoWSQiIiIiIiIlITGeBOpTnklg8zs/cD3gG6gDfhMXNZcYCxwn7tfXKQYRUREREREpAZpjDeR6pRvz6C/AR9y93YzOxU4DZgAXOXuN5vZnWY2xd2fLVqkKfSuW01/Z1vBy+nvzn9MB5F6V86rObqCIyIiIiIiUnp5JYPcfVnSn91ALzADmBPfuwOYCZQ0GdTf2UbbkgUFL2fUTtOLEI1IfSrn1RxdwRERERERESm9gsYMMrMtgS8CBwOfcfeO+NEqYMcCYxMRERERESmZYt1pALrbQERqS97JIDMbDdwMnOHuy82s3cxGunsnMB5YWawgRUSktmhsORERqQXFutMAdLeBiNSWfAeQHg7cCPzQ3R+Jb88HDgVuAw4BLihKhFJ3urp76erpy3v+gfZuOvqyG8Omqzv/9Uj5aXyiulIVY8uJiEh1SnPR4Czg08BbwBvufkyp4+jq7mPNuuL06Bnem/2y+voGirJOEZF85dszaDZhTKBxZnYmcCfwHeA6MzsHuN/dlxQpRqkzXT19vPJ6/if8E7ZqY/GK7CravW2rvNcj5afxiepHtYwtJyIiVSvVRQOAi9z9lnIF0d1bWLs02YgderJe1qSJo4uyThGRfOU7gPRcQlf/wQ4pLByRoW3e2sJUerKadkL/KqZumXrazuaRvPCWeg6JlJLGlhMRkVTSXDRoAf4jXmy+3N1/VZHgREQaQEEDSItUwrC+LlY9nd293aMmjmbV8vaUn02YNpPQ5hCRUtDYciIiMpRBFw0G3P1iMxsP3G9mC9z975WNUESkPjVXOgAREak/Q4wtB6En6fxKxCYiItVh8EUDd18B4O6rgf8FplYyPhGReqaeQSIiUgoaW05ERNJKddHAzMa7++r42fuBn1cyRglyGaIhleRhGzRMg0j1UDJIGlahFVuCKjWRTWlsORERGUKqiwZmZlOAYcAN7v7XSgYoQS5DNKSSPGyDhmkQqR5KBknDKrRiS1ClJiIiIpKbDBcNRESkDDRmkIiIiIiIiIhIA1EySERERERERESkgeg2MRERkTR6162mv7Mt5/lGdXUzYXQTq0oQk4iIiIhIoZQMEhERSaO/s422JbmPLbZ2XRcjdtizBBGJiIiIiBROySARaUh9/f28+ubasqxr7OgRjG/drCzrEhEREZHytvWSqd0ntULJIBFpSB2dvTzhb5VlXTP32k6NApEcbN7awlR6Mk4zoX8VU7fcdJrO5pG88FZfqUITEZEaUc62XjK1+6RWKBkkIiIiVWVYXxerns58e96oiaNZtbx9k/cnTJsJtJQoMhEREZH6oGSQiIiISJYqddvB8GHN9Pb1l329ut1BRIopm56f2VAvUJHCKRkk63V199LVs2mhOry3jzXrugpefmI5fX0DBS9LRESkEip128HetpVudxCRmpdNz89sqBeoSOGUDJL1unr6eOX1Ta92jtihJ+X7uUosZ9LE0QUvS0RERERERETy01zpAEREREREREREpHzUM0hEROpW77rV9He25T1/f3fht8iKiIhI49Aj7aVWKBkkIiJ1q2PNalYsmpf3/Fvs9p68xkzT2GgiIiKNSY+0l1qhZJCIiNSt7t7UY6FlK98x0zQ2moiISOlk81SyCf2rmLpl5mn0VDJpZEoGiRQoURllU+FkospIRERERGRo2TyVbNTE0axa3p5xGj2VTBpZ0ZNBZvZ54ESgGzjJ3V8s9jpEqkmiMsqmwslElZE0CtUTIrVDY19IJaiekHLJpodRNnRRV2pRUZNBZrYFcDJwALAX8C3gqGKuo9F0dffS1ZNfwTK8ty+nsS40xoWIlJrqCal2O281jJH9nWk/H6oXaL2dEGjsCyk31RNSTtn0MMqGLupKLSp2z6D9gHnu3gs8ZmZW5OVXhXQJmlyTL+kkL6evb4DX3szvSTi5jnWhMS5EpAwaop6Qyin0Ku+Ekb28/OhDaT8fqheoTgiKo1I9koYPa6a3r7+s61QvqE2onhARKYNiJ4O2AN5O+ru5yMuvCl09qQckzXeg0UzLUYJGROpMQ9QTUjmFXuWd+N4ZRYxG8lWpHkl721ZlX696QW1C9YTUnOQLEYWOI9o6dhRtaztynm9U13K6V6xZ/3fzyFaGjxmfdxzZWt3Wxdr27pKvZ7BKJNLrbVubBgaKd2uQmR0CfMDdvxL/XuTu04eYR/cmiYik4e5NlY6hmFRPiIgUl+oJ1RMiIpmkqyeK3TNoITDHzIYBewLP5xuYiIjUJdUTIiKSieoJEZEyKGoyyN1Xmtm1wHygB/hcMZcvIiK1TfWEiIhkonpCRKQ8inqbmIiIiIiIiIiIVDcNyCYiIiIiIiIi0kCUDBIRERERERERaSBKBomIiIiIiIiINBAlg0REREREREREGkixHy1fVGb2eeBEoBs4yd1fTPpsX+D7QBNwibvfEd+/GPgwsBY4zt2XlzvuVCciSQQAAAgISURBVHLdFjN7J/AX4Jk42Vnuvqi8Uac3xPb8CDgSuNXdT4/vNQM/BqYCy4DPuntH2QNPIY9tmQVcz4ZHnR7j7m+UNeg00m2LmU0AbgVGEPazM9z9iVr8XTJsyyyq9HeBIfezW4GJwGjgO+5+c3z/YqqwPGtkue6XFQu0CPLZZ2tVpm2Nn48H/g841d1vqUCIRTPE7zoPaCE8welud/9WRYIskiG2dWtC/bcl8Ia7H1uZKKVa5HnecRhwDjAM+IW7z62G2MysBbgB2Da+f6q7L65AbBU9J8gjtlOAk4Be4DF3P6sUceUTW9JnFwKz3X1qNcVmZicDxxDyC5e4+73VEFtsn/2a0D7rB/7V3ZeVIrZM8VXruVjV9gwysy2Ak4GZwHnA4AbJZcCngX8GvmZmw8xsd2Bfd58BXAV8qYwhp5XPtsT3/+Tus+K/akoEDbU93wAGN6oOAfrd/UDgccJBUnF5bgvA7Um/TVUkHIbYli7g+Pj9nwx8O75fi79Lum2BKvxdIKv9bLa7fwA4CPhqnKcqy7NGVsB+WXPy2WdrVRbbCvDvwMJyxlUKWW7rEbEMrfVE0FDb+l3gK+5+kBJBkud5x1aEdtNH4jFTqkRQPucRs4A1sU46nxK1Iar5nCDP2O4D3ufuBwDbmtkBVRRbYr4ppYipkNjMbA9Cm/XD8VgoVSIon+/t08B8d58FXA2cWorYsoivKs/FqjYZBOwHzHP3Xnd/DLDEB2Y2Ehju7q+5exvwV+DdwIHAHXGyO+Lf1SCfbQHY18zmm9nlZjaq/GGnlXZ7AGK2dWDQPIN/m5kljzI7+WwLwMfMbIGZXWJmTeUINAtpt8XdO9z9tfhnN+GKB9Tg75JhW6A6fxcYej/rji9bgSXxdbWWZ40s3/2yFuWzz9aqjNtqZtsAOwGPVSK4Isu4rYT67nYzu8fMppc/vKLK1PYaBuwKXGxmD5rZ0ZUKUqpGPm31Q4E24A9m9jsz26GKYnuR0AMBYALwVrljg4qfE+Qcm7v/n7sn3itlXZ7v+cdXCInsUsontk8CfWZ2n5ndYGabV1FsS4Gx8XUpj4WM8VXruVg1J4O2AN5O+rt50Gerkv5eFd9bP0/sXtVa4hizlc+2vA7sHLOEy4EzSx1kDjJtTzbzJLaxGuSzLY8TDu6ZwGSgWhqRQ25LTJB8D/hOinlq6ndJsS3V+rtAdttzL/AUcPfgeaqsPGtk+eyXtSqffbZWDbWtF1LjPb2SDLWtn47tjnMIV1BrWaZt3RrYE/g64YT+K/GKrjSufNrqk4F/IuxDPwAuraLYXgVazWwpcHn8V+7YspmnlG3PfGIDwMz+f3v37yJXFYZx/JsttjASUyWuCqKRvARMkRgIQiCFihb+B/4qQoiIjYUpXANqk06MIqlsJEHBzkKCyMomaYKdroEXBU0RrDRCkjWSNWvxnsXZuHN37mHOzLnc5wPLDLNzd993zr0z97xzzzlPAjvdvdQVoa1jM7OHgDkvPwQ953V7ANjq7k8DC0TRqoSc2JaAQ2b2A/A6cKZEYEnn+mI1F4OuEdW7Nf80/O4+4I/Bx1Ol/EbhGEfVOhd3/9vdr6fHPgf2lQ2xlaZ8Rtlmrb1q0DoXd7+R2ucO8AX1tM0ouZwiKtbfbrBN19plXS4VtwuMkI+7PwPsBt5M85PU+n7WZznHWFfl7LNdNTRXM3sE2O7u3088qjIa29XTvGTu/iNwu7Krktva7NzrioebxByNj00yOKlObr9jIV0F8A2wp6LYXgF+cvc9wHPA6SnENso2Jc89c2LDzHYTnfUXSwSV5MR2AjhZJpx1ctv063T/HLB33EEN/J+2sR0Hzrj7XuANyn5Z17m+WM3FoEvA4TQmdz//TQq79i35ipnNmdlW4nLIn4HzxLg7iCr9xQnHPEzrXMxs28D2hwe3qcDQfBqcJ9qEdHuhVHAttc7lrk5PTW3TmIuZvQWsuPsHAw93sl02yqXidoGGfMxsJk30CLAM3Eo/tb6f9VnOMdZVOftsVzW16z5gl5mdIzoG82k+r67abB/elm7vJ77lrWJBgUxN5163gKtmtjMNGXscuDKlOKUOOf2ORWA/QBpWWWofyolthhhZAP/voE4ktgaTOvfMOcefI64cecndSw4nynndHgXeT59HD5tZqcJQTmyLwBPp/gFiwYVaYhs8Fn4HSg1hgw72xbasrm40HLEOZvYq8DKxqsUR4BDwi7svmtlBYsK0LcBJd/8ybfMeMaHldWKSppIH8sja5mJmzwPvElcDXCNmFv9z478+eZvkM0+MHd0BXHb3Zy1mSj9NTHr2G5HP8nSiXy8jl2PAUeAv4s3uqLvfnk706w3LhRg7/itRUFgFrrr7C11sF4bnUm27QGM+3wFfpafNAh+7+9m0TZXvZ33Wdr+cUphjkbPPdlXT58DAc94Blrz7q4kNa9cLRNsuE6vBzLv7wrTiHIdNPt8PEN/+zwJn3f2j6UUqNcjsd7xNTNw8Axxz98s1xGZm9wKfEXOl3EMcz6Um9a22T5AR2yfAU8TnOZRdFatVbHdtu+RlVxPLadMPicL6CnHOWmTFrozYHgQ+JVb8mwVe84ILM3WtL1Z1MUhERERERERERMar5mFiIiIiIiIiIiIyZioGiYiIiIiIiIj0iIpBIiIiIiIiIiI9omKQiIiIiIiIiEiPqBgkIiIiIiIiItIjKgaJiIiIiIiIiPSIikEiIiIiIiIiIj2iYpCIiIiIiIiISI/8C48CTR0RpePTAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 1440x2160 with 30 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Xgnm8f9ZfOCi"
      },
      "source": [
        "So we can see several interesting aspects to the data from these comparitive histograms.\n",
        "\n",
        "Firstly, there are some features that seem to have a significant difference between the distributions of malignant v benign, these are:\n",
        "\n",
        "- perimeter_mean\n",
        "- area_mean\n",
        "- concave points_mean\n",
        "- radius_worst\n",
        "- perimeter_worst\n",
        "- concave points_worst\n",
        "\n",
        "These might proove useful features for our model to help better predict & distinguish between and predict malignant v benign tumors. \n",
        "\n",
        "### 5.2 Check normality of key distributions\n",
        "\n",
        "Let us just check the normality of one of the distributions that appears to be more normal and one of the more skewed distributions."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2fq6yJHqDzjE",
        "outputId": "f8c0402f-490e-491d-c6c0-e728297e17c7"
      },
      "source": [
        "features = ['area_mean', 'symmetry_mean']\n",
        "\n",
        "for feature_name in features:\n",
        "\n",
        "  print(' ')\n",
        "  print('Testing ' + feature_name + ' normality')\n",
        "  print(' ')\n",
        "\n",
        "  print('Benign subgroup')\n",
        "  print(' ')\n",
        "  feature = benign_df[feature_name]\n",
        "  print('shapiro-wilk test')\n",
        "  # normality test\n",
        "  stat, p = shapiro(feature) \n",
        "  print('Statistics=%.3f, p=%.3f' % (stat, p)) \n",
        "  # interpret\n",
        "  alpha = 0.05\n",
        "  if p > alpha:\n",
        "    print('Sample looks Gaussian (fail to reject H0)') \n",
        "  else:\n",
        "    print('Sample does not look Gaussian (reject H0)')\n",
        "\n",
        "  print('dagostino and pearsons test')\n",
        "  # normality test\n",
        "  stat, p = normaltest(feature) \n",
        "  print('Statistics=%.3f, p=%.3f' % (stat, p)) \n",
        "  if p > alpha:\n",
        "    print('Sample looks Gaussian (fail to reject H0)') \n",
        "  else:\n",
        "    print('Sample does not look Gaussian (reject H0)')\n",
        "\n",
        "\n",
        "  print('anderson test')\n",
        "  result = anderson(feature)\n",
        "  print('Statistic: %.3f' % result.statistic) \n",
        "  p=0\n",
        "  for i in range(len(result.critical_values)):\n",
        "    sl, cv = result.significance_level[i], result.critical_values[i]\n",
        "    if result.statistic < result.critical_values[i]:\n",
        "      print('%.3f: %.3f, data looks normal (fail to reject H0)' % (sl, cv)) \n",
        "    else:\n",
        "      print('%.3f: %.3f, data does not look normal (reject H0)' % (sl, cv))\n",
        "\n",
        "  print(' ')\n",
        "  print('Malignant subgroup')\n",
        "  print(' ')\n",
        "  feature = malignent_df[feature_name]\n",
        "  print('shapiro-wilk test')\n",
        "  # normality test\n",
        "  stat, p = shapiro(feature) \n",
        "  print('Statistics=%.3f, p=%.3f' % (stat, p)) \n",
        "  # interpret\n",
        "  alpha = 0.05\n",
        "  if p > alpha:\n",
        "    print('Sample looks Gaussian (fail to reject H0)') \n",
        "  else:\n",
        "    print('Sample does not look Gaussian (reject H0)')\n",
        "\n",
        "  print('dagostino and pearsons test')\n",
        "  # normality test\n",
        "  stat, p = normaltest(feature) \n",
        "  print('Statistics=%.3f, p=%.3f' % (stat, p)) \n",
        "  if p > alpha:\n",
        "    print('Sample looks Gaussian (fail to reject H0)') \n",
        "  else:\n",
        "    print('Sample does not look Gaussian (reject H0)')\n",
        "\n",
        "\n",
        "  print('anderson test')\n",
        "  result = anderson(feature)\n",
        "  print('Statistic: %.3f' % result.statistic) \n",
        "  p=0\n",
        "  for i in range(len(result.critical_values)):\n",
        "    sl, cv = result.significance_level[i], result.critical_values[i]\n",
        "    if result.statistic < result.critical_values[i]:\n",
        "      print('%.3f: %.3f, data looks normal (fail to reject H0)' % (sl, cv)) \n",
        "    else:\n",
        "      print('%.3f: %.3f, data does not look normal (reject H0)' % (sl, cv))"
      ],
      "execution_count": 141,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            " \n",
            "Testing area_mean normality\n",
            " \n",
            "Benign subgroup\n",
            " \n",
            "shapiro-wilk test\n",
            "Statistics=0.991, p=0.023\n",
            "Sample does not look Gaussian (reject H0)\n",
            "dagostino and pearsons test\n",
            "Statistics=8.255, p=0.016\n",
            "Sample does not look Gaussian (reject H0)\n",
            "anderson test\n",
            "Statistic: 0.460\n",
            "15.000: 0.570, data looks normal (fail to reject H0)\n",
            "10.000: 0.649, data looks normal (fail to reject H0)\n",
            "5.000: 0.778, data looks normal (fail to reject H0)\n",
            "2.500: 0.908, data looks normal (fail to reject H0)\n",
            "1.000: 1.080, data looks normal (fail to reject H0)\n",
            " \n",
            "Malignant subgroup\n",
            " \n",
            "shapiro-wilk test\n",
            "Statistics=0.933, p=0.000\n",
            "Sample does not look Gaussian (reject H0)\n",
            "dagostino and pearsons test\n",
            "Statistics=46.927, p=0.000\n",
            "Sample does not look Gaussian (reject H0)\n",
            "anderson test\n",
            "Statistic: 2.267\n",
            "15.000: 0.566, data does not look normal (reject H0)\n",
            "10.000: 0.644, data does not look normal (reject H0)\n",
            "5.000: 0.773, data does not look normal (reject H0)\n",
            "2.500: 0.901, data does not look normal (reject H0)\n",
            "1.000: 1.072, data does not look normal (reject H0)\n",
            " \n",
            "Testing symmetry_mean normality\n",
            " \n",
            "Benign subgroup\n",
            " \n",
            "shapiro-wilk test\n",
            "Statistics=0.974, p=0.000\n",
            "Sample does not look Gaussian (reject H0)\n",
            "dagostino and pearsons test\n",
            "Statistics=34.254, p=0.000\n",
            "Sample does not look Gaussian (reject H0)\n",
            "anderson test\n",
            "Statistic: 1.534\n",
            "15.000: 0.570, data does not look normal (reject H0)\n",
            "10.000: 0.649, data does not look normal (reject H0)\n",
            "5.000: 0.778, data does not look normal (reject H0)\n",
            "2.500: 0.908, data does not look normal (reject H0)\n",
            "1.000: 1.080, data does not look normal (reject H0)\n",
            " \n",
            "Malignant subgroup\n",
            " \n",
            "shapiro-wilk test\n",
            "Statistics=0.965, p=0.000\n",
            "Sample does not look Gaussian (reject H0)\n",
            "dagostino and pearsons test\n",
            "Statistics=27.185, p=0.000\n",
            "Sample does not look Gaussian (reject H0)\n",
            "anderson test\n",
            "Statistic: 1.452\n",
            "15.000: 0.566, data does not look normal (reject H0)\n",
            "10.000: 0.644, data does not look normal (reject H0)\n",
            "5.000: 0.773, data does not look normal (reject H0)\n",
            "2.500: 0.901, data does not look normal (reject H0)\n",
            "1.000: 1.072, data does not look normal (reject H0)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "c0tWiqkBtO_I",
        "outputId": "6ad1686f-0dd7-4b59-89dd-b8d25bcbdd23"
      },
      "source": [
        "features = ['area_mean', 'symmetry_mean']\n",
        "\n",
        "for feature_name in features:\n",
        "\n",
        "  print(' ')\n",
        "  print('Testing ' + feature_name + ' normality with QQ Plot')\n",
        "  print(' ')\n",
        "\n",
        "  print('Benign subgroup')\n",
        "  print(' ')\n",
        "  feature = benign_df[feature_name]\n",
        "  # q-q plot\n",
        "  qqplot(feature, line='s')\n",
        "  plt.show()\n",
        "\n",
        "  print(' ')\n",
        "  print('Malignant subgroup')\n",
        "  print(' ')\n",
        "  feature = malignent_df[feature_name]\n",
        "  # q-q plot\n",
        "  qqplot(feature, line='s')\n",
        "  plt.show()"
      ],
      "execution_count": 142,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            " \n",
            "Testing area_mean normality with QQ Plot\n",
            " \n",
            "Benign subgroup\n",
            " \n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            " \n",
            "Malignant subgroup\n",
            " \n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            " \n",
            "Testing symmetry_mean normality with QQ Plot\n",
            " \n",
            "Benign subgroup\n",
            " \n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEHCAYAAAC9TnFRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXhV9bn3//dOSJgTBESGyCjcCIiCSlXAoWqdh+rjUD3twam2PT129NfHau1zjj3W0/PU1qp9jtpWPKfHIji0VqlWrFgUhKDIFLhRiYLMGZiRkGT//lhrwybsJCuQlfHzuq5e7L3WXmvfoV758B3W95tIJpOIiIjUJ6u5CxARkdZBgSEiIpEoMEREJBIFhoiIRKLAEBGRSBQYIiISSYfmLiBOZqY5wyIiDeTuiUzH23RgALh7c5cgItJqmFmt59QlJSIikSgwREQkEgWGiIhEosAQEZFIYh30NrPbgJuACuBmd18dHu8BPAfkAgngDnd/z8xuAL4aXl4AvOju3zWzqcAJwA7gfXf/dpx1i4jIoWILDDPrCdwKTATGAQ8A14an9wJfcfd1ZjYSeBg4392fBp4Or59BECopt7v7wrjqFRFpL4qKS9lQsov+vbtx/JCeka+Ls4UxAZjt7pVAoaXN1XL3PcC68G0FUJl+oZl1B8YAc9MOP2pme4D73P31GOsWEWmzHnt+Ca8VriEBJIHzTx3I7VeNjXRtnGMYPYHyur7LzBLAg8DPapy6CviTu6cevPu+u38OuAF42Mw6x1CviEibVlRcymuFa9hbUcVnFVXsrajitcI1rCgui3R9nIFRDvRIe1+V4TMPEbRC3qhx/EvAH1Jv3L0k/HM9sAwY3KiVioi0AxtKdlHzEe4EsL5kZ6Tr4wyM+cBZZpZtZuOBD9JPmtkPgUp3/2WN432Afu6+OO1YfvhnV2AU8GmMdYuItEn9enel5npJSaB/726Rro9tDMPdy8zsKWAOsA+4xcymAMXAauA+4C0zmw2sc/cbw0uvAWbUuN0fwtDoQDCGsSOuukVE2qpRQ3px/qkDDx7DmDAw8sB3oi3v6W1mSa0lJSJysBXFZawv2ZlxlpSZtd/FB0VE5GDHD+nZoOm0KXrSW0REIlFgiIhIJAoMERGJRIEhIiKRKDBERCQSBYaIiESiwBARkUgUGCIiEokCQ0REItGT3iIirdThboR0uBQYIiKt0JFshHS41CUlItLKHOlGSIdLgSEi0soc6UZIh0uBISLSyhzpRkiHS4EhItLKpDZC6pibTafcbDrmZjdoI6TDpUFvEZEWqq5ZULdfNZYzxxXUuhFSHBQYIiJNKFMIZDoWZRbU4W6EdLgUGCIiTSRTCACHHJs8bsD+WVAprxWu4cxxBU0aEDUpMEREmkD6VNiUV+d/AsC+yur9x14rXEOXTjm1zoJqzsDQoLeISBPINBWWZJKa052CzySbZRZUfWJtYZjZbcBNQAVws7uvDo/3AJ4Dcgn+fu5w9/fM7Gzgf4APwltc7+4bzWwY8Nvw80+6+xNx1i0i0tgyTYUlcUiEkAROOb4vuz+rPLirqglmQdUntsAws57ArcBEYBzwAHBteHov8BV3X2dmI4GHgfPDcy+4+zdr3O4B4E7gPWCumT3n7vE+0igi0ohSU2FrhgBJMgbD8UN6NvksqPrE2cKYAMx290qg0MwsdcLd9wDrwrcVQGXadZea2UnAm8A97p4ERrh7IYCZzQ7v/UqMtYuINLrapsLWFgxNPQuqPnEGRk+gPO39IeMlZpYAHgR+Fh5aCBiwj6AL6jpgGhzU9bc1vLeISKuTKQRaWjDUJs5B73KgR9r7qgyfeYigFfIGgLvvdPe97l4NzCDoyoKDh4XyAXVHiYg0sTgDYz5wlpllm9l4DgxkA2BmPwQq3f2Xacfy0z5yVto1q8xsvJllh8cXxFi3iEijKCou5fUmWEW2qcTWJeXuZWb2FDCHoIvpFjObAhQDq4H7gLfCMYl17n4jcH04s2oP8BFwT3i7uwi6qHKAqRrwFpGWrjn2q4hbIpk8ZKJXm2FmSXdv7jJEpJ0pKi7l3sfnHfSQXsfcbO776hktfqzCzHD3Q+f7ogf3REQaXXPtVxE3BYaISCNrrv0q4qbAEBFpZM21X0XctPigiEgMmmO/iih2frSadX98kb4Xnk/+6NENulaBISLSiGrubdGSgmLttOmULXyX3hNPp/OAggbfQ4EhItJIWuJU2ppBMe6hB+ky8NjDupcCQ0SkEWTa76I5Nz06KCgmnXFEQZGiwBAROQzpXU9JkrxeuIaaz7U1x6ZHOz/8iDXTplP+7ntBUPzqF3Q5tuHdT5koMEREGii966misooECbKzElSk7ZwHTTuVNs6gSFFgiIg0QKauJ0hSVX2gdZGbk0UikWiSqbRNERQpCgwRkQbIuNVqmtwOWZw9voDzTh0Ua1gcHBQTYw2KFAWGiEhERcWlrN+y86DWRE2JrESsYbHjgw9Z+8x0yt9d1GRBkaLAEBGpR1FxKU+/upKi4jKysxJUVVeTnZUgp0PW/jGMnA5Zse693ZxBkaLAEBGpw2PPL+HV+Z+wLxzQ3hcez+mQ4PLJwzjl+GMAYnui+5CgePgXdClo2qBIUWCIiNQiNcC9r8bsJ4DsrAT9j+66PyDiDIqjJ09q1qBIUWCIiNSirgHuuKbM7vjgQ9ZOm075ey0nKFIUGCIitci0TDkEM6Eae6yiJQdFigJDRKQWqWXKUw/pVVUnGTWkJzdecHyjhUVrCIoUBYaISB3iWqa8NQVFigJDRKQejblM+UFBceYkxj38S7oUDGiUe8dNgSEi0gR2rPogCIpF77e6oEiJNTDM7DbgJqACuNndV4fHewDPAbkECzre4e7vmdntwM1AJVDo7t8OPz8byCGYAv2Kuz8QZ90iIo2lLQRFSmyBYWY9gVuBicA44AHg2vD0XuAr7r7OzEYCDwPnA7OAx909aWbTzGyiu78dXnOFu5fEVa+ISGNqS0GREmcLYwIw290rgUIzs9QJd98DrAvfVhC0KHD3j9Ku33+cYMrzC2a2G/iBu78fY90iIodstRrVwUExmfGPPETnAf1jrLTp1BsYZnYm8K677zKzfwROBB5JdS/VoSdQnvY+K8O9E8CDwM9qHD8dOMbd54eHrnH3EjMbDfyeoMUiIhKLw9lqdYevCp7MXrS4zQVFyiG/xDN4OAyLU4GvA+8DT0W4rhzokfa+KsNnHiJohbyROmBmIwhC5B9Sx1JdUe6+HNhnZp0jfL+ISIPNnFvMK+98wt6KKj6rqGJvRRWvFa5hRXFZxs/v8FUU/etPWPK/76ZD9zzGP/IQI75zR5sLC4jWJZV60PEq4Nfu/l9m9p0I180H7jWzbIJWyQfpJ83sh0Clu/8y7Vg/ghbEDe6+Je14nrtvN7O+QNewS0tEpNGkVqRd9lHpIcuXZ9pqtT20KGqKEhgbzOxJ4BxgvJl1BLLru8jdy8zsKWAOweymW8xsClAMrAbuA94KZ0Ctc/cbgZ8AfYDfhEMe/wa8DrwRjl90AP65QT+hiEg9aq5IW1P6ulEHBcVZZ7aLoEhJ1Ny0vCYz6wpcACx19w/CVsAJ7v7XpijwSJhZ0t2buwwRacFmzi3miT8uo7Iqc1h0yM7iwtMHccOoTqyZNp2t7wdBcey1V9O5f9sLCjPD3TOuuRilhVEFDAHOAr4FdGnE2kREms39Uxcwf9lGqmv5h3OH7ARfPzWPAUteZMl/hS2KRx9qk0ERRZTAeAooBM4N328geOiuxbcwRERq89OpC5i3dEOt5wftK+HSPavo+vRqctp5UKRECYxh7n6dmd0I4O670x6pEBFpdYqKS1lQtDHjuYK9JVywp4ijy9bQ5+wzKbjnW+0+KFKiBMZuM8snnC1lZmOAnbFWJSISo3dXbKJmL1T/PVuYVL6YIXs2csw5Z1JwzfcUFDVECYzvAX8CBpvZKwTjGTfEWpWISExSM6JSU2dTQTF49wY2DxzNyXfdpaCoRb2B4e6FZnY+YATTkVe6+756LhMRaXHS9+hOD4qivKEs/MJtfOefLmjuElu0WgPDzL5Qy6l+4bQrDXqLSKuyoWQX/Xdv5tTNixi8ewPLuw/lySFXcs4F47n9olHNXV6LV1cL40t1nEuiWVIi0oos+dsCPntmBtdsLGZZ96E8PvAKtubm0TE3m1NG9m3u8lqFWgPD3W9qykJERBpbUXEpy94oJPvvr9KnfC1rug9lbhgU2VkJOnbI4vwJAxttN722rq4uqW+6+yNm9lMOrCe1n7v/MNbKREQOQ2pZ8oWvvcOAJX9n0J6NLOs+lD+GQZHulsvGcNEZg5un0Faori6pT8I/V2Y4V/d6IiIizeD+qQtY9+5STt+yiDPCoHg8Q1AAZCUS5OZEWbBbUurqkvpz+LK3u/88/ZyZfS/WqkREGujRnz/PsYV/Y3IqKAZdydac7rV+PpE4sKCgRBPlOYwbgZ/XOPblDMdERJpUUXEpGxcto/K1lxm7fnWkoADIzkrwhdMGaeyigeoaw7iFYE/uEWY2N+1Ud2Bp3IWJiNSmqLiU55/6K0P8bQbu3oh3H8rc+loUwLCCfMbbMZxy/DEKi8NQVwvjWYK9KH4C3J12fIe7Z956SkQkZv/9ny+RePMVztpdf9dTIgFZCRjSP5+vXjlWIXGE6t0PAyDcErU/aRsnufuqGOtqFNoPQ6Tt2L5iJSue/D0VvpKl3Ycxr+cJdbYoBhzdlWvOHUH/3t0UFA1wRPthmNldBHt5r+bA7Kgk8PlGq1BEpBbbV6xkzR+eYdvSZRT3Gckrg65kWz1jFB2yE3zruvEKikYWZdB7CmDaR1tEmlJ6UPT5/NksOO82/ra6/l9D2VkJLjx9sMIiBlECw4GOgAJDRGK3vWgFa6ZNZ9vSZWwffhLFF32NTXRh0arNtV4zZlgvRg3pCcmEBrRjFCUwcoGVZvY2sDd10N21xLmINJpUUGxftpz1BaP507FXsLWqO/hOatuCJ5GAr191op7WbiJRAuOnsVchIu3W9qIVrHjy9+z7cBXrBozmjeO+yLqqLpGuPX1MP4VFE4qyH8abTVGIiLQvqaCo+GAVS7sPZe6xVwSD2VXRrj+uIJ+7pkyIt0g5SJRZUmcADwKpxeK7AhvdfUCEa28DbgIqgJvdfXV4vAfwHEF3VwK4w93fM7Ms4FFgDLAemOLue8xsGPDb8PNPuvsTDfsxRaSl2La8iLXTplO+dHkQFAOvqHfWU00dshN89cqxMVUotYmy8tbDwNXAR+6eB1wHzKjvIjPrSfCk+JnAncADaaf3Al9x98nhZ/49PH4RUB0eX0gQNoTX3glMBm4N7y0irci25UXMv/Nult7zYxas38djA6/gL33OaHBYaBZU84kyhpF093Vmlg3g7s+a2T0RrpsAzHb3SqDQzCx1Ipyiuy58WwFUhq8nAy+Fr18Cfgz8Ghjh7oUAZjY7vPcrEWoQkWaW3qJY0n0o845tWIuid49OjB12dPi6s2ZBNaMogbHNzLoCb5nZE8Bm4LMI1/UEytPeH9KaMbMEQXfXzzJcszV8D0G3FRmOi0gLlQqKbcuWs7LnCN44jK6nM8b2465/1DhFSxElMK4k6EL6FsHKtXnApRGuKwfSOxkzDWU9RNAKeSPtmh7h63wgtWZV+vol6cdFpIXZtryI9x97iqw1q1l51HDebECLIr9bLv17d6Vf725ceJq6nVqaKIGxlwOtg6cbcO/5wL1hV9aJwAfpJ83sh0Clu/8y7fDfgYsJupsuBuaEx1eZ2XhgMXAWB1okItJCLJ41j3XPzKDb5jUsyxvGvAa2KNSaaPmiPumdJOgW6gj0AdYCQ+q8yL3MzJ4i+KW/D7jFzKYAxQTrUt1H0M01G1jn7jcCfwEuN7M5wAaCZUkA7iKYJZUDTNVquSItx7bly5nzi9+St2UNxXnHMW/QF9mWE31jouMKtJJsaxFptdp0ZjYBuMndvx5PSY1Hq9WKxOfVZ2ZR8deXyS9Zy9K845h31AkKijbgiFarrcndF5jZ7468LBFpTYqKS3l13sdUfLiKgSveYsDuTXx4GC2K/K653H3T5xQUrVCUB/d+yoFB5yyC8YiVcRYlIi1DKiQWrNjEUaVrmVS2hII9m1iadxwvDzqjQUHROTebmy4bo6U8WrEoLYz0cKgCXgbeiqccEWkp7p+6gHlLN3Dsno1cFgbFkrzjmNknelDkd8vFBh7FKcf3VVC0AVEC41lgWPja3X1vXR8Wkdbvjp+/QeVHq/hS2WIK9mxucFCMH9mH688zdTu1MbUGhpl1Inio7jrgY4JZUn3M7Nfufr+ZjXf395qmTBFpKt//zm85bcO7aUExsUFdT5dOGsLtX9Q6T21RXS2Mhwme6B7o7rsAzKwb8BMzmwpMAo6LvUIRiV1RcSlz//QmXd+ZxeVhULzcZyLbIwRFbocshhXk62G7dqCuwDjb3YenH3D3nWb2HYLlQc6MtTIRiV1RcSnPPP4SVjyfEXu2sCTvOF6KEBQam2ifDmdabdLMtrn7ijgKEpF4zZxbzBvvrqXyQ2fCxkWc+9mWSC2KY3p1YfSQXmpFtGN1BcbrZvYr4Afh6rKYWWeCHfj+1hTFiUjjuuFHMzmqZC2TyhYzIAyKl46pv0WhcQmBugPjW8B/AGvN7OPw2CDgD8Ad8ZYlIo2pqLiUX//Hs1xW2rCgyO2Q4Cdfm6QWhQB1BEY4ffYOM7uLA9NqP0oNgItIy5Z66G5d4fucvP5drv1sC4sjBgVAz7yOPPXjC+MvVFqNKHt67wKWNEEtItJIHntuMUtmvcOkssWc3MCg6NE9lxu+cLwGs+UQDR70FpGWK5lMMu2xl8h74y9c18CgGHB0V7513Xh1P0mtFBgibUAymeSv02ax9cUXGLA7elAM6tudYQU9NPNJIomy+ODRwL3AAHe/ysxGAhPc/b9ir05E6jTz7dUsnTWPQSvepv+eLazKO44XB9UfFF06ZvPM/VE2zhQ5IEoL4yngd8CPwvcfAjMABYZIM0kmk9x155OcuHYhp4ddT3+O+GT2mGG9+Ok3JjVBldLWRAmMo939WTO7G8DdK80s0/7cIhKzoOvpNcr/9AIX7ymJ3PWUlYCzTz5WXU9yRKIERrmZHUu4J4aZnU+wNIiINJFkMsm2pcuY9e//SZ+dm1mVdxx/7jNJ02OlSUUJjG8AvwFGmNlKoBz4h1irEhHgQIti98wX6bV9ExvyjuP5QRPZkdO13mtzshPcduVYTY+VRhPlOYwPgQvClWqz3H17/GWJtG+poNj6pxfou6cEzx/OjEFnRAoK0DiFxKOu/TC+WstxANz98ZhqEmm3kskk25YsZda//ydH79rCpvzh/KnP5EhB0a1LByaM6qdxColNXS2Mfkd6czO7DbgJqABudvfVaeceAa4GnnP3b4bHbgBSQVUAvOju3w333zgB2AG87+7fPtLaRFqSVFCsnTad8hXO+rzhPD9oUqSg6JSbzb9+9QyFhMSurrWk/iX9vZnlAUl33xHlxmbWE7gVmAiMAx4Ark37yP3AcwShkfrOp4Gnw+tnhOdTbnf3hVG+W6S1SA+KHas+4LMTJvCbgV9U15O0SFEe3DuVYNA7O3y/D7gtwi/vCcBsd68ECi3VlxVy9/VmNqKW7+wOjAHmph1+1Mz2APe5++v11S3SkqUHxXb/gE1DTmT2mOso3pENOfVf37VTB3586+lqVUiTijJL6nfAl939fQAzOxH4PUEXUV16EsyoSslqQF1XAX9y92T4/vvuXmJm/YFZZnZyao8OkdYkmUyybfES1kybzs4PPuTTgWP547FXsIOusLP+6wf1684lZwzVzCdpFlECY18qLADcfbGZVUS4rhxI33GlIQ/7fQn4Qdp3loR/rjezZcBgQDv+SauRCorCX0+l4+ZPWdF7JG8NvJKtWV0i/VPqi2cfx82XjY6/UJE6RAmMeeF4wjMED+9dGx77AoC7/7WW6+YD95pZNnAi8EGUgsysD9DP3RenHct3921m1hUYBXwa5V4izS09KHI3fcqK/OG8M+iL7OgQbYwCFBbSckQJjC4EjeVLwve7ga4ErYAkkDEw3L3MzJ4C5gD7gFvMbApQ7O5vhkuNXAX0MbPh7n5BeOk1BGtVpfuDmeWH9d4XdeBdpLlkDIrBDQsKgI45WQoLaTESyWSy/k+1UmaWdPfmLkPakWQyyZJX32LttOl0L9/I4vzhvHPUmAYHBUD3Ljk8fd/FMVQpUjszw90Tmc5FmSU1ELgdGEg4UwrA3W9otApFWrlUUHz0X3+g164trM4fzjuDT29wUCSAgRrYlhYqSpfUnwmm1c4CquMtR6R1SXU9zfnFE+Rt3cSn+cOZMXgyOzt0afC99EyFtHRRAmOPuz8ceyUircjy1SXM+sNfKVj+Fr13baE4fzjzBk+MHBRZCcjv1pGsrASD+uVx/XmmZyqkxYsSGPeZ2YPA3wiW+ADqnB0l0mYlk0kevO9pCpbP4aS9pbyfN4JnG9Ci0Aqy0ppFCYxrgdFALw50SdU6O0qkLVq+uoR3/vgG3ee/zilhUPyx71kKCmlXogTG59x9ZOyViLQwRcWlvDq3mHXz3+XUTYs4LgyKFxoQFN275PDli0YpKKRNiBIYs8zsc+4+P/ZqRJrZzLnFvPHuWlZ/upX+29cxqWwx4w+jRXF0j07c+Q+nalxC2pQogXE18A0zKyMYw0gQrFrbP9bKRJrQzLnF/PbFZVRUVDF4zwauLVtM38MIig7ZCe7/+iQFhbRJUXbcO+J9MURaolRrYtUn5VRXJxmyez2TyhZzTEVZg4IiJztB76O6cNqYfnoqW9q0KC2M1N4Ww4GOqWPu/ve4ihKJ2233v8bG0t2QPDQoXuh3dqSgUGtC2psoT3p/nWDXvGEE+1OcBbwFKDCkVUm1KD5cu5XKyurDDoounbK54LQhak1IuxOlhfENgh3z3nX3y8xsMPBIrFWJNIKi4lJenfcxH63bxqebd1BVzSEtikV5FikoOuZmMXpobz1gJ+1a1Ce9K82sysy6uPvHZjYs9spEDtPMucX84a8r2bojbduWBgZFVgKOyuukp7BF0kQJjEVm1oNg5713zGw7UBhvWSINt3+m0760Jc/SgqJPRXm9XU9ZCfjaVSfquQmRDKLMkro9fPmImb0MdHf3JfGWJRJdUXEp//qbd9j1WeWBgxmC4vl+Z7OrlqDold+JM8cVaFxCpA61BoaZDQK2uvu28P25wBXAGjNzd9/bRDWK1Or+qQuYt3TDgQMNDAqASycN4fYvjq31vIgE6mphzACuBLaZ2cnAdOB+YAzwGDAl9upEalFUXMp//H4hJVs/Cw4kkwwNg+LoeoIiOwsK+nRnWEEPLjxtsMYnRCKqKzA6ufv68PWXgcfd/edmlgDUJSVNLjXracGKTezcvS84mCEonssQFJ07ZjO4Xz7nnHysxidEDlNdgZFlZlnuXg2cC3wLwN2TZtYkxYlA7bOe6guKDtkJzhxXoFaESCOpKzCeAV41s80Ey5nPBjCzocCO+EsTSXsiOyVCUGgpcZF41BoY7n6fmZ0O9AVuD1saqWu+2RTFSfszc24xC1dsYtuOvXy8cfuBKbI1gmJRLV1PQ/vn8dD3zmmGykXavjqn1br7vAzHVkW9uZndRrCsSAVws7uvTjv3CMFKuM+5+zfDY2cD/wN8EH7senffGD4o+FsgF3jS3Z+IWoO0DjPnFjP1z8vZU1F18IlkkqG714VBsTUMinPY1aHzIfc4Y2w/7vrHCU1UsUj7E2nxwcMRLlh4KzCRYGmRBwh270u5H3iOIDTSvZAKkDQPAHcC7wFzzew5dy+LpXBpMqlB7LeWrD/4YTuoJSg+f0hQ5HbIYtJJAzROIdIEYgsMYAIw290rgUKrMVLu7uvNbESG6y41s5OAN4F73D0JjHD3QgAzmx3e+5UYa5cYFBWX8u6KTXz46TaK12+lPH0QOyVCUHTITjD82KM040mkicUZGD2B8rT3WRGuWQgYsI+gC+o6YBrBpk0pW8N7Syty/9QFvLN0A8naPlAzKPLtkKDo0jGbKZeOUUiINJM4A6McSH98tqq2D6a4+87UazObQbCU+jQ46PdMPqDuqFYg1aJ4df4nbNuZoTUBkEwyLAyK3rUEBUDfXl144ofnN0HVIlKbOANjPnCvmWUDJ3JgILtWZpafWoqEICxS16wys/HA4vD4z2KoVxpBKiTeXrKBdVt21v7BDEHxbFpQ5HfLpUe3jnTplKOuJ5EWIrbAcPcyM3sKmEPQxXSLmU0Bit39TTO7G7gK6GNmw939AuD6cGbVHuAj4J7wdncRdFHlAFM14N3yFBWX8sQfl/Lhp9vq/mA9QTHg6K5867rxGsAWaYESyWStvcqtnpkl3b25y2jTDjco3ss35vcYze4OnTmqey7j7BjNdBJpAcwMd09kOhdnl5S0cY89v4SX5xZT5785MgTFzCFfoHdBH07r3U0hIdKKKDDksMycW8zMusIiQ1DM6Pd58o7pxe80eC3SKikwpEHq7YLKEBR/G3kR/Qb1Y8rxfTV4LdKKKTCkXkXFpWwo2cWswjUs+6g084cyBMXCk6/kH689TV1OIm2EAkNqlWpNFK/fTnV1MvNDd8kkx+3+lIllS+hdsZVVfcdQcd5tXHTyMAWFSBujwJCMDtn6tKYwKCaVLaZXxTYW5Rvdbvtnvnr+CU1XpIg0KQVGO5bqaqrYV0XJ1j2UbP2M3j06U7hiI8Xrt2e+qEZQvJdvPNv/PM4+63guVFiItGkKjHYmfTxi5cflVFdXUx3lUZy0oOhZsZ338o3p/c6l/+C+3HvlWHU/ibQDCox2oqi4lKdfXUlRcRlVVRFDAg4JikU9jOcGnMsxA/vyfxQUIu2KAqMNS63r9J5vpnjDdqqqGvBUf82gyDeWnH4NZ04ayUW9uykoRNohBUYrl+pi6t+7G0mS+8ck3l6yniUfltT9FHYmGbqe5pxwGbf/wySFhEg7p8BoRWqGQ6qLKTsrQUVlFQkSQJKq6npvdagMXU/PF5zHmBOH8EtteyoiKDBavFRILFi+kXd9MwmgorIKkuwfh9i3/9MNa05kJYL/ndV1K9gCTYMAAA18SURBVGPXFtKhvIQdY05j7+RzObFHvrqeROQgCowW7LHnl/Ba4RqS1UkqKg+n2XCoDtkJRg/txcQT+tOpuIjO82aR/GQT/S65iAFXXk5Ofn6jfI+ItD0KjBaqqLiU1wrXsLei3o0KI8lKwAnH9eaGL4zkmC0fsXbao+xZt56+CgoRiUiB0QIVFZfyeuEaGrJXSVYCEokDYxjZWQmyshIM6tud8XYMJ4/sEwTFwz9l1br19LvkIkb/nx8pKEQkMgVGC1NfN1RuThaVVdUkSJDTIYuq6iSjhvTkxguOB2B9yU4q9lWTm5NF/97dGDn4KMoWFLL2V79l1YYN9Lv4QgWFiBwWBUYLUls3VG5OFolEgpNH9mHCqL70790NCMKhf42B6dTrZDJJ2fwFLH5oBntSQfEvCgoROXwKjBZkQ8kuau6LmNshi7PHF3DeqYMOmbGUaQZTKijWTlNQiEjjUmC0AOmLANYctUhkJTKGRU0KChGJmwKjmaXGLILhaujbqwsbS3fvf3/+hIF1hsUhQXHJRYy+8l5y8vKa6CcQkfYi1sAws9uAm4AK4GZ3X5127hHgauA5d/9meOx24GagEih092+Hx2cDOQTPqL3i7g/EWXdTyTRmsbF0N7dcNmb/oHVtYZGsrqZsfiFrn5nOng0bFRQiErvYAsPMegK3AhOBccADwLVpH7kfeI4gNFJmAY+7e9LMppnZRHd/Ozx3hbuXxFVvc8g0ZpEgGOQ+99SBGa/JHBSXKyhEJHZxtjAmALPdvRIoNDNLP+nu681sRI1jH6W9rSBoaUDQO/OCme0GfuDu78dYd5Pp17vrIWMWSdg/C+qg4woKEWlmcQZGT6A87X1W1AvN7HTgGHefHx66xt1LzGw08HuCFkurN2pIL84/deBBYxg1xyxqBkX/Sy9m9BWXKShEpMnFGRjlwNi095HWuAhbHQ8Cl6eOpbqi3H25me0zs87uvqcxi21K6avO3n7VWM4cV3DIMxVBUCxgzbTpfLZxk4JCRJpdnIExH7jXzLKBE4EP6rvAzPoRtCBucPctacfz3H27mfUFuramsEgPh+OH9DxkVtT5pw7k9qvG1hkU/RUUItICxBYY7l5mZk8BcwhmN91iZlOAYnd/08zuBq4C+pjZcHe/APgJ0Af4TTjk8W/A68Ab4fhFB+Cf46r5SNQMBjh0yuzJ1od3ffNBs6JeK1zDmeMKGDmoh4JCRFq0REMWuGttzCzp7rF/T6ZWw+RxA7j38XkHhUOH7GBBwIp9B9aI6pSTxTdGJ+ix8G8KChFpdmaGu9ecwAnowb0jlulZitcK19ClU84hU2azEgmqU1mRTDJi1xomly8hd/Vuel52Cf2vuJycvO5NVruISEMoMI5Qbc9SQDLjMh8TRhzNtoWFnLblffL37WDrmNOYcOetCgoRafEUGEeotmcpTjm+L7s/qzzQVZVMcnWfnYxc9Dq7t2yi8pTJ9L74Es4Zk/kBPRGRlkaBEVGmQW2o+1mK44f0ZPKJ/dn09jw6z32NZHEpPS+9mDHqehKRVkiBEUFtU2FTMj1LkayupvSd+eydNp1OmzbT77JL6H/5ZQoKEWm1FBj1qG1Q+8xxBYdsXJQKipK357H2mel8tmkz/RUUItJGKDDqUdug9vqSnYcs4VE6bz5rn5nO3s1b6HfpxQoKEWlTFBj1qG+BQAWFiLQXCox61DaoPXJQj/1dT3s3bwnHKC4lp7uCQkTaJgVGBOmD2v16dqHPxlW8/+3vKShEpF1RYEQ0clAPjl6/krW/nM6HW0oUFCLS7igw6hGMUbzD2mnT2augEJF2TIFRCwWFiMjBFBg1JKurKZ07j7XPzFBQiIikUWCEDgqKktJweqyCQkQkpd0FRs01oRQUIiLRtKvAOGhNqHD1WFv9DhWlZQoKEZF6tJvA2L8m1N5KRu78hInli8nz3STPv4BTvnINHbp1a+4SRURatHYTGKk1oSaVLeaUbStYmH88S/uM4aZTT1VYiIhE0G4CI7Um1IKjRrHgqFFUZOXSMTt7/5pQIiJSt6zmLqCppNaESnTqTFanznTMzd6/0ZGIiNSv3bQwIPNGRyIiEk2sgWFmtwE3ARXAze6+Ou3cI8DVwHPu/s3wWBbwKDAGWA9Mcfc9ZjYM+C2QCzzp7k8cbk2pjY5ERKRhYuuSMrOewK3AmcCdwAM1PnI/cEONYxcB1e4+GVhIEDaE194JTAZuDe8tIiJNKM4xjAnAbHevdPdCwNJPuvt6OGRvosnAS+HrlwjCBmCEuxe6exUwO7y3iIg0oTgDoydQ3sDvSr9ma/geOGiX1PTjIiLSROIMjHKgR9r7qgZekw+Uha/TWyLpx0VEpInEGRjzgbPMLNvMxgMfRLjm78DF4euLgTnh61VmNt7MsoGzgAWNXq2IiNQptsBw9zLgKYJf+r8A7jKzKWZ2FoCZ3Q38HLjCzF4NL/sL0NHM5gCnAU+Gx+8K7zEHmBreW0REmlAimaw57tx2mFnb/eFERGLi7olMx9t0YIiISONpN0uDiIjIkVFgiIhIJAoMERGJRIEhIiKRKDBERCQSBYaIiETSrvbDaO3M7GWCpVOygXvd/a/NXFIszOyHwO3AUne/tLnraWx1LfvfVphZDsFCoaOBW9392eatKB5mdjrwIMH/lzuBG919a/NWFR+1MFqXb7v7ROAS4D+au5gY/Q74fHMXEYcIy/63FZXA/wJ+2dyFxOwT4Fx3Pwv4M/BPzVxPrBQYrYi7p9bj+gyobs5a4uTuG4m2WGVrVOey/22FuyfdfUNz1xE3d1/v7rvDtxUEQdlmqUuqdfp32v6/3Nqqw1n2X1o4M+sFfAO4sLlriZMCo4UJuyxmZjh1n7u/bGbfBarc/akmLq1R1fdzNnU9TagcGJv2vq22pNoNM+sCzADucPeS5q4nTgqMFiZcife0TOfM7Mbw3PVNWlQM6vo527j5wL3hUv0nEm3Zf2mhzKwDMA142N3nNnc9cdPig61E+AtmN7CIYAyjyt3Pbd6q4mFmtxDMIjJgMXB9W/qXm5l9DfgKsA+4xd0/bOaSYmFm04FTCGYPveLu/18zl9TozOzLwMPA++Ghl929zU5IUWCIiEgkGnATEZFIFBgiIhKJAkNERCJRYIiISCQKDBERiUTPYUirYWYLCf6b7UswJbUUKAF+AnzN3Zvk+RQzmwK8lJrqa2az3P28w7jPRnfvm+H4ScCvgV5AEvg3d//vI6v6kO84G9jp7gvD978BfuLuH9dWl4haGNJquPsp7n4S8J8Ev0RPOpxf1FGEz73UZgrQO62uRqvBzLoCLwA/cHcjWITxR2bW2A85nk3wjAQA7n6ru3/cyN8hbYxaGNJWHGVmfwZGAlPd/d8AzOxm4OtALvC8u/+LmWUBvyL4pbkX+Cd3fydsOVwKHA2sM7PvAI8DA4A9wC0Ey3WfArxgZiXuPjn9X+Rm9mPgGoLFIR9198fM7HFgPNAZ+LW7P1rHz3EDweKEcyBY3M7M7ge+DVxvZrMJWlMrw1bC19z9ejO7ErgL6EiwguqN7r7TzKYC24DTge7AlwhaZl8D9oUPEV4HPJa6b3oxZnY3cGV434fd/QkzOyf8+6sCdoUrKEs7oBaGtBXjCP7lfyLwNTPramajCBaDOy08f7KZnUKw7HYBcEJ4zdS0+4wFLnX3G4BfAP/i7qcA3wUedPcXgIXAF919cnoBZnYpMBEY7+5jCdYXAvjf4T1OAqaYWW9qN4rgaf50iwiCsC5vuvvnwhbY2wThltLN3ScA9wB3uvtaDm6leaYbmtmFwNHufipBSN5qZv0I/i7uCL/r4nrqkjZELQxpK95291IAMysG+gPnEvzL+t3wM92A4cDngKfdPQksNbPdZnZM+JlX3H1H+PrzwCiz/SuQ17dQ4OeB37l7BexfLwvghnC5k2zg2LCGxl7qZKCZzQCOAboAs9LOvRj+uYjgl31U5wOXhS0ZgHxgGDAX+JmZPQVMP5KipXVRYEhbsTftdTXBL+cE8P/c/f70D5rZ5+q4z+6010mC1sJh7z1iZkMJdg88w913mNlLBN07tVlBEHLpTiJo1UAQWqmegfT7/Ar4sbvPNrP/RdC1lpL6u0n9vUSVAH7k7k/XOP6Wmf0FuAyYb2anpMJa2jZ1SUlb9jeCfv8eAGZWEO5bMBe4xswSZjYa6OzumzJcP4dgdzzMLMvMTgiP7yAYD6hpFnCzmeWG1/QMP7cD2Glmg4Gz6qn5f4BzzGxieI9+BOMX/zc8/wlBgABcnnZdHrAhHJ/5cj3fUdfPUPPnucXMOoW1mJl1MrOh7v6+u98HfEzQapJ2QIEhbZa7LyP4RfummS0h6D7pCjwLbAKWAv9NsDJuJv8MXGRmi4FlBFvjQjDm8d9mNqfG980E3gIWhddc7e6LgY+AlQRTZd+up+ZdwNXA/zWzVQS/kL+fNhj9C+CecIpx+u5u9wF/IVg+vbiu7wj9GfiymS2ytD63DD/Pq0ChmS0D/h9BC+W7ZrY8/DtdSbCisLQDWq1WpAUzs+8DVwDnufve+j4vEicFhoiIRKIuKRERiUSBISIikSgwREQkEgWGiIhEosAQEZFIFBgiIhKJAkNERCL5/wEjF85XoIlTXAAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            " \n",
            "Malignant subgroup\n",
            " \n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SVZ4zfESDak8"
      },
      "source": [
        "So the normality tests for both the more symmetrical radius_mean and symmetry_mean both fail. However we will still attempt to log-transform the more skewed features to assist with better modelling later. \n",
        "\n",
        "### 5.3 Test log transformation of skewed distributions for normality\n",
        "\n",
        "Let is see for now how well we can transform one of the skewed distributions such as radius_se."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 922
        },
        "id": "KUMnVYBQ2FaM",
        "outputId": "a5e5d07c-0bb2-43c5-8750-bccd2dd0fc48"
      },
      "source": [
        "# Plot skewed distributions of radius_se sub-groups\n",
        "malignent_df['radius_se'].hist()\n",
        "benign_df['radius_se'].hist()\n",
        "\n",
        "print('Malignant radius_se normality test')\n",
        "stat, p = shapiro(malignent_df['radius_se']) \n",
        "print('Statistics=%.3f, p=%.3f' % (stat, p)) \n",
        "# interpret\n",
        "alpha = 0.05\n",
        "if p > alpha:\n",
        "  print('Sample looks Gaussian (fail to reject H0)') \n",
        "else:\n",
        "  print('Sample does not look Gaussian (reject H0)')\n",
        "\n",
        "print('Benign radius_se normality test')\n",
        "stat, p = shapiro(benign_df['radius_se']) \n",
        "print('Statistics=%.3f, p=%.3f' % (stat, p)) \n",
        "# interpret\n",
        "alpha = 0.05\n",
        "if p > alpha:\n",
        "  print('Sample looks Gaussian (fail to reject H0)') \n",
        "else:\n",
        "  print('Sample does not look Gaussian (reject H0)')\n",
        "\n",
        "# q-q plot\n",
        "qqplot(malignent_df['radius_se'], line='s')\n",
        "plt.title('Malignant radius_se QQ plot')\n",
        "plt.show()\n",
        "\n",
        "qqplot(benign_df['radius_se'], line='s')\n",
        "plt.title('Benign radius_se QQ plot')\n",
        "plt.show()"
      ],
      "execution_count": 143,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Malignant radius_se normality test\n",
            "Statistics=0.816, p=0.000\n",
            "Sample does not look Gaussian (reject H0)\n",
            "Benign radius_se normality test\n",
            "Statistics=0.903, p=0.000\n",
            "Sample does not look Gaussian (reject H0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 922
        },
        "id": "ZE7ujZMA3GZ9",
        "outputId": "70271048-85cc-4097-cc7c-271f7eeb38cd"
      },
      "source": [
        "# Apply box-cox transforms to both sub-groups\n",
        "transformed_radius_se_malignant = boxcox(malignent_df['radius_se'], 0)\n",
        "transformed_radius_se_benign = boxcox(benign_df['radius_se'], 0)\n",
        "plt.hist(transformed_radius_se_malignant)\n",
        "plt.hist(transformed_radius_se_benign)\n",
        "plt.show()\n",
        "\n",
        "print('Malignant radius_se normality test')\n",
        "stat, p = shapiro(transformed_radius_se_malignant) \n",
        "print('Statistics=%.3f, p=%.3f' % (stat, p)) \n",
        "# interpret\n",
        "alpha = 0.05\n",
        "if p > alpha:\n",
        "  print('Sample looks Gaussian (fail to reject H0)') \n",
        "else:\n",
        "  print('Sample does not look Gaussian (reject H0)')\n",
        "\n",
        "print('Benign radius_se normality test')\n",
        "stat, p = shapiro(transformed_radius_se_benign) \n",
        "print('Statistics=%.3f, p=%.3f' % (stat, p)) \n",
        "# interpret\n",
        "alpha = 0.05\n",
        "if p > alpha:\n",
        "  print('Sample looks Gaussian (fail to reject H0)') \n",
        "else:\n",
        "  print('Sample does not look Gaussian (reject H0)')\n",
        "\n",
        "# q-q plot\n",
        "qqplot(transformed_radius_se_malignant, line='s')\n",
        "plt.title('Malignant radius_se QQ plot')\n",
        "plt.show()\n",
        "\n",
        "qqplot(transformed_radius_se_benign, line='s')\n",
        "plt.title('Benign radius_se QQ plot')\n",
        "plt.show()"
      ],
      "execution_count": 144,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Malignant radius_se normality test\n",
            "Statistics=0.988, p=0.060\n",
            "Sample looks Gaussian (fail to reject H0)\n",
            "Benign radius_se normality test\n",
            "Statistics=0.995, p=0.285\n",
            "Sample looks Gaussian (fail to reject H0)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Hj8sm6zc1iOw"
      },
      "source": [
        "So the box-cox method can be very effective at transforming skewed data to make it normal. This can be useful for applying further parametric statistical tests as well as helping with model training.\n",
        "\n",
        "### 5.4 Statisitcal test for difference between cancer and non-cancer distributions\n",
        "\n",
        "Let us check if there is a statisitcally significant difference between the two groups for these features, using a non-parametric test (now that we know the distributions are not normal without transformation). We will compare symmetry_mean as that appears to have very similar distributions with only a small difference - but is it statiscally significant?\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qs_NCco7fkx0",
        "outputId": "b0dd1822-e9ab-44d3-ed0f-ab7abaf2901f"
      },
      "source": [
        "# example of the mann-whitney u non-parametric hypothesis test\n",
        "\n",
        "print('radius_se non-parametric hypothesis test')\n",
        "print(' ')\n",
        "# compare samples\n",
        "stat, p = mannwhitneyu(benign_df['radius_se'], malignent_df['radius_se']) \n",
        "print('Statistics=%.3f, p=%.3f' % (stat, p)) \n",
        "\n",
        "# interpret    \n",
        "alpha = 0.05\n",
        "if p > alpha: \n",
        "  print('Same distributions (fail to reject H0)') \n",
        "else:\n",
        "  print('Different distributions (reject H0)')\n",
        "\n",
        "# Student t test parametric\n",
        "print(' ')\n",
        "print('radius_se normalised data parametric hypothesis test')\n",
        "print(' ')\n",
        "# compare samples\n",
        "stat, p = ttest_ind(transformed_radius_se_benign, transformed_radius_se_malignant) \n",
        "print('Statistics=%.3f, p=%.3f' % (stat, p)) \n",
        "\n",
        "# interpret    \n",
        "alpha = 0.05\n",
        "if p > alpha: \n",
        "  print('Same distributions (fail to reject H0)') \n",
        "else:\n",
        "  print('Different distributions (reject H0)')"
      ],
      "execution_count": 145,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "radius_se non-parametric hypothesis test\n",
            " \n",
            "Statistics=9965.000, p=0.000\n",
            "Different distributions (reject H0)\n",
            " \n",
            "radius_se normalised data parametric hypothesis test\n",
            " \n",
            "Statistics=-19.236, p=0.000\n",
            "Different distributions (reject H0)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LTChGbdKmBbX",
        "outputId": "e48f2a77-b85d-4924-8c18-e174fa736f14"
      },
      "source": [
        "# Calculate effect size, assume normality here and use Cohen's d\n",
        "\n",
        "# function to calculate cohen's d for independent samples \n",
        "def cohend(d1, d2):\n",
        "  # calculate the size of samples\n",
        "  n1, n2 = len(d1), len(d2)\n",
        "  # calculate the variance of the samples\n",
        "  s1, s2 = var(d1, ddof=1), var(d2, ddof=1)\n",
        "  # calculate the pooled standard deviation\n",
        "  s = sqrt(((n1 - 1) * s1 + (n2 - 1) * s2) / (n1 + n2 - 2))\n",
        "  # calculate the means of the samples\n",
        "  u1, u2 = mean(d1), mean(d2)\n",
        "  # calculate the effect size\n",
        "  return (u1 - u2) / s\n",
        "\n",
        "print(' ')\n",
        "print('Effect size for normalised radius_se difference between malignant & benign groups')\n",
        "print(' ')\n",
        "# calculate cohen's d\n",
        "d = cohend(transformed_radius_se_malignant, transformed_radius_se_benign) \n",
        "print('Cohens d: %.3f' % d)\n",
        "\n"
      ],
      "execution_count": 146,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            " \n",
            "Effect size for normalised radius_se difference between malignant & benign groups\n",
            " \n",
            "Cohens d: 1.668\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aotnq__DMJug",
        "outputId": "2c96dce0-a098-46c1-8132-0c053347b213"
      },
      "source": [
        "# Perform power analysis to estimate statisitcal power of this effect size, for given sample size and alpha\n",
        "# parameters for power analysis\n",
        "effect = 1.67\n",
        "alpha = 0.05\n",
        "# power = 0.8\n",
        "sample_size = 569\n",
        "# perform power analysis\n",
        "analysis = TTestIndPower()\n",
        "result = analysis.solve_power(effect, power=None, nobs1=sample_size, ratio=1.0, alpha=alpha) \n",
        "print('Statistical power of test: %.3f' % result)"
      ],
      "execution_count": 147,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Statistical power of test: 1.000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7JK9YrGjNxj6",
        "outputId": "6a5ef0ce-5b33-494b-b083-071ba22a2728"
      },
      "source": [
        "# Perform power analysis \n",
        "# What would be the minimum sample size needed for given effect size, alpha and power?\n",
        "\n",
        "effect = 0.8\n",
        "alpha = 0.05\n",
        "power = 0.9\n",
        "# perform power analysis\n",
        "analysis = TTestIndPower()\n",
        "result = analysis.solve_power(effect, power=power, nobs1=None, ratio=1.0, alpha=alpha) \n",
        "print('Sample Size: %.3f' % result)"
      ],
      "execution_count": 148,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Sample Size: 33.826\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 294
        },
        "id": "ZsILFojsOuH6",
        "outputId": "2a9d147c-71f8-45ba-f3ff-775257b174ec"
      },
      "source": [
        "# Plot power curve\n",
        "# See statistical power for a range of effect & sample sizes\n",
        "\n",
        "# parameters for power analysis\n",
        "effect_sizes = array([0.2, 0.5, 0.8])\n",
        "sample_sizes = array(range(5, 100))\n",
        "# calculate power curves from multiple power analyses\n",
        "analysis = TTestIndPower()\n",
        "analysis.plot_power(dep_var='nobs', nobs=sample_sizes, effect_size=effect_sizes) \n",
        "plt.show()"
      ],
      "execution_count": 149,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bcAuADuUL3LK"
      },
      "source": [
        "So we can see from out hypothesis test there is an indication that radius_se  have significant differences between the benign and malignent groups. Furthermore, the effect size between the two groups for the normalised radius_se is high. This is confirmed by what we observe in the histograms above.\n",
        "\n",
        "We can also see from our power analysis that we only need to have 50 observations to have high statisical power for our significance tests between the malignant and benign groups for the features.\n",
        "\n",
        "Secondly, we can see some of the distributions are particularly skewed i.e. for features:\n",
        "\n",
        "- radius_se\n",
        "- perimeter_se\n",
        "- area_se\n",
        "\n",
        "These are going to be more difficult for the ML model to work with, so we will probably want to log-transform them to spread out their values more to make them of maximum use to our model."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LI96kpqz9Hkg"
      },
      "source": [
        "Thirdly, we can see the units of the features and scale vary widely, with many such as 'symmetry_mean' being fractional decimals below 1, and others such as 'area_worst' has values ranging from a few hundred to over 3500. ML models are very sensitive to these differences of scale so to ensure our model treats all features equally, we will need to normalise them all so each feature is a value between 0 and 1.\n",
        "\n",
        "Finally more generally we can see comapring the malignant v benign sub-groups accross all features, we can see there are some features where there is clearly much overlap between the groups, and other features where there is a big difference between the values for each groups - so both areas of commonality and difference. \n",
        "\n",
        "Let us now move from a univariate perspective to combining multiple features to build up a bigger picture of the dataset."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "n_0WJWGdkRG8"
      },
      "source": [
        "## 6 Normalise numerical features before modelling\n",
        "\n",
        "We also noticed earlier what differing scales of values each of the features has (which again can be problematic for ML models). This will be useful to do before our dimensonality techniques for further analysis, as well as for our predictive modelling later. Lets normalise the values for all features to be betweeen 0 and 1."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 334
        },
        "id": "2GQOIE3VlkqP",
        "outputId": "2d89ab7d-f3f8-414c-fe3b-c6fea445bf7b"
      },
      "source": [
        "bcancer_clean_df.describe()"
      ],
      "execution_count": 150,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>diagnosis</th>\n",
              "      <th>radius_mean</th>\n",
              "      <th>texture_mean</th>\n",
              "      <th>perimeter_mean</th>\n",
              "      <th>area_mean</th>\n",
              "      <th>smoothness_mean</th>\n",
              "      <th>compactness_mean</th>\n",
              "      <th>concavity_mean</th>\n",
              "      <th>concave points_mean</th>\n",
              "      <th>symmetry_mean</th>\n",
              "      <th>fractal_dimension_mean</th>\n",
              "      <th>radius_se</th>\n",
              "      <th>texture_se</th>\n",
              "      <th>perimeter_se</th>\n",
              "      <th>area_se</th>\n",
              "      <th>smoothness_se</th>\n",
              "      <th>compactness_se</th>\n",
              "      <th>concavity_se</th>\n",
              "      <th>concave points_se</th>\n",
              "      <th>symmetry_se</th>\n",
              "      <th>fractal_dimension_se</th>\n",
              "      <th>radius_worst</th>\n",
              "      <th>texture_worst</th>\n",
              "      <th>perimeter_worst</th>\n",
              "      <th>area_worst</th>\n",
              "      <th>smoothness_worst</th>\n",
              "      <th>compactness_worst</th>\n",
              "      <th>concavity_worst</th>\n",
              "      <th>concave points_worst</th>\n",
              "      <th>symmetry_worst</th>\n",
              "      <th>fractal_dimension_worst</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>0.372583</td>\n",
              "      <td>14.127292</td>\n",
              "      <td>19.289649</td>\n",
              "      <td>91.969033</td>\n",
              "      <td>654.889104</td>\n",
              "      <td>0.096360</td>\n",
              "      <td>0.104341</td>\n",
              "      <td>0.088799</td>\n",
              "      <td>0.048919</td>\n",
              "      <td>0.181162</td>\n",
              "      <td>0.062798</td>\n",
              "      <td>0.405172</td>\n",
              "      <td>1.216853</td>\n",
              "      <td>2.866059</td>\n",
              "      <td>40.337079</td>\n",
              "      <td>0.007041</td>\n",
              "      <td>0.025478</td>\n",
              "      <td>0.031894</td>\n",
              "      <td>0.011796</td>\n",
              "      <td>0.020542</td>\n",
              "      <td>0.003795</td>\n",
              "      <td>16.269190</td>\n",
              "      <td>25.677223</td>\n",
              "      <td>107.261213</td>\n",
              "      <td>880.583128</td>\n",
              "      <td>0.132369</td>\n",
              "      <td>0.254265</td>\n",
              "      <td>0.272188</td>\n",
              "      <td>0.114606</td>\n",
              "      <td>0.290076</td>\n",
              "      <td>0.083946</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.483918</td>\n",
              "      <td>3.524049</td>\n",
              "      <td>4.301036</td>\n",
              "      <td>24.298981</td>\n",
              "      <td>351.914129</td>\n",
              "      <td>0.014064</td>\n",
              "      <td>0.052813</td>\n",
              "      <td>0.079720</td>\n",
              "      <td>0.038803</td>\n",
              "      <td>0.027414</td>\n",
              "      <td>0.007060</td>\n",
              "      <td>0.277313</td>\n",
              "      <td>0.551648</td>\n",
              "      <td>2.021855</td>\n",
              "      <td>45.491006</td>\n",
              "      <td>0.003003</td>\n",
              "      <td>0.017908</td>\n",
              "      <td>0.030186</td>\n",
              "      <td>0.006170</td>\n",
              "      <td>0.008266</td>\n",
              "      <td>0.002646</td>\n",
              "      <td>4.833242</td>\n",
              "      <td>6.146258</td>\n",
              "      <td>33.602542</td>\n",
              "      <td>569.356993</td>\n",
              "      <td>0.022832</td>\n",
              "      <td>0.157336</td>\n",
              "      <td>0.208624</td>\n",
              "      <td>0.065732</td>\n",
              "      <td>0.061867</td>\n",
              "      <td>0.018061</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>6.981000</td>\n",
              "      <td>9.710000</td>\n",
              "      <td>43.790000</td>\n",
              "      <td>143.500000</td>\n",
              "      <td>0.052630</td>\n",
              "      <td>0.019380</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.106000</td>\n",
              "      <td>0.049960</td>\n",
              "      <td>0.111500</td>\n",
              "      <td>0.360200</td>\n",
              "      <td>0.757000</td>\n",
              "      <td>6.802000</td>\n",
              "      <td>0.001713</td>\n",
              "      <td>0.002252</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.007882</td>\n",
              "      <td>0.000895</td>\n",
              "      <td>7.930000</td>\n",
              "      <td>12.020000</td>\n",
              "      <td>50.410000</td>\n",
              "      <td>185.200000</td>\n",
              "      <td>0.071170</td>\n",
              "      <td>0.027290</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.156500</td>\n",
              "      <td>0.055040</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>11.700000</td>\n",
              "      <td>16.170000</td>\n",
              "      <td>75.170000</td>\n",
              "      <td>420.300000</td>\n",
              "      <td>0.086370</td>\n",
              "      <td>0.064920</td>\n",
              "      <td>0.029560</td>\n",
              "      <td>0.020310</td>\n",
              "      <td>0.161900</td>\n",
              "      <td>0.057700</td>\n",
              "      <td>0.232400</td>\n",
              "      <td>0.833900</td>\n",
              "      <td>1.606000</td>\n",
              "      <td>17.850000</td>\n",
              "      <td>0.005169</td>\n",
              "      <td>0.013080</td>\n",
              "      <td>0.015090</td>\n",
              "      <td>0.007638</td>\n",
              "      <td>0.015160</td>\n",
              "      <td>0.002248</td>\n",
              "      <td>13.010000</td>\n",
              "      <td>21.080000</td>\n",
              "      <td>84.110000</td>\n",
              "      <td>515.300000</td>\n",
              "      <td>0.116600</td>\n",
              "      <td>0.147200</td>\n",
              "      <td>0.114500</td>\n",
              "      <td>0.064930</td>\n",
              "      <td>0.250400</td>\n",
              "      <td>0.071460</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>13.370000</td>\n",
              "      <td>18.840000</td>\n",
              "      <td>86.240000</td>\n",
              "      <td>551.100000</td>\n",
              "      <td>0.095870</td>\n",
              "      <td>0.092630</td>\n",
              "      <td>0.061540</td>\n",
              "      <td>0.033500</td>\n",
              "      <td>0.179200</td>\n",
              "      <td>0.061540</td>\n",
              "      <td>0.324200</td>\n",
              "      <td>1.108000</td>\n",
              "      <td>2.287000</td>\n",
              "      <td>24.530000</td>\n",
              "      <td>0.006380</td>\n",
              "      <td>0.020450</td>\n",
              "      <td>0.025890</td>\n",
              "      <td>0.010930</td>\n",
              "      <td>0.018730</td>\n",
              "      <td>0.003187</td>\n",
              "      <td>14.970000</td>\n",
              "      <td>25.410000</td>\n",
              "      <td>97.660000</td>\n",
              "      <td>686.500000</td>\n",
              "      <td>0.131300</td>\n",
              "      <td>0.211900</td>\n",
              "      <td>0.226700</td>\n",
              "      <td>0.099930</td>\n",
              "      <td>0.282200</td>\n",
              "      <td>0.080040</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>1.000000</td>\n",
              "      <td>15.780000</td>\n",
              "      <td>21.800000</td>\n",
              "      <td>104.100000</td>\n",
              "      <td>782.700000</td>\n",
              "      <td>0.105300</td>\n",
              "      <td>0.130400</td>\n",
              "      <td>0.130700</td>\n",
              "      <td>0.074000</td>\n",
              "      <td>0.195700</td>\n",
              "      <td>0.066120</td>\n",
              "      <td>0.478900</td>\n",
              "      <td>1.474000</td>\n",
              "      <td>3.357000</td>\n",
              "      <td>45.190000</td>\n",
              "      <td>0.008146</td>\n",
              "      <td>0.032450</td>\n",
              "      <td>0.042050</td>\n",
              "      <td>0.014710</td>\n",
              "      <td>0.023480</td>\n",
              "      <td>0.004558</td>\n",
              "      <td>18.790000</td>\n",
              "      <td>29.720000</td>\n",
              "      <td>125.400000</td>\n",
              "      <td>1084.000000</td>\n",
              "      <td>0.146000</td>\n",
              "      <td>0.339100</td>\n",
              "      <td>0.382900</td>\n",
              "      <td>0.161400</td>\n",
              "      <td>0.317900</td>\n",
              "      <td>0.092080</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.000000</td>\n",
              "      <td>28.110000</td>\n",
              "      <td>39.280000</td>\n",
              "      <td>188.500000</td>\n",
              "      <td>2501.000000</td>\n",
              "      <td>0.163400</td>\n",
              "      <td>0.345400</td>\n",
              "      <td>0.426800</td>\n",
              "      <td>0.201200</td>\n",
              "      <td>0.304000</td>\n",
              "      <td>0.097440</td>\n",
              "      <td>2.873000</td>\n",
              "      <td>4.885000</td>\n",
              "      <td>21.980000</td>\n",
              "      <td>542.200000</td>\n",
              "      <td>0.031130</td>\n",
              "      <td>0.135400</td>\n",
              "      <td>0.396000</td>\n",
              "      <td>0.052790</td>\n",
              "      <td>0.078950</td>\n",
              "      <td>0.029840</td>\n",
              "      <td>36.040000</td>\n",
              "      <td>49.540000</td>\n",
              "      <td>251.200000</td>\n",
              "      <td>4254.000000</td>\n",
              "      <td>0.222600</td>\n",
              "      <td>1.058000</td>\n",
              "      <td>1.252000</td>\n",
              "      <td>0.291000</td>\n",
              "      <td>0.663800</td>\n",
              "      <td>0.207500</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "        diagnosis  radius_mean  ...  symmetry_worst  fractal_dimension_worst\n",
              "count  569.000000   569.000000  ...      569.000000               569.000000\n",
              "mean     0.372583    14.127292  ...        0.290076                 0.083946\n",
              "std      0.483918     3.524049  ...        0.061867                 0.018061\n",
              "min      0.000000     6.981000  ...        0.156500                 0.055040\n",
              "25%      0.000000    11.700000  ...        0.250400                 0.071460\n",
              "50%      0.000000    13.370000  ...        0.282200                 0.080040\n",
              "75%      1.000000    15.780000  ...        0.317900                 0.092080\n",
              "max      1.000000    28.110000  ...        0.663800                 0.207500\n",
              "\n",
              "[8 rows x 31 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 150
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 334
        },
        "id": "hYWIVsHrk9Up",
        "outputId": "50059d3c-9bbf-40f6-98ff-d2d4beeb8221"
      },
      "source": [
        "# Initialize a scaler, then apply it to the features to normalise\n",
        "scaler = MinMaxScaler() # default=(0, 1)\n",
        "\n",
        "numerical_features = bcancer_clean_df.columns\n",
        "numerical_features = numerical_features[1:]\n",
        "\n",
        "bcancer_clean_df = pd.DataFrame(data = bcancer_clean_df)\n",
        "bcancer_clean_df[numerical_features] = scaler.fit_transform(bcancer_clean_df[numerical_features])\n",
        "\n",
        "bcancer_clean_df.describe()"
      ],
      "execution_count": 151,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>diagnosis</th>\n",
              "      <th>radius_mean</th>\n",
              "      <th>texture_mean</th>\n",
              "      <th>perimeter_mean</th>\n",
              "      <th>area_mean</th>\n",
              "      <th>smoothness_mean</th>\n",
              "      <th>compactness_mean</th>\n",
              "      <th>concavity_mean</th>\n",
              "      <th>concave points_mean</th>\n",
              "      <th>symmetry_mean</th>\n",
              "      <th>fractal_dimension_mean</th>\n",
              "      <th>radius_se</th>\n",
              "      <th>texture_se</th>\n",
              "      <th>perimeter_se</th>\n",
              "      <th>area_se</th>\n",
              "      <th>smoothness_se</th>\n",
              "      <th>compactness_se</th>\n",
              "      <th>concavity_se</th>\n",
              "      <th>concave points_se</th>\n",
              "      <th>symmetry_se</th>\n",
              "      <th>fractal_dimension_se</th>\n",
              "      <th>radius_worst</th>\n",
              "      <th>texture_worst</th>\n",
              "      <th>perimeter_worst</th>\n",
              "      <th>area_worst</th>\n",
              "      <th>smoothness_worst</th>\n",
              "      <th>compactness_worst</th>\n",
              "      <th>concavity_worst</th>\n",
              "      <th>concave points_worst</th>\n",
              "      <th>symmetry_worst</th>\n",
              "      <th>fractal_dimension_worst</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "      <td>569.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>0.372583</td>\n",
              "      <td>0.338222</td>\n",
              "      <td>0.323965</td>\n",
              "      <td>0.332935</td>\n",
              "      <td>0.216920</td>\n",
              "      <td>0.394785</td>\n",
              "      <td>0.260601</td>\n",
              "      <td>0.208058</td>\n",
              "      <td>0.243137</td>\n",
              "      <td>0.379605</td>\n",
              "      <td>0.270379</td>\n",
              "      <td>0.106345</td>\n",
              "      <td>0.189324</td>\n",
              "      <td>0.099376</td>\n",
              "      <td>0.062636</td>\n",
              "      <td>0.181119</td>\n",
              "      <td>0.174439</td>\n",
              "      <td>0.080540</td>\n",
              "      <td>0.223454</td>\n",
              "      <td>0.178143</td>\n",
              "      <td>0.100193</td>\n",
              "      <td>0.296663</td>\n",
              "      <td>0.363998</td>\n",
              "      <td>0.283138</td>\n",
              "      <td>0.170906</td>\n",
              "      <td>0.404138</td>\n",
              "      <td>0.220212</td>\n",
              "      <td>0.217403</td>\n",
              "      <td>0.393836</td>\n",
              "      <td>0.263307</td>\n",
              "      <td>0.189596</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.483918</td>\n",
              "      <td>0.166787</td>\n",
              "      <td>0.145453</td>\n",
              "      <td>0.167915</td>\n",
              "      <td>0.149274</td>\n",
              "      <td>0.126967</td>\n",
              "      <td>0.161992</td>\n",
              "      <td>0.186785</td>\n",
              "      <td>0.192857</td>\n",
              "      <td>0.138456</td>\n",
              "      <td>0.148702</td>\n",
              "      <td>0.100421</td>\n",
              "      <td>0.121917</td>\n",
              "      <td>0.095267</td>\n",
              "      <td>0.084967</td>\n",
              "      <td>0.102067</td>\n",
              "      <td>0.134498</td>\n",
              "      <td>0.076227</td>\n",
              "      <td>0.116884</td>\n",
              "      <td>0.116316</td>\n",
              "      <td>0.091417</td>\n",
              "      <td>0.171940</td>\n",
              "      <td>0.163813</td>\n",
              "      <td>0.167352</td>\n",
              "      <td>0.139932</td>\n",
              "      <td>0.150779</td>\n",
              "      <td>0.152649</td>\n",
              "      <td>0.166633</td>\n",
              "      <td>0.225884</td>\n",
              "      <td>0.121954</td>\n",
              "      <td>0.118466</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.223342</td>\n",
              "      <td>0.218465</td>\n",
              "      <td>0.216847</td>\n",
              "      <td>0.117413</td>\n",
              "      <td>0.304595</td>\n",
              "      <td>0.139685</td>\n",
              "      <td>0.069260</td>\n",
              "      <td>0.100944</td>\n",
              "      <td>0.282323</td>\n",
              "      <td>0.163016</td>\n",
              "      <td>0.043781</td>\n",
              "      <td>0.104690</td>\n",
              "      <td>0.040004</td>\n",
              "      <td>0.020635</td>\n",
              "      <td>0.117483</td>\n",
              "      <td>0.081323</td>\n",
              "      <td>0.038106</td>\n",
              "      <td>0.144686</td>\n",
              "      <td>0.102409</td>\n",
              "      <td>0.046750</td>\n",
              "      <td>0.180719</td>\n",
              "      <td>0.241471</td>\n",
              "      <td>0.167837</td>\n",
              "      <td>0.081130</td>\n",
              "      <td>0.300007</td>\n",
              "      <td>0.116337</td>\n",
              "      <td>0.091454</td>\n",
              "      <td>0.223127</td>\n",
              "      <td>0.185098</td>\n",
              "      <td>0.107700</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.302381</td>\n",
              "      <td>0.308759</td>\n",
              "      <td>0.293345</td>\n",
              "      <td>0.172895</td>\n",
              "      <td>0.390358</td>\n",
              "      <td>0.224679</td>\n",
              "      <td>0.144189</td>\n",
              "      <td>0.166501</td>\n",
              "      <td>0.369697</td>\n",
              "      <td>0.243892</td>\n",
              "      <td>0.077023</td>\n",
              "      <td>0.165267</td>\n",
              "      <td>0.072092</td>\n",
              "      <td>0.033112</td>\n",
              "      <td>0.158650</td>\n",
              "      <td>0.136675</td>\n",
              "      <td>0.065379</td>\n",
              "      <td>0.207047</td>\n",
              "      <td>0.152643</td>\n",
              "      <td>0.079191</td>\n",
              "      <td>0.250445</td>\n",
              "      <td>0.356876</td>\n",
              "      <td>0.235320</td>\n",
              "      <td>0.123206</td>\n",
              "      <td>0.397081</td>\n",
              "      <td>0.179110</td>\n",
              "      <td>0.181070</td>\n",
              "      <td>0.343402</td>\n",
              "      <td>0.247782</td>\n",
              "      <td>0.163977</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>1.000000</td>\n",
              "      <td>0.416442</td>\n",
              "      <td>0.408860</td>\n",
              "      <td>0.416765</td>\n",
              "      <td>0.271135</td>\n",
              "      <td>0.475490</td>\n",
              "      <td>0.340531</td>\n",
              "      <td>0.306232</td>\n",
              "      <td>0.367793</td>\n",
              "      <td>0.453030</td>\n",
              "      <td>0.340354</td>\n",
              "      <td>0.133044</td>\n",
              "      <td>0.246155</td>\n",
              "      <td>0.122509</td>\n",
              "      <td>0.071700</td>\n",
              "      <td>0.218683</td>\n",
              "      <td>0.226800</td>\n",
              "      <td>0.106187</td>\n",
              "      <td>0.278651</td>\n",
              "      <td>0.219480</td>\n",
              "      <td>0.126556</td>\n",
              "      <td>0.386339</td>\n",
              "      <td>0.471748</td>\n",
              "      <td>0.373475</td>\n",
              "      <td>0.220901</td>\n",
              "      <td>0.494156</td>\n",
              "      <td>0.302520</td>\n",
              "      <td>0.305831</td>\n",
              "      <td>0.554639</td>\n",
              "      <td>0.318155</td>\n",
              "      <td>0.242949</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "        diagnosis  radius_mean  ...  symmetry_worst  fractal_dimension_worst\n",
              "count  569.000000   569.000000  ...      569.000000               569.000000\n",
              "mean     0.372583     0.338222  ...        0.263307                 0.189596\n",
              "std      0.483918     0.166787  ...        0.121954                 0.118466\n",
              "min      0.000000     0.000000  ...        0.000000                 0.000000\n",
              "25%      0.000000     0.223342  ...        0.185098                 0.107700\n",
              "50%      0.000000     0.302381  ...        0.247782                 0.163977\n",
              "75%      1.000000     0.416442  ...        0.318155                 0.242949\n",
              "max      1.000000     1.000000  ...        1.000000                 1.000000\n",
              "\n",
              "[8 rows x 31 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 151
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pFXcJYzakPEk"
      },
      "source": [
        "## 7 Dimensionality reduction analysis \n",
        "\n",
        "### 7.1 Using t-SNE to compare cancer sub-groups\n",
        "\n",
        "[t-SNE](https://www.datacamp.com/community/tutorials/introduction-t-sne) is a recent dimensionality reduction technique that can be used to compress a multi-dimensional dataset into fewer dimensions while aiming to retain the maximum variance of the original dimensions. \n",
        "\n",
        "It is a more recent dimensionality reduction technique, say compared to [PCA](https://towardsdatascience.com/a-step-by-step-explanation-of-principal-component-analysis-b836fb9c97e2) that has various differences that I will not go into here, but are explored further in an earlier project of mine [here](https://github.com/pranath/high_dim_data_vis).\n",
        "\n",
        "We will use t-SNE on this dataset to reduce the dimensionality to 2 dimensions, and then plot these on scatterplots dividing the data into malignant & benign sub-groups, to see how the values between these sub-groups compare, and see if we can get a better overall picture of how similar or different these sub-groups are."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "x342TP3X9Hkg",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 759
        },
        "outputId": "6e88b7bb-2b8a-46a5-835b-431854d0f3ec"
      },
      "source": [
        "# Create 2 components using t-SNE\n",
        "time_start = time.time()\n",
        "tsne = TSNE(n_components=2, verbose=1, perplexity=40, n_iter=300)\n",
        "features = bcancer_clean_df.drop('diagnosis', 1)\n",
        "tsne_results = tsne.fit_transform(features)\n",
        "print('t-SNE done! Time elapsed: {} seconds'.format(time.time()-time_start))\n",
        "\n",
        "# Plot dataset with 2 t-SNE features, using same colour for data for benign v malignent tumors\n",
        "bcancer_clean_df['tsne-2d-one'] = tsne_results[:,0]\n",
        "bcancer_clean_df['tsne-2d-two'] = tsne_results[:,1]\n",
        "plt.figure(figsize=(16,10))\n",
        "sns.scatterplot(\n",
        "    x=\"tsne-2d-one\", y=\"tsne-2d-two\",\n",
        "    hue=\"diagnosis\",\n",
        "    palette=sns.color_palette(\"hls\", 2),\n",
        "    data=bcancer_clean_df,\n",
        "    legend=\"full\",\n",
        "    alpha=0.9\n",
        ")"
      ],
      "execution_count": 152,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[t-SNE] Computing 121 nearest neighbors...\n",
            "[t-SNE] Indexed 569 samples in 0.001s...\n",
            "[t-SNE] Computed neighbors for 569 samples in 0.033s...\n",
            "[t-SNE] Computed conditional probabilities for sample 569 / 569\n",
            "[t-SNE] Mean sigma: 0.215072\n",
            "[t-SNE] KL divergence after 250 iterations with early exaggeration: 63.284069\n",
            "[t-SNE] KL divergence after 300 iterations: 0.894506\n",
            "t-SNE done! Time elapsed: 1.2689051628112793 seconds\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f19df0a3358>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 152
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1152x720 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uY8rj-u_9Hkg"
      },
      "source": [
        "So we can observe various characteristics of the dataset and the subgroups.\n",
        "\n",
        "We can see that there are clear areas for the benign and malignant groups only, but also common areas of overlap between the two groups.\n",
        "\n",
        "These indicate the potential for a model using the dataset to be able to distinguish between the benign and malignant groups/cases."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "F_UwNbL09Hkh"
      },
      "source": [
        "### 7.2 Using UMAP to compare cancer sub-groups\n",
        "\n",
        "While t-sne has various advantages over the 100 year old PCA technique, it does have its own disadvantages. These include:\n",
        "\n",
        "- T-sne does not scale well for performance with increasing dataset sizes\n",
        "- T-sne does not preseve global structure and relationships well\n",
        "- T-sne can consume large amounts of memory when executing\n",
        "\n",
        "[UMAP](https://umap-learn.readthedocs.io/en/latest/index.html) is a dimensionality technique that aims to [improve upon these weaknesses](https://medium.com/@dan.allison/dimensionality-reduction-with-umap-b081837354dd) of t-sne. Read more about how UMAP works [here](https://towardsdatascience.com/how-exactly-umap-works-13e3040e1668).\n",
        "\n",
        "We will now apply UMAP to render both a 2D and a 3D representation of our cancer dataset."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "Kvbtx4F-fnXz",
        "outputId": "d5ea707a-611e-4d19-c66d-816df5ea95ee"
      },
      "source": [
        "# Separate feature columns\n",
        "features = bcancer_clean_df.drop('diagnosis', 1)\n",
        "\n",
        "# Initialise 2D & 3D UMAP models\n",
        "umap_2d = UMAP(n_components=2, init='random', random_state=0)\n",
        "umap_3d = UMAP(n_components=3, init='random', random_state=0)\n",
        "\n",
        "# Fit 2D & 3D UMAP models to our cancer data\n",
        "proj_2d = umap_2d.fit_transform(features)\n",
        "proj_3d = umap_3d.fit_transform(features)\n",
        "\n",
        "# Use plotly to plot our projections\n",
        "fig_2d = px.scatter(\n",
        "    proj_2d, x=0, y=1, \n",
        "    color=bcancer_clean_df.diagnosis, labels={'color': 'diagnosis'}\n",
        ")\n",
        "fig_3d = px.scatter_3d(\n",
        "    proj_3d, x=0, y=1, z=2,\n",
        "    color=bcancer_clean_df.diagnosis, labels={'color': 'diagnosis'}\n",
        ")\n",
        "# Configure projections\n",
        "fig_3d.update_traces(marker_size=5)\n",
        "# Hide axes colouring for 3D\n",
        "fig_3d.update_scenes(xaxis_visible=False, yaxis_visible=False,zaxis_visible=False )\n",
        "fig_2d.update_layout(\n",
        "    autosize=False,\n",
        "    width=800,\n",
        "    height=800)\n",
        "fig_3d.update_layout(\n",
        "    autosize=False,\n",
        "    width=800,\n",
        "    height=800)\n",
        "# Render projections\n",
        "fig_2d.show()\n",
        "fig_3d.show()"
      ],
      "execution_count": 153,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<html>\n",
              "<head><meta charset=\"utf-8\" /></head>\n",
              "<body>\n",
              "    <div>\n",
              "            <script src=\"https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/MathJax.js?config=TeX-AMS-MML_SVG\"></script><script type=\"text/javascript\">if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script>\n",
              "                <script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n",
              "        <script src=\"https://cdn.plot.ly/plotly-latest.min.js\"></script>    \n",
              "            <div id=\"0493233e-e6a5-4ecf-8001-da119dcd437f\" class=\"plotly-graph-div\" style=\"height:800px; width:800px;\"></div>\n",
              "            <script type=\"text/javascript\">\n",
              "                \n",
              "                    window.PLOTLYENV=window.PLOTLYENV || {};\n",
              "                    \n",
              "                if (document.getElementById(\"0493233e-e6a5-4ecf-8001-da119dcd437f\")) {\n",
              "                    Plotly.newPlot(\n",
              "                        '0493233e-e6a5-4ecf-8001-da119dcd437f',\n",
              "                        [{\"hoverlabel\": {\"namelength\": 0}, \"hovertemplate\": \"0=%{x}<br>1=%{y}<br>diagnosis=%{marker.color}\", \"legendgroup\": \"\", \"marker\": {\"color\": [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 1, 0, 1, 1, 0, 1, 0, 1, 1, 0, 0, 0, 1, 1, 0, 1, 1, 1, 0, 0, 0, 1, 0, 0, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 1, 1, 0, 0, 0, 1, 1, 0, 1, 0, 1, 1, 0, 1, 1, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 1, 1, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1, 0, 0, 1, 0, 1, 1, 1, 1, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 1, 1, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 0], \"coloraxis\": \"coloraxis\", \"symbol\": \"circle\"}, \"mode\": \"markers\", \"name\": \"\", \"showlegend\": false, \"type\": \"scatter\", \"x\": [-2.871018886566162, 0.4388813376426697, -2.0428781509399414, 15.132161140441895, 0.10361234843730927, 14.57568645477295, 0.7843625545501709, 14.16335678100586, 14.700784683227539, 15.110175132751465, 3.6629698276519775, 13.523475646972656, -2.7820000648498535, 2.6736767292022705, 14.994651794433594, 14.773147583007812, 3.3618931770324707, 14.421850204467773, -0.9946038722991943, -7.254260540008545, 7.5811076164245605, 6.906799793243408, 14.782857894897461, -0.9405256509780884, -1.7142258882522583, -2.8408286571502686, 14.663042068481445, -0.006264746189117432, 14.402164459228516, 1.3896697759628296, -2.1642770767211914, 14.47429370880127, -2.3580145835876465, -2.1154510974884033, 13.650628089904785, 13.557504653930664, 13.465210914611816, 9.698444366455078, 6.2153191566467285, 13.579538345336914, 3.826315402984619, 6.698270320892334, -2.647368907928467, 13.691703796386719, 13.455958366394043, -2.5494866371154785, 8.127718925476074, 14.233662605285645, 7.450293064117432, 3.825035810470581, 9.336186408996582, 10.435441970825195, 8.439093589782715, -0.05194775387644768, 3.3445966243743896, 7.930617332458496, -1.1026430130004883, 13.790923118591309, 9.645753860473633, 6.715727806091309, 7.380450248718262, 6.948193550109863, 14.806305885314941, 7.7286529541015625, 14.039348602294922, 13.865836143493652, 6.552426815032959, 8.599312782287598, 7.2442731857299805, 8.737997055053711, 0.46152883768081665, 7.098171710968018, 14.295158386230469, -7.772524833679199, 8.813432693481445, 1.561864972114563, 6.943836212158203, -2.7772912979125977, -2.5909042358398438, 8.526105880737305, 6.540641784667969, 13.928173065185547, -1.959776759147644, -2.3285043239593506, 7.858416557312012, 0.017171960324048996, 2.7069919109344482, -1.1744400262832642, 6.336781978607178, -7.732083320617676, 6.826783180236816, 2.6357548236846924, 10.617047309875488, 8.427681922912598, 13.667936325073242, -0.6598007082939148, 6.237144470214844, 6.409815788269043, 7.436576843261719, 2.8375442028045654, 3.5334537029266357, 7.603107452392578, 6.572993278503418, 6.495553970336914, 7.424954891204834, 14.594247817993164, 6.445587158203125, 8.714512825012207, -2.3895936012268066, 6.444178581237793, 6.967440128326416, 7.0400543212890625, 7.2723708152771, 7.029110908508301, 6.780109405517578, 6.425015449523926, -7.652756690979004, 14.285653114318848, 14.61624813079834, 1.6942468881607056, 7.45559549331665, -0.11330617219209671, -2.445556640625, -7.4670491218566895, 8.70724105834961, 10.341280937194824, 3.3257534503936768, 1.4152026176452637, -7.707159996032715, -2.022770404815674, 7.435182094573975, 1.1544344425201416, 1.2438428401947021, -7.154496669769287, 0.7203966975212097, 6.762577056884766, 6.214611053466797, 8.608418464660645, 1.369095802307129, 6.560862064361572, 8.65315055847168, 1.4884977340698242, 6.232783794403076, 8.9829683303833, 10.020332336425781, 6.972513675689697, 14.415773391723633, 6.962371349334717, -7.7021098136901855, 10.75783920288086, 7.343925476074219, 7.03959846496582, 7.299952030181885, 7.006407737731934, 7.588275909423828, 8.120488166809082, -0.291069895029068, -7.200498104095459, 9.625394821166992, 10.05850601196289, 6.95193338394165, 0.27479037642478943, -2.0256166458129883, 6.388875961303711, -1.0420703887939453, 10.681750297546387, 7.117461681365967, 1.5970230102539062, -2.171394109725952, 9.848308563232422, 7.273548603057861, 3.4561715126037598, 13.932007789611816, 7.064109802246094, 8.895830154418945, 8.574220657348633, 7.131933689117432, 13.626495361328125, 10.43896198272705, 10.330604553222656, -1.7489217519760132, -2.414266347885132, 3.2401740550994873, 6.749701499938965, 3.0941126346588135, 7.90279483795166, 1.2086728811264038, 8.50760555267334, 6.818399906158447, 9.890243530273438, 15.086627960205078, 6.211310386657715, 9.577631950378418, 14.283855438232422, 13.630316734313965, 9.295001983642578, 6.571330547332764, 1.346287488937378, -1.2155133485794067, 13.545226097106934, 6.34782600402832, 1.0492033958435059, -2.2268261909484863, 14.759235382080078, 6.488901615142822, -7.806366443634033, 6.921740531921387, 1.6035816669464111, 13.726323127746582, -6.971192359924316, -0.4138951003551483, 8.602763175964355, -1.6147187948226929, 1.9934295415878296, 14.351722717285156, 13.82508659362793, 7.027727127075195, 7.6633453369140625, -0.6898751258850098, -1.37373685836792, 9.121283531188965, -7.4509453773498535, 6.887499809265137, 13.52487850189209, 10.062641143798828, -7.145938396453857, 6.845681667327881, -7.673521518707275, 3.237545967102051, 14.578004837036133, 13.88511848449707, 6.017209053039551, 5.899812698364258, -1.2703449726104736, 6.7939653396606445, 6.823314666748047, -1.6751409769058228, 0.5296874642372131, 5.998839855194092, -1.3818988800048828, 9.858820915222168, 10.226036071777344, 7.1787614822387695, 6.802081108093262, -0.2645123302936554, 6.924002170562744, 6.5542521476745605, 7.352736473083496, 6.397611141204834, 7.2785844802856445, -1.9574575424194336, 8.36418628692627, -2.6332032680511475, 0.9517704248428345, -0.8953321576118469, -7.780147552490234, -2.163053274154663, 14.759176254272461, -2.8443543910980225, 14.707320213317871, -1.3045865297317505, 3.665403366088867, 0.5273827910423279, 3.71163272857666, 0.9152514338493347, -1.3503402471542358, 7.259063720703125, 6.790016174316406, 7.902483940124512, 7.046363830566406, 10.507686614990234, 7.373562335968018, -1.7098852396011353, 7.745702266693115, 3.641235113143921, 6.562802314758301, 8.908368110656738, 1.615328073501587, 10.710309028625488, 9.479954719543457, -1.8470782041549683, 8.264266967773438, -1.202014446258545, 13.707442283630371, 8.64285945892334, 9.542067527770996, 7.065338134765625, 10.511636734008789, 7.001969337463379, 7.625124931335449, 7.1512346267700195, 2.9915528297424316, 7.490354061126709, 8.030034065246582, 9.64097785949707, 9.810577392578125, 10.004535675048828, 6.219202041625977, 10.715811729431152, 7.424714088439941, -2.058943271636963, 8.607786178588867, -2.428840160369873, 6.269988536834717, 8.794548988342285, 6.3197126388549805, 10.143881797790527, 8.700421333312988, 10.470264434814453, 10.482549667358398, 7.742701530456543, 10.739598274230957, 8.98869514465332, 9.521886825561523, 7.240296840667725, 10.368749618530273, 10.204654693603516, 0.9026027321815491, 7.168976783752441, 9.738885879516602, 7.062613010406494, 1.1436067819595337, 7.356151580810547, -2.368579864501953, 9.57328987121582, 8.493727684020996, 9.967802047729492, 10.289021492004395, 1.1926547288894653, 1.3424636125564575, 13.64143180847168, 7.997675895690918, 7.143918037414551, 9.075204849243164, 6.491727352142334, -0.19032050669193268, 8.905466079711914, -1.1945412158966064, 6.624124050140381, -1.5514732599258423, -7.766822814941406, 7.4337477684021, 7.337843894958496, -0.9831737279891968, 7.039582252502441, 6.927689552307129, 8.444642066955566, -7.051167964935303, 6.8605804443359375, 7.276669979095459, 9.997711181640625, 14.777616500854492, -1.7565499544143677, 1.4060418605804443, 9.746705055236816, 8.631242752075195, 7.0370917320251465, 10.463733673095703, 7.823729991912842, 6.341503620147705, 10.33910083770752, 6.744711875915527, 8.433073997497559, 3.550858974456787, 9.874295234680176, 0.3988375961780548, -1.046040415763855, 8.63628101348877, -1.1065592765808105, -1.719969391822815, 13.685298919677734, 10.179312705993652, -1.592015027999878, -0.8667829036712646, 9.389472007751465, -7.59422492980957, 7.241970539093018, 6.155146598815918, 9.008370399475098, 15.027610778808594, 7.236214637756348, 8.192572593688965, 7.084733009338379, 7.415927410125732, 9.169605255126953, 3.5132431983947754, 8.871159553527832, 10.784748077392578, 7.107335567474365, -0.40607255697250366, 7.084897041320801, 7.634235858917236, 14.26550579071045, -2.2803492546081543, 6.689552307128906, 10.461997032165527, 2.884526491165161, 8.657809257507324, 9.628214836120605, 8.259615898132324, 14.617714881896973, 9.880496978759766, 8.62523365020752, 8.860508918762207, 10.002365112304688, 6.243504047393799, -7.05561637878418, 6.989813327789307, 0.5969406962394714, 7.796088695526123, 6.645824432373047, 6.302118301391602, 8.760086059570312, 2.9767425060272217, 6.460657596588135, 6.711520671844482, 7.00267219543457, -2.6222753524780273, 7.684745788574219, 6.731116771697998, 7.829660415649414, -7.719557762145996, 7.198794841766357, 3.019371747970581, 7.36558723449707, 8.89859676361084, 7.369990825653076, 6.383822917938232, 9.990866661071777, 10.290257453918457, 14.826529502868652, 7.060043811798096, -1.8219455480575562, -1.1505897045135498, 10.003043174743652, 13.621362686157227, 8.13591194152832, 9.943227767944336, 6.7864089012146, 10.475676536560059, 7.005962371826172, 2.0521085262298584, 10.417001724243164, 7.694455623626709, 1.2368733882904053, 6.2891950607299805, -1.4598170518875122, 9.36068344116211, 3.4051458835601807, -0.975776195526123, 7.070476055145264, 0.37016865611076355, 5.967078685760498, -7.317153453826904, 9.138036727905273, 6.00955867767334, 5.887032985687256, 6.336053371429443, 6.315256118774414, 5.991295337677002, 0.7407580614089966, -1.6258493661880493, 6.376094341278076, 8.576647758483887, 10.12857723236084, 7.291077613830566, 3.0620944499969482, 8.003541946411133, -2.54970121383667, 6.786853313446045, 7.445936679840088, 5.9561567306518555, -7.068813323974609, 9.585966110229492, 7.494814395904541, 9.11490535736084, 3.076767683029175, 10.726910591125488, 7.35984992980957, 13.763833999633789, 8.605511665344238, 10.61301326751709, -7.51236629486084, 9.395983695983887, -7.648815155029297, 7.209524154663086, 10.540234565734863, -1.8935903310775757, 6.990942001342773, -6.980566024780273, 6.52692985534668, 10.549281120300293, 0.07273968309164047, 9.947211265563965, 6.623125076293945, 3.550461769104004, -7.791592121124268, 8.812898635864258, -0.16511745750904083, -1.8439090251922607, -7.592961311340332, 14.280760765075684, 7.096899032592773, -1.3681060075759888, 6.881974697113037, 6.891568660736084, 6.718873977661133, 7.005202293395996, -7.062305927276611, 14.712048530578613, 8.828021049499512, 10.250944137573242, 14.262808799743652, -7.103879928588867, 3.540095090866089, 6.878340244293213, 0.5039257407188416, -0.3975401520729065, 6.60114860534668, 7.098932266235352, 7.1024169921875, -1.7640321254730225, 8.97040843963623, 6.485232830047607, 6.249810218811035, 6.727346420288086, -7.800436019897461, 9.113037109375, 6.527957916259766, 7.3304443359375, 6.1981353759765625, 6.465376377105713, 9.511240005493164, -0.46945610642433167, 6.461258411407471, -1.9005166292190552, 2.8063361644744873, 6.789132118225098, 7.8854570388793945, 7.000997543334961, 7.075256824493408, 13.495193481445312, 3.7970070838928223, 6.073390960693359, 8.42492961883545, 6.857657432556152, 8.669413566589355, 6.9685139656066895, 8.555603981018066, 6.352323055267334, 9.376934051513672, 7.5588698387146, 6.058804988861084, 7.431726455688477, 6.027827739715576, 5.9137372970581055, 7.392441749572754, 9.599349975585938, 3.0737881660461426, 5.993553161621094, 3.8097407817840576, 9.614864349365234, 14.953315734863281, -2.2594940662384033, -1.5415993928909302, -1.2431851625442505, 2.2140657901763916, -2.437441110610962, 9.345711708068848], \"xaxis\": \"x\", \"y\": [0.48793599009513855, -1.5452311038970947, 0.42881500720977783, 0.426017701625824, -1.2909859418869019, -0.14745834469795227, -0.8531692028045654, -0.4167153239250183, 0.08424302190542221, 0.423944354057312, -0.3538450300693512, 0.20034360885620117, 0.7272301316261292, -1.3028547763824463, 0.4764725863933563, 0.31166598200798035, -0.7447296380996704, 0.7530995011329651, 0.42940667271614075, 1.3611148595809937, 5.53173303604126, 7.073292255401611, 0.9176209568977356, 0.38388729095458984, -0.32305705547332764, 0.47896069288253784, 0.20065538585186005, -1.1686500310897827, 0.6138613820075989, -1.3624051809310913, 0.16262368857860565, -0.2490970641374588, -0.1422220766544342, -0.25939276814460754, 0.262382447719574, 0.16498003900051117, -0.4396313726902008, 10.292525291442871, 11.430914878845215, -0.5034114718437195, -0.12472032010555267, 0.6859406232833862, 0.9055253267288208, -0.4821527302265167, -0.5497678518295288, 0.30912333726882935, 10.393247604370117, -0.3670876622200012, 6.275266170501709, -0.1693858951330185, 10.180597305297852, 7.33576774597168, 8.662717819213867, -1.0657885074615479, -0.7612611651420593, 8.900816917419434, -0.2443198710680008, -0.33193063735961914, 10.247383117675781, 7.036693572998047, 10.110567092895508, 10.797184944152832, 0.8117994666099548, 9.756526947021484, -0.3867999315261841, -0.30542632937431335, 10.915205955505371, 8.572098731994629, 0.6142484545707703, 8.258354187011719, -1.569041132926941, 1.262422800064087, 1.0022097826004028, 0.8330694437026978, 7.097772598266602, -1.5089728832244873, 5.769745826721191, 0.3719044625759125, 0.9946895837783813, 7.106398105621338, 10.956310272216797, -0.47600215673446655, 1.303473711013794, -0.0555417463183403, 8.494006156921387, -1.1207537651062012, -1.3333481550216675, -0.6290522217750549, 4.382893085479736, 0.6173334121704102, 13.680119514465332, -1.333234429359436, 7.21936559677124, 7.050798416137695, 0.6279211640357971, -0.3731462359428406, 11.200569152832031, 10.918225288391113, 6.56321382522583, -1.3092148303985596, -0.5917897820472717, 10.556314468383789, 13.374911308288574, 4.570043087005615, 9.836128234863281, -0.11910675466060638, 4.345412254333496, 7.197591304779053, 1.1705265045166016, 4.820868492126465, 4.22307825088501, 2.391639471054077, 0.5954300761222839, 3.3681323528289795, 6.953230857849121, 4.688107967376709, 1.0023270845413208, 0.49839475750923157, 0.8103750348091125, -1.6330697536468506, 6.620508670806885, -1.0372341871261597, 1.14015793800354, 1.2093318700790405, 5.083164215087891, 7.51092529296875, -0.7960591316223145, -1.6587718725204468, 0.5721498727798462, 0.1480509489774704, 5.644845008850098, -0.7467046976089478, -0.758892297744751, 1.3222339153289795, -0.8892146944999695, 13.387002944946289, 11.250093460083008, 7.454916954040527, -0.6483687162399292, 6.0026140213012695, 10.426116943359375, -1.4279354810714722, 10.999178886413574, 6.1367597579956055, 9.815289497375488, 3.47391676902771, -0.35553988814353943, 2.2004153728485107, 0.9926466345787048, 7.437402248382568, 9.706755638122559, 1.0229439735412598, 0.6179124116897583, 7.192514896392822, 5.396235942840576, 8.349411010742188, -0.8879777193069458, 1.2553374767303467, 8.316267013549805, 9.780818939208984, 3.972139358520508, -1.4371471405029297, 0.4687436521053314, 4.435866832733154, 0.5516431331634521, 7.5634074211120605, 6.942897319793701, -1.5690339803695679, 0.004225464072078466, 6.799490928649902, 6.475052356719971, -0.6404564380645752, 0.6631082892417908, 10.786191940307617, 10.374382019042969, 10.41006851196289, 1.0592470169067383, 0.6227421164512634, 9.358311653137207, 9.172981262207031, 1.3181841373443604, 1.0930951833724976, -0.8525405526161194, 6.444741725921631, -1.1182607412338257, 10.457610130310059, -1.1038864850997925, 7.392288684844971, 10.356142044067383, 9.002006530761719, 0.46634477376937866, 11.361794471740723, 10.511531829833984, -0.32074424624443054, 0.6225264668464661, 6.887243270874023, 0.600061297416687, -1.6822072267532349, -0.634355366230011, -0.3133559823036194, 4.51413631439209, -0.9360080361366272, 1.1905533075332642, 0.26878416538238525, 4.250972270965576, 0.8068145513534546, 10.259856224060059, -1.6003726720809937, -0.6176830530166626, 1.4047160148620605, -0.5318691730499268, 7.717662334442139, 1.2657700777053833, -1.5044620037078857, -0.40486496686935425, -0.49206480383872986, 2.5591304302215576, 9.727156639099121, -0.30434107780456543, -0.9136313796043396, 5.93593168258667, 1.237768530845642, 10.241353988647461, -0.19210858643054962, 7.026796340942383, 1.3168543577194214, 7.398721694946289, 0.9343555569648743, -0.9237654209136963, -0.039387837052345276, 0.773474931716919, 13.601417541503906, 13.707921028137207, -0.9281313419342041, 7.26752233505249, 13.55929183959961, 1.2732328176498413, -1.57001531124115, 14.180320739746094, -0.8848802447319031, 6.7902021408081055, 9.562844276428223, 0.6910272240638733, 13.670600891113281, -0.9316262602806091, 10.823725700378418, 13.199018478393555, 0.49810120463371277, 12.86065673828125, 6.296367168426514, 0.826812207698822, 8.687305450439453, 0.3679589033126831, -0.9193539619445801, -0.3267610967159271, 0.6911585330963135, 0.11806946992874146, 0.8729304075241089, 0.6074195504188538, 0.42643359303474426, -0.7879854440689087, -0.3698439300060272, -1.6199817657470703, -0.27739381790161133, -0.8524459600448608, -0.8907879590988159, 3.8923323154449463, 13.582197189331055, 8.481605529785156, 2.9307150840759277, 9.155315399169922, 6.602994918823242, 1.1013588905334473, 10.276154518127441, -0.4127325713634491, 6.012455940246582, 10.38570785522461, -1.5949881076812744, 7.586503028869629, 6.51371431350708, -0.33135655522346497, 8.971946716308594, -0.25992289185523987, 0.6336514353752136, 5.0244975090026855, 10.305957794189453, 1.9913843870162964, 9.188385009765625, 2.260573148727417, 9.4456148147583, 1.0371850728988647, -1.2336256504058838, 5.512850761413574, 8.34124755859375, 8.319735527038574, 8.32563304901123, 10.083661079406738, 11.2650728225708, 7.52648401260376, 10.490169525146484, 0.44359275698661804, 4.988302230834961, 0.6782179474830627, 10.934152603149414, 8.373157501220703, 13.067133903503418, 9.067683219909668, 10.3931303024292, 9.225625991821289, 9.305472373962402, 9.049790382385254, 7.615149021148682, 5.6996917724609375, 9.304085731506348, 10.670286178588867, 9.430870056152344, 9.69528865814209, -0.8879130482673645, 0.7596744298934937, 10.281906127929688, 3.4669501781463623, -1.664016604423523, 6.124094486236572, 1.028402328491211, 8.279532432556152, 7.9839067459106445, 8.22460651397705, 9.597415924072266, -0.6695278286933899, -0.6678286790847778, 0.578485906124115, 4.801232814788818, 10.639878273010254, 10.304850578308105, 13.0855712890625, -0.8373318910598755, 5.619734287261963, -0.5279978513717651, 10.566629409790039, 1.1831254959106445, 0.7810271382331848, 4.108271598815918, 5.593693256378174, -0.0008494447683915496, 6.0000176429748535, 4.083144187927246, 8.71723461151123, 1.3902535438537598, 7.20408821105957, 9.817147254943848, 9.916234970092773, 0.8707109689712524, 1.2990562915802002, -0.5912322402000427, 9.838094711303711, 4.991353511810303, 2.399548292160034, 7.558347225189209, 9.784781455993652, 10.880108833312988, 9.54172134399414, 13.503247261047363, 7.190859794616699, -0.5219281911849976, 7.146030426025391, -1.5361649990081787, 0.03173055499792099, 7.060883045196533, 0.6202570199966431, 1.228576898574829, 0.48749351501464844, 6.934429168701172, 0.20099706947803497, 0.15682268142700195, 6.816528797149658, 1.0975652933120728, 0.6301451325416565, 13.906280517578125, 5.751008033752441, 0.3980858325958252, 5.584066390991211, 8.48827075958252, 1.9546359777450562, 3.9981651306152344, 5.996786594390869, -0.5746351480484009, 5.4405646324157715, 7.78877067565918, 1.7448986768722534, -0.5596184730529785, 6.972391128540039, 10.554252624511719, 0.8007581233978271, 1.1056697368621826, 6.337009906768799, 7.453500270843506, -1.3053303956985474, 5.0329060554504395, 8.4481782913208, 8.222089767456055, 0.8990382552146912, 8.484827995300293, 5.026040554046631, 6.84797477722168, 10.008411407470703, 11.034387588500977, 1.3902357816696167, 2.2055766582489014, -0.823835015296936, 8.614340782165527, 13.301740646362305, 10.849873542785645, 10.281271934509277, -1.2423136234283447, 13.888411521911621, 6.0755615234375, 10.748187065124512, 0.2249225527048111, 6.7489824295043945, 10.382070541381836, 8.613752365112305, 0.5297871828079224, 5.723329544067383, -1.2379645109176636, 10.213062286376953, 10.262688636779785, 4.024892330169678, 12.959173202514648, 9.890789031982422, 9.590620040893555, 0.7498526573181152, 3.5807769298553467, 0.22463418543338776, -0.6694529056549072, 6.899946212768555, -0.5669720768928528, 8.393693923950195, 6.92329216003418, 13.51544189453125, 7.2147932052612305, 2.028538227081299, -1.4682203531265259, 9.167340278625488, 9.867060661315918, -1.123052954673767, 4.370176792144775, -0.7811148762702942, 6.433421611785889, -0.7042597532272339, 0.35067233443260193, 1.9613159894943237, -0.8625229597091675, 14.012819290161133, 1.3103315830230713, 6.788893222808838, 13.99601936340332, 14.0335054397583, 13.759176254272461, 13.774553298950195, 13.760664939880371, -0.7182263731956482, 1.2643874883651733, 13.857757568359375, 7.399825096130371, 7.067091941833496, 0.5076172351837158, -1.1894323825836182, 10.385092735290527, 0.16903002560138702, 0.7744143605232239, 9.785737037658691, 13.780832290649414, 1.3502458333969116, 10.5431547164917, 4.230116844177246, 5.9357218742370605, -1.1873884201049805, 7.562447547912598, 5.728317737579346, 0.6634204387664795, 7.706634044647217, 7.41428804397583, 1.2050533294677734, 6.637863636016846, 1.0227599143981934, 0.7112186551094055, 7.25976037979126, 0.19432303309440613, 5.91511869430542, 1.377578616142273, 13.378388404846191, 7.087973117828369, -1.192227840423584, 10.1219482421875, 13.333117485046387, -0.5145602822303772, 0.7320784330368042, 6.981478214263916, -0.9490258097648621, 0.27824702858924866, 1.0968482494354248, -0.36203014850616455, 5.782588481903076, 1.0360331535339355, 0.9611036777496338, 0.9673284292221069, 4.476019382476807, 3.5338783264160156, 1.3840810060501099, 0.7155217528343201, 5.343787670135498, 6.952510356903076, -0.24617189168930054, 1.354537010192871, -0.5518269538879395, 6.111108303070068, -0.8632914423942566, -0.8296515345573425, 4.034552574157715, 5.751128673553467, 3.5091795921325684, 1.2900118827819824, 10.231402397155762, 4.218331336975098, 10.935223579406738, 7.008031845092773, 0.764197587966919, 6.6887407302856445, 6.0037150382995605, 6.415095329284668, 11.076583862304688, 4.725776672363281, 6.775763511657715, -0.516122579574585, 4.53118896484375, 0.34137606620788574, -1.3278976678848267, 0.7813781499862671, 10.474380493164062, 1.0026403665542603, 3.5920331478118896, -0.48887908458709717, -0.06616827100515366, 13.977555274963379, 6.897221565246582, 13.64533805847168, 10.334099769592285, 4.369426250457764, 10.285785675048828, 12.98662281036377, 10.326045036315918, 9.498224258422852, 13.979467391967773, 10.535978317260742, 13.992179870605469, 14.079889297485352, 10.543787002563477, 10.548989295959473, -1.171920657157898, 14.009977340698242, -0.08211696892976761, 10.5724515914917, 0.5463722348213196, 1.1017487049102783, 1.1453951597213745, -0.8508827090263367, -1.451305866241455, 1.1216661930084229, 10.389026641845703], \"yaxis\": \"y\"}],\n",
              "                        {\"autosize\": false, \"coloraxis\": {\"colorbar\": {\"title\": {\"text\": \"diagnosis\"}}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]]}, \"height\": 800, \"legend\": {\"tracegroupgap\": 0}, \"margin\": {\"t\": 60}, \"template\": {\"data\": {\"bar\": [{\"error_x\": {\"color\": \"#2a3f5f\"}, \"error_y\": {\"color\": \"#2a3f5f\"}, \"marker\": {\"line\": {\"color\": \"#E5ECF6\", \"width\": 0.5}}, \"type\": \"bar\"}], \"barpolar\": [{\"marker\": {\"line\": {\"color\": \"#E5ECF6\", \"width\": 0.5}}, \"type\": \"barpolar\"}], \"carpet\": [{\"aaxis\": {\"endlinecolor\": \"#2a3f5f\", \"gridcolor\": \"white\", \"linecolor\": \"white\", \"minorgridcolor\": \"white\", \"startlinecolor\": \"#2a3f5f\"}, \"baxis\": {\"endlinecolor\": \"#2a3f5f\", \"gridcolor\": \"white\", \"linecolor\": \"white\", \"minorgridcolor\": \"white\", \"startlinecolor\": \"#2a3f5f\"}, \"type\": \"carpet\"}], \"choropleth\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"choropleth\"}], \"contour\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"contour\"}], \"contourcarpet\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"contourcarpet\"}], \"heatmap\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"heatmap\"}], \"heatmapgl\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"heatmapgl\"}], \"histogram\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"histogram\"}], \"histogram2d\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"histogram2d\"}], \"histogram2dcontour\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"histogram2dcontour\"}], \"mesh3d\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"mesh3d\"}], \"parcoords\": [{\"line\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"parcoords\"}], \"pie\": [{\"automargin\": true, \"type\": \"pie\"}], \"scatter\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatter\"}], \"scatter3d\": [{\"line\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatter3d\"}], \"scattercarpet\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattercarpet\"}], \"scattergeo\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattergeo\"}], \"scattergl\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattergl\"}], \"scattermapbox\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattermapbox\"}], \"scatterpolar\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterpolar\"}], \"scatterpolargl\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterpolargl\"}], \"scatterternary\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterternary\"}], \"surface\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"surface\"}], \"table\": [{\"cells\": {\"fill\": {\"color\": \"#EBF0F8\"}, \"line\": {\"color\": \"white\"}}, \"header\": {\"fill\": {\"color\": \"#C8D4E3\"}, \"line\": {\"color\": \"white\"}}, \"type\": \"table\"}]}, \"layout\": {\"annotationdefaults\": {\"arrowcolor\": \"#2a3f5f\", \"arrowhead\": 0, \"arrowwidth\": 1}, \"coloraxis\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"colorscale\": {\"diverging\": [[0, \"#8e0152\"], [0.1, \"#c51b7d\"], [0.2, \"#de77ae\"], [0.3, \"#f1b6da\"], [0.4, \"#fde0ef\"], [0.5, \"#f7f7f7\"], [0.6, \"#e6f5d0\"], [0.7, \"#b8e186\"], [0.8, \"#7fbc41\"], [0.9, \"#4d9221\"], [1, \"#276419\"]], \"sequential\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"sequentialminus\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]]}, \"colorway\": [\"#636efa\", \"#EF553B\", \"#00cc96\", \"#ab63fa\", \"#FFA15A\", \"#19d3f3\", \"#FF6692\", \"#B6E880\", \"#FF97FF\", \"#FECB52\"], \"font\": {\"color\": \"#2a3f5f\"}, \"geo\": {\"bgcolor\": \"white\", \"lakecolor\": \"white\", \"landcolor\": \"#E5ECF6\", \"showlakes\": true, \"showland\": true, \"subunitcolor\": \"white\"}, \"hoverlabel\": {\"align\": \"left\"}, \"hovermode\": \"closest\", \"mapbox\": {\"style\": \"light\"}, \"paper_bgcolor\": \"white\", \"plot_bgcolor\": \"#E5ECF6\", \"polar\": {\"angularaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"bgcolor\": \"#E5ECF6\", \"radialaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}}, \"scene\": {\"xaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}, \"yaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}, \"zaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}}, \"shapedefaults\": {\"line\": {\"color\": \"#2a3f5f\"}}, \"ternary\": {\"aaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"baxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"bgcolor\": \"#E5ECF6\", \"caxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}}, \"title\": {\"x\": 0.05}, \"xaxis\": {\"automargin\": true, \"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\", \"title\": {\"standoff\": 15}, \"zerolinecolor\": \"white\", \"zerolinewidth\": 2}, \"yaxis\": {\"automargin\": true, \"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\", \"title\": {\"standoff\": 15}, \"zerolinecolor\": \"white\", \"zerolinewidth\": 2}}}, \"width\": 800, \"xaxis\": {\"anchor\": \"y\", \"domain\": [0.0, 1.0], \"title\": {\"text\": \"0\"}}, \"yaxis\": {\"anchor\": \"x\", \"domain\": [0.0, 1.0], \"title\": {\"text\": \"1\"}}},\n",
              "                        {\"responsive\": true}\n",
              "                    ).then(function(){\n",
              "                            \n",
              "var gd = document.getElementById('0493233e-e6a5-4ecf-8001-da119dcd437f');\n",
              "var x = new MutationObserver(function (mutations, observer) {{\n",
              "        var display = window.getComputedStyle(gd).display;\n",
              "        if (!display || display === 'none') {{\n",
              "            console.log([gd, 'removed!']);\n",
              "            Plotly.purge(gd);\n",
              "            observer.disconnect();\n",
              "        }}\n",
              "}});\n",
              "\n",
              "// Listen for the removal of the full notebook cells\n",
              "var notebookContainer = gd.closest('#notebook-container');\n",
              "if (notebookContainer) {{\n",
              "    x.observe(notebookContainer, {childList: true});\n",
              "}}\n",
              "\n",
              "// Listen for the clearing of the current output cell\n",
              "var outputEl = gd.closest('.output');\n",
              "if (outputEl) {{\n",
              "    x.observe(outputEl, {childList: true});\n",
              "}}\n",
              "\n",
              "                        })\n",
              "                };\n",
              "                \n",
              "            </script>\n",
              "        </div>\n",
              "</body>\n",
              "</html>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<html>\n",
              "<head><meta charset=\"utf-8\" /></head>\n",
              "<body>\n",
              "    <div>\n",
              "            <script src=\"https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/MathJax.js?config=TeX-AMS-MML_SVG\"></script><script type=\"text/javascript\">if (window.MathJax) {MathJax.Hub.Config({SVG: {font: \"STIX-Web\"}});}</script>\n",
              "                <script type=\"text/javascript\">window.PlotlyConfig = {MathJaxConfig: 'local'};</script>\n",
              "        <script src=\"https://cdn.plot.ly/plotly-latest.min.js\"></script>    \n",
              "            <div id=\"c59b2ecc-e477-4a92-bc77-09ea6032765f\" class=\"plotly-graph-div\" style=\"height:800px; width:800px;\"></div>\n",
              "            <script type=\"text/javascript\">\n",
              "                \n",
              "                    window.PLOTLYENV=window.PLOTLYENV || {};\n",
              "                    \n",
              "                if (document.getElementById(\"c59b2ecc-e477-4a92-bc77-09ea6032765f\")) {\n",
              "                    Plotly.newPlot(\n",
              "                        'c59b2ecc-e477-4a92-bc77-09ea6032765f',\n",
              "                        [{\"hoverlabel\": {\"namelength\": 0}, \"hovertemplate\": \"0=%{x}<br>1=%{y}<br>2=%{z}<br>diagnosis=%{marker.color}\", \"legendgroup\": \"\", \"marker\": {\"color\": [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 1, 0, 1, 1, 0, 1, 0, 1, 1, 0, 0, 0, 1, 1, 0, 1, 1, 1, 0, 0, 0, 1, 0, 0, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 1, 1, 0, 0, 0, 1, 1, 0, 1, 0, 1, 1, 0, 1, 1, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 1, 1, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1, 0, 0, 1, 0, 1, 1, 1, 1, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 1, 1, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 0], \"coloraxis\": \"coloraxis\", \"size\": 5, \"symbol\": \"circle\"}, \"mode\": \"markers\", \"name\": \"\", \"scene\": \"scene\", \"showlegend\": false, \"type\": \"scatter3d\", \"x\": [-2.86033034324646, 1.001710057258606, -2.0994575023651123, -3.416468620300293, 0.5483139753341675, -3.0961527824401855, 1.1601094007492065, -2.8714263439178467, -3.1449732780456543, -3.393594980239868, 4.319056034088135, -2.6815474033355713, -2.9132637977600098, 3.036375045776367, -3.3497204780578613, -3.2855241298675537, 3.8036725521087646, -3.0971591472625732, -1.1151926517486572, 3.7312519550323486, 5.682247638702393, 6.266336917877197, -3.253650188446045, -1.199439525604248, -1.5271825790405273, -2.80816388130188, -3.2056984901428223, 0.45075106620788574, -3.156269073486328, 1.6891164779663086, -2.0007007122039795, -3.0919687747955322, -2.263254165649414, -1.8993988037109375, -2.668431282043457, -2.6215343475341797, -2.4883127212524414, 10.254443168640137, 7.466062545776367, -2.481015920639038, 4.583946704864502, 2.357806444168091, -2.923902750015259, -2.5361146926879883, -2.433443069458008, -2.51310396194458, 9.392982482910156, -2.987891912460327, 5.982173442840576, 4.526190757751465, 9.960403442382812, 9.175015449523926, 8.512411117553711, 0.4465941786766052, 3.874948263168335, 8.28681468963623, -0.9712494015693665, -2.6938343048095703, 10.135273933410645, 6.123859405517578, 8.593109130859375, 8.435705184936523, -3.299072504043579, 8.767200469970703, -2.8048994541168213, -2.6760928630828857, 8.123929977416992, 8.654565811157227, 2.707157611846924, 8.508742332458496, 0.97251957654953, 2.9859161376953125, -3.0458831787109375, 3.1614577770233154, 7.789201736450195, 1.9872187376022339, 5.346060276031494, -2.6557695865631104, -3.0381922721862793, 7.444102764129639, 8.095805168151855, -2.613429069519043, -2.583634376525879, -2.146104097366333, 7.884896755218506, 0.40730157494544983, 3.1582140922546387, -0.8348960280418396, 3.941080093383789, 2.9800496101379395, 5.11523962020874, 3.070321559906006, 9.145557403564453, 7.38519287109375, -2.751925230026245, -0.40863505005836487, 7.574934005737305, 7.95187520980835, 6.294837474822998, 3.238954544067383, 4.076730728149414, 8.901715278625488, 5.456803321838379, 4.134198188781738, 8.465851783752441, -3.1984353065490723, 4.029675483703613, 7.773410320281982, -3.0411956310272217, 4.2508440017700195, 4.208566665649414, 3.455784797668457, 2.7568395137786865, 3.929500102996826, 6.001187324523926, 4.262729644775391, 3.3234989643096924, -3.077967643737793, -3.2313082218170166, 1.9179885387420654, 6.276680946350098, 0.37171968817710876, -2.965052604675293, 3.5184125900268555, 6.183507919311523, 9.210317611694336, 3.7653071880340576, 1.6134065389633179, 2.9393696784973145, -1.9335570335388184, 5.591219425201416, 1.3727935552597046, 1.4605568647384644, 3.8861353397369385, 1.1391907930374146, 5.194340229034424, 7.568573951721191, 7.773401737213135, 1.4939736127853394, 5.105597972869873, 9.70831298828125, 1.8144781589508057, 7.805683612823486, 7.161498069763184, 10.141023635864258, 3.883957862854004, -3.047093152999878, 3.3790087699890137, 3.22204327583313, 9.356182098388672, 8.351038932800293, 2.669965982437134, 2.7618467807769775, 6.399940490722656, 5.585775375366211, 7.950855731964111, 0.09407100826501846, 3.8769352436065674, 9.06053352355957, 10.128366470336914, 4.084866523742676, 0.810878336429596, -2.1888904571533203, 4.010176658630371, -1.302199363708496, 9.502012252807617, 6.289163112640381, 1.9158254861831665, -1.9799436330795288, 8.360233306884766, 6.112257480621338, 3.99808931350708, -2.883493661880493, 8.574701309204102, 9.90886116027832, 9.666041374206543, 2.8773977756500244, -2.7295281887054443, 10.294376373291016, 10.045721054077148, -2.3957066535949707, -2.912672996520996, 3.600996255874634, 5.638329029083252, 3.5328519344329834, 9.106531143188477, 1.5799813270568848, 7.7012410163879395, 8.216019630432129, 9.580497741699219, -3.4234206676483154, 7.553493976593018, 10.26326847076416, -2.911241292953491, -2.7809133529663086, 7.940094470977783, 2.0060722827911377, 1.7336281538009644, -0.9193949699401855, -2.55880069732666, 4.01445198059082, 1.4010125398635864, -2.7831602096557617, -3.2479186058044434, 3.9857242107391357, 3.161332845687866, 8.303244590759277, 1.8772616386413574, -2.481799840927124, 3.9919891357421875, -0.07556437700986862, 7.95417594909668, -2.1476008892059326, 2.379074811935425, -2.9368250370025635, -2.6439101696014404, 3.5678420066833496, 8.565545082092285, -0.4874568283557892, -0.878042459487915, 7.155594825744629, 3.501573324203491, 8.244972229003906, -2.590562343597412, 8.662515640258789, 3.8454229831695557, 6.397367477416992, 3.3078973293304443, 3.8255653381347656, -3.166585683822632, -2.886845827102661, 5.593207359313965, 5.613577365875244, -0.8924382925033569, 6.235696792602539, 5.107388496398926, -2.2523691654205322, 1.0473442077636719, 5.2611565589904785, -0.9201093316078186, 8.343658447265625, 10.23475456237793, 2.7354419231414795, 5.113016605377197, 0.1311921626329422, 8.455437660217285, 5.5602216720581055, 2.78109073638916, 5.844277858734131, 5.963306903839111, -2.286457061767578, 8.345451354980469, -2.6171298027038574, 1.3871251344680786, -0.7489988803863525, 3.040736675262451, -2.0086281299591064, -3.2694339752197266, -2.8715860843658447, -3.286738157272339, -0.9627606868743896, 4.303734302520752, 1.0125676393508911, 4.433174133300781, 1.4208238124847412, -0.8734204173088074, 4.25884485244751, 5.063185214996338, 7.900373935699463, 3.706141948699951, 10.255509376525879, 6.231386661529541, -2.2202279567718506, 8.945951461791992, 4.250635623931885, 5.137105941772461, 9.890742301940918, 1.909804105758667, 9.43259334564209, 7.872532844543457, -1.6131073236465454, 8.569825172424316, -1.0548385381698608, -2.755580425262451, 6.131004333496094, 10.115643501281738, 3.308098793029785, 10.33829116821289, 3.3562543392181396, 8.411463737487793, 2.8997175693511963, 3.408060312271118, 5.552235126495361, 7.855210781097412, 9.087295532226562, 9.239826202392578, 10.249107360839844, 7.642376899719238, 9.390121459960938, 8.754630088806152, -2.199646234512329, 6.095958232879639, -2.5310332775115967, 7.835602283477783, 8.625125885009766, 5.7396321296691895, 9.883273124694824, 9.740543365478516, 10.255486488342285, 10.344849586486816, 8.210861206054688, 9.57674503326416, 6.912332534790039, 9.53919506072998, 8.643000602722168, 10.341703414916992, 10.269251823425293, 1.3297169208526611, 2.7935967445373535, 10.238212585449219, 3.9059441089630127, 1.4355170726776123, 5.884456634521484, -2.8229823112487793, 9.013012886047363, 8.068039894104004, 9.269216537475586, 10.269000053405762, 1.3616729974746704, 1.506206750869751, -2.714561700820923, 5.457807540893555, 8.63849925994873, 9.879310607910156, 5.594261646270752, 0.1833803355693817, 6.730654239654541, -0.9682711958885193, 8.123896598815918, -2.071781635284424, 3.1496052742004395, 4.630393028259277, 5.516786098480225, -0.9186223149299622, 5.517317771911621, 4.068914413452148, 8.470163345336914, 3.913034677505493, 6.219438552856445, 8.3622465133667, 10.155352592468262, -3.228229284286499, -2.456102132797241, 1.5181143283843994, 10.034930229187012, 6.04117488861084, 3.495090961456299, 9.275955200195312, 8.762245178222656, 7.860315799713135, 10.332494735717773, 5.1775736808776855, 7.530772686004639, 4.205385684967041, 8.666284561157227, 0.9174660444259644, -0.9871084094047546, 7.550625324249268, -1.3349130153656006, -2.2671399116516113, -2.7960715293884277, 8.683871269226074, -1.5913264751434326, -0.8996227979660034, 7.996294975280762, 3.340686559677124, 2.7527341842651367, 5.2901692390441895, 6.974185943603516, -3.41483211517334, 5.416343688964844, 8.15241527557373, 3.231187343597412, 4.518407821655273, 7.114490032196045, 4.081201553344727, 6.6237945556640625, 9.637335777282715, 3.1295173168182373, -0.04267698898911476, 6.2943644523620605, 8.929035186767578, -3.0025522708892822, -2.792184829711914, 5.503355503082275, 9.198920249938965, 3.312028646469116, 6.153950214385986, 9.103046417236328, 8.028931617736816, -3.1702640056610107, 9.269536972045898, 6.104498863220215, 7.601557731628418, 10.219655990600586, 7.696996212005615, 3.873138666152954, 3.413984775543213, 0.9962894916534424, 7.892282485961914, 5.364288330078125, 7.82063627243042, 9.693979263305664, 3.4723122119903564, 5.134345054626465, 5.247305870056152, 8.43701171875, -2.5447757244110107, 6.45717191696167, 8.212892532348633, 7.897321701049805, 2.9150259494781494, 5.44515323638916, 3.437992572784424, 8.56588363647461, 9.787589073181152, 4.442204475402832, 5.802074909210205, 10.143617630004883, 10.276666641235352, -3.3067431449890137, 4.00310754776001, -1.798657774925232, -0.8025072813034058, 8.547417640686035, -2.46592378616333, 8.029029846191406, 8.492691040039062, 5.1294779777526855, 9.09478759765625, 3.2877230644226074, 2.4552252292633057, 10.230592727661133, 8.759652137756348, 1.5815038681030273, 3.8881001472473145, -1.0021374225616455, 7.737822532653809, 4.064568519592285, -1.1420234441757202, 3.2354674339294434, 0.6827773451805115, 5.38158655166626, 3.675194501876831, 7.808527946472168, 5.317652225494385, 5.426402568817139, 5.223071098327637, 5.306667804718018, 5.53346586227417, 1.0311102867126465, -2.2225801944732666, 5.206117630004883, 7.737478733062744, 8.86092758178711, 2.817781925201416, 3.485908269882202, 9.178605079650879, -2.480154275894165, 2.4477217197418213, 8.51678466796875, 5.588367462158203, 3.9850780963897705, 10.257084846496582, 4.809813499450684, 7.1086883544921875, 3.574145793914795, 9.470614433288574, 5.63514518737793, -2.7847790718078613, 7.958466053009033, 9.286389350891113, 3.4880309104919434, 7.915488243103027, 3.2829713821411133, 2.771357297897339, 9.120394706726074, -1.856128454208374, 5.436006546020508, 3.9666833877563477, 5.422667980194092, 8.981321334838867, 0.5063862204551697, 10.229572296142578, 5.348793029785156, 4.208568096160889, 3.0023353099823, 7.6369500160217285, 0.31654515862464905, -1.873915672302246, 3.415498733520508, -2.97400164604187, 5.399914264678955, -1.838797688484192, 2.5891973972320557, 2.5980916023254395, 4.298734664916992, 3.9683074951171875, 3.9158031940460205, -3.2531697750091553, 6.466464042663574, 8.729290962219238, -2.9388294219970703, 3.903348684310913, 4.123203754425049, 5.402464389801025, 0.9791235327720642, -0.02464156784117222, 3.921692132949829, 5.38488245010376, 4.058453559875488, -2.3190059661865234, 9.834254264831543, 3.9531214237213135, 7.779791831970215, 6.031684398651123, 3.0592637062072754, 7.773969650268555, 5.1326069831848145, 6.0987725257873535, 7.714286804199219, 4.269192695617676, 8.070961952209473, -0.012986613437533379, 4.098285675048828, -1.9422475099563599, 3.2327418327331543, 2.47141170501709, 9.168487548828125, 2.7180426120758057, 4.041884422302246, -2.3929662704467773, 4.616462707519531, 5.2467546463012695, 7.247101306915283, 5.111072063446045, 9.630743026733398, 4.377971172332764, 9.542019844055176, 5.829288959503174, 10.10844612121582, 8.402318000793457, 5.330064296722412, 8.81738567352295, 5.346924781799316, 5.416744232177734, 8.745107650756836, 10.203930854797363, 3.563910484313965, 5.320544719696045, 4.650362491607666, 10.254496574401855, -3.3592214584350586, -2.7672677040100098, -1.919179916381836, -0.7489309310913086, 2.636056900024414, -2.9704227447509766, 10.117410659790039], \"y\": [2.0281982421875, 2.024493455886841, 1.4629228115081787, 4.951525688171387, 1.819256067276001, 5.8493170738220215, 2.907670497894287, 6.206664085388184, 5.613591194152832, 4.970948696136475, 2.313070774078369, 5.742432594299316, 1.7961918115615845, 2.461164951324463, 4.976020812988281, 5.2795186042785645, 2.2834184169769287, 4.897169589996338, 0.971183717250824, 8.321795463562012, 5.938363075256348, 5.640936851501465, 4.5215535163879395, 0.8989933729171753, 1.9329392910003662, 2.123931646347046, 5.435695648193359, 2.0155420303344727, 4.980131149291992, 2.680210590362549, 1.7101465463638306, 5.763540744781494, 2.2061052322387695, 2.0870347023010254, 5.8454203605651855, 5.9008989334106445, 6.435052394866943, 4.541885852813721, 5.257188320159912, 6.517399787902832, 2.4146742820739746, 5.6222710609436035, 1.6746773719787598, 6.501496315002441, 6.5415873527526855, 1.9045926332473755, 4.559465408325195, 6.016314506530762, 5.892645835876465, 2.3727927207946777, 4.563896179199219, 5.427707195281982, 5.315176486968994, 2.0567731857299805, 2.275076389312744, 5.292001247406006, 1.5748780965805054, 6.259632110595703, 4.56760311126709, 5.627898693084717, 4.904993534088135, 4.896672248840332, 4.707564353942871, 4.909095764160156, 6.282721996307373, 6.295764923095703, 4.9914960861206055, 5.287172317504883, 6.327345371246338, 5.375999450683594, 2.0226821899414062, 6.072786808013916, 4.646247386932373, 7.673611164093018, 5.622596740722656, 2.4791505336761475, 5.860335826873779, 2.0613439083099365, 1.4615198373794556, 5.724172115325928, 4.998956203460693, 6.477722644805908, 0.8934712409973145, 2.040785074234009, 5.490987300872803, 2.0215842723846436, 2.5038516521453857, 1.9134423732757568, 5.6797709465026855, 7.384728908538818, 5.534671783447266, 2.4382870197296143, 5.476228713989258, 5.7328877449035645, 5.383542060852051, 1.6079049110412598, 5.223074913024902, 5.044475555419922, 5.8098907470703125, 2.53842830657959, 2.29491925239563, 4.717820644378662, 5.654617786407471, 5.654147624969482, 4.988352298736572, 5.671962738037109, 5.648568630218506, 5.6395769119262695, 1.3791608810424805, 5.643033504486084, 5.7275824546813965, 5.810116291046143, 6.398591995239258, 5.795466899871826, 5.635751247406006, 5.643283843994141, 7.896315097808838, 5.249232292175293, 4.736836910247803, 2.3484065532684326, 5.8207783699035645, 2.0752735137939453, 1.3220163583755493, 8.107723236083984, 5.947668552398682, 5.407575607299805, 2.2786059379577637, 2.249164342880249, 7.261606216430664, 1.7425520420074463, 6.001363754272461, 3.2941410541534424, 3.326934576034546, 8.446836471557617, 2.9339518547058105, 5.472588062286377, 5.208961009979248, 5.613396644592285, 3.4728429317474365, 5.692798137664795, 4.448867321014404, 2.5673413276672363, 5.10014009475708, 5.830433368682861, 4.764292240142822, 5.7695488929748535, 6.063958168029785, 5.768062591552734, 7.7731804847717285, 5.398154258728027, 5.055805206298828, 5.982251167297363, 6.3138957023620605, 5.648232936859131, 5.944860458374023, 5.521167278289795, 2.0326809883117676, 8.426432609558105, 5.248061180114746, 4.79599666595459, 5.733616828918457, 1.918269395828247, 1.3630897998809814, 5.647495746612549, 0.84022057056427, 5.295196533203125, 5.717221260070801, 2.412451982498169, 1.8583067655563354, 5.676621437072754, 5.817334175109863, 2.274446964263916, 5.2744035720825195, 4.830721855163574, 4.411782741546631, 4.453697681427002, 6.102783203125, 5.367859840393066, 4.8467888832092285, 4.857787132263184, 0.7460822463035583, 1.2637101411819458, 2.295125961303711, 5.722262859344482, 2.4258556365966797, 4.664398670196533, 2.9108526706695557, 5.630008697509766, 5.046818256378174, 5.04509162902832, 4.950100898742676, 5.194746971130371, 4.399614334106445, 6.179681301116943, 5.40525484085083, 5.6852126121521, 5.144185543060303, 2.298426866531372, 1.9861805438995361, 6.335288047790527, 5.611847877502441, 3.0153138637542725, 1.1472328901290894, 5.3571457862854, 5.662734031677246, 7.676368713378906, 5.03038215637207, 2.3888208866119385, 6.57177209854126, 8.50075626373291, 1.5717085599899292, 5.531695365905762, 0.6413320899009705, 2.419189929962158, 6.155582427978516, 6.422114372253418, 5.788280963897705, 5.018467426300049, 1.5227349996566772, 2.3011436462402344, 5.809711456298828, 8.07191276550293, 5.032275199890137, 6.25347375869751, 5.573572158813477, 8.427337646484375, 5.5836639404296875, 7.860910892486572, 2.492560863494873, 5.623416900634766, 5.084853649139404, 5.848064422607422, 5.904232978820801, 2.297358274459839, 5.620500087738037, 5.444415092468262, 0.7134695053100586, 2.0498905181884766, 5.89426851272583, 2.3136866092681885, 5.690305709838867, 4.796238422393799, 6.261064529418945, 5.557446002960205, 1.9447972774505615, 4.876189708709717, 5.608051300048828, 6.738003253936768, 5.617213726043701, 5.85187292098999, 1.1224578619003296, 5.3726983070373535, 1.8314085006713867, 3.0932908058166504, 1.6332533359527588, 7.468328952789307, 1.769141674041748, 4.535568714141846, 1.9314846992492676, 5.071163654327393, 2.1683661937713623, 2.3055636882781982, 2.0878512859344482, 2.3383162021636963, 3.056411027908325, 2.2658791542053223, 5.783656120300293, 5.464430809020996, 5.504628658294678, 5.800779342651367, 4.878124237060547, 5.799253463745117, 0.7843397855758667, 4.743893146514893, 2.267395257949829, 5.696560382843018, 4.403985977172852, 2.372361660003662, 5.342061519622803, 5.743143081665039, 2.002744197845459, 5.221711158752441, 1.6414387226104736, 5.392172813415527, 5.91682767868042, 4.51915979385376, 5.862895488739014, 4.8147454261779785, 5.816915035247803, 5.112392902374268, 6.166357517242432, 2.5475692749023438, 5.972390174865723, 5.505509853363037, 5.259632587432861, 5.220202445983887, 4.630622863769531, 5.168177604675293, 5.377807140350342, 4.809309482574463, 1.4167224168777466, 5.946244239807129, 1.4940775632858276, 5.108719825744629, 5.307571887969971, 5.688802719116211, 4.9629340171813965, 4.4339094161987305, 4.879934787750244, 4.790249824523926, 5.278910160064697, 5.2505106925964355, 5.880641460418701, 4.973315238952637, 4.811169624328613, 4.776472091674805, 4.770730972290039, 2.995741367340088, 6.216947555541992, 4.518594264984131, 5.840506553649902, 2.1825873851776123, 5.904170513153076, 1.3395965099334717, 5.287189483642578, 5.4935383796691895, 5.248785018920898, 4.777257919311523, 3.3354220390319824, 3.487539529800415, 5.47268009185791, 5.896217346191406, 4.827913284301758, 4.534594535827637, 5.5831685066223145, 2.204517364501953, 5.88881254196167, 1.8509242534637451, 5.056319713592529, 0.6860949993133545, 7.65020751953125, 5.828574180603027, 5.963394641876221, 1.3526420593261719, 5.828185081481934, 5.706432342529297, 5.319921493530273, 8.428104400634766, 5.580775737762451, 5.032719612121582, 4.738737106323242, 4.501551628112793, 0.7383899092674255, 3.728128671646118, 4.755970478057861, 6.004078388214111, 5.795276641845703, 5.3867034912109375, 4.9273529052734375, 5.079994201660156, 4.788541793823242, 5.552335262298584, 5.679013252258301, 2.330146551132202, 5.5545854568481445, 2.041213274002075, 1.2939547300338745, 5.696969032287598, 0.8269582390785217, 0.7971106171607971, 5.458467960357666, 5.617489337921143, 1.3746310472488403, 1.0907343626022339, 5.699527263641357, 7.911667823791504, 6.346795558929443, 5.870376110076904, 5.8733229637146, 4.9933905601501465, 5.936906814575195, 5.429734230041504, 5.946542739868164, 5.823299407958984, 5.842685222625732, 2.3034231662750244, 5.903293132781982, 5.207875728607178, 5.99885368347168, 1.5899492502212524, 5.678659915924072, 4.689005374908447, 4.990255355834961, 1.2371137142181396, 5.726591110229492, 5.430578708648682, 2.567915201187134, 5.947303295135498, 5.231675624847412, 5.495565414428711, 4.569997787475586, 5.2136311531066895, 5.933168411254883, 5.710550308227539, 4.663186073303223, 5.171750068664551, 8.428814888000488, 5.759331226348877, 2.8859715461730957, 5.449172019958496, 5.529959678649902, 5.12800931930542, 4.5250396728515625, 2.5041346549987793, 5.694319248199463, 5.741184234619141, 4.8963823318481445, 2.038616180419922, 5.803277969360352, 5.004720687866211, 5.460947036743164, 7.0894951820373535, 5.915504455566406, 2.601367473602295, 4.92162561416626, 4.511317729949951, 5.80855655670166, 5.618354320526123, 4.7499589920043945, 4.776318550109863, 4.754566192626953, 5.791283130645752, 1.5672699213027954, 2.0090038776397705, 5.6327080726623535, 6.522928237915039, 5.478829383850098, 5.6642608642578125, 5.485386848449707, 5.481222152709961, 5.850682735443115, 2.4603819847106934, 4.86735725402832, 4.88516092300415, 2.9318926334381104, 5.692088603973389, 2.2560536861419678, 5.758404731750488, 2.3628089427948, 0.9697510004043579, 5.914269924163818, 2.6169676780700684, 5.941591739654541, 8.265555381774902, 5.708977699279785, 5.937975883483887, 6.025681495666504, 5.748917579650879, 5.757392883300781, 5.875559329986572, 3.0851972103118896, 0.6641057729721069, 5.747582912445068, 5.623376846313477, 5.532130718231201, 6.785458087921143, 2.6874940395355225, 4.630245208740234, 2.0107953548431396, 5.700977802276611, 4.998167991638184, 5.9331374168396, 8.518680572509766, 4.387196063995361, 5.847194194793701, 5.856264114379883, 2.5938305854797363, 5.321661949157715, 5.963446617126465, 5.285303115844727, 5.546899318695068, 5.411121845245361, 8.058271408081055, 5.723628044128418, 7.84012508392334, 6.31354284286499, 5.502166271209717, 1.5327461957931519, 5.855234622955322, 8.474869728088379, 5.593247413635254, 5.545948505401611, 2.075284719467163, 4.630777359008789, 5.5278143882751465, 2.328275680541992, 7.3751325607299805, 5.708590030670166, 2.0463061332702637, 1.5066838264465332, 7.993497371673584, 6.08047342300415, 5.870571136474609, 0.733788251876831, 5.809926986694336, 5.808352947235107, 5.6954665184021, 5.788963794708252, 8.495503425598145, 4.804171085357666, 5.92659854888916, 5.61293363571167, 5.974811553955078, 8.456863403320312, 2.276569128036499, 5.785958290100098, 2.854961633682251, 1.9560080766677856, 5.6778364181518555, 5.874251842498779, 5.8157501220703125, 0.7610512971878052, 4.492734432220459, 5.663146018981934, 5.131646156311035, 5.616086959838867, 7.4484453201293945, 5.7222418785095215, 5.6956257820129395, 5.829310417175293, 5.169118404388428, 5.6557440757751465, 5.695373058319092, 1.6270414590835571, 5.655323505401611, 1.4065873622894287, 2.527446746826172, 5.7191996574401855, 4.624314308166504, 5.972428321838379, 5.774099826812744, 6.474634170532227, 2.428100347518921, 5.859563827514648, 5.761529922485352, 5.502089023590088, 4.5303192138671875, 5.741293907165527, 4.600226402282715, 5.643549919128418, 4.478128433227539, 5.095979690551758, 5.889841556549072, 4.776607036590576, 5.8994832038879395, 5.996215343475342, 4.802554130554199, 4.4307122230529785, 2.635342597961426, 5.958472728729248, 2.4180588722229004, 4.404966354370117, 4.873924732208252, 1.1885523796081543, 0.7336114645004272, 2.185274362564087, 2.448465585708618, 1.36483895778656, 4.462133407592773], \"z\": [9.326181411743164, 9.375383377075195, 9.335861206054688, 9.569923400878906, 9.375012397766113, 9.498595237731934, 8.954334259033203, 9.303597450256348, 9.436783790588379, 9.543028831481934, 9.39670181274414, 8.580662727355957, 9.308122634887695, 9.21727466583252, 9.492839813232422, 9.431163787841797, 9.482022285461426, 8.907357215881348, 9.235164642333984, 7.666891098022461, 4.049073219299316, 2.409804105758667, 9.124194145202637, 9.203761100769043, 9.333290100097656, 9.314970016479492, 9.403321266174316, 9.29357624053955, 9.026799201965332, 9.124820709228516, 9.348017692565918, 9.549850463867188, 9.308829307556152, 9.33767318725586, 8.654152870178223, 8.642374992370605, 8.912751197814941, 2.594715118408203, -0.5150102972984314, 8.991101264953613, 9.2473726272583, 7.3552727699279785, 9.307543754577637, 9.028051376342773, 8.965988159179688, 9.317241668701172, 1.146653413772583, 9.330606460571289, 3.503488063812256, 9.292309761047363, 2.2382495403289795, 5.198614597320557, 2.724790096282959, 9.290214538574219, 9.49739933013916, 2.1368041038513184, 9.28041934967041, 9.011270523071289, 2.5143253803253174, 2.349860429763794, 0.7146493196487427, 0.008597971871495247, 9.19507122039795, 1.3444514274597168, 9.20899772644043, 9.032424926757812, -0.05820915475487709, 2.9612343311309814, 7.21016788482666, 3.1922879219055176, 9.370464324951172, 6.675629615783691, 8.877981185913086, 7.739486217498779, 3.855670928955078, 9.248820304870605, 3.358177661895752, 9.323358535766602, 9.284213066101074, 3.717280626296997, -0.08680406212806702, 9.151104927062988, 9.311568260192871, 9.33171272277832, 2.4508886337280273, 9.289505958557129, 9.143508911132812, 9.283568382263184, 4.008776664733887, 7.588979244232178, -2.449887752532959, 9.242095947265625, 5.336538314819336, 3.704197406768799, 8.460066795349121, 9.265905380249023, -0.392306387424469, -0.08523904532194138, 3.2126805782318115, 9.110300064086914, 9.429753303527832, 0.4834348261356354, -2.3315141201019287, 3.9364540576934814, 0.9700291752815247, 9.52585506439209, 4.0468430519104, 3.8495852947235107, 9.281246185302734, 3.708892583847046, 4.462997913360596, 5.832554817199707, 7.236009120941162, 5.093343257904053, 2.4273831844329834, 3.710383892059326, 7.819478988647461, 8.99254035949707, 9.088052749633789, 9.328357696533203, 3.2022132873535156, 9.251441955566406, 9.294126510620117, 7.728145599365234, 5.059607028961182, 4.9614434242248535, 9.502988815307617, 9.358345985412598, 7.544238090515137, 9.355534553527832, 3.7929346561431885, 8.77006721496582, 8.748391151428223, 7.638487339019775, 8.935589790344238, -2.235806703567505, -0.4105677902698517, 3.596435070037842, 8.67250919342041, 2.902830123901367, 1.4836363792419434, 9.19333553314209, -0.1076047495007515, 4.642059326171875, 3.112586259841919, 5.055593490600586, 9.479392051696777, 5.960798740386963, 7.764686107635498, 5.206193447113037, 1.0702006816864014, 7.100655555725098, 7.2078118324279785, 2.5079503059387207, 4.093955993652344, 2.647273540496826, 9.238755226135254, 7.71982479095459, 3.6893723011016846, 3.2259204387664795, 4.616464138031006, 9.386590957641602, 9.336480140686035, 3.9212887287139893, 9.215506553649902, 5.14254903793335, 2.7592530250549316, 9.299897193908691, 9.344002723693848, 5.099119186401367, 3.166048765182495, 9.464707374572754, 8.637677192687988, 0.11422031372785568, 1.7825144529342651, 1.407228708267212, 6.862109184265137, 8.440227508544922, 3.7728312015533447, 3.6281468868255615, 9.301043510437012, 9.317879676818848, 9.446722030639648, 2.7501895427703857, 9.24372386932373, 0.8083931803703308, 8.984607696533203, 3.550116777420044, 0.4760226011276245, 3.4792768955230713, 9.564000129699707, -0.4286499321460724, 2.2924644947052, 9.346543312072754, 8.5054292678833, 4.4831156730651855, 7.7072553634643555, 9.342219352722168, 9.308499336242676, 8.913060188293457, 3.8903002738952637, 8.923683166503906, 9.310774803161621, 9.439773559570312, 4.1357316970825195, 7.708360195159912, 0.5709878206253052, 9.298081398010254, 9.059061050415039, 7.610373020172119, 9.32030200958252, 3.428600311279297, 9.301426887512207, 9.276805877685547, 9.389969825744629, 9.128644943237305, 5.656898021697998, 1.290280818939209, 9.280427932739258, 9.340431213378906, 4.892039775848389, 7.71256160736084, 0.604839563369751, 8.881815910339355, 5.0992608070373535, 7.587255954742432, 2.2531394958496094, 7.82257604598999, 9.14981746673584, 9.43072509765625, 8.60118579864502, -2.6913902759552, -2.7591567039489746, 9.356162071228027, 2.303591728210449, -2.340257406234741, 9.249942779541016, 9.366068840026855, -3.041811227798462, 9.342390060424805, 5.105335712432861, 3.452479600906372, 7.179145812988281, -2.492967128753662, 9.266079902648926, 0.03744293749332428, -2.0495030879974365, 7.404592514038086, -1.9755934476852417, 3.276979923248291, 9.323341369628906, 2.6303985118865967, 9.291192054748535, 8.875823974609375, 9.275434494018555, 7.654347896575928, 9.346678733825684, 9.098692893981934, 9.314776420593262, 9.366190910339355, 9.308903694152832, 9.406208038330078, 9.32055377960205, 9.351085662841797, 8.91357135772705, 9.332023620605469, 4.83028507232666, -2.3605387210845947, 2.4300613403320312, 5.435883045196533, 3.9463858604431152, 3.1395437717437744, 9.293776512145996, 0.8304648399353027, 9.459480285644531, 2.854959011077881, 1.776555061340332, 9.317509651184082, 5.130721569061279, 4.888373374938965, 9.340231895446777, 2.3797824382781982, 9.290018081665039, 8.49467658996582, 5.026937484741211, 2.3833718299865723, 6.1009063720703125, 3.782846689224243, 5.973367691040039, 1.4832552671432495, 6.86343240737915, 9.073163032531738, 3.961575746536255, 2.5231659412384033, 3.7073605060577393, 3.8570621013641357, 2.8851478099823, -0.32684770226478577, 5.208263397216797, 0.36570921540260315, 9.336753845214844, 5.046684265136719, 9.324067115783691, -0.0518876276910305, 3.1504015922546387, -2.1363275051116943, 3.671560287475586, 1.4946742057800293, 3.843749523162842, 3.747938632965088, 1.9436787366867065, 5.069278240203857, 4.8746514320373535, 3.091365337371826, 0.21257224678993225, 3.598705768585205, 3.3822150230407715, 8.936309814453125, 7.102111339569092, 2.5958926677703857, 5.086032867431641, 9.35895824432373, 3.494725227355957, 9.310620307922363, 3.6636674404144287, 3.115290403366089, 3.9818434715270996, 3.4267475605010986, 8.740822792053223, 8.665040969848633, 8.465554237365723, 4.763996601104736, 0.24798887968063354, 2.008310556411743, -2.0266129970550537, 9.184867858886719, 4.849735260009766, 9.293876647949219, 0.36174657940864563, 9.261594772338867, 7.718995094299316, 4.791592597961426, 3.7378625869750977, 9.260168075561523, 3.153398036956787, 4.5326457023620605, 2.6334824562072754, 7.6514201164245605, 2.2716586589813232, 0.9338598847389221, 3.062871217727661, 9.11655044555664, 9.288124084472656, 8.530963897705078, 2.9608497619628906, 5.121572017669678, 5.764952182769775, 4.968245983123779, 1.4174647331237793, -0.07755064964294434, 3.5664772987365723, -2.372786521911621, 3.6048269271850586, 9.34248161315918, 4.911813259124756, 9.3482027053833, 9.257648468017578, 3.816819906234741, 9.217015266418457, 9.296873092651367, 8.591590881347656, 5.266180038452148, 9.302834510803223, 9.231351852416992, 4.701958179473877, 7.777426719665527, 7.206053733825684, -2.951446533203125, 4.850978374481201, 9.57535171508789, 3.738741159439087, 2.6445374488830566, 6.210389137268066, 4.875147342681885, 4.852382183074951, 9.399919509887695, 4.954708576202393, 4.974368572235107, 6.385321617126465, 9.330530166625977, 2.594301462173462, 0.5214361548423767, 8.772405624389648, 9.328339576721191, 2.7951581478118896, 5.089591979980469, 9.062837600708008, 5.054052829742432, 3.560358762741089, 2.89786958694458, 9.013409614562988, 3.823063611984253, 5.020763874053955, 4.048819541931152, 2.949361562728882, -0.21722567081451416, 7.609896183013916, 5.905818462371826, 8.955436706542969, 2.2779369354248047, -2.1547775268554688, 0.034845639020204544, 1.6887333393096924, 9.120737075805664, -2.7522428035736084, 2.9479758739471436, 0.034257955849170685, 9.316607475280762, 3.2668099403381348, 0.44271090626716614, 2.303321361541748, 7.476970672607422, 3.550727367401123, 8.999693870544434, 0.664082407951355, 1.811622142791748, 4.858519554138184, -1.976544737815857, 3.067265272140503, 3.4551897048950195, 9.240426063537598, 4.992180347442627, 9.333700180053711, 9.29209041595459, 5.180436134338379, 8.983756065368652, 2.802518606185913, 5.2027974128723145, -2.342683792114258, 5.317488670349121, 6.1233229637146, 9.240580558776855, 3.82305645942688, 1.159900188446045, 8.989054679870605, 3.8496603965759277, 9.342877388000488, 4.807809829711914, 9.314313888549805, 9.223267555236816, 6.197115421295166, 9.041695594787598, -3.0392253398895264, 7.6813812255859375, 4.409923076629639, -3.0985567569732666, -3.1427552700042725, -2.7871901988983154, -2.7797162532806396, -2.8278746604919434, 8.845965385437012, 9.285943031311035, -2.808671712875366, 3.5822501182556152, 5.1948561668396, 7.372023582458496, 8.915057182312012, 0.8972824811935425, 9.315895080566406, 7.300824165344238, 1.116434931755066, -2.84226655960083, 7.592831134796143, 2.2293660640716553, 4.745712757110596, 4.801469802856445, 9.01081371307373, 5.169038772583008, 3.719489097595215, 8.48254108428955, 3.434255838394165, 5.228640079498291, 7.715271472930908, 4.753030776977539, 7.794229984283447, 7.159590244293213, 5.293515205383301, 9.333525657653809, 3.283851146697998, 7.654035568237305, -2.2367467880249023, 5.359278678894043, 9.271611213684082, 2.8650476932525635, -2.191467046737671, 9.359210014343262, 7.614920139312744, 4.064919471740723, 9.260287284851074, 9.330924987792969, 7.765391826629639, 9.391373634338379, 3.3529176712036133, 9.252711296081543, 7.162842750549316, 7.140867710113525, 4.104308605194092, 4.996796607971191, 7.584543228149414, 9.166267395019531, 4.9779510498046875, 5.289971828460693, 9.283756256103516, 7.631771087646484, 9.345836639404297, 3.0219438076019287, 8.955877304077148, 9.250802993774414, 4.307499408721924, 3.430781841278076, 5.0425333976745605, 9.292859077453613, 1.9053789377212524, 4.203989028930664, -0.04750059172511101, 2.3809759616851807, 7.661910057067871, 4.472822189331055, 2.8623034954071045, 3.2454190254211426, -0.22513903677463531, 3.730097532272339, 4.790256977081299, 9.270329475402832, 3.9824559688568115, 9.32429027557373, 9.129199028015137, 7.278326511383057, 0.8420422077178955, 7.037677764892578, 4.9206132888793945, 8.910903930664062, 9.214127540588379, -2.962287664413452, 3.8298120498657227, -2.3974690437316895, 1.5722365379333496, 4.316435813903809, 1.6167988777160645, -2.0104119777679443, 2.2572216987609863, 1.3350731134414673, -2.98465633392334, 0.46208178997039795, -2.988672971725464, -3.1002604961395264, 0.3782808184623718, 2.225386381149292, 8.965280532836914, -3.126033306121826, 9.219942092895508, 2.278672456741333, 9.434990882873535, 9.31115436553955, 9.258614540100098, 9.291431427001953, 9.242691040039062, 9.285379409790039, 2.204198122024536]}],\n",
              "                        {\"autosize\": false, \"coloraxis\": {\"colorbar\": {\"title\": {\"text\": \"diagnosis\"}}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]]}, \"height\": 800, \"legend\": {\"tracegroupgap\": 0}, \"margin\": {\"t\": 60}, \"scene\": {\"domain\": {\"x\": [0.0, 1.0], \"y\": [0.0, 1.0]}, \"xaxis\": {\"title\": {\"text\": \"0\"}, \"visible\": false}, \"yaxis\": {\"title\": {\"text\": \"1\"}, \"visible\": false}, \"zaxis\": {\"title\": {\"text\": \"2\"}, \"visible\": false}}, \"template\": {\"data\": {\"bar\": [{\"error_x\": {\"color\": \"#2a3f5f\"}, \"error_y\": {\"color\": \"#2a3f5f\"}, \"marker\": {\"line\": {\"color\": \"#E5ECF6\", \"width\": 0.5}}, \"type\": \"bar\"}], \"barpolar\": [{\"marker\": {\"line\": {\"color\": \"#E5ECF6\", \"width\": 0.5}}, \"type\": \"barpolar\"}], \"carpet\": [{\"aaxis\": {\"endlinecolor\": \"#2a3f5f\", \"gridcolor\": \"white\", \"linecolor\": \"white\", \"minorgridcolor\": \"white\", \"startlinecolor\": \"#2a3f5f\"}, \"baxis\": {\"endlinecolor\": \"#2a3f5f\", \"gridcolor\": \"white\", \"linecolor\": \"white\", \"minorgridcolor\": \"white\", \"startlinecolor\": \"#2a3f5f\"}, \"type\": \"carpet\"}], \"choropleth\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"choropleth\"}], \"contour\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"contour\"}], \"contourcarpet\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"contourcarpet\"}], \"heatmap\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"heatmap\"}], \"heatmapgl\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"heatmapgl\"}], \"histogram\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"histogram\"}], \"histogram2d\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"histogram2d\"}], \"histogram2dcontour\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"histogram2dcontour\"}], \"mesh3d\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"type\": \"mesh3d\"}], \"parcoords\": [{\"line\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"parcoords\"}], \"pie\": [{\"automargin\": true, \"type\": \"pie\"}], \"scatter\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatter\"}], \"scatter3d\": [{\"line\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatter3d\"}], \"scattercarpet\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattercarpet\"}], \"scattergeo\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattergeo\"}], \"scattergl\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattergl\"}], \"scattermapbox\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scattermapbox\"}], \"scatterpolar\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterpolar\"}], \"scatterpolargl\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterpolargl\"}], \"scatterternary\": [{\"marker\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"type\": \"scatterternary\"}], \"surface\": [{\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}, \"colorscale\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"type\": \"surface\"}], \"table\": [{\"cells\": {\"fill\": {\"color\": \"#EBF0F8\"}, \"line\": {\"color\": \"white\"}}, \"header\": {\"fill\": {\"color\": \"#C8D4E3\"}, \"line\": {\"color\": \"white\"}}, \"type\": \"table\"}]}, \"layout\": {\"annotationdefaults\": {\"arrowcolor\": \"#2a3f5f\", \"arrowhead\": 0, \"arrowwidth\": 1}, \"coloraxis\": {\"colorbar\": {\"outlinewidth\": 0, \"ticks\": \"\"}}, \"colorscale\": {\"diverging\": [[0, \"#8e0152\"], [0.1, \"#c51b7d\"], [0.2, \"#de77ae\"], [0.3, \"#f1b6da\"], [0.4, \"#fde0ef\"], [0.5, \"#f7f7f7\"], [0.6, \"#e6f5d0\"], [0.7, \"#b8e186\"], [0.8, \"#7fbc41\"], [0.9, \"#4d9221\"], [1, \"#276419\"]], \"sequential\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]], \"sequentialminus\": [[0.0, \"#0d0887\"], [0.1111111111111111, \"#46039f\"], [0.2222222222222222, \"#7201a8\"], [0.3333333333333333, \"#9c179e\"], [0.4444444444444444, \"#bd3786\"], [0.5555555555555556, \"#d8576b\"], [0.6666666666666666, \"#ed7953\"], [0.7777777777777778, \"#fb9f3a\"], [0.8888888888888888, \"#fdca26\"], [1.0, \"#f0f921\"]]}, \"colorway\": [\"#636efa\", \"#EF553B\", \"#00cc96\", \"#ab63fa\", \"#FFA15A\", \"#19d3f3\", \"#FF6692\", \"#B6E880\", \"#FF97FF\", \"#FECB52\"], \"font\": {\"color\": \"#2a3f5f\"}, \"geo\": {\"bgcolor\": \"white\", \"lakecolor\": \"white\", \"landcolor\": \"#E5ECF6\", \"showlakes\": true, \"showland\": true, \"subunitcolor\": \"white\"}, \"hoverlabel\": {\"align\": \"left\"}, \"hovermode\": \"closest\", \"mapbox\": {\"style\": \"light\"}, \"paper_bgcolor\": \"white\", \"plot_bgcolor\": \"#E5ECF6\", \"polar\": {\"angularaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"bgcolor\": \"#E5ECF6\", \"radialaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}}, \"scene\": {\"xaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}, \"yaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}, \"zaxis\": {\"backgroundcolor\": \"#E5ECF6\", \"gridcolor\": \"white\", \"gridwidth\": 2, \"linecolor\": \"white\", \"showbackground\": true, \"ticks\": \"\", \"zerolinecolor\": \"white\"}}, \"shapedefaults\": {\"line\": {\"color\": \"#2a3f5f\"}}, \"ternary\": {\"aaxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"baxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}, \"bgcolor\": \"#E5ECF6\", \"caxis\": {\"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\"}}, \"title\": {\"x\": 0.05}, \"xaxis\": {\"automargin\": true, \"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\", \"title\": {\"standoff\": 15}, \"zerolinecolor\": \"white\", \"zerolinewidth\": 2}, \"yaxis\": {\"automargin\": true, \"gridcolor\": \"white\", \"linecolor\": \"white\", \"ticks\": \"\", \"title\": {\"standoff\": 15}, \"zerolinecolor\": \"white\", \"zerolinewidth\": 2}}}, \"width\": 800},\n",
              "                        {\"responsive\": true}\n",
              "                    ).then(function(){\n",
              "                            \n",
              "var gd = document.getElementById('c59b2ecc-e477-4a92-bc77-09ea6032765f');\n",
              "var x = new MutationObserver(function (mutations, observer) {{\n",
              "        var display = window.getComputedStyle(gd).display;\n",
              "        if (!display || display === 'none') {{\n",
              "            console.log([gd, 'removed!']);\n",
              "            Plotly.purge(gd);\n",
              "            observer.disconnect();\n",
              "        }}\n",
              "}});\n",
              "\n",
              "// Listen for the removal of the full notebook cells\n",
              "var notebookContainer = gd.closest('#notebook-container');\n",
              "if (notebookContainer) {{\n",
              "    x.observe(notebookContainer, {childList: true});\n",
              "}}\n",
              "\n",
              "// Listen for the clearing of the current output cell\n",
              "var outputEl = gd.closest('.output');\n",
              "if (outputEl) {{\n",
              "    x.observe(outputEl, {childList: true});\n",
              "}}\n",
              "\n",
              "                        })\n",
              "                };\n",
              "                \n",
              "            </script>\n",
              "        </div>\n",
              "</body>\n",
              "</html>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "b4p2j5FcJArU"
      },
      "source": [
        "![title](img/umap1.png)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rRsM0xV-JCzi"
      },
      "source": [
        "![title](img/umap2.png)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sXu-bcQ46cpR"
      },
      "source": [
        "The UMAP 2D projection suggests a clearer separation between the cancer and non-cancer groups. The 3D projection also shows a clear separation, but also with a clearer morphology of each group as a shape."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mQQ7iKJY9Hkh"
      },
      "source": [
        "## 8 Prepare data for ML model\n",
        "\n",
        "### 8.1 Log transform skewed features\n",
        "\n",
        "We identified earier some features that seem particularly skewed (which can be problematic for ML models). Lets log-transform these to try and even their distributions out."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4W0kgi0h9Hkh",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        },
        "outputId": "73bd7f89-0819-4e9a-abf1-3154a5f3cc5b"
      },
      "source": [
        "# Skewed features\n",
        "skewed = ['radius_se', 'perimeter_se', 'area_se']\n",
        "\n",
        "# Split dataset into those with malignent and those with benign tumors\n",
        "malignent_df = bcancer_clean_df[bcancer_clean_df['diagnosis']==1]\n",
        "benign_df = bcancer_clean_df[bcancer_clean_df['diagnosis']==0]\n",
        "\n",
        "# Display the distributions of skewed features\n",
        "compare_features_benign_malignent(skewed, malignent_df, benign_df, 1, 3, 20, 2)"
      ],
      "execution_count": 154,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x144 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        },
        "id": "WKnwBegrml38",
        "outputId": "f6913efc-31b6-40a4-a2a4-6ea4e4ecc9f5"
      },
      "source": [
        "# Log transform Skewed features\n",
        "features_log_transformed = pd.DataFrame(data = bcancer_clean_df)\n",
        "features_log_transformed[skewed] = features_log_transformed[skewed].apply(lambda x: np.log(x + 1))\n",
        "\n",
        "# Split dataset into those with malignent and those with benign tumors\n",
        "malignent_log_transformed_df = features_log_transformed[features_log_transformed['diagnosis']==1]\n",
        "benign_log_transformed_df = features_log_transformed[features_log_transformed['diagnosis']==0]\n",
        "\n",
        "# Display the distributions of previously skewed features\n",
        "compare_features_benign_malignent(skewed, malignent_log_transformed_df, benign_log_transformed_df, 1, 3, 20, 2)"
      ],
      "execution_count": 155,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x144 with 3 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZrIxRuVS9Hki"
      },
      "source": [
        "Ok the log transform seems to have improved these distributions."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ehY2Sc5h9Hki"
      },
      "source": [
        "## 9 Metrics and naive predictor\n",
        "\n",
        "In this section we will define our measure of success for our models and also define a naive predictor to act as a baseline measure.\n",
        "\n",
        "So in this project, what is most important? we are going to define the key priority as being diagnosing those cases that actually have a malignant tumor most accurately. Of course, as a secondary priority - we do also want to minimise the number of cases where the tumor is actually benign, but we predict it to be malignant.\n",
        "\n",
        "To get a measure of how well our models are doing for both of these at the same time, we can use the F1 score. We will also calculate an accuarcy score, although the F1 score will be the most important measure. The F1 score is also a good measure to use for our relatively small dataset (570 rows only).\n",
        "\n",
        "We also want to establish some kind of benchmark to judge our models against, a baseline that we hope to do much better than with our sophisticated ML models! A common baseline is to define a 'naive' predictor - in this case that would be a predictor that classifies all tumors as malignant.\n",
        "\n",
        "We will now define a naive predictor and calculate an F1 and accuracy score for it - this will be our baseline model and score that we will judge all our other models against."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6gvxVEaB9Hki",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2201b99f-4f0d-4c89-d632-d304b5b3c936"
      },
      "source": [
        "# Naive predictor\n",
        "'''\n",
        "TP = np.sum(features_log_minmax_transform['diagnosis']) # Counting the ones as this is the naive case. \n",
        "FP = features_log_minmax_transform['diagnosis'].count() - TP # Specific to the naive case\n",
        "\n",
        "TN = 0 # No predicted negatives in the naive case\n",
        "FN = 0 # No predicted negatives in the naive case\n",
        "'''\n",
        "# Calculate accuracy, precision and recall\n",
        "accuracy = np.sum(features_log_minmax_transform['diagnosis']) / features_log_minmax_transform.shape[0]\n",
        "recall = 1\n",
        "precision = accuracy\n",
        "\n",
        "# Calculate F-score using the formula above for beta = 0.5 and correct values for precision and recall.\n",
        "beta_squared = 0.5 * 0.5\n",
        "fscore = (1 + beta_squared) * ((precision * recall) / ((beta_squared * precision) + recall))\n",
        "\n",
        "# Print the results \n",
        "print(\"Naive Predictor: [Accuracy score: {:.4f}, F-score: {:.4f}]\".format(accuracy, fscore))"
      ],
      "execution_count": 156,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Naive Predictor: [Accuracy score: 0.3726, F-score: 0.4260]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "PbV6VscI9Hki"
      },
      "source": [
        "## 10 ML model selection & development\n",
        "\n",
        "For this problem - I have judged the following features are appropriate & desirable in guiding the selection of a machine learning model.\n",
        "\n",
        "## 10.1 Model criteria\n",
        "\n",
        "- **Memory & speed:** not primarily important but for what is feasible to be run within a notebook in a reasonable time, the aim is assumed to be a one off run of the model to predict rather than a model run continously where significant computational resources might be more an issue. \n",
        "- **Not overfitting:** I would regard this as desirable/important, we want a model that generalises well, its going to be run most likely periodically if used and we want to maximise the predictive power each time its run.\n",
        "- **Time for learning/fitting:** like memory & speed, not so important beyond that which is reasonable to be run in the notebook without the browser timing out!\n",
        "- **Time for predicting:** again like time for learning/fitting, not so important beyond what can be done in the notebook/browser. However, if we want to deploy and it's useful, we may want to run it on bigger datsets of patient data - so it is a factor worth taking into account.\n",
        "- **Parameterisation:** slightly important. We will be using grid search later to optimise our solution, this will multiply the computational requirements used for a single model. Given the limit of our computatioal resources available in the notebook/browser a more simple parameterisation would allow the model to scale & deliver within this limited environment for a grid search.\n",
        "- **Feature tolerance:** not so important, we have just under 40 features and there is no indication these will grow.\n",
        "\n",
        "With these in mind, these are my selected models, their features - and the rationale for choosing them.\n",
        "\n",
        "**Model 1 - LogisticRegression**\n",
        "\n",
        "- **Strengths of model:** Fast to train, no parameter tuning, features don't need scaling, more tolerant to correlated features, excellent for 2-class classification problems\n",
        "- **Weaknesses of model:** Not so good for large number of features, not good with non-linear decision boundries, not flexible enough to capture more complex relationships, adversely effected by features not relevent to output variable\n",
        "- **Why model is a good candidate for this problem:** 1. The problem is a binary classification one, logistic regression is one of the best & simplest models for this kind of problem - worth considering occams razor! 2. The computational resources required for this model are low and ensure I can be certain this model will have no problem running in the notebook/browser on this dataset! I would like at least one of my models to be likely to run well with given computational resources. \n",
        "\n",
        "**Model 2 - GradientBoostingClassifier**\n",
        "\n",
        "- **Strengths of model:** Can handle big datasets, very accurate, can approximate most non-linear classification boundries, one of best boosting models for many classification problems\n",
        "- **Weaknesses of model:** Slow to train (concern over available resources in notebook/browser sufficient?), prone to overfitting, not easy to understand, sensitive to outliers & noisy data, parameterisation can make a real difference to performance - therefore could be more work/fiddly to optimise & get best results.\n",
        "- **Why model is a good candidate for this problem:** 1. This model is one of the most popular (getting some of the best results) models for classification problems - worth serious consideration! 2. Paramaterisation can make a real difference to performance with this model, good chance of improvement in optimisation/grid search phase for one of the best solutions (assuming computational resources of notebook/browser sufficient).\n",
        "\n",
        "\n",
        "**Model 3 - XGBoost**\n",
        "\n",
        "- **Strengths of model:** Can handle big datasets, very accurate, can approximate most non-linear classification boundries, the best boosting model for many classification problems (best in class), faster than Gradient Boosting.\n",
        "- **Weaknesses of model:** Prone to overfitting, not easy to understand, sensitive to outliers & noisy data, parameterisation can make a real difference to performance - therefore could be more work/fiddly to optimise & get best results.\n",
        "- **Why model is a good candidate for this problem:** 1. This model is one of the most popular (getting some of the best results) models for classification problems - worth serious consideration! 2. Paramaterisation can make a real difference to performance with this model, good chance of improvement in optimisation/grid search phase for one of the best solutions (assuming computational resources of notebook/browser sufficient).\n",
        "\n",
        "References\n",
        "\n",
        "I found the following tables & cheatsheets particularly useful in selecting my models:\n",
        "\n",
        "https://scikit-learn.org/stable/tutorial/machine_learning_map/index.html\n",
        "\n",
        "https://blogs.sas.com/content/subconsciousmusings/2017/04/12/machine-learning-algorithm-use/\n",
        "\n",
        "https://www.dataschool.io/comparing-supervised-learning-algorithms/\n",
        "\n",
        "https://recast.ai/blog/machine-learning-algorithms/2/\n",
        "\n",
        "We will now split our dataset into training and testing subsets, and develop functions that will act as a training & testing pipleline, as well as displaying the results from all our models."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5WTlFC1q9Hki"
      },
      "source": [
        "# Split data into train & test\n",
        "X = features_log_minmax_transform[numerical_features]\n",
        "y = features_log_minmax_transform['diagnosis']\n",
        "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)                                                                                                                                                                                              "
      ],
      "execution_count": 157,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-xANaykI9Hki"
      },
      "source": [
        "# Define function that will act as pipeline for training and testing our models (and returning the results)\n",
        "def train_predict(learner, sample_size, X_train, y_train, X_test, y_test): \n",
        "    '''\n",
        "    inputs:\n",
        "       - learner: the learning algorithm to be trained and predicted on\n",
        "       - sample_size: the size of samples (number) to be drawn from training set\n",
        "       - X_train: features training set\n",
        "       - y_train: income training set\n",
        "       - X_test: features testing set\n",
        "       - y_test: income testing set\n",
        "    '''\n",
        "    \n",
        "    results = {}\n",
        "    \n",
        "    # Fit the learner to the training data using slicing with 'sample_size' using .fit(training_features[:], training_labels[:])\n",
        "    start = time.time() # Get start time\n",
        "    learner.fit(X_train[:sample_size], y_train[:sample_size])\n",
        "    end = time.time() # Get end time\n",
        "    \n",
        "    # Calculate the training time\n",
        "    results['train_time'] = end - start\n",
        "        \n",
        "    # Get the predictions on the test set(X_test),\n",
        "    # then get predictions on the first 300 training samples(X_train) using .predict()\n",
        "    start = time.time() # Get start time\n",
        "    predictions_test = learner.predict(X_test)\n",
        "    predictions_train = learner.predict(X_train[:300])\n",
        "    end = time.time() # Get end time\n",
        "    \n",
        "    # Calculate the total prediction time\n",
        "    results['pred_time'] = end - start\n",
        "            \n",
        "    # Compute accuracy on the first 300 training samples which is y_train[:300]\n",
        "    results['acc_train'] = accuracy_score(y_train[:300], predictions_train)\n",
        "        \n",
        "    # Compute accuracy on test set using accuracy_score()\n",
        "    results['acc_test'] = accuracy_score(y_test, predictions_test)\n",
        "    \n",
        "    # Compute F-score on the the first 300 training samples using fbeta_score()\n",
        "    results['f_train'] = fbeta_score(y_train[:300], predictions_train, beta=0.5)\n",
        "        \n",
        "    # Compute F-score on the test set which is y_test\n",
        "    results['f_test'] = fbeta_score(y_test, predictions_test, beta=0.5)\n",
        "       \n",
        "    # Success\n",
        "    print(\"{} trained on {} samples.\".format(learner.__class__.__name__, sample_size))\n",
        "        \n",
        "    # Return the results\n",
        "    return results"
      ],
      "execution_count": 158,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "W5lujqLr9Hki"
      },
      "source": [
        "# Define function that will take the results of our model training & testing and plot the results\n",
        "def evaluate(results, accuracy, f1):\n",
        "    \"\"\"\n",
        "    Visualization code to display results of various learners.\n",
        "\n",
        "    inputs:\n",
        "      - learners: a list of supervised learners\n",
        "      - stats: a list of dictionaries of the statistic results from 'train_predict()'\n",
        "      - accuracy: The score for the naive predictor\n",
        "      - f1: The score for the naive predictor\n",
        "    \"\"\"\n",
        "\n",
        "    # Create figure\n",
        "    fig, ax = plt.subplots(2, 3, figsize = (13,9))\n",
        "\n",
        "    # Constants\n",
        "    bar_width = 0.3\n",
        "    colors = ['#A00000','#00A0A0','#00A000']\n",
        "\n",
        "    # Super loop to plot four panels of data\n",
        "    for k, learner in enumerate(results.keys()):\n",
        "        for j, metric in enumerate(['train_time', 'acc_train', 'f_train', 'pred_time', 'acc_test', 'f_test']):\n",
        "            for i in np.arange(3):\n",
        "                # Creative plot code\n",
        "                ax[j//3, j%3].bar(i+k*bar_width, results[learner][i][metric], width = bar_width, color = colors[k])\n",
        "                ax[j//3, j%3].set_xticks([0.45, 1.45, 2.45])\n",
        "                ax[j//3, j%3].set_xticklabels([\"10%\", \"50%\", \"100%\"])\n",
        "                ax[j//3, j%3].set_xlabel(\"Training Set Size\")\n",
        "                ax[j//3, j%3].set_xlim((-0.1, 3.0))\n",
        "\n",
        "    # Add unique y-labels\n",
        "    ax[0, 0].set_ylabel(\"Time (in seconds)\")\n",
        "    ax[0, 1].set_ylabel(\"Accuracy Score\")\n",
        "    ax[0, 2].set_ylabel(\"F-score\")\n",
        "    ax[1, 0].set_ylabel(\"Time (in seconds)\")\n",
        "    ax[1, 1].set_ylabel(\"Accuracy Score\")\n",
        "    ax[1, 2].set_ylabel(\"F-score\")\n",
        "\n",
        "    # Add titles\n",
        "    ax[0, 0].set_title(\"Model Training\")\n",
        "    ax[0, 1].set_title(\"Accuracy Score on Training Subset\")\n",
        "    ax[0, 2].set_title(\"F-score on Training Subset\")\n",
        "    ax[1, 0].set_title(\"Model Predicting\")\n",
        "    ax[1, 1].set_title(\"Accuracy Score on Testing Set\")\n",
        "    ax[1, 2].set_title(\"F-score on Testing Set\")\n",
        "\n",
        "    # Add horizontal lines for naive predictors\n",
        "    ax[0, 1].axhline(y = accuracy, xmin = -0.1, xmax = 3.0, linewidth = 1, color = 'k', linestyle = 'dashed')\n",
        "    ax[1, 1].axhline(y = accuracy, xmin = -0.1, xmax = 3.0, linewidth = 1, color = 'k', linestyle = 'dashed')\n",
        "    ax[0, 2].axhline(y = f1, xmin = -0.1, xmax = 3.0, linewidth = 1, color = 'k', linestyle = 'dashed')\n",
        "    ax[1, 2].axhline(y = f1, xmin = -0.1, xmax = 3.0, linewidth = 1, color = 'k', linestyle = 'dashed')\n",
        "\n",
        "    # Set y-limits for score panels\n",
        "    ax[0, 1].set_ylim((0, 1))\n",
        "    ax[0, 2].set_ylim((0, 1))\n",
        "    ax[1, 1].set_ylim((0, 1))\n",
        "    ax[1, 2].set_ylim((0, 1))\n",
        "\n",
        "    # Create patches for the legend\n",
        "    patches = []\n",
        "    for i, learner in enumerate(results.keys()):\n",
        "        patches.append(mpatches.Patch(color = colors[i], label = learner))\n",
        "    plt.legend(handles = patches, bbox_to_anchor = (-.80, 2.53), \\\n",
        "               loc = 'upper center', borderaxespad = 0., ncol = 3, fontsize = 'x-large')\n",
        "\n",
        "    # Aesthetics\n",
        "    plt.suptitle(\"Performance Metrics for Three Supervised Learning Models\", fontsize = 16, y = 1.10)\n",
        "    plt.tight_layout()\n",
        "    plt.show()\n"
      ],
      "execution_count": 159,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XS_zFJzl9Hki",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "11499b0c-6fbd-408c-f511-57eb3c4ecb8c"
      },
      "source": [
        "# Initialize the three models\n",
        "clf_A = linear_model.LogisticRegression()\n",
        "clf_B = GradientBoostingClassifier()\n",
        "clf_C = XGBClassifier()\n",
        "\n",
        "# Calculate the number of samples for 10%, 50% and 100% of the training data\n",
        "samples_100 = len(y_train)\n",
        "samples_50 = int((samples_100 / 100) * 50)\n",
        "samples_10 = int((samples_100 / 100) * 10)\n",
        "\n",
        "# Collect results on the learners\n",
        "results = {}\n",
        "for clf in [clf_A, clf_B, clf_C]:\n",
        "    clf_name = clf.__class__.__name__\n",
        "    results[clf_name] = {}\n",
        "    for i, samples in enumerate([samples_10, samples_50, samples_100]):\n",
        "        results[clf_name][i] = \\\n",
        "        train_predict(clf, samples, X_train, y_train, X_test, y_test)"
      ],
      "execution_count": 160,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "LogisticRegression trained on 45 samples.\n",
            "LogisticRegression trained on 227 samples.\n",
            "LogisticRegression trained on 455 samples.\n",
            "GradientBoostingClassifier trained on 45 samples.\n",
            "GradientBoostingClassifier trained on 227 samples.\n",
            "GradientBoostingClassifier trained on 455 samples.\n",
            "XGBClassifier trained on 45 samples.\n",
            "XGBClassifier trained on 227 samples.\n",
            "XGBClassifier trained on 455 samples.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ofSZ1ts39Hkj",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 694
        },
        "outputId": "d368d790-e844-4d9d-d05d-6b0e7254f00d"
      },
      "source": [
        "# Run metrics visualization for the three supervised learning models chosen\n",
        "evaluate(results, accuracy, fscore)"
      ],
      "execution_count": 161,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 936x648 with 6 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mx2Ve_am9Hkj"
      },
      "source": [
        "We can observe the results of our model training and evaluation in the above plots of the metric scores for each model type.\n",
        "\n",
        "Firstly we can see that all 3 models have done quite well, with all models getting at least 0.80 on the testing dataset, even when only 10% or 50% of that testing set was used.\n",
        "\n",
        "For the model training, we can see that the GradientBoostingClassifier took by far the most time to train, and in terms of speed of predicting unsurprisingly LogisticRegression was the fastest. However, in terms of the key metric using 100% of the test dataset, we can see that the XGBoost seems to perform the best, and trains much faster than normal gradient boosting.\n",
        "\n",
        "Therefore we will select the XGBoost as our best model and the one we want to optimise going forward."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2TA1jnDon5Ux"
      },
      "source": [
        "### 10.2 Model cross-validation and tuning\n",
        "\n",
        "We could of course simply split our data into training and test sets, train and optimise our model on the training data, and then get a final performance score on the test set to estimate how well the model will do on new data. However this is not a very rigerous methodology, it only gives us one single estimate of the model performance on the test data, and there is a danger of overfitting our model on our training data. To improve on this, we can employ two techniques, nested and repeated cross validation.\n",
        "\n",
        "With [nested cross validation](https://machinelearningmastery.com/nested-cross-validation-for-machine-learning-with-python/) we can separate out the model optimisation process from the cross validation process of the data. This can give us a less biased estimation of how our optmised model will perform on new data and which hyper-parameters might be best.\n",
        "\n",
        "With [repeated cross validation](https://machinelearningmastery.com/repeated-k-fold-cross-validation-with-python/), we don't just run a cross validation procedure of k-folds once, but multiple times where on each run, the folds are randomised. Again, this can give us a less biased estimation of how our model will perform on new data.\n",
        "\n",
        "We will now perform nested cross validation of the model using these 2 techniques together, to get a better estimate of how the model might perform on new data, as well as helping us to optimise the best model hyper-parameters using a cross-validated grid search."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DfgvF9_uoMid",
        "outputId": "115468d5-0275-49f0-9aa8-fe3145f3e8da"
      },
      "source": [
        "# Configure outer cross-validation procedure\n",
        "# (Used on whole dataset, splits into training & test set folds)\n",
        "# (Also perform repeated k-fold to get better estimate of model performance)\n",
        "# (Repeated k-fold will perform kfold x n_repeats times, randomising folds each time)\n",
        "cv_outer = RepeatedKFold(n_splits=10, n_repeats=10, random_state=1)\n",
        "# Initialise vars to store results\n",
        "outer_results_accuracy = list()\n",
        "outer_results_f1 = list()\n",
        "best_result = None\n",
        "\n",
        "# For each outer loop train and test folds\n",
        "for train_ix, test_ix in cv_outer.split(X):\n",
        "  # Split whole dataset into current train and test folds\n",
        "  X_train, X_test = X.iloc[train_ix, :], X.iloc[test_ix, :]\n",
        "  y_train, y_test = y.iloc[train_ix], y.iloc[test_ix]\n",
        "  # Configure inner cross-validation procedure\n",
        "  # (For model training, splits train into train-train and train-validation folds)\n",
        "  cv_inner = KFold(n_splits=3, shuffle=True, random_state=1)\n",
        "  # Define the model\n",
        "  model = XGBClassifier(random_state=1)\n",
        "  # Define grid search space\n",
        "  space = dict()\n",
        "  space['n_estimators'] = [10, 50, 300]\n",
        "  space['max_depth'] = [2,3,5]\n",
        "  # Define metrics\n",
        "  scoring = {'F1': make_scorer(fbeta_score, beta=0.5, average='macro'), 'Accuracy': make_scorer(accuracy_score)}\n",
        "  # Define cv grid search for model on training set\n",
        "  # (Once search over, take best model and refit on whole training set using F1)\n",
        "  search = GridSearchCV(model, space, scoring=scoring, cv=cv_inner, refit='F1')\n",
        "  # Execute grid search\n",
        "  result = search.fit(X_train, y_train)\n",
        "  # Get the best performing model (refit on the whole training set)\n",
        "  best_model_train = result.best_estimator_\n",
        "  # Generate model predictions on the outer loop test set (never seen by model)\n",
        "  yhat = best_model_train.predict(X_test)\n",
        "  # Evaluate model predictions on the outer loop test set \n",
        "  acc = accuracy_score(y_test, yhat)\n",
        "  f1 = fbeta_score(y_test, yhat,beta=0.5)\n",
        "  # Store the results\n",
        "  outer_results_accuracy.append(acc)\n",
        "  outer_results_f1.append(f1)\n",
        "  # Save model if its the best so far\n",
        "  if best_result == None:\n",
        "    best_result = result\n",
        "  else:\n",
        "    if best_result.best_score_ < result.best_score_:\n",
        "      best_result = result\n",
        "  # Report progress\n",
        "  print('>acc=%.3f, >f1=%.3f, est=%.3f, cfg=%s' % (acc, f1, result.best_score_, result.best_params_))\n",
        "# Summarize the estimated performance of the model\n",
        "print('Mean Accuracy: %.3f (%.3f), Mean F1: %.3f (%.3f)' % (mean(outer_results_accuracy), std(outer_results_accuracy), mean(outer_results_f1), std(outer_results_f1)))\n",
        "  "
      ],
      "execution_count": 162,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            ">acc=0.982, >f1=0.991, est=0.951, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.964, est=0.963, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.930, >f1=0.885, est=0.962, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.955, est=0.952, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.981, est=0.976, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.968, est=0.965, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.981, est=0.966, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.955, est=0.970, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.991, est=0.971, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.964, >f1=0.980, est=0.967, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.991, est=0.963, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.990, est=0.960, cfg={'max_depth': 5, 'n_estimators': 50}\n",
            ">acc=0.947, >f1=0.937, est=0.953, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.912, >f1=0.879, est=0.970, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.980, est=0.974, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.991, est=0.968, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.949, est=0.958, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.990, est=0.967, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.981, est=0.966, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.946, >f1=0.973, est=0.966, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.975, est=0.965, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.991, est=0.966, cfg={'max_depth': 2, 'n_estimators': 50}\n",
            ">acc=0.965, >f1=0.963, est=0.964, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.947, est=0.961, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.912, >f1=0.887, est=0.975, cfg={'max_depth': 5, 'n_estimators': 50}\n",
            ">acc=0.947, >f1=0.968, est=0.957, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.930, >f1=0.946, est=0.965, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=1.000, >f1=1.000, est=0.966, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.987, est=0.961, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.990, est=0.957, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.982, est=0.956, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.991, est=0.952, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.949, est=0.953, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=1.000, >f1=1.000, est=0.942, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.992, est=0.966, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.966, est=0.965, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.941, est=0.959, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.930, >f1=0.931, est=0.973, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.888, est=0.969, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.988, est=0.960, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.990, est=0.958, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.969, est=0.952, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.944, est=0.971, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=1.000, >f1=1.000, est=0.960, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.930, >f1=0.867, est=0.957, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.978, est=0.964, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.960, est=0.966, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.975, est=0.968, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=1.000, >f1=1.000, est=0.956, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.964, >f1=0.983, est=0.962, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.991, est=0.956, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.912, >f1=0.929, est=0.969, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.991, est=0.965, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.952, est=0.958, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=1.000, >f1=1.000, est=0.956, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.988, est=0.964, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.984, est=0.963, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.897, est=0.965, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.960, est=0.966, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.962, est=0.974, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=1.000, >f1=1.000, est=0.960, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.946, est=0.956, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.984, est=0.957, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.930, >f1=0.909, est=0.972, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=1.000, >f1=1.000, est=0.967, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.990, est=0.958, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.980, est=0.958, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.912, >f1=0.940, est=0.963, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.933, est=0.963, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.964, >f1=0.957, est=0.961, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.888, est=0.966, cfg={'max_depth': 2, 'n_estimators': 50}\n",
            ">acc=1.000, >f1=1.000, est=0.964, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.955, est=0.952, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.964, est=0.977, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.950, est=0.969, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.952, est=0.946, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.992, est=0.959, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.941, est=0.961, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.976, est=0.964, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.992, est=0.963, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.992, est=0.963, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.912, >f1=0.914, est=0.953, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.960, est=0.963, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.979, est=0.965, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.943, est=0.955, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=1.000, >f1=1.000, est=0.951, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.914, est=0.960, cfg={'max_depth': 5, 'n_estimators': 50}\n",
            ">acc=0.982, >f1=0.991, est=0.969, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=1.000, >f1=1.000, est=0.957, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.964, >f1=0.952, est=0.964, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.944, est=0.966, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.937, est=0.957, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.965, est=0.971, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.965, >f1=0.960, est=0.966, cfg={'max_depth': 2, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.965, est=0.962, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.947, >f1=0.971, est=0.968, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=1.000, >f1=1.000, est=0.960, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.991, est=0.970, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            ">acc=0.912, >f1=0.865, est=0.964, cfg={'max_depth': 3, 'n_estimators': 300}\n",
            ">acc=0.982, >f1=0.993, est=0.965, cfg={'max_depth': 5, 'n_estimators': 300}\n",
            "Mean Accuracy: 0.967 (0.023), Mean F1: 0.964 (0.033)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "X55jAQSR86H9"
      },
      "source": [
        "Our nested cross validation procedure has given us 100 measures of model performance (both F1 and accuracy). \n",
        "\n",
        "#### 10.2.1 Plot distributions of F1 & accuracy scores for cross validation\n",
        "\n",
        "Lets plot the distributions and means for F1 and accuracy scores."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 311
        },
        "id": "cG-ikLDl9Qh0",
        "outputId": "a7825455-87e8-4781-f272-2a13d3fd671a"
      },
      "source": [
        "# Plot distribution of model F1 scores\n",
        "plt.hist(outer_results_f1)\n",
        "# Also plot the mean\n",
        "mean_f1 = mean(outer_results_f1)\n",
        "plt.axvline(x=mean_f1, color = 'red');\n",
        "plt.title('Distribution of F1 scores for cross-validation of model')\n",
        "plt.xlabel('F1 Score')\n",
        "plt.ylabel('Count')"
      ],
      "execution_count": 163,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0, 0.5, 'Count')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 163
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 311
        },
        "id": "llYUaFRs9S4o",
        "outputId": "11820523-2d3a-44ba-a344-9286eb1ab972"
      },
      "source": [
        "# Plot distribution of model accuracy scores\n",
        "plt.hist(outer_results_accuracy)\n",
        "# Also plot the mean\n",
        "mean_accuracy = mean(outer_results_accuracy)\n",
        "plt.axvline(x=mean_accuracy, color = 'red');\n",
        "plt.title('Distribution of Accuracy scores for cross-validation of model')\n",
        "plt.xlabel('Accuracy Score')\n",
        "plt.ylabel('Count')"
      ],
      "execution_count": 164,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0, 0.5, 'Count')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 164
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AevLHQ5x-LcD"
      },
      "source": [
        "We know the mean accuracy and F1 scores for our cross validated models is around 0.96, and we can see the distributions of the values in each case. We can use these to better estimate the actual performance on the model on new data, by using these to build confidence intervals for both scores.\n",
        "\n",
        "#### 10.2.2 Build confidence intervals for mean F1 & accuracy scores \n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 379
        },
        "id": "IEllOpusKWKJ",
        "outputId": "7bf98fb3-6c22-44ac-c4c2-4a3070ebaa39"
      },
      "source": [
        "# Build confidence interval for mean F1 score using bootstrap sampling distribution \n",
        "# (Non-parametric - assumes no distribution of mean F1 score)\n",
        "# seed the random number generator\n",
        "seed(1)\n",
        "# Convert list into array\n",
        "outer_results_f1 = np.array(outer_results_f1)\n",
        "\n",
        "# bootstrap\n",
        "scores = list()\n",
        "for _ in range(100):\n",
        "\t# bootstrap sample (with replacement)\n",
        "\tindices = randint(0, 100, 100)\n",
        "\tsample = outer_results_f1[indices]\n",
        "\t# calculate and store statistic\n",
        "\tstatistic = mean(sample)\n",
        "\tscores.append(statistic)\n",
        "\n",
        "print('50th percentile (median) = %.3f' % median(scores))\n",
        "# calculate 95% confidence intervals (100 - alpha)\n",
        "alpha = 5.0\n",
        "# calculate lower percentile (e.g. 2.5)\n",
        "lower_p = alpha / 2.0\n",
        "# retrieve observation at lower percentile\n",
        "lower = max(0.0, percentile(scores, lower_p))\n",
        "print('%.1fth percentile = %.3f' % (lower_p, lower))\n",
        "# calculate upper percentile (e.g. 97.5)\n",
        "upper_p = (100 - alpha) + (alpha / 2.0)\n",
        "# retrieve observation at upper percentile\n",
        "upper = min(1.0, percentile(scores, upper_p))\n",
        "print('%.1fth percentile = %.3f' % (upper_p, upper))\n",
        "print(' ')\n",
        "\n",
        "plt.title('Sampling distribution of mean F1 scores via bootstrap')\n",
        "plt.xlabel('Mean F1 Score')\n",
        "plt.ylabel('Count')\n",
        "plt.hist(scores)\n",
        "plt.axvline(x=median(scores), color = 'red')\n",
        "plt.axvline(x=upper, color = 'green')\n",
        "plt.axvline(x=lower, color = 'green')"
      ],
      "execution_count": 165,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "50th percentile (median) = 0.963\n",
            "2.5th percentile = 0.959\n",
            "97.5th percentile = 0.970\n",
            " \n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.lines.Line2D at 0x7f19df1676a0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 165
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAEVCAYAAADn6Y5lAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAcyElEQVR4nO3deZhcVZ3G8W8nIaAhBNERQVAE9ZcJW9hFVmUZg7ihIMuMIqIgYRBQBwYwssiIOCNCQGVABVlkdUVhICACogwgEAnyKoxBTUCWLCQhpOl0zx/nFCk6vVQ6dau7c97P8+RJ16177/lVddVb555b93RbV1cXZmZWjhGDXYCZmbWWg9/MrDAOfjOzwjj4zcwK4+A3MyuMg9/MrDCjBruA4SoipgAfBTqBJcD+kv5cUVunAk9J+nZEXAx8WdLMJu7/PcCBkg6NiNOBmyTd3cu6xwLnS+ro4b71gbMkfSwiDgXGSzqxwRrWBj4i6eJ8+/3ARpLOG9ij6re93YBvAnMk7VJFGysqImYC84Had6z3AdYDLgS2BN4n6aZBKW4lRMQvgA9Kam9g3Y2AqyS9YyXb3AjYVtJ1fayzO7BQ0n0r09Zw5OAfgIh4J7AbMFHSSxGxAbCoFW1LOrzi/U/pZ5VjgW8Drwj+iBgpaTbwsQE2vTZwOHBxruOnA9xPow4GTpb044rbWVE7SHqxdiMi2kjPy/GtaDz/Hpc2c5+S9mnm/hq0EfARoNfgB3YHngKWC/4qnoehxME/MG8Anpb0EoCkv9XuiIj/BrYGXgV8U9IFeflTwPXAnsA9wDXAF4FXk3pyMyPiEuAFYEdgNHCYpHvqG46I24EjJT2a93l13udjwH6SlkbEB4GzgXnAI8BMSad228+OpJBdAvymbvklwFXALcClwFako5qvAK8B1gfuiYhHgROAHwN/BLaMiH/ilb21TSLiLuD1wH9JujD3so6UdGD948nPxeYR8SBwBfAM+YghIt4KfA8YBzwKfELSorztPfnxjwDeL+mv3R7nctsCHwb2B/aMiL0kTa5b/1BST/sfgLcAn8373xP4jaTD8nrvzTWvAfwWOEpSZ0TcQOqljwa+JOmHuff5Q+APwDbALyV9hgZImgXMiojO3taJiHcB5wFLgUWSdoqI1YBzSB2ULtKH3M8i4hTgQNLvtLZsd+AU0of56hHxPuBbwPjcxL9K+m0+2juS9Jr5paRj62pYB7hb0vh8ewLpyPDd+UhmvKQXe3p+enhIq0fE9cAE4E7S66UzIj4BfC6vc56k/85tLfeYgDOBCfn19LX83EwBXiK9Xo/Pj+WliDiSdPR+IfAAsCvwtXwU+klgdeB+4JO5jtvzeu8G2oGDJD3W2+9nKPIY/8DcAmwWETMi4tyI2LbuvhMlbQtMBA6NiNfl5euSAm08sCmwp6QdgIuAI+q2X5f0wfHP+b6+rAtcL2lT0u/y3RHxKtIbfndgJyB62fYi4BBSEL2hh/snAhtK2lTS5sDP84fYbFKv9KN5vU2B0yT11M52pBDdFvhcHgrqzcnA7yVNlPS1bvedB5wjaQtgJnBc3X0LJW0DXE56I3e33LaSvg/8FJhcH/p1xgOTgA8AP8j/JpA+3CL/To8Fdpc0kRQq++VtP5br2Qk4I/fYydufSnq+doyIt/XyPNwTEQ9GxNW93N+T44Fjci213vURwBhgC9Iw0V0RsT3wftLv/J+AqRGxZl5/G+BQSe8ifQhcL2k74EPABXmdk4GtJG0JfKm+AElzgCciYmJetD8997Z7e37qbQGcQXrOXgfsl4+qTwZ2JnWMjo+Ijfp4TCcD/5NfT1fk2+/NtR+eOwjfBs7M6yi3/aKkbSRdBVwjabv82lkMvK+uxq68ry8D3+jhMQxpDv4BkLSA1BP+LOkFcUtE7JXvPjgiHiD1EN4K1N7gcyXdLakLmAHclpc/DLypbvdXS+qS9AAwKvc6ejNP0h355wdIh7cBPCJpdj4iWa5Hlfc5QtKDuZ6eQub/gDdFxNSI2EPS/F5q+IOkGb3c9wtJz0t6HpgGbN/HY+nLxLqe4eWk0KipDQnVHv+KbNubaXm45fekIKj/vb2JFDxbAL/NPco9gY3ztsdFxEOknupGLPtQ/YOkP+Xhg+m91ArpQ3Vi3QdrI+4Gzo6Io0m9U0i90Qvza6lL0lzgncB1kpZIehL4HemDCOBOSU/ln/cCTsuP7Qbg9RExivSavjwiDiT1+ru7ljS8AukDo6fefG/PTz3VvTavIoX9tsDNkubl998vgB36eUz1fg1cFBGHsewcSk+urft5y4j4dUT8nhT6E+ruuyoX+hNSFgwrHuoZIKWTm9OAaRHxLPCBiHic1NN6p6QF+bC29kasP7HVWXe7ExhZd19Xt5/7epHWv/lq++mpB9WT7u28gqS5EbEFqQd5QkTs0n24KHthBdroIvWO6zscq7Nyas9B9+dxZbQD5MP67r+32nP8E0mfrt8oD7lsD2wvaUlEPMyyx9fT76opJH0lIm4khdM93Y5AG1X/e2wDJuVzNi/Lw1vvIg2VHUk6qqz3Y+CXEXEZML/ug6S2fV/PT70VeQ806jOkD+z3k56jzXtZr/55uIj0PPwpIj4PrFl3X5/vn6HOPf4ByIf7m+Sf24DNgL8AY4EFwMI8rrvbAHZ/QN7vlkBHHz3t3jxKGtt8Qx7n/WD3FSTNAzojYotc/wHd18nDGW2SriaNl9YO4ReQHmcjJkXE2IgYC+wB3Et6njaNiFER8WZSz7m//T6Yx50hnZS9q8H2V3bb3vwW2CMPPxARr80/r0X6ltCSPATxj01oq18RsXHuIZ9BGs7akNQpOSIi2vK/tUlHBh+KiNUi4g2knmpPR2vTgPrzHltGxAhgA0nTSOPsb+++kaRnSSdLTyWdz+qu0ednfLfX5l2k184eEbFWHsqZRDq/09tj6v56eovSN9VOAlZj2Xu1r9fyGOCZiFiDdA6g3kcBImJf0tHmsOIe/8CsCZwfEWvl2/cDUyUtzr3+R4HHSYeXK+qZiPgdqSd02IpunGv4HHAH6eTuY8DzPaz6adLY9RLSm2fNbvdvAHwvv/k6gKPz8ouAO/Ph+gn9lHM/cCPLTu7OBoiIW0lvzumkoS4kPRcR0yNiOnAZ6eRuzTG5ljNZdoK2USuzbY8kPR0Rk4Gf5A/Xl4BPAf8DTI6I2mN7aGXbiogAbiWdWH9vRMyQtHu31Y7PvemlpNfcQ6TndUL+vwM4SdLPI+LnpKDqJJ0XWJiaeIXTSWPl00kZcSvpPMKV+UO8jXSitCfXksbOP9/DfY0+P9NJ5xA2Jb2Of5SPvr5Keq1COm8zE5jZy2OaDryq7uTuRyKd6G8DLs1HtD8DrouIg0knh7s7kzR09HeWD/eR+T3QDhzUy+MYsto8LfPQEfkbNVrJ72pHxBilb72MIo3RniXp9iaUaFa8qPtm3WDXMlAe6lk1HZV7Or8HHnbom1k99/jNzArjHr+ZWWEc/GZmhXHwm5kVprKvc0bEusCPSF91G0m64ONx4BLSfC8Pky6Z73UOkrp9+USEmdkASFruos4qv8f/LLBz/v7t7sCJpMnA7pP0tYi4AHgP6dLrfi2bSmPVMXvBbN749Tcy6/hZrD+2r2lsrGVmz4Y3vhFmzYL1/TuxZYbj+7WHazSACod6JC2t682vTbpYYxfS98rJ/+9aVftmZtazSq/cjTQ168WkS8g/TJr8aW6+ex6wTpXtm5nZ8io9uSvpEUnvBPYFppJCvzbb5DhgTpXtm5nZ8ioL/oion3VvHmnWuztYNl/4PqSpWc3MrIWqHOrZOiLOIk2c1Eaa5OlR4JKIuIP014hurLB9MzPrQWXBL+k39Dwt8XJTAJuZWev4Ai4zs8I4+M3MCuM/xGKrhAWL2mnvWLrS+2l7fjHrAHOeX0zXmMX9rj961EjGjhm90u2atZKD31YJ7R1LOfT0m1d6P+ssnMOlwHHn3MGcNfu/zOSSKXuvdJtmreahHjOzwjj4zcwK4+A3MyuMg9/MrDAOfjOzwjj4zcwK4+A3MyuMg9/MrDAOfjOzwjj4zcwK4+A3MyuMg9/MrDAOfjOzwjj4zcwK4+A3MyuMg9/MrDAOfjOzwjj4zcwK4+A3MyuMg9/MrDAOfjOzwjj4zcwK4+A3MyuMg9/MrDCjqtpxROwIfB1oBxYChwDHAvsDzwBPSTqwqvbNzKxnlQU/8ASwh6QXIuJIYHJe/iVJ11XYrpmZ9aGyoR5JsyW9kG+2Ax3551Mi4s6IOKiqts3MrHeVj/FHxGuBo4DvAFMlTQT2BT4fERtW3b6Zmb1SpcEfEa8GrgWOkfSspOcAJM0HbgU2q7J9MzNbXmXBHxGjgKtIvfy787JxdfftCDxeVftmZtazKk/uHgTsCqwVEZ8Ffg5EREwARgJXSvpjhe2bmVkPKgt+SZcBl1W1fzMzGxhfwGVmVhgHv5lZYRz8ZmaFcfCbmRXGwW9mVhgHv5lZYRz8ZmaFcfCbmRXGwW9mVhgHv5lZYRz8ZmaFcfCbmRXGwW9mVhgHv5lZYRz8ZmaFcfCbmRXGwW9mVhgHv5lZYRz8ZmaFcfCbmRXGwW9mVhgHv5lZYRz8ZmaFcfCbmRXGwW9mVhgHv5lZYRz8ZmaFGVXVjiNiR+DrQDuwEDgkt3cZMBaYJunUqto3M7OeVdnjfwLYQ9JuwM+AycAJwHcl7QxsFxETKmzfzMx6UFmPX9LsupvtQAewMzAlL7sB2BV4pKoazMxseZWP8UfEa4GjgO8AYyQtznfNA9apun0zM3ulSoM/Il4NXAscI+lZ4IWIWCPfPQ6YU2X7Zma2vMqCPyJGAVcBUyXdnRffCeyTf56Ub5uZWQtVNsYPHEQaw18rIj4L/Bw4G/h+RBwP3CZpRoXtm5lZD6o8uXsZ6aub3U2qqk0zM+ufL+AyMyuMg9/MrDAOfjOzwjj4zcwK4+A3MyuMg9/MrDAOfjOzwjj4zcwK4+A3MyuMg9/MrDAOfjOzwjj4zcwK4+A3MyuMg9/MrDAOfjOzwjj4zcwK4+A3MytMQ8EfEQc1sszMzIa+Rnv8X2hwmZmZDXF9/s3diPgw8BHgzRFxZd1dawFzqyzMzMyq0d8fW78PeA7YELiwbvkCYHpVRZmZWXX6DH5JTwBPADsDRMRIYGS+2yeGzcyGof56/ABExL8ApwNtQGf+vwvYuLrSzMysCg0FPzAFeIekv1dZjJmZVa/R4H8CmFdlIWa2YhYsaqe9Y2nL2x09aiRjx4xuebvWPI0G/zPA/0bENGBJbaGkkyqpysz61d6xlENPv7nl7V4yZe+Wt2nN1Wjw35T/mZnZMNdQ8Eu6tOpCzMysNRr9Vs+TpG/xvIKk9fvYZjXgdmBT4HBJ10XEqcD+pKGjpyQdOICazcxsJTTa41+v9nNEjAY+AEzoZ7MO0lW/R3Rb/iVJ161IkWZm1jwrfBGWpHZJ1wIf6me9LklP9nDXKRFxpyd5MzMbHI0O9Xy67uYIYCsGNlfPVEmnRsQ44LaIuEvSXwewHzMzG6BGe/zr1f17HXAXsO+KNibpufz/fOBWYLMV3YeZma2cRsf4T2tGYxExTtL8iBgF7Ahc3Iz9mplZ4xod6nkr8C1gfF70CDBZ0mP9bHcNsC2wMCK2B9aJiAmkid6ulPTHAVduZmYD0ugFXN8D/kPSjQAR8Z68bJe+NpJ0wMqVZ2ZmzdboGP+4WugDSLqJ9MdYzMxsmGm0x/9wRHwDuDzfPhCYUU1JZmZWpf7+9OI44DXAJ4GjgZPzXb8G/rPa0szMrAr99fjPBy6VNBP4Wm1hROwBfBX4eHWlmZlZFfob4x8vaVr3hZJuBbaupiQzM6tSf8G/dh/3rdHMQszMrDX6C/6HI+KQ7gsj4mDSd/nNzGyY6W+M/1+BH0XEYcCDpKmZtyZ9lfODFddmZmYV6DP4Jf0N2C6fzK1Nw3xTT+P+ZmY2PDQ6V8+tpEnVzMxsmGv0Ai4bRhYsaqe9Y2nL2x09aiRjx4xuebtmtmIc/Kug9o6lHHr6zS1v95Ipe7e8TTNbcSv8F7jMzGx4c/CbmRXGwW9mVhgHv5lZYRz8ZmaFcfCbmRXGwW9mVhgHv5lZYRz8ZmaFcfCbmRXGwW9mVhgHv5lZYTxJm9lKGNHWxnPzFw9K211dg9KsrQIc/GYroaOzk8POuGVQ2v7uF/calHZt+PNQj5lZYRz8ZmaFqWyoJyJWA24HNgUOl3RdRLwOuAwYC0yTdGpV7ZuZWc+q7PF3AB8BvlG37ATgu5J2Jv0R9wk9bmlmZpWpLPgldUl6stvinYEb8s83ALtW1b6ZmfWs1d/qGSOp9t23ecBbWty+VchfbTQbHlod/C9ExBqSXgTGAXNa3L5VyF9tNBseWv2tnjuBffLPk/JtMzNroUp7/BFxDbAtsDAitgfOBr4fEccDt0maUWX7Zma2vEqDX9IBPSyeVGWbZmbWN1/AZWZWGAe/mVlhHPxmZoVx8JuZFcbBb2ZWGAe/mVlhHPxmZoVx8JuZFcbBb2ZWGAe/mVlhHPxmZoVx8JuZFcbBb2ZWGAe/mVlhHPxmZoVx8JuZFcbBb2ZWmFb/sXUzswFZsKid9o6lg9L26FEjB6Xdqjj4zWxYaO9YyqGn3zwobV8yZe9VanxkFXooZmbWCAe/mVlhHPxmZoVx8JuZFcbBb2ZWGAe/mVlhHPxmZoVx8JuZFWZQLuCKiEXAvfnmWZJuGow6zMxKNFhX7v5Z0u6D1LaZWdEGK/g3jIg7gL8Cx0h6bpDqMDMrzmCN8W8iaVfgVuDMQarBzKxIgxL8kp7NP14FbDUYNZiZlarlwR8RYyKiNsfpbsCfWl2DmVnJBmOMfzxwcUQ8DywBDh+EGszMitXy4Jd0Px7eMTMbNL6Ay8ysMA5+M7PCOPjNzArjv7lbkUb+MPSchYvT/88vZvXOxU1ru6urabsyW86Itjaem9+812ujBvN1PaKtjTnPV/N+7cvoUSMZO2Z00/fr4K9II38Y+sW2OTAWjjvnDtboWqdpbX/3i3s1bV9m3XV0dnLYGbe0vN3BfF13dHZy3Dl3VPJ+7cslU/auZL8e6jEzK4yD38ysMA5+M7PCOPjNzArj4DczK4yD38ysMA5+M7PCOPjNzArj4DczK4yD38ysMA5+M7PCOPjNzAqzyk/S1sgsmVXwDJlmNlSt8sHfyCyZVfAMmWY2VHmox8ysMA5+M7PCOPjNzArj4DczK4yD38ysMA5+M7PCOPjNzArj4DczK4yD38ysMC2/cjciPgV8AmgHDpP0f62uwcysZC3t8UfEOsDhwK7AF4CzWtm+mZm1fqhne+B2SR2S7gWixe2bmRWv1cG/DjB3ENs3MyteW1cL5w+OiEnAbpJOzLcflDSxge08ybGZ2QBIauu+rNUnd+8BpkTESGBL4E+NbNRT4WZmNjAtDX5JcyLiUuBO4CXgk61s38zMWjzUY2Zmg88nV83MCuPgNzMrjIPfzKwwDn4zs8K0fK6e4aavuYUiYjLwz/m+4yT9LiJ2B65g2VdVD5T0VES8BziD9G2muyT922DWmZcfDhxIeh2cKemWiNgOOBdoy8tuaGadzaoVuBe4Hhidaz2mtu5QqlPSLXn5OOBx4EhJ1w3FOiPi9cAFwGuBpyQd3Mw6m1zrx4GjgU7gSknnDmadEXEw8Om8ygbATyUdHxGbAN8hvU6/J+miZtY5UA7+PtTNLbQTsBVpbqED8n3rAocAOwOvA64C3pU3/ZGko7vt7ovAfpL+GhG3RsTGzZqgbiB1RsTmwHaS9uy2u3OA/YH5wJ0RcaOkpc2os5m1RsSrgI9JmhUR44GpwF5Drc46nyddx9JUTa7zv4ATJT3e7DorqPXfgW2AxcDDEfEtSe2DVaekK4Er8zrXkjol5G2/APwOuDsirpc0pxl1rgwHf99enlsIuDci6ucW2giYkUPx7xGxQUSsnu/bNyImAr8CTpHUBcwA1o6IJ4HVgOcHuc79gKURMQ14GphMehONkjQLICL+CLwNeHSo1SppLjArb9cOdDSxxqbWmcNiY9JRSrM163f/PDAeODUi3gR8U9LVQ7HW/Lt/FFgzb7sYaFrnZCB1SloCEBFjgc2Au/P6b8/zkhERt+d939TEWgfEY/x962tuoceAbSLi1RHxduDNwGuA+0iTz+0KrA98NK9/NekXLuA+Sc8Ocp3rA2NyT+o24MS8n3l1287Ly5qpWbUCEBFtwNeBs4dwnScDX21yfc2u8/Wkq+m/DOwDnJh7vkOxVoDrgAdIHwCXNvOodIB11uwH/CR39iANQ9ZU8X4aEAd/3+YCa9fdfvnFJek50mHcjcApwHTgWUkLJS2R1AlcSzpUBDgf2JbUg14/InYazDrzNjfn1W4CNu9hP+OAZh+WNqvWmnNJvbNfDsU6I+ItwNqSpje5vqbWmZc9oWQRcD/w1qFYa+5Vn0TqYG0CfDgfpQxmnTUHAT+ou11/hWwV76cBcfD37R5gt4gYGRFb021uIUnXSNoNOA14VFJHPolXs1vdNkuB+fkDYS6v7CW0vE7SMNQ2eZVtgcclLQY6ImK9iBhD+pB6rIl1Nq1WgIg4CeiQ9I0m19jMOrcCNomIm0gnBE+OiE2HWp2SXgRmRcS6kebS2gx4ool1Nq1W0gnddmBRHtd/AVhrkOsknxxfT9JDdav/MSK2zs/pbsD/NrHOAfOUDf2IiCOBj7FsbqGdgT9L+lVEXAGsRwryyfnbO0cAnyKNOz4OfErSSxHxUeD4vJ+/kE5MNm1cegB1jgDOI73BO3I9syNiB9IJ3jbgK5J+2qwam1krMBKYCdxF6lXNknTIUKtT0uy6/Z0KPFzBt3qa9bvfljRsNhq4QtLUZtbZ5FqPJfWuu4DfSDpuMOvM20wGXiPpy3X7eSvpWz2rkYakLmxmnQPl4DczK4yHeszMCuPgNzMrjIPfzKwwDn4zs8I4+M3MCuMpG2xYioguYKqkY/LtIF3F+RlJ366ozUuAXYAFedEU4BbgBmAH4HxJJ/ay7RTSVdydwBJgf0l/rqJOs/44+G24egbYKSLa8uXx+wMPt6DdyZJenmslIlYDTgcmkC7fX05EvJN08c7EfE3HBsCilSkiIkY2eZoCK4iD34arTtIVljuSJsTaB/hF7c6IeBvwTdLcKM8AH5f094g4La/7KtLUuSfl9WcClwIfJM1Muq+kfifSk/QS8Ks8NUNv3gA8nddF0t/q6nw/6YOjDbhX0uGRpsb+FrA66QK1oyUtzTX+AJgEHJ3nivkM6YKrH0o6rb96zcBj/Da8XQvsnwPwr6SrpWsuAA6XtA3pyskv5uXnStoO2ALYIiK2rNvmMUlbAg+xbHK97i6IiAfzv7EN1nkLsFlEzIiIc/MVsrUpfr8OTMrt1v5Gw/eAIyRtTvrgOqhuXzMlTSTN+fIe4B2kaSG2qe3XrD/u8dtwdgdp1svnSLM1bgovT427E/CTPKPuSPL8PsAeEfFvpN70uqQhmtrcKrXpKR4gTf7Vk1cM9TRC0oKI2ArYHdgTuCUiDiAddUyT9GReb05ErA2MkHR/3vwKUg//8nz72trjIB3t1NZbkzS30n0rUpuVycFvw1Ye/rgPOJI0U2Nt8rMRpLl7JtavHxFrkHrY20h6OiLOJ30A1CzJ/3eSPiyaWWsHMA2YFhHPAh9g2ayTK+KF/H8b8C1J/9GkEq0gHuqx4W4qcEKeShgASfOBuRGxN6QTsBHxj8AapFCfm+eaf18rCoxkk/xzG2nCsb8AvwX2jIj18n3rSJpH+sMjtSGog0jj/N3dBhyYjxDIfxDktRU/FFtFOPhtWJP0B0lX9HDXIcAXIuIh4EFghxyqPwD+APwQ+E0zaoiI6aQjiaMi4m8R8Q/dVlkTuDwiZpC+eTSC9FXUp4HjgJtynWfl9Q8DvhMRvyedaL6qe5uSHgb+k3RieTpwDTCmGY/HVn2endPMrDDu8ZuZFcbBb2ZWGAe/mVlhHPxmZoVx8JuZFcbBb2ZWGAe/mVlhHPxmZoX5f9GyY56RYsW0AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 379
        },
        "id": "OJq-Ozdc_r70",
        "outputId": "8b97d026-bbc0-454e-cfba-b6aa310be01b"
      },
      "source": [
        "# Build confidence interval for mean accuracy score using bootstrap sampling distribution \n",
        "# (Non-parametric - assumes no distribution of mean accuracy score)\n",
        "# seed the random number generator\n",
        "seed(1)\n",
        "# Convert list into array\n",
        "outer_results_accuracy = np.array(outer_results_accuracy)\n",
        "\n",
        "# bootstrap\n",
        "scores = list()\n",
        "for _ in range(100):\n",
        "\t# bootstrap sample (with replacement)\n",
        "\tindices = randint(0, 100, 100)\n",
        "\tsample = outer_results_accuracy[indices]\n",
        "\t# calculate and store statistic\n",
        "\tstatistic = mean(sample)\n",
        "\tscores.append(statistic)\n",
        "\n",
        "print('50th percentile (median) = %.3f' % median(scores))\n",
        "# calculate 95% confidence intervals (100 - alpha)\n",
        "alpha = 5.0\n",
        "# calculate lower percentile (e.g. 2.5)\n",
        "lower_p = alpha / 2.0\n",
        "# retrieve observation at lower percentile\n",
        "lower = max(0.0, percentile(scores, lower_p))\n",
        "print('%.1fth percentile = %.3f' % (lower_p, lower))\n",
        "# calculate upper percentile (e.g. 97.5)\n",
        "upper_p = (100 - alpha) + (alpha / 2.0)\n",
        "# retrieve observation at upper percentile\n",
        "upper = min(1.0, percentile(scores, upper_p))\n",
        "print('%.1fth percentile = %.3f' % (upper_p, upper))\n",
        "print(' ')\n",
        "\n",
        "plt.title('Sampling distribution of mean Accuracy scores via bootstrap')\n",
        "plt.xlabel('Mean Accuracy Score')\n",
        "plt.ylabel('Count')\n",
        "plt.hist(scores)\n",
        "plt.axvline(x=median(scores), color = 'red')\n",
        "plt.axvline(x=upper, color = 'green')\n",
        "plt.axvline(x=lower, color = 'green')"
      ],
      "execution_count": 166,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "50th percentile (median) = 0.967\n",
            "2.5th percentile = 0.963\n",
            "97.5th percentile = 0.971\n",
            " \n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.lines.Line2D at 0x7f19de19fb38>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 166
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kmOxJene9Hkj"
      },
      "source": [
        "## 11 ML model evaluation\n",
        "\n",
        "Below is a summary table of the key results from developing our ML model."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9XQYaRRg9Hkj"
      },
      "source": [
        "| Metric | Benchmark Model | Optimized Model |\n",
        "| --- | --- | --- | \n",
        "| Accuracy Score | 0.3726 | 0.9670 |\n",
        "| F-score | 0.4260 | 0.9630 |"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qMLchzuU9Hkj"
      },
      "source": [
        "We will take the F1 score as the better indicator of model performance. We can then say, as a result of our cross-validation procedure, that we estimate the model F1 score performance to be 0.96, within a range of 0.95-0.97 with a 95% confidence level.\n",
        "\n",
        "Let us now examine the model to see which features it found most useful for prediction.\n",
        "\n",
        "## 12 Feature importance\n",
        "\n",
        "We will now define a function to extract and plot the top most useful features for the model."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KHYvKQ0Q9Hkj"
      },
      "source": [
        "def feature_plot(importances, X_train, y_train):\n",
        "\n",
        "    # Display the five most important features\n",
        "    indices = np.argsort(importances)[::-1]\n",
        "    columns = X_train.columns.values[indices[:5]]\n",
        "    values = importances[indices][:5]\n",
        "\n",
        "    # Creat the plot\n",
        "    fig = plt.figure(figsize = (9,5))\n",
        "    plt.title(\"Normalized Weights for First Five Most Predictive Features\", fontsize = 16)\n",
        "    plt.bar(np.arange(5), values, width = 0.6, align=\"center\", color = '#00A000', \\\n",
        "          label = \"Feature Weight\")\n",
        "    plt.bar(np.arange(5) - 0.3, np.cumsum(values), width = 0.2, align = \"center\", color = '#00A0A0', \\\n",
        "          label = \"Cumulative Feature Weight\")\n",
        "    plt.xticks(np.arange(5), columns)\n",
        "    plt.xlim((-0.5, 4.5))\n",
        "    plt.ylabel(\"Weight\", fontsize = 12)\n",
        "    plt.xlabel(\"Feature\", fontsize = 12)\n",
        "\n",
        "    plt.legend(loc = 'upper center')\n",
        "    plt.tight_layout()\n",
        "    plt.show()"
      ],
      "execution_count": 167,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "A5WBpc5r9Hkj",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 369
        },
        "outputId": "c257cc12-e55b-447d-b029-6a176a53baf8"
      },
      "source": [
        "# Extract the feature importances \n",
        "importances = best_result.best_estimator_.feature_importances_\n",
        "\n",
        "# Plot\n",
        "feature_plot(importances, X_train, y_train)"
      ],
      "execution_count": 168,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 648x360 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IwxmJhLz9Hkj"
      },
      "source": [
        "We can see from the extracted feature importances that the top 3 features used by the model are:\n",
        "\n",
        "- perimeter worst\n",
        "- area_worst\n",
        "- concave points_worst\n",
        "\n",
        "Interestingly, these were features we highlighted earlier in the univariate analysis as features that had a very different distribution of values for the malignant and benign tumor sub-groups.\n",
        "\n",
        "## 13 Conclusion"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2LPz3zBM9Hkj"
      },
      "source": [
        "In this project we have looked at a dataset relating to breast cancer, and have \n",
        "developed a model able to predict malignant tumors with a very high degree of accuracy (an F-score of 0.963).\n",
        "\n",
        "We have also gained an understanding of the reasons why our model is able to predict this well. Our earlier analysis showed the difference in morphology beween cell metrics for malignant v benign tumors, which could be seen visually in the images and were expressed in different distributions of values for particular feature measurements of the cells that we observed.\n",
        "\n",
        "Our use of higher level analytical tools such as UMAP also allowed us to gain a much better understanding of the dataset, in particular a better idea of the range of feature values that were typical for malignant and benign tumors as 'groups' within the dataset."
      ]
    }
  ]
}