train_test_EDA.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Imports for EDA and beyond\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import re\n",
    "\n",
    "import time\n",
    "\n",
    "from collections import Counter\n",
    "\n",
    "from sklearn.feature_extraction.text import CountVectorizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Imports: NOT USED\n",
    "from geopy.distance import vincenty\n",
    "from geopy.geocoders import Nominatim"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Testing some Geopy stuff\n",
    "# newport_ri = (41.49008, -71.312796)\n",
    "# cleveland_oh = (41.499498, -81.695391)\n",
    "# print(vincenty(newport_ri, cleveland_oh).miles)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "# EDA and Organization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Read in dataset and holdout dataset\n",
    "df = pd.read_csv('./Original_Data/train.csv')\n",
    "df_holdout = pd.read_csv('./Original_Data/test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Separate our X and y from dataset, and set our X from holdout dataset\n",
    "X = df.drop('WnvPresent', axis=1)\n",
    "X_holdout = df_holdout\n",
    "y = df['WnvPresent']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10506, 11)\n",
      "(116293, 11)\n",
      "(10506,)\n"
     ]
    }
   ],
   "source": [
    "# Looks like our target is included in the Train set\n",
    "print(X.shape)\n",
    "print(X_holdout.shape)\n",
    "print(y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Counter({0: 9955, 1: 551})"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Going to have VERY unbalanced classes\n",
    "Counter(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Date', 'Address', 'Species', 'Block', 'Street', 'Trap',\n",
       "       'AddressNumberAndStreet', 'Latitude', 'Longitude', 'AddressAccuracy',\n",
       "       'NumMosquitos'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Id', 'Date', 'Address', 'Species', 'Block', 'Street', 'Trap',\n",
       "       'AddressNumberAndStreet', 'Latitude', 'Longitude', 'AddressAccuracy'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_holdout.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Looks like our dataset has a column our houldout dataset does not ('NumMosquitos'), and our holdout dataset has a column our\n",
    "# dataset does not ('Id'). We'll drop 'NumMosquitos' from dataset, since we won't have that information to work\n",
    "# from in our holdout dataset, and add a sequential Id\n",
    "\n",
    "X['Id'] = range(1,len(X)+1)\n",
    "X.drop('NumMosquitos', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10506, 11)\n",
      "(116293, 11)\n",
      "(10506,)\n"
     ]
    }
   ],
   "source": [
    "# Shapes/Columns look good now\n",
    "print(X.shape)\n",
    "print(X_holdout.shape)\n",
    "print(y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Convert Date columns to datetime type\n",
    "X['Date'] = pd.to_datetime(X['Date'])\n",
    "X_holdout['Date'] = pd.to_datetime(X_holdout['Date'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "# New Feature Generation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Create new features out of Date\n",
    "X['Year'] = X['Date'].map(lambda x: x.year)\n",
    "X['Month'] = X['Date'].map(lambda x: x.month)\n",
    "X['Day'] = X['Date'].map(lambda x: x.day)\n",
    "X['DayofYear'] = X['Date'].map(lambda x: x.dayofyear)\n",
    "\n",
    "X_holdout['Year'] = X_holdout['Date'].map(lambda x: x.year)\n",
    "X_holdout['Month'] = X_holdout['Date'].map(lambda x: x.month)\n",
    "X_holdout['Day'] = X_holdout['Date'].map(lambda x: x.day)\n",
    "X_holdout['DayofYear'] = X_holdout['Date'].map(lambda x: x.dayofyear)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Create new feature of Zip Code\n",
    "\n",
    "zipcodes = []\n",
    "for i in X['Address']:\n",
    "    try:\n",
    "        zipcode = re.findall('[\\d][\\d][\\d][\\d][\\d]', i)[0]\n",
    "    except:\n",
    "        zipcode = ''\n",
    "    zipcodes.append(zipcode)\n",
    "    \n",
    "zipcodes_holdout = []\n",
    "for i in X_holdout['Address']:\n",
    "    try:\n",
    "        zipcode = re.findall('[\\d][\\d][\\d][\\d][\\d]', i)[0]\n",
    "    except:\n",
    "        zipcode = ''\n",
    "    zipcodes_holdout.append(zipcode)\n",
    "    \n",
    "X['Zipcode'] = zipcodes\n",
    "X_holdout['Zipcode'] = zipcodes_holdout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Convert Block to string type\n",
    "X['Block'] = X['Block'].map(lambda x: str(x))\n",
    "X_holdout['Block'] = X_holdout['Block'].map(lambda x: str(x))\n",
    "\n",
    "X['Year'] = X['Year'].map(lambda x: str(x))\n",
    "X_holdout['Year'] = X_holdout['Year'].map(lambda x: str(x))\n",
    "\n",
    "X['Month'] = X['Month'].map(lambda x: str(x))\n",
    "X_holdout['Month'] = X_holdout['Month'].map(lambda x: str(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Count Vectorize some features for X_train\n",
    "\n",
    "cvec_spec = CountVectorizer()\n",
    "X_spec = cvec_spec.fit_transform(X['Species'])\n",
    "X_spec = pd.DataFrame(X_spec.todense(), columns=cvec_spec.get_feature_names())\n",
    "\n",
    "cvec_zip = CountVectorizer()\n",
    "X_zip = cvec_zip.fit_transform(X['Zipcode'])\n",
    "X_zip = pd.DataFrame(X_zip.todense(), columns=cvec_zip.get_feature_names())\n",
    "\n",
    "#cvec_block = CountVectorizer()\n",
    "#X_block = cvec_block.fit_transform(X['Block'])\n",
    "#X_block = pd.DataFrame(X_block.todense(), columns=cvec_block.get_feature_names())\n",
    "\n",
    "cvec_trap = CountVectorizer()\n",
    "X_trap = cvec_trap.fit_transform(X['Trap'])\n",
    "X_trap = pd.DataFrame(X_trap.todense(), columns=cvec_trap.get_feature_names())\n",
    "\n",
    "#cvec_year = CountVectorizer()\n",
    "#X_year = cvec_year.fit_transform(X['Year'])\n",
    "#X_year = pd.DataFrame(X_year.todense(), columns=cvec_year.get_feature_names())\n",
    "\n",
    "cvec_month = CountVectorizer()\n",
    "X_month = cvec_month.fit_transform(X['Month'])\n",
    "X_month = pd.DataFrame(X_month.todense(), columns=cvec_month.get_feature_names())\n",
    "\n",
    "#X = pd.concat([X, X_spec, X_zip, X_block, X_trap, X_year, X_month], axis=1)\n",
    "X = pd.concat([X, X_spec, X_zip, X_trap, X_month], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Count Vectorize some features for X_test\n",
    "\n",
    "X_holdout_spec = cvec_spec.transform(X_holdout['Species'])\n",
    "X_holdout_spec = pd.DataFrame(X_holdout_spec.todense(), columns=cvec_spec.get_feature_names())\n",
    "\n",
    "X_holdout_zip = cvec_zip.transform(X_holdout['Zipcode'])\n",
    "X_holdout_zip = pd.DataFrame(X_holdout_zip.todense(), columns=cvec_zip.get_feature_names())\n",
    "\n",
    "#X_holdout_block = cvec_block.transform(X_holdout['Block'])\n",
    "#X_holdout_block = pd.DataFrame(X_holdout_block.todense(), columns=cvec_block.get_feature_names())\n",
    "\n",
    "X_holdout_trap = cvec_trap.transform(X_holdout['Trap'])\n",
    "X_holdout_trap = pd.DataFrame(X_holdout_trap.todense(), columns=cvec_trap.get_feature_names())\n",
    "\n",
    "#X_holdout_year = cvec_year.transform(X_holdout['Year'])\n",
    "#X_holdout_year = pd.DataFrame(X_holdout_year.todense(), columns=cvec_year.get_feature_names())\n",
    "\n",
    "X_holdout_month = cvec_month.transform(X_holdout['Trap'])\n",
    "X_holdout_month = pd.DataFrame(X_holdout_month.todense(), columns=cvec_month.get_feature_names())\n",
    "\n",
    "#X_holdout = pd.concat([X_holdout, X_holdout_spec, X_holdout_zip, X_holdout_block, X_holdout_trap,\n",
    "#                      X_holdout_year, X_holdout_month], axis=1)\n",
    "X_holdout = pd.concat([X_holdout, X_holdout_spec, X_holdout_zip, X_holdout_trap, X_holdout_month], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Incorporate Weather Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "weather = pd.read_csv('./weather_odd.csv')\n",
    "weather.drop(['Unnamed: 0', 'Unnamed: 0.1'], axis=1, inplace=True)\n",
    "\n",
    "weather_holdout = pd.read_csv('./weather_even.csv')\n",
    "weather_holdout.drop(['Unnamed: 0', 'Unnamed: 0.1'], axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "weather['Date'] = pd.to_datetime(weather['Date'])\n",
    "weather_holdout['Date'] = pd.to_datetime(weather_holdout['Date'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(729, 40)\n",
      "(735, 40)\n"
     ]
    }
   ],
   "source": [
    "print(weather.shape)\n",
    "print(weather_holdout.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10506, 210)\n",
      "(116293, 210)\n"
     ]
    }
   ],
   "source": [
    "print(X.shape)\n",
    "print(X_holdout.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Engineer an 'Hours_in_Day' feature\n",
    "\n",
    "sunrise_hrs = []\n",
    "sunrise_mins = []\n",
    "\n",
    "for i in range(len(weather)):\n",
    "    time_tokens = list(str(weather['Sunrise'][i]))\n",
    "    minute = int(time_tokens[-2] + time_tokens[-1])\n",
    "    sunrise_mins.append( minute )\n",
    "    try:\n",
    "        hour = int(time_tokens[-4] + time_tokens[-3])\n",
    "        sunrise_hrs.append( hour )\n",
    "    except:\n",
    "        hour = int(time_tokens[0])\n",
    "        sunrise_hrs.append( hour )\n",
    "        \n",
    "sunset_hrs = []\n",
    "sunset_mins = []\n",
    "\n",
    "for i in range(len(weather)):\n",
    "    time_tokens = list(str(weather['Sunset'][i]))\n",
    "    minute = int(time_tokens[-2] + time_tokens[-1])\n",
    "    sunset_mins.append( minute )\n",
    "    try:\n",
    "        hour = int(time_tokens[-4] + time_tokens[-3])\n",
    "        sunset_hrs.append( hour )\n",
    "    except:\n",
    "        hour = int(time_tokens[0])\n",
    "        sunset_hrs.append( hour )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "weather['Hours_in_Day'] = (sunset_hrs + (np.array(sunset_mins) / 60) - sunrise_hrs + (np.array(sunrise_mins) / 60))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Engineer an 'Hours_in_Day' feature in Holdout\n",
    "\n",
    "sunrise_hrs = []\n",
    "sunrise_mins = []\n",
    "\n",
    "for i in range(len(weather_holdout)):\n",
    "    time_tokens = list(str(weather_holdout['Sunrise'][i]))\n",
    "    minute = int(time_tokens[-2] + time_tokens[-1])\n",
    "    sunrise_mins.append( minute )\n",
    "    try:\n",
    "        hour = int(time_tokens[-4] + time_tokens[-3])\n",
    "        sunrise_hrs.append( hour )\n",
    "    except:\n",
    "        hour = int(time_tokens[0])\n",
    "        sunrise_hrs.append( hour )\n",
    "        \n",
    "sunset_hrs = []\n",
    "sunset_mins = []\n",
    "\n",
    "for i in range(len(weather_holdout)):\n",
    "    time_tokens = list(str(weather_holdout['Sunset'][i]))\n",
    "    minute = int(time_tokens[-2] + time_tokens[-1])\n",
    "    sunset_mins.append( minute )\n",
    "    try:\n",
    "        hour = int(time_tokens[-4] + time_tokens[-3])\n",
    "        sunset_hrs.append( hour )\n",
    "    except:\n",
    "        hour = int(time_tokens[0])\n",
    "        sunset_hrs.append( hour )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "weather_holdout['Hours_in_Day'] = (sunset_hrs + (np.array(sunset_mins) / 60) - sunrise_hrs + (np.array(sunrise_mins) / 60))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Merge our X's with the weather data\n",
    "X = X.merge(weather, how='left', on='Date')\n",
    "X_holdout = X_holdout.merge(weather_holdout, how='left', on='Date')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10506, 250)\n",
      "(116293, 250)\n"
     ]
    }
   ],
   "source": [
    "print(X.shape)\n",
    "print(X_holdout.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Let's drop some X (See README for explanations)\n",
    "drop_cols = ['Date', 'Address', 'Species', 'Street', 'AddressNumberAndStreet', 'AddressAccuracy', 'CodeSum', 'Year_y', \n",
    "             'Month_y', 'Day_y', 'Unnamed: 24', 'Zipcode', 'Block', 'Trap']\n",
    "X.drop(drop_cols, axis=1, inplace=True)\n",
    "X_holdout.drop(drop_cols, axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10506, 236)\n",
      "(116293, 236)\n",
      "(10506,)\n"
     ]
    }
   ],
   "source": [
    "print(X.shape)\n",
    "print(X_holdout.shape)\n",
    "print(y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Will have to balance our classes!!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X.to_csv('./X_train_clean_v2.csv')\n",
    "X_holdout.to_csv('./X_test_clean_v2.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "# Modeling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Imports for Modeling\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.metrics import roc_curve\n",
    "from sklearn.metrics import auc\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "from operator import itemgetter\n",
    "\n",
    "from imblearn.over_sampling import RandomOverSampler\n",
    "from imblearn.over_sampling import SMOTE\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Baseline Accuracy (Although, we're optimizing for Recall, so this isn't really useful)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Counter({0: 9955, 1: 551})"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Counter(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9475537787930707"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "baseline = 9955/(551+9955)\n",
    "baseline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Final EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# ID and Day_x is erroneous, Year_x is useless since there's no overlap between train and test set, Month_X will have\n",
    "# strong collinearity with Day_of_Year\n",
    "drop_cols = ['Id', 'Year_x', 'Month_x', 'Day_x']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X.drop(drop_cols, axis=1, inplace=True)\n",
    "X_holdout.drop(drop_cols, axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10506, 232)\n",
      "(116293, 232)\n",
      "(10506,)\n"
     ]
    }
   ],
   "source": [
    "print(X.shape)\n",
    "print(X_holdout.shape)\n",
    "print(y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10506, 233)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Latitude</th>\n",
       "      <th>Longitude</th>\n",
       "      <th>DayofYear</th>\n",
       "      <th>culex</th>\n",
       "      <th>erraticus</th>\n",
       "      <th>pipiens</th>\n",
       "      <th>restuans</th>\n",
       "      <th>salinarius</th>\n",
       "      <th>tarsalis</th>\n",
       "      <th>territans</th>\n",
       "      <th>...</th>\n",
       "      <th>MIFG</th>\n",
       "      <th>RA</th>\n",
       "      <th>SN</th>\n",
       "      <th>SQ</th>\n",
       "      <th>TS</th>\n",
       "      <th>TSRA</th>\n",
       "      <th>VCFG</th>\n",
       "      <th>VCTS</th>\n",
       "      <th>Hours_in_Day</th>\n",
       "      <th>WnvPresent</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>41.954690</td>\n",
       "      <td>-87.800991</td>\n",
       "      <td>149</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>15.633333</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>41.954690</td>\n",
       "      <td>-87.800991</td>\n",
       "      <td>149</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>15.633333</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>41.994991</td>\n",
       "      <td>-87.769279</td>\n",
       "      <td>149</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>15.633333</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>41.974089</td>\n",
       "      <td>-87.824812</td>\n",
       "      <td>149</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>15.633333</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>41.974089</td>\n",
       "      <td>-87.824812</td>\n",
       "      <td>149</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>15.633333</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 233 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    Latitude  Longitude  DayofYear  culex  erraticus  pipiens  restuans  \\\n",
       "0  41.954690 -87.800991        149      1          0        1         1   \n",
       "1  41.954690 -87.800991        149      1          0        0         1   \n",
       "2  41.994991 -87.769279        149      1          0        0         1   \n",
       "3  41.974089 -87.824812        149      1          0        1         1   \n",
       "4  41.974089 -87.824812        149      1          0        0         1   \n",
       "\n",
       "   salinarius  tarsalis  territans     ...      MIFG   RA   SN   SQ   TS  \\\n",
       "0           0         0          0     ...       0.0  0.0  0.0  0.0  0.0   \n",
       "1           0         0          0     ...       0.0  0.0  0.0  0.0  0.0   \n",
       "2           0         0          0     ...       0.0  0.0  0.0  0.0  0.0   \n",
       "3           0         0          0     ...       0.0  0.0  0.0  0.0  0.0   \n",
       "4           0         0          0     ...       0.0  0.0  0.0  0.0  0.0   \n",
       "\n",
       "   TSRA  VCFG  VCTS  Hours_in_Day  WnvPresent  \n",
       "0   0.0   0.0   0.0     15.633333           0  \n",
       "1   0.0   0.0   0.0     15.633333           0  \n",
       "2   0.0   0.0   0.0     15.633333           0  \n",
       "3   0.0   0.0   0.0     15.633333           0  \n",
       "4   0.0   0.0   0.0     15.633333           0  \n",
       "\n",
       "[5 rows x 233 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Join X and y momentarily so we can drop some null observations\n",
    "df = X.join(pd.DataFrame(y))\n",
    "print(df.shape)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Split X and y back out after dropping null observations\n",
    "df.dropna(inplace=True)\n",
    "X = df.drop('WnvPresent', axis=1)\n",
    "y = df['WnvPresent']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10413, 232)\n",
      "(10413,)\n",
      "(116293, 232)\n"
     ]
    }
   ],
   "source": [
    "print(X.shape)\n",
    "print(y.shape)\n",
    "print(X_holdout.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Play with Subsetting Features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#features = ['Latitude', 'Longitude','pipiens', 'restuans', 'salinarius',\n",
    "#            'territans', 'Tmax', 'Tmin', 'Tavg', 'Depart', 'DewPoint', 'WetBulb', 'Heat', 'Cool',\n",
    "#            'PrecipTotal', 'StnPressure', 'SeaLevel', 'ResultSpeed', 'ResultDir', 'AvgSpeed', 'BR', 'DZ',\n",
    "#            'FG', 'HZ', 'RA', 'TS', 'TSRA']\n",
    "\n",
    "#features = [item[0] for item in sorted_importance][0:20]\n",
    "#X = X[features]\n",
    "#X_holdout = X_holdout[features]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test/Train Split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42, stratify=y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "ss = StandardScaler()\n",
    "Xs_train = ss.fit_transform(X_train)\n",
    "Xs_test = ss.transform(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Balance Classes by: Oversampling (SMOTE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Oversampling on training data only\n",
    "sm = SMOTE(random_state=42, ratio = 'minority')\n",
    "X_res, y_res = sm.fit_sample(Xs_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(13806, 232)\n",
      "(13806,)\n"
     ]
    }
   ],
   "source": [
    "print(X_res.shape)\n",
    "print(y_res.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Test/Train Split again on Training Data, and Oversample on Training_2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# An additional test/train split, since we can't evaluate cross-val on original X_test if modeling was trained on\n",
    "# a corresponing X_train of artificially inflated data\n",
    "X_train_2, X_test_2, y_train_2, y_test_2 = train_test_split(X_res, y_res, test_size=0.3, random_state=42, stratify=y_res)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(9664, 232)\n",
      "(9664,)\n",
      "(4142, 232)\n",
      "(4142,)\n"
     ]
    }
   ],
   "source": [
    "print(X_train_2.shape)\n",
    "print(y_train_2.shape)\n",
    "print(X_test_2.shape)\n",
    "print(y_test_2.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=10, n_jobs=1,\n",
       "            oob_score=False, random_state=42, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf = RandomForestClassifier(random_state=42)\n",
    "rf.fit(X_train_2, y_train_2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.96      0.97      0.96      2959\n",
      "          1       0.25      0.19      0.22       165\n",
      "\n",
      "avg / total       0.92      0.93      0.92      3124\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Pred -</th>\n",
       "      <th>Pred +</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Act -</th>\n",
       "      <td>2862</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Act +</th>\n",
       "      <td>133</td>\n",
       "      <td>32</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Pred -  Pred +\n",
       "Act -    2862      97\n",
       "Act +     133      32"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_preds = rf.predict(Xs_test)\n",
    "print(classification_report(y_test, y_preds))\n",
    "pd.DataFrame(confusion_matrix(y_test, y_preds), columns=['Pred -', 'Pred +'], index=['Act -', 'Act +'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Sunset', 0.08381325068431573),\n",
       " ('DayofYear', 0.08155804816546172),\n",
       " ('pipiens', 0.07532999447955298),\n",
       " ('Hours_in_Day', 0.05566806830980986),\n",
       " ('restuans', 0.0471256988115619),\n",
       " ('Sunrise', 0.04055205919003505),\n",
       " ('Depart', 0.03702912600022774),\n",
       " ('Longitude', 0.03547963500910475),\n",
       " ('WetBulb', 0.035472487335804215),\n",
       " ('Latitude', 0.03422500463165259),\n",
       " ('Tmin', 0.03415991181837901),\n",
       " ('ResultSpeed', 0.03004490807520786),\n",
       " ('Cool', 0.02904194702257588),\n",
       " ('StnPressure', 0.02608463253769263),\n",
       " ('Tavg', 0.02403075502059367),\n",
       " ('SeaLevel', 0.021692164576107297),\n",
       " ('ResultDir', 0.02165632691282622),\n",
       " ('t900', 0.020509022716558703),\n",
       " ('Tmax', 0.02015355939148219),\n",
       " ('DewPoint', 0.019808134126902),\n",
       " ('AvgSpeed', 0.019739932115690948),\n",
       " ('PrecipTotal', 0.0154439931964862),\n",
       " ('t115', 0.01444888235484674),\n",
       " ('BR', 0.01406683974629368),\n",
       " ('RA', 0.013972531937608163),\n",
       " ('60666', 0.011079090555423252),\n",
       " ('TSRA', 0.009980876543461564),\n",
       " ('HZ', 0.0060296807750488575),\n",
       " ('TS', 0.004506523634741648),\n",
       " ('60617', 0.004003968237748268),\n",
       " ('t002', 0.0033639182127360853),\n",
       " ('60638', 0.0027203556458303014),\n",
       " ('60609', 0.002609482972302373),\n",
       " ('60634', 0.002552021558749438),\n",
       " ('60651', 0.0024047418223189323),\n",
       " ('t138', 0.0021771500791982175),\n",
       " ('60646', 0.0021403476821961623),\n",
       " ('territans', 0.002102375496183222),\n",
       " ('60655', 0.002090882123558145),\n",
       " ('10', 0.002070453431812477),\n",
       " ('t013', 0.0020696295965847906),\n",
       " ('t228', 0.002049121848394565),\n",
       " ('60639', 0.0020164370941902717),\n",
       " ('Heat', 0.0019222316441387618),\n",
       " ('t065', 0.0018972846349875829),\n",
       " ('t225', 0.0018921570849084208),\n",
       " ('60652', 0.0018236277950165712),\n",
       " ('t008', 0.0016960489718737218),\n",
       " ('t003', 0.0016215296549753593),\n",
       " ('t230', 0.0015431855863698028),\n",
       " ('t143', 0.001542522735494589),\n",
       " ('t009', 0.0015226661238231872),\n",
       " ('t235', 0.001451716428220225),\n",
       " ('t212', 0.0014220439738608438),\n",
       " ('t155', 0.0013931432136272668),\n",
       " ('60632', 0.001277395655785602),\n",
       " ('t158', 0.0012683886863478577),\n",
       " ('60106', 0.00126231976820926),\n",
       " ('60618', 0.0011831405450531502),\n",
       " ('60633', 0.0011454308654982623),\n",
       " ('60622', 0.0011113794526239381),\n",
       " ('t223', 0.0011077102260594475),\n",
       " ('60631', 0.001105358544761732),\n",
       " ('t147', 0.001077187144656008),\n",
       " ('VCTS', 0.0010755174317629668),\n",
       " ('60643', 0.0010106401670293666),\n",
       " ('60608', 0.0009881506564523063),\n",
       " ('60656', 0.0009142424639541974),\n",
       " ('t086', 0.0009091548277298815),\n",
       " ('t070', 0.0008861229813453695),\n",
       " ('60623', 0.0008850004873895863),\n",
       " ('60707', 0.0008738820537695417),\n",
       " ('60612', 0.0008734377274677452),\n",
       " ('t082', 0.0008656097212553369),\n",
       " ('t027', 0.0008332482312411198),\n",
       " ('t011', 0.0008304108029399306),\n",
       " ('t226', 0.000812995885123378),\n",
       " ('t031', 0.0007979581253681293),\n",
       " ('t094', 0.0007836975531440313),\n",
       " ('t030', 0.0007774372910430709),\n",
       " ('t095', 0.0007703891429275821),\n",
       " ('60659', 0.0007650477555200062),\n",
       " ('t096', 0.000742535737689002),\n",
       " ('t061', 0.0007362546093589953),\n",
       " ('t233', 0.0007310763805210047),\n",
       " ('t039', 0.0007135919306198472),\n",
       " ('t081', 0.0007006138047811976),\n",
       " ('t215', 0.0006922963130470818),\n",
       " ('t073', 0.0006895877805365925),\n",
       " ('60621', 0.0006752224393874872),\n",
       " ('60803', 0.0006721275654316518),\n",
       " ('t220', 0.0006708366280196475),\n",
       " ('60637', 0.0006682969526440293),\n",
       " ('60630', 0.0006515162987804134),\n",
       " ('t903', 0.0006347293506484781),\n",
       " ('60614', 0.0006330630547508251),\n",
       " ('t048', 0.0006307907410184387),\n",
       " ('t151', 0.0006229062114944722),\n",
       " ('t227', 0.000613430709155062),\n",
       " ('t066', 0.0006084696324569973),\n",
       " ('t054', 0.0005980164063606006),\n",
       " ('t128', 0.0005797278334392726),\n",
       " ('t159', 0.0005794360715458362),\n",
       " ('t033', 0.0005718707224991879),\n",
       " ('60629', 0.0005638566179120633),\n",
       " ('60654', 0.000563757595877877),\n",
       " ('t200', 0.0005548184751342914),\n",
       " ('DZ', 0.0005515951080882159),\n",
       " ('t221', 0.0005326700544229414),\n",
       " ('t114', 0.0005208048333528004),\n",
       " ('t232', 0.0005099265858099497),\n",
       " ('t085', 0.0005008194284751077),\n",
       " ('t145', 0.0004953347704220581),\n",
       " ('60620', 0.00048785209650148225),\n",
       " ('t135', 0.00048508533486029456),\n",
       " ('t015', 0.0004678747435162463),\n",
       " ('t074', 0.0004654976105112387),\n",
       " ('t102', 0.0004646501901288777),\n",
       " ('t014', 0.0004596372027453026),\n",
       " ('t028', 0.00043680809561213644),\n",
       " ('t054c', 0.00042620280334496166),\n",
       " ('t152', 0.00042597750236327556),\n",
       " ('60625', 0.00042570624332882083),\n",
       " ('t209', 0.0004244512595278951),\n",
       " ('t103', 0.00041834835100741294),\n",
       " ('salinarius', 0.0004005394693467086),\n",
       " ('t156', 0.00038251797455136644),\n",
       " ('t107', 0.00038160038970092296),\n",
       " ('t160', 0.00037491085015098076),\n",
       " ('t047', 0.0003734818686394602),\n",
       " ('t035', 0.00036462762207953824),\n",
       " ('60626', 0.00035183724168988216),\n",
       " ('t153', 0.00033764671913699663),\n",
       " ('t012', 0.0003318221718443186),\n",
       " ('t144', 0.0003259137528853931),\n",
       " ('60641', 0.0003029022679352697),\n",
       " ('t062', 0.00029692817898110084),\n",
       " ('t222', 0.0002803600370392),\n",
       " ('t236', 0.00026993670677148864),\n",
       " ('60607', 0.0002696341703345473),\n",
       " ('t080', 0.00026701246718028855),\n",
       " ('t089', 0.0002656870352668904),\n",
       " ('t083', 0.00026391892985027195),\n",
       " ('t162', 0.00025850604762687774),\n",
       " ('t077', 0.0002563582588353265),\n",
       " ('t224', 0.0002475468288833522),\n",
       " ('60018', 0.0002473967552951594),\n",
       " ('t063', 0.00023654622036326607),\n",
       " ('t097', 0.0002317275514255684),\n",
       " ('t231', 0.0002072278301523438),\n",
       " ('60628', 0.00020114100981463264),\n",
       " ('60613', 0.00019440389672691988),\n",
       " ('t016', 0.00018927700586636497),\n",
       " ('60619', 0.00018431893432580845),\n",
       " ('t218', 0.0001812612457808609),\n",
       " ('t079', 0.00018036756376821962),\n",
       " ('t043', 0.00017865678448761394),\n",
       " ('60640', 0.00017243514994811867),\n",
       " ('t091', 0.00016986979932586855),\n",
       " ('t067', 0.0001584997514458263),\n",
       " ('60660', 0.00015646713001020627),\n",
       " ('t090', 0.00015452587382972402),\n",
       " ('60827', 0.00015298761038125267),\n",
       " ('t099', 0.00014698904033723362),\n",
       " ('t161', 0.00013912557847570606),\n",
       " ('t219', 0.00013424824903832908),\n",
       " ('t025', 0.00012288464962787666),\n",
       " ('t092', 0.0001207258721738355),\n",
       " ('FG', 0.00011720068671295201),\n",
       " ('60605', 0.00010435001924551218),\n",
       " ('t060', 0.0001039343101787554),\n",
       " ('t206', 9.998349310670463e-05),\n",
       " ('t017', 9.641517739746696e-05),\n",
       " ('t142', 9.130879084350005e-05),\n",
       " ('t084', 9.130688496201272e-05),\n",
       " ('t100', 8.995986095627884e-05),\n",
       " ('t069', 8.286653511444087e-05),\n",
       " ('t045', 8.250244808944726e-05),\n",
       " ('t149', 7.77715697955562e-05),\n",
       " ('60642', 7.207442501194946e-05),\n",
       " ('60636', 6.598758699225538e-05),\n",
       " ('t049', 6.358548982584401e-05),\n",
       " ('t019', 6.157381391266613e-05),\n",
       " ('t141', 6.130708571464872e-05),\n",
       " ('t071', 5.7076557816986904e-05),\n",
       " ('60649', 5.5145312213488706e-05),\n",
       " ('t154', 5.259273585986351e-05),\n",
       " ('t129', 5.249744963843884e-05),\n",
       " ('t037', 5.0890792308837004e-05),\n",
       " ('t046', 5.05409554170402e-05),\n",
       " ('60645', 4.2068080158318315e-05),\n",
       " ('t148', 3.39228157754571e-05),\n",
       " ('t005', 2.7853320681611232e-05),\n",
       " ('t150', 2.7318767687049704e-05),\n",
       " ('60615', 2.2106535222166298e-05),\n",
       " ('t075', 1.806278622073564e-05),\n",
       " ('t146', 1.7366221179382636e-05),\n",
       " ('t036', 1.5151318568366452e-05),\n",
       " ('t004', 1.4963678183381783e-05),\n",
       " ('60611', 8.7691954484636e-06),\n",
       " ('t078', 7.907047396692139e-06),\n",
       " ('t018', 6.603983867468559e-06),\n",
       " ('t072', 5.336111809146273e-06),\n",
       " ('t094b', 4.652806815716495e-06),\n",
       " ('t088', 3.816815221482531e-06),\n",
       " ('60647', 2.235037629928625e-06),\n",
       " ('t051', 1.339639073736012e-06),\n",
       " ('t157', 1.105592976262368e-06),\n",
       " ('t076', 1.015831457235355e-06),\n",
       " ('t044', 7.108248213146953e-07),\n",
       " ('t050', 3.4699189498469015e-07),\n",
       " ('t006', 1.972005332891435e-07),\n",
       " ('t001', 4.9328791479180164e-08),\n",
       " ('culex', 0.0),\n",
       " ('erraticus', 0.0),\n",
       " ('tarsalis', 0.0),\n",
       " ('t007', 0.0),\n",
       " ('t034', 0.0),\n",
       " ('t040', 0.0),\n",
       " ('t229', 0.0),\n",
       " ('t237', 0.0),\n",
       " ('t238', 0.0),\n",
       " ('Station', 0.0),\n",
       " ('SnowFall', 0.0),\n",
       " ('BCFG', 0.0),\n",
       " ('FG+', 0.0),\n",
       " ('FU', 0.0),\n",
       " ('GR', 0.0),\n",
       " ('MIFG', 0.0),\n",
       " ('SN', 0.0),\n",
       " ('SQ', 0.0),\n",
       " ('VCFG', 0.0)]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "zipped = list(zip(X_train.columns, rf.feature_importances_))\n",
    "sorted_importance = sorted(zipped, key=itemgetter(1), reverse=True)\n",
    "sorted_importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Feature</th>\n",
       "      <th>Importance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Sunset</td>\n",
       "      <td>0.083813</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DayofYear</td>\n",
       "      <td>0.081558</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>pipiens</td>\n",
       "      <td>0.075330</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Hours_in_Day</td>\n",
       "      <td>0.055668</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>restuans</td>\n",
       "      <td>0.047126</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Feature  Importance\n",
       "0        Sunset    0.083813\n",
       "1     DayofYear    0.081558\n",
       "2       pipiens    0.075330\n",
       "3  Hours_in_Day    0.055668\n",
       "4      restuans    0.047126"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(sorted_importance[0:5], columns=['Feature', 'Importance'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.730761825760136"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "proba_pairs = rf.predict_proba(Xs_test)\n",
    "probas = [item[1] for item in proba_pairs]\n",
    "roc_auc_score(y_test, probas)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Forest with Gridsearch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import make_scorer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def custom_recall(y_true, y_pred, labels=None, pos_label=1, average='binary', sample_weight=None):\n",
    "    TP = 0\n",
    "    FN = 0\n",
    "    FP = 0\n",
    "    TN = 0\n",
    "    y_pred = list(y_pred)\n",
    "    y_true = list(y_true)\n",
    "    \n",
    "    for i in range(len(y_true)):\n",
    "        if y_true[i] == 1 and y_pred[i] == 1:\n",
    "            TP += 1\n",
    "        if y_true[i] == 1 and y_pred[i] == 0:\n",
    "            FN += 1\n",
    "        if y_true[i] == 0 and y_pred[i] == 1:\n",
    "            FP += 1\n",
    "        if y_pred[i] == 0 and y_pred[i] == 0:\n",
    "            TN += 1\n",
    "    return(TP / (TP + FN))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "my_scorer = make_scorer(custom_recall)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score='raise',\n",
       "       estimator=RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=10, n_jobs=1,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False),\n",
       "       fit_params=None, iid=True, n_jobs=1,\n",
       "       param_grid={'n_estimators': [2, 5, 10, 15, 25], 'criterion': ['gini', 'entropy'], 'max_depth': [2, 5, 10, 20, None], 'min_samples_split': [2, 5, 10], 'random_state': [42]},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring=make_scorer(custom_recall), verbose=0)"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rt_params = {'n_estimators':[2,5,10,15,25],\n",
    "            'criterion':['gini','entropy'],\n",
    "            'max_depth':[2,5,10,20,None],\n",
    "            'min_samples_split':[2,5,10],\n",
    "            'random_state':[42]}\n",
    "\n",
    "rt_grid = GridSearchCV(RandomForestClassifier(), rt_params, scoring=my_scorer, cv=5)\n",
    "rt_grid.fit(X_train_2, y_train_2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.97      0.84      0.90      2959\n",
      "          1       0.17      0.59      0.27       165\n",
      "\n",
      "avg / total       0.93      0.83      0.87      3124\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Pred -</th>\n",
       "      <th>Pred +</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Act -</th>\n",
       "      <td>2498</td>\n",
       "      <td>461</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Act +</th>\n",
       "      <td>68</td>\n",
       "      <td>97</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Pred -  Pred +\n",
       "Act -    2498     461\n",
       "Act +      68      97"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_preds = rt_grid.predict(Xs_test)\n",
    "print(classification_report(y_test, y_preds))\n",
    "pd.DataFrame(confusion_matrix(y_test, y_preds), columns=['Pred -', 'Pred +'], index=['Act -', 'Act +'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'criterion': 'entropy',\n",
       " 'max_depth': 10,\n",
       " 'min_samples_split': 5,\n",
       " 'n_estimators': 15,\n",
       " 'random_state': 42}"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rt_grid.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8077483179206735"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "proba_pairs = rt_grid.predict_proba(Xs_test)\n",
    "probas = [item[1] for item in proba_pairs]\n",
    "roc_auc_score(y_test, probas)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random Trees with Random Search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomizedSearchCV(cv=5, error_score='raise',\n",
       "          estimator=RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=10, n_jobs=1,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False),\n",
       "          fit_params=None, iid=True, n_iter=50, n_jobs=1,\n",
       "          param_distributions={'n_estimators': [2, 5, 10, 15, 25], 'criterion': ['gini', 'entropy'], 'max_depth': [2, 5, 10, 20, None], 'min_samples_split': [2, 5, 10], 'random_state': [42]},\n",
       "          pre_dispatch='2*n_jobs', random_state=42, refit=True,\n",
       "          return_train_score='warn', scoring=make_scorer(custom_recall),\n",
       "          verbose=0)"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rt_rand = RandomizedSearchCV(RandomForestClassifier(), rt_params, scoring=my_scorer, cv=5, n_iter=50, random_state=42)\n",
    "rt_rand.fit(X_train_2, y_train_2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.97      0.87      0.92      2959\n",
      "          1       0.19      0.52      0.27       165\n",
      "\n",
      "avg / total       0.93      0.85      0.88      3124\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Pred -</th>\n",
       "      <th>Pred +</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Act -</th>\n",
       "      <td>2582</td>\n",
       "      <td>377</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Act +</th>\n",
       "      <td>79</td>\n",
       "      <td>86</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Pred -  Pred +\n",
       "Act -    2582     377\n",
       "Act +      79      86"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_preds = rt_rand.predict(Xs_test)\n",
    "print(classification_report(y_test, y_preds))\n",
    "pd.DataFrame(confusion_matrix(y_test, y_preds), columns=['Pred -', 'Pred +'], index=['Act -', 'Act +'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'criterion': 'entropy',\n",
       " 'max_depth': 10,\n",
       " 'min_samples_split': 2,\n",
       " 'n_estimators': 25,\n",
       " 'random_state': 42}"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rt_rand.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.8089649451596055"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "proba_pairs = rt_rand.predict_proba(Xs_test)\n",
    "probas = [item[1] for item in proba_pairs]\n",
    "roc_auc_score(y_test, probas)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## AUC-ROC Curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA00AAAJnCAYAAABLQ7jIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJzs3XeYVNX5wPHvAQQEFUVEETUaKzF2\nsKPYUOyKGrso1tiixtgVuzFRE03sYsHYfvbEFhv2BmIlFhRRFBQsKNKX8/vjzMqwzC6zsLN3dvb7\neZ557s69d+68U3fee855T4gxIkmSJEkqrEXWAUiSJElSOTNpkiRJkqQ6mDRJkiRJUh1MmiRJkiSp\nDiZNkiRJklQHkyZJkiRJqoNJk6RmK4QwIIQQQwi9so5F866SXscQwvK5x3JL1rFUkhDCLbnndfkS\nHf+zEMJnpTi2pPJg0iRpnuR+gORfqkII34UQBocQ+oUQQtYxVrIQwiohhH+GED4IIUwMIfwcQvgw\nhHB1CGHVrONrSLn3Uwwh9Ms6lnkRQlgthHBVCOG9EMKEEMK0EMJXIYRHQgj9Qwhts46xMYUQeuVe\nzwFZx1Ks3PeaE1tKzVirrAOQ1OSdm1suAKwE7AZsDnQHjskqqCL9A7gL+DzrQOojhHAccDnpxNfz\nwH+ACKwHHAkcHkI4McZ4ZXZRNqqyfR1DCGcD55Beq1eBW4GJwJJAL+BG4CjS50WlcxpwCfBliY6/\nVYmOK6lMmDRJmi8xxgH510MIm5B+yP8+hHBZjHFkJoEVIcY4HhifdRz1EUI4EPg78B2wW4zx+Rrb\newIPAn8PIfwQY7wtgzAbVbm+jiGE00knFb4A9owxvlZgnx2Bkxo7tuYmxjgGGFPC439SqmNLKg92\nz5PUoGKMLwEfAIHU8jGHEMK2IYRHQwjjQwhTQwifhBD+EkJYtJb9lwkhXBlC+DiEMCXXDfD1EMJZ\ntez7jxDCp7ljfxtCeDiE0KPAvrONhQkhdM11M3yztscXQng8d5vf1li/QQjh3hDC2Fz3qy9CCNeF\nEJYucIzBuWO0DiGcnetWN3Vu41hCCAsDf8td3bdmwgQQY3wB2C939Yrcbapv/0u3qBDCRiGEp3Ld\nxX4KITwRQijY2hFCaBVC+H0I4dUQwo8hhEkhhGEhhGNCCC1q7PvLmJxcF8K7QwjfhBBm5j3P64UQ\n/h5CeDv3Wk7JvbaXhRAWq/lcATfnrt5co0vo8rl9Co5pyq0bHELoFEK4PoQwJvc8vx9COLiWx9om\nd7zq98/IEMIFufUxF89c5WIbAEwHti+UMAHEGP8DbFfbMUIId+U+J1NCCENySVbN/TqEEE4OITwT\nQhide/+Ny73vN6zl2NXPzVIhhBtDCF/m3vv9cttXCSFckrvPcbnnYlTueVymjsfdO4Tw79xrPjX3\nOXgohLB1bvstwLO53c+p8Xr2qnGsfUIIz4YQvs89/v+FEM4MIbSZh8dTcExTCGHnEMLTee+Nr0II\nz4UQfp/3GkRS63nNbsmD845T65imEMLvcvdR/V7/LIRwZ6jl8yapPNnSJKkUqsczTZ9jQ+qudC6p\npeQ/wDfAmsAfge1DCBvFGH/M27878ATQkdSCdT/QDvgN6Ufp+Xn7rgv8N7fvE7l9OwG7Ai+GEHaL\nMT5aW9Axxi9DCE8BvUMIa8QY360Rexdga2BojPG9vPUHAzcAU4GHSS0LKwOHAjuFEDaMMRbqOnYf\n0AN4jNQ69E1tseXsASwGvB5jfKKOx/F4COGN3LH3YFbSUW0DUnelp4B/krpV7g5sFkLonUu8qh/b\nAsC/gW2BD4E7gCnAFsBVuWMdUCCMFYHXgI+AfwELAtWv62GkbpzP5WJoCawLnAj0CSFsEGP8Kbfv\nLcAPwC7AQ8BbeffxQ23PQZ5FgZeAacC9QNvcczIwhDAzxnhr3mMNpNdkB+BjUre/BYB+wOpF3Fe+\ng3O3vSv/vVJIjHFqgdW/Al4HPgUGkd7TvwMeCiFsHWN8Nm/fbsCFpM/HI8D3wHLAzqTnc6cY4+MF\n7qMjqcvgRNJnZSbwdW7b7qSuns8CL5Oev9WZ9Z7uHmOcratbCOFc4Ozc8R4kfQ6WBjYG9ie91g/m\ndj+I9PoPzjvEZ3nHugk4BBidi+0HYEPS532rEMI2McYZ9Xg8cwghHA5cB4wlvcfHA51J30cHA1fn\n7vdc0nvgV8zqjjxbvLUcP5A+ewfljn0/MA5YhvT5+RAYUtcxJJWRGKMXL1681PtCGkMTC6zfDKgi\nJRBdamzbIne7l4FFa2zrl9t2Rd661sDI3Pp9C9zXsnl/twJGkH7Qb15jv6VJYxnGAG3y1g/IHbtX\n3rp9cuv+WuD+Ts5tOzZv3SqkH5QjgK419t8y91w8UGP94Nxx3gE61eM5vyl3uwuL2PfC3L435q3r\nVf26AcfU2H+X3PqPgRYFnqOrgJZ561vmxbNL3vrl8+7jolpi+1X+sfLW98/d7pRa3hv9ajneHK9j\n/nuUNG4oP/bfADOA4TX2PyC3//NA67z1i5JaTyMwuMjX6unc/ofW83OV//ydU2Pbtrn1j9ZY36HQ\n+4j04/wr4H8FtlXfx21AqwLbu5L3Wclb3zv3nr6mwPpISvK6FoqlwPtwQC3PQfXrfT+wYC2v9fH1\nfDy35LYvn7duKOl7qnOB/TvVuD6YAt93eds/Az6rse7w3H2+DnSosa0lNb4fvXjxUt4Xu+dJmi+5\nrkwDQggXhhDuJp1NDsAfYxpHkO+43PKwGONsrQQxxltIrQj75a3eifQj8uEY4x017zvG+EXe1R1I\nrRtXxRifq7HfV8ClwFLMfcD2g8AEYL8QQssa2w4itZ7dmbfuKFKLwvGxxpn3GOMzpJannUJeN7k8\nZ8U0HqdYXXLLL+rca/Z95ugeSErwrs5fEWN8iHTmfyWgJ0BIXe+OIZ2JPyHGWJW3fxVpLE5k9tes\n2tfMflY+/75G5R8rz0BSa9S2tT6q+psEnFgj9uGk1qduNV6Xg3LLM2OM0/L2/4G8Fs0iVb9Wo+sf\nMgCjgAvyV8TUuvg5sH6N9RMKvY9ijKNJrWurhRCWK3Af00if05otNsQYv4wFWsBijP8F3mfO1+jY\n3PKkmp+DvFiKdTwpqT0kxji5xrbzgW8p/J6r9fHUYQYFWsTr+bmsTfVzckSMcUKN41cV+H6UVMbs\nnidpfp1T43oE+scYa3YJA9iI9ANlzxDCngW2twaWCCEsHmP8ltQdB1L3tbnZKLf8VShcynjl3LIb\nUFcXvckhhHtIXci2rd43hLAeqXvSAzV+UFXf7+ahwLgpUneflqQWqaE1tr1e66MprLrbYzGlj+va\n94UY48wC6weTxm6sQ0qgVgEWJ7U+nRkKV5GfTHpOa3q70I9u+KXL3xHA3qRWnw7MPsa2a6HbzaOP\nY153zzzVSeWiQHVXwHVIXbpeLrD/i/W83/q8VoW8VUti+QWz3nOz7iwVYDk+t60z6bOUrytzVhf8\nLMZYsEtormvZfqRWn7VI3ULzTyJMq3GTDUmPtVA3wKKFENrl7m888Ida3nNTKfyeq/Xx1OJfwGXA\n+7kTPs8BL8UYx9Uv6jmFENoDvwW+jjEOm9/jScqeSZOk+RJjDPDLj4SNSF22rg0hjMq1tORbnPS9\nUzPRqmkh0tnk6sIQxZQJXjy3LJSM1Tz23NxCSpoOYlaCVd0KcWuNfavv9+R5uN+xRcSSr/rMdKFW\ng5qqB+sXOptd2ziP6ng65JbVj21l6n7N6vvY7iaNafqUNE5pLOmHMMAfgDkG+s+H2sY9VbdG5CcC\nHYDvammpqHVsTC2+AlZj1utQX3XFXbP4xm6kFqUpwJPAJ8DPpASwFykRLvSc1vUaXU56LcaQxgd+\nSUqQYdb4nnyLAt8XaBmqr8VICecSzP17oqZ6fZ5ijJeHEMYDvye1gv8BiCGE54CTY4zzM96oPt9d\nkpoAkyZJDSLG+DPwVAhhJ+BN4NYQwqoxxkl5u00gjZfpWORhq384FtPyUN39ZZcY48NFHr+gGOPL\nIYSPgV1Cquj3M2ms03jmbKWqvt8OtbRo1HU/9W2FeJE0QH1r4Iy57Lt1bvlSgW1L1nKbpXLLCTWW\nD8QYdy82yJyCjy1X2GM3UjfO7WOM0/O2tQD+VM/7aUg/Ah1DCK0KJE61PWe1eZE0pm0r0omEUjqf\n1PLTPcb4v/wNIYTryFV+K6C216gzKYl4D9g4zirKUb19nwI3+wFYPISw4HwmTtXvuWExxnXredt6\nt+rFVJL/ttznfGPSe/MQ4IkQQrd6tlzlq893l6QmwDFNkhpUjPEdUiW5ZYATamx+FVgshFBsJbJX\nc8s+9di3Z5HHnptbSWfnf0caL9UJuCP/R36J7rcu95J+jK0fQtimtp1y29YnVVG7t8Aum4YapcJz\neuWW1d2JPsjd34a5LnUNYaXc8uECz+X6pCp7NVV3U6s5xqyhDSP9X9y4wLZN63msm0ldUfuGEH5T\n146FSmjX00qkohY1E6YW1D9ugF+Tnof/FkiYlsltr+lVUgtRwfLpNdT6esYYJ5LGTK0eQij25Mp8\nizH+EGN8NMZ4GKmluSOzf6arAAqMc6zteD+Tks4lQwjrNHC4kjJg0iSpFC4gdRX6Y5h93p0rcssb\nQuH5i9qH2eeV+TepKtXOhc5uhxDyz+I+ROqWdHQIYftCQYU0N1G7Ih/DbaTuTQfmLpB+TNX0D9KP\n4ytCCKsUuM/WIU04O99yLVnVE6HekRvHUvP+NiaVBQf4Q80fvTkrk7ok5d9uF1KLxAjghdz9zSBV\nzesCXBlCmCOhCSF0mVtSUMNnuWWvGsfpTCp/Xsi3uWUx3RLnR/VEwBeEEH4ZExRC6ADMMSdYXWKM\nn5EqvbUGHqltTp4QwnYUN2avLp8BK+d/pnJjks4hjRmbl+NBSq5/SRJCCAuRTogU6qVyVW55WY3P\nZfVt89fN7fW8nPS8DQwF5m4LISyWm15gvoQQtgshFHosnXPL/FbyeXkPXplbXpd7D+Xfd4vcFAaS\nmgi750lqcDHNd3QdaWD6n0hzAhFjfDqEcCpwMfBxCOFRUknxhUhjJDYndWvaLrf/tFzBiP+SkoQj\nSGe025IGgm9F7nssxjg9hLA7afzFIyGEl0nV+CYBy5LmLPo1KQHI/zFU22P4IoTwbO4+ZgDvFhrQ\nHWP8IIRwCKny2/shhMdJcxMtQPqB1ZM0N8tqxT+DdcZV/UPyUuCFkCbYHErqmrQeqaz7TFLCdFst\nh3mc9OO2D/A2s+ZpmkIq4pFfJOJ80sD8I0lVAJ8hjdPoTEq+NiF1FRxe5EN4g9RlcPfca/Qiqetb\nH9K8NV8VuM0rpNfsD7nWh+rxRVfVrEo2n24jFafYDngvhPAw6XXsS5pPZ1XSc1uUGONFuR/l5wBv\n5B7vENI8QkuSyvOvzPzP1XMFcC0wLIRwHymJ34SUMP2bVIWyaDHGsSGEu0jPxVshhP+SxnttQ3qP\nvAWsXeM2/w0hnE9KLv8XQqiep2lJUmvXq6SxUJBe5y+BvUMI00gFKiIwKFdZcWCu8MrvgU9CCNVV\nAzsCK5Cet5tJ78n5cRcwJYTwIilRDKTPaw/SZ+qpvH2fJo2XvD/3vTUZGBVjHFTH8W8kPfYDSd93\nD5G+C5Ymdd0cSEqsJTUFjVnf3IsXL5VzoZZ5mvK2L0kaC/QzsGSNbZsC95B+IE8j/ZB4i3SGuXuB\nYy1HKpE9Mrf/t6SJU88osG9n4BJS15hJpB+oH5O6qe1P3hwu1DK/T972/asfJ6mUcl3PxxqklqhR\npKIG3+ViuA7Yssa+g+t67op8/lcDriH9AJ2Uu3yUW7daLbfplXssA0hFO54ijeP5iZSY9qjldoE0\nh9HTucc1jfSj90XgdGafL2v53H3cUkfsHXOv52ekH+GfABeRJi3+jBrz3eRusx0peZqY95osX9fr\nSB3zKlFg3p7c+rbAebn32tRcPBeSxqZE4MF5eK26kVpi3ss939NIBRYeI81NlT93WJ3PX23vHVJC\n8hbp8zYeeCD3nqz3c5Pb3i73uKvnPvuC1BK4eF3vX2B7UlL+Xe75+yIXS83PQI/c+2kCKREtFOOO\nzJoAexqp0MPrpJbs1er5eOZ4vUlJ1wOkgiSTcjEPI53oWbjG7Vvm3qOfkpLS2e6vtvdtbtt+pMp8\nE3LP5UhS5b515+c7wIsXL417CTHOazVUSVJTEkLoBTwLnBtjHJBtNE1LbpzYf4FLYoynZR2PJKlx\nOaZJkqScWsbaLU5qvYTUMiFJamYadUxTCGEgqbn9mxjjbwtsD8DfSc37k4B+McY3GzNGSVKzdnkI\nYS3SBLfjSFUg+5C6FF4XY6zvhMSSpArQ2IUgbiFVmqptcHIf0qDYlYENSH3zN2iUyCRJgvtJ4/F2\nIk1QOoVUAnsgaWC/JKkZavQxTSGE5YH/1NLSdB1pYOWduesfkgaGFprRXpIkSZJKrtzGNHUlVdqp\nNhpn05YkSZKUoXKbpykUWFewKSyEcDhwOED79u3XW221BpkCRZIkSWUkRnj3XZgxA0KhX4pSHQKR\nznEsXeJXvAnjY4xLzMtxyi1pGk2ahLLaMhSe6JAY4/XA9QDdu3ePQ4bM79yAkiRJKjc33giHHQaP\nPw7bbpt1NGpSPvkEDjgAXvkK9t6bcNddo+b1UOWWND0MHJObiXwDYILjmSRJkpq2Dz6AMfP4i+6S\nS2C99aB374aNSRVu0CA46ihYYAG44w7YZx+46655Plxjlxy/kzQjfacQwmjgHGABgBjjtcCjpHLj\nI0glxw9uzPgkSZLUsKZOhbXWgmnT5v0YDzxg1zzVU+vWsOGGcPPNsOyyc99/Lhq9el4p2D1PkiSp\nPE2cCAsvDEceCXvvXf/bt28P3bs3fFyqQA89BOPGwaGHpusxzpZthxCGxhjn6d1Ubt3zJEmSVIFW\nXBE23zzrKFSRfvoJTjgBbroJNt4YDjkEWrRo0OZJkyZJkiTNoaoK1l0XPv54/o5T3ampRblNdKPK\n8OKLcOCBMGoUnHYaDBhQkjebSZMkSZLm8Pbb8M470Lcv/PrX83esVq1gr70aJi7pF599Br16wXLL\nwXPPwaabluyuTJokSZI0h+efT8u//Q2WWSbbWKTZjB8PnTrB8svDnXfCdtulgXMlZNIkSZJU4QYM\ngPffr99thg6FFVYwYVIZmTkTrrwSzjgjTdzVsyfsuWej3LVJkyRJUoU77zxYfHHo3Ln42yy4IPTv\nX7qYpHoZPRr69YOnn4Ydd4RVVmnUuzdpkiRJagaOOiolT1KTc/fdqWb99Olwww0pm2/kibtMmiRJ\nkpq4SZPgwQfTb8pCKmBaTjVno0dDt25w222w0kqZhGDSJEmS1MTdeCMcf3zd+9Sna56UuaeegilT\nUle8E05Ib/BW2aUuJk2SJElN3ODBqZDYM88U3t6yJSy7bGNGJM2jyZPh1FNTwYdNN4UddkjzLmU8\n0ZdJkyRJUpmIEb79tv63eeGF9NtyhRVKE5fUKIYOhf33hw8+gGOPhUsuafSxS7UxaZIkSSoTp50G\nf/7zvN12s80aNhapUQ0fDhtumPqRPvEE9O6ddUSzMWmSJEkqE6NHp9LgAwbU73Zt28K++5YkJKm0\nJk2Cdu1SoYfLL4f99oOOHbOOag4mTZIkSWVk0UXhmGOyjkIqsRjhppvg9NPh+edhtdVSl7wyZdIk\nSZKUkVNPTUUcqo0YkZImqaJ98w0cdhg8/DBsuSUstFDWEc2VSZMkSVIGJk2Cyy6DFVdMle8AuneH\nrbfONCyptP797zQ57Y8/whVXwHHHZV4ZrxgmTZIkSRl49VWYMSP9buzTJ+topEby7LPQtWtarr56\n1tEUzaRJkqRm5N134cEHs45CAC+/nKopb7xx1pFIJfbyy2mysA02gIsuSi1LrVtnHVW9mDRJktSM\nXHwx3Hln1lGo2rbbQocOWUchlci0aXDuuWm+pS22gKeeSqUemyCTJkmSmpGqKlh1VXj//awjETSJ\noRzSvBk+PE1UO2wYHHJI6ofahJk0SZLUzISQespIUkkMHQqbbpqq4j3wAOy6a9YRzTfPb0iSJEma\nf1VVabn22nDSSfDeexWRMIFJkyRJkqT5dddd0K0bjB2bmrIvuACWXDLrqBqMSZMkSZKkefP997Df\nfrDPPrD44jBlStYRlYRjmiRJaqLeeAP23DMVqCrWd9/Br39dupgkNSNPPw39+qXWpfPPh1NPhVaV\nmV5U5qOSJKkZuOyydJL3d7+r3+222KI08UhqZq6/Htq3h1dege7ds46mpEyaJElqgsaNg/vvh6OO\ngr//PetoJDUbw4alqngrrwzXXZcmqW3XLuuoSs6kSZKkMvHgg/DII8Xt++mnMH06HHFEaWOSJCBV\nxrv0UjjnHOjTBx56CBZdNOuoGo1JkyRJZWDqVDjssDSGepFFirvN/vvDb35T2rgkiU8/hQMPhJde\nSgMpr7km64ganUmTJEll4P77Yfx4eOIJ6N0762gkKeeVV9KXUsuWcPvtsO++aYbsZsakSZKkBjR6\nNAwdWv/b/f3vqard1ls3fEySVG8xpuRo7bVh773hrLNgueWyjiozJk2SJDWQMWNgnXVSi9G8uPRS\naOEMipKy9u9/wyWXpKbvhRaCG27IOqLMmTRJktQAZs5MXf5//jn9zlhiifrdvmVLWH310sQmSUWZ\nOBFOPDElSWutlc4ALbRQ1lGVBZMmSZJyYkwFoubFFVfAU0+lCryOSZLU5LzyChxwQCr6cMopcO65\n0KZN1lGVDZMmSZJyevSYt/FI1XbfPVXAk6QmJUY47bR01ui556Bnz6wjKjsmTZIk5Xz4IWy8cZqC\npL7at4f+/ZtlUSlJTdX//gcdO8KSS8K//gULL1z8nAfNjEmTJEl5NtwQzjwz6ygkqYRmzoR//hP+\n9Kc079Jtt0HXrllHVdZMmiRJzdb06al16Ouv0/VJk7KNR5JK7ssv4eCD4cknYfvtU9lOzZVJkySp\n2Ro+HAYNglVXhcUWS61M226bdVSSVCIvvAC77AJTp8K118Lhh9unuEgmTZKkZmvUqLS87TZYf/1s\nY5GkkuvWDTbbDP7yF1h55ayjaVKcQk+S1Gx9/nla/upX2cYhSSXzzDOwxx4wYwZ06gQPPmjCNA9M\nmiRJzdaoUdC2LXTunHUkktTApkxJE9VutRW89x6MGZN1RE2aSZMkqdkaNQqWW84u/ZIqzLBhsN56\nadbto4+GN9+EZZfNOqomzTFNkqSy9u23MGFCaY49YoRd8yRVmJkz4aCD4Pvv4bHHYLvtso6oIpg0\nSZLK1rhxsPTSqSt+qRx5ZOmOLUmNZuTI1Ne4fXu45x5YYglYfPGso6oYJk2SpLL1ww8pYTriCNh4\n44Y/fgjQu3fDH1eSGk2McPPNcPzxcOihqUveaqtlHVXFMWmSJJW9nj1hv/2yjkKSysy4cWmupQcf\nhF694IQTso6oYlkIQpJUdh57DH77W7viS1KtXnghfVE++ihcdhk8/XSqbKOSsKVJklR2XnwR3n8f\n9twTNtkknUCVJOVZeuk039I118Aaa2QdTcUzaZIklaVWrdJYZklSzquvwp13wt/+BiuumFqbnDOh\nUZg0SZIyN20aXHrprNLiL7yQbTySVFamT4fzzoOLLkrzLZ12Giy1lAlTIzJpkiRl7u234ayzoHXr\n1MIE0L17tjFJUln44APYf38YOjTNv3TllbDIIllH1eyYNEmSMjdzZlo++CD06ZNtLJJUNmbMSF+K\nP/0E994LfftmHVGzZdIkSfrF8OHwzTfZ3K8kKWfMmDQ5batWcMcdsPzy0KVL1lE1ayZNkiQAvv8+\nFWCqbvXJwsILZ3ffklQW7rkHjjwSTj45jV3aaKOsIxImTZKknEmTUsJ00kmw446Nf//t2zuOSVIz\n9sMPcOyxcPvtsP76dsUrMyZNkqTZrLqq8yJJUqN66SXYZx/46is491w4/fRZVXFUFnw1JKmZGTcO\n1l03dcfLV90tr0WLxo9Jkpq1Vq1S/+RXXoEePbKORgWYNElSM/PBBzB6NOy1V5ruI1/r1rDTTtnE\nJUnNyltvwZNPprFLG2wA77wDLVtmHZVqYdIkSc3MuHFpedppsPba2cYiSc1OVRX89a9pcrpOneDQ\nQ2GxxUyYypydMCSpmRk/Pi07dco2DklqdkaOTINGTz0Vdt45tS4ttljWUakItjRJUjNT3dJk0iRJ\njWjKFNh4Y/j5Z7j1VjjgAAgh66hUJJMmSWpmxo+HhRaCtm2zjkSSmoEJE2CRRdKX7vXXpwnxll8+\n66hUTyZNklTmvv8eHnssdYNvCEOGpInmJUkl9sgj0L8/XHIJ9OtnpZ0mzKRJksrcNdfAGWc07DF7\n927Y40mS8vz8M/zxj3DttbDmmrDeellHpPlk0iRJZW7q1LQcMaLhjtm1a8MdS5KU5/XXYb/94JNP\nUjnx88+HNm2yjkrzyaRJkhrQ9Onw008Ne8zJk9NyxRUb9riSpBIYOzb9M3j2Wdh886yjUQMxaZKk\nBrT55mlC94a2wAINf0xJUgP58MPUwnTAAamUeO/eVtupMCZNktSAPv8cNtwQ9tmnYY+78soNezxJ\nUgOIEa6+OnXD69ABdt8d2rc3YapAJk2S1MBWXx2OOy7rKCRJJfXVV3DIIfDEE7DddjBwYEqYVJFM\nmiSpFhdckKrF1sc335QmFklSGfnxR1h7bZg4MbU0HXmkE9VWOJMmSarFjTfCjBmp5ahYW24Je+xR\nupgkSRmaOjVVwltkEbj4YujZE1ZZJeuo1AhMmiSpFuPHw+GHw+WXZx2JJClzgwenCWqvvz4Veujf\nP+uI1IhaZB2AJJWjyZPT3IRLLJF1JJKkTE2Zkiaq3XJLaN0aFlss64iUAVuaJKmA8ePTslOnbOOQ\nJGXonXdg//3h3XfhqKPgL3+x2EMzZdIkSQWMG5eWtjRJUjP20kvpH8Kjj0KfPllHowyZNEkSafL2\n4cPTlBuQ5igEW5okqdn57LM0We2226aqeHvvbZc8mTRJEsB556US4zV17dr4sUiSMhAj3HYbHHss\nLLoojBjhGCb9wqRJkoDvv4eFFoJBg2atW3xxWGGF7GKSJDWS8ePhiCPg/vtTGfHbbksJk5Rj0iRJ\nOW3awK67Zh2FJKlRjR8Pa6wWiRJNAAAgAElEQVQB330Hl14KJ54ILVtmHZXKjEmTpIpSVQVbbAGf\nflq/233/PbRrV5qYJEllaOZMaNEiDV49+mjYeWdYc82so1KZMmmSVFGmTIEXXoAePer/v69Hj9LE\nJEkqM6+9BoceCrffDmutBWeemXVEKnMmTZIq0p57wsknZx2FJKmsTJ8OF16YKv907QqTJmUdkZoI\nkyZJkiRVvg8/hAMOgDfeSMurroIOHbKOSk2ESZMkSZIq3+23wyefwD33pO4IUj20yDoASZIkqSTG\njIEhQ9LfZ50F771nwqR5YkuTpCZv9GgYNiz9PWVKtrFIksrEfffB4Yen6njDh6d5l7p0yToqNVEm\nTZKavH32gRdfnH3dootmE4skKWMTJsBxx6UJanv0SLOWO++S5pNJk6QmbdIkePXVVDn2yCPTulat\n0jyFkqRmZvRo2GQT+PJLOOccOOMMWGCBrKNSBTBpktTkzJw56+9XXoEZM2C33WC99bKLSZJUBpZe\nGvr0gYMPhg02yDoaVRALQUhqUu68M7UktWyZLltvnSZ033jjrCOTJGXinXdg883hiy/SP4RrrzVh\nUoOzpUlSk/LxxxBj6nXRInfaZ7XVHMMkSc1OVRVccUXqgrfYYvD557DssllHpQpl0iSpSTr77FlJ\nkySpmRk1Cg46CJ57LvXPvu46WGKJrKNSBTNpktQkzJwJRxwBgwdnHYkkKXMXXwxvvgk335ySpxCy\njkgVzvO0kpqE0aPhxhtT8nTggf5/lKRmZ/x4GDEi/f3nP8Pbb0O/fv5DUKMwaZLUJIwbl5aXXw63\n3ur/SElqVh57LM0lsc8+aWBrhw6wwgpZR6VmxKRJUtn76adUNQ/SxO6SpGbi55/h6KNh++3TP4Ab\nb/SsmTLhmCZJZe++++Cyy9Lfv/51trFIkhrJZ5/BttumsqknnQQXXABt22YdlZopkyZJZW/69LT8\n8EPo0iXbWCRJjWTppdOcEtdeC1tskXU0aubsniepLI0Zk04yfvZZGvsL0L59lhFJkkruo4+gb1/4\n4Qdo3RoeesiESWWh0ZOmEMJ2IYQPQwgjQginFti+XAjh2RDCsBDCOyGE7Rs7RknZevLJdIJxhRXS\n5fTT0/rWrbONS5JUIjHCNdfAOuvAs8/C//6XdUTSbBq1e14IoSXwT2AbYDTwRgjh4Rjj8LzdzgTu\niTFeE0L4DfAosHxjxikpW998k5YXXwxLLZX+Xmop5y2UpIo0Zgz0758q5PXuDQMHQteuWUclzaax\nxzStD4yIMX4KEEK4C9gFyE+aIrBI7u8OwFeNGqGkstG3L6y8ctZRSJJK6vjjU+vSVVelSnlWx1MZ\nauzueV2BL/Kuj86tyzcA2D+EMJrUynRs44QmqVw89VTWEUiSSmrCBPj66/T3FVfAm2/CMceYMKls\nNXbSVOiTEGtc3we4Jca4DLA9MCiEMEecIYTDQwhDQghDxlXPeimpojhvoSRVoOefh7XWgn790vWu\nXaFbt0xDkuamsZOm0cCyedeXYc7ud/2BewBijK8AbYE5prOMMV4fY+weY+y+hAMdpIqz3HLQykkR\nJKlyTJ0Kf/oT9OqVvuDPPjvriKSiNfZPkjeAlUMIKwBfAnsD+9bY53NgK+CWEEI3UtJkU5JUgV5/\nPU1cW9OQIY0fiySphD79FHbbDd55B444Av76V1hooayjkorWqElTjHFGCOEY4AmgJTAwxvh+COE8\nYEiM8WHgJOCGEMIJpK57/WKMNbvwSWrivv0Wtt8+TcWxwAJzbt9228aPSZJUIh07Qps28J//wA47\nZB2NVG+N3vklxvgoqcBD/rqz8/4eDmzS2HFJalxnnJESpjffhDXXzDoaSVKDGzUKLrkE/vY3WHRR\neO01Cz2oyWr0yW0lacgQuP56OPZYEyZJqjgxwqBB6Qv+9ttTlzwwYVKTZtIkqVHNnAm//z107gwD\nBmQdjSSpQX37Ley1Fxx4IKyxRkqYevTIOippvlmbSlKjuukmeOONdBKyQ4eso5EkNaj99oNnnoGL\nL4aTT4aWLbOOSGoQJk2SGs2338Jpp0HPnun/qiSpAkyaBFVVsPDCcNllMG0arLNO1lFJDcrueZIa\nzMCBqfVokUUKX5ZZJhV/+Oc/7douSRXhjTdSgnTssen66qubMKki2dIkqcG89Vaau/D3v699n803\nT93cJUlN2IwZcNFFcN55sPTSaQyTVMFMmiQ1qHbt4PLLs45CklQyI0fCPvukEuL77Qf/+EcqKS5V\nMJMmSfNtxIh0svHVV7OORJJUcq1awTffwF13we9+l3U0UqNwTJOk+fbAA6kaHsDOO2cbiySpBMaO\nhXPPTXMwLbssfPSRCZOaFVuaJM238eOhTRv48EMLPEhSxbn/fjj8cPj5Z9httzRpbSt/Qqp5saVJ\n0jyrqoKHHkrFkzp1MmGSpIry449w8MHQty/86lfw5pspYZKaIU8TSJpnL78Mu+6a/u7VK9NQJEkN\nbccd4aWX4Mwz4ayzoHXrrCOSMmPSJGmeTZmSlnfe6VgmSaoIU6embgOtW8OFF6ZueBttlHVUUubs\nnidpnkycmC6QxgS3a5dtPJKk+fTee7DBBqngA0DPniZMUo5Jk6R6e/JJWGQR2H33dH2BBbKNR5I0\nH2bOTBPsrbcejBljoiQVYPc8SfU2enSqOnvWWbD88rDuullHJEmaJ198AQcdBM8+C7vsAjfcAEss\nkXVUUtkxaZI0z/r3TwWVJElN1HffwTvvwE03pUp5lkGVCjJpklRvzz6bdQSSpHn23Xdw771p7qW1\n1oJRo6B9+6yjksqaY5ok1VtVVVp27ZptHJKkenriCVhjDTjmGBgxIq0zYZLmyqRJ0jxZaSUnhJek\nJmPSJDj2WNhuO1h0UXjttfRFLqkoJk2S6uWtt+COO1KxJUlSExAjbLkl/OMf8Ic/wJAhsM46WUcl\nNSmeJ5ZUL6edlparrJJtHJKkuZgxA1q2TMUdTj45tTBttVXWUUlNki1Nkurl669T745HH806EklS\nrUaMSJPTXnttut63rwmTNB9saZJUp/Hj4auvZl0fOxbWXNOqtJJUlmJMcy2dcAK0bg2dOmUdkVQR\nTJok1WrmzFSNNj9pAqvmSVJZGjsWDj0UHnkEtt4abr4Zllkm66ikimDSJKlWw4enhOnEE2GTTdK6\nFi1giy2yjUuSVMC778Izz8Df/55KirdwFIbUUEyaJNXq+efT8uij4de/zjYWSVIBP/2UEqVddoFt\ntoHPPoPOnbOOSqo4Jk1SM/T443DUUbMmqa3N99+nrngrrNA4cUmS6uHFF+HAA+HLL2HkSFh6aRMm\nqURMmqRmaMiQdDLyoIPm3ntju+0s+iBJZWXaNDjnHPjzn9NZrWefTQmTpJIxaZKasRtvhFZ+C0hS\n01FVlUqJv/46HHYYXH45LLRQ1lFJFc+fS1IFmjkTBgyAceMKbx86tFHDkSTNrxhTs3/LlqlL3hln\nwM47Zx2V1GyYNEkVaORIOP98WHhhWHDBwvv07Jn+90qSytwXX0C/fnD88SlROvrorCOSmh2TJqkC\nxZiWV18N+++fbSySpHkUI9x5J/z+9zBjBkycmHVEUrNl0iRVkIkTU5nw0aOzjkSSNF+++y4lS3ff\nDRtvDLfdBiuumHVUUrNl0iRVkMsuS2OZqnXokFkokqT58dhjcN99cOGF8Kc/WbVHypifQKmCjB4N\nnTql/7Vt28Lqq2cdkSSpaJMmwZtvwqabwr77wgYbwEorZR2VJEyapIoyfjx06QLdu2cdiSSpXoYO\nTYNQR49OE+ktvrgJk1RG5jKtpaSmYvfd4cEHnQxekpqUGTPgggtgww3hp5/ggQdSwiSprNjSJFWI\nYcNgrbXgoouyjkSSVJSpU2HLLeHll2HvvVPJ08UWyzoqSQXY0iRViEmTYKONYP31s45EklSUNm1g\n883hjjtSaXETJqlsmTRJFWLSJGjXLusoJEl1+vpr2G03eO21dP2ii2CffbKNSdJcmTRJFSBGkyZJ\nKnsPPgi//W0qcfrRR1lHI6keTJqkCjBtGsycadIkSWXpp5+gf//UwrTMMqms+AEHZB2VpHqwEITU\nBE2Zkrq/T5mSrk+enJbt22cXkySpFgMHwi23wGmnpRnIW7fOOiJJ9WTSJDVB//kPHHLInOtXWKHx\nY5EkFTBtGnz8cZpl/Oij04S1662XdVSS5pFJk9QEjRmTlu+/P2s6jwUWgI4ds4tJkpTz/vtpotox\nY2DECFhoIRMmqYkzaZKaoPHjIQRYdVVo2TLraCRJQBpceuWVcOqpsMgicMMNKWGS1OSZNElN0Lhx\nqVXJhEmSysTEibDrrvD007DjjnDjjbDkkllHJamBWD1PaoLGj4dOnbKOQpL0i/btoUsXuP56ePhh\nEyapwpg0SU3QuHGwxBJZRyFJzdx338HBB8Mnn6Q+04MGwWGHpb8lVRSTJqmJeecdGDx4VgEISVIG\nnnwS1lwTbr8dXn4562gklZhJk9TEDBmSlttvn20cktQsTZ4Mxx8PvXvDwgvDq686Ua3UDJg0SU3U\ndttlHYEkNUN//nOqkHfssfDmm5YSl5oJq+dJTcigQXDvvVlHIUnNzIwZMHYsLLMMnHwybLYZbLll\n1lFJakQmTVITMXky9OuXxhcvu6xjmiSpUXzyCRx4IHz7Lbz9dqqSZ8IkNTt2z5OaiI8+SvMm/utf\n8Pnn6f+2JKlEYkxzLa21Frz/PpxzDrRpk3VUkjJiS5NURn76KZ3ILGTw4LTs1q3RwpGk5mnChNS6\n9PDDqVXplltSE7+kZsukSSojRx2VWpJqs+CCsMoqjRePJDVL7dvDjz/CFVfAccdBCzvmSM2dSZNU\nJqZMgYcegt13T8lTIcssA23bNm5cktQs/PQTDBgAp50GnTrB00+bLEn6hUmTVCaeegomToTDD4et\nt846GklqRl5+Oc21NHJkKiG+774mTJJm4zeClLFPP00tSHvsAR06wBZbZB2RJDUT06bBGWdAz56p\n0s7zz6eESZJqMGmSMjZyJHz5Jey0EwwcCK1bZx2RJDUTp58OF12U5nN4+23YdNOsI5JUpuyeJ5WJ\n445LJzslSSU0c2aqjrfYYrMmqt1556yjklTmTJqkDA0cWHe1PElSAxo9Gg4+OM0W/txzsOSSJkyS\nimLSJGXo0kvT//B11oGVV846GkmqYHfdlUqTTp+eSolb6EFSPfiNIWVsxx3hzTdhqaWyjkSSKtCE\nCam4wz77wGqrwVtvwWGHQQhZRyapCbGlSWokTz4J48bNvm7ChGxikaRmo2VLGDYMzj8fTj0VWvnT\nR1L9+c0hNYIRI6B378LbOnVq3FgkqeJNnpy64J1wAiy0UGpdatMm66gkNWEmTVIjePfdtLz/flh9\n9dm3/frXjR+PJFWsYcNg//1h+HBYdVXo29eESdJ8M2mSSqCqCn7+edb1t99Oy623hoUXziYmSapo\nVVWpus4558ASS8ATT9TexC9J9WTSJJVAr17w4ouzr1tuORMmSSqZ446Dq6+GPfeEa6+Fjh2zjkhS\nBTFpkkrgrbdS4rTTTrPW9eiRWTiSVJlihKlToW3blDRtvHGqlGdlPEkNzKRJamBTpsDEibDNNnDi\niVlHI0kV6ptv4PDDYYEF4J570vilVVfNOipJFcp5mqQGMm1amvqjV6903ap4klQi//43rLEGPP54\nal2SpBIzaZIayFlnwY03piJNffrAVltlHZEkVZiJE1Pr0s47Q5cuMGRIKitudzxJJWb3PKkBPPVU\nKtp0xBFp/LEkqQQmToSHHoJTToFzz7WUuKRGY9IkzadHHoH99oNu3eDyy7OORpIqzPTpcPPN0L8/\nLLUUfPQRdOiQdVSSmhmTJmk+HXkkTJgAzz0H7dplHY0kVZD//Q8OOACGDoVlloHttzdhkpQJxzRJ\n82nGDOjXD9ZaK+tIJKlCzJwJV10F664Lo0bB/fenhEmSMmJLkzSPfv4ZRo5Mk9C38pMkSQ3nyCPh\nhhtghx1ShZ2llso6IknNnD/1pHm0xx6p2i3AggtmG4skVYSqKmjZEg46CNZbL1XKszKepDJg0iTN\ng6lTYfBg2H33NPn85ptnHZEkNWE//ABHHw1LLpkq6myySbpIUplwTJM0D4YMgSlTYP/9oW9fJ7KV\npHn2zDNpotq774aOHSHGrCOSpDmYNEn1dPvtaV5FgE03zTYWSWqypkyBE09MM4G3awevvAJnnml3\nPEllyaRJqqeXXoLvvoO//AWWWCLraCSpiRo1Ks0GfvTRMGwY9OiRdUSSVCvHNEnzoHNn+OMfs45C\nkpqYqip4+GHYbTdYdVX4+GPo2jXrqCRprurV0hRCWDWEcGgI4ZQQwpK5dcuGEJzSUxXv6afTCdHn\nn886Eklqgj79NFXN2X33WV+kJkySmoiiWppCCAsAA4F9gQBE4Enga+AfwPvA6SWKUSoL55wDr7+e\nJqO3Wp4kFSlGuPlmOP54aNECBg2Cnj2zjkqS6qXYlqbzgZ2Bw4BfkRKnao8C2zZwXFJZiRGGD4dD\nDoFx4+Dee7OOSJKaiMMOg/79oXt3ePfdVHbUYg+SmphixzTtB5wVYxwYQmhZY9unwAoNG5ZUHl5/\nHcaPh4kT4fvvoVu3rCOSpCYixpQcbb99+vI84YTU0iRJTVCxSdMSwHt1bG/bALFIZWXsWNhgg9nX\nrbtuNrFIUpMxcSKcdFIq9HDiiWkMkyQ1ccWe8hkF1FYLtDvwccOEI5WPSZPS8uyz4bXX4L33nJdJ\nkur0yiuw9tpwww2peV6SKkSxSdPtwBkhhL5Adfe8GELYCDgRuKUEsUllYcUVYf31YfXV7YYvSQVN\nnw5nnZXOLM2YAYMHw/nnZx2VJDWYYpOmi4Gngf8DxufWPQu8CDwH/K3hQ5Max003QatWqat9/mXF\nFdP2ljVH8UmSZjdsGFx4IRxwALz9Nmy2WdYRSVKDKmpMU4xxBrBbCGEbUqW8zsC3wOMxxidKGJ9U\nUlOnplLiv/kN7LrrnNvbtoUddmj8uCSp7M2cCS+/nFqX1l8/VcZbffWso5Kkkih2nqbOwLcxxidJ\n8zPlb2sBdIoxflOC+KSSuu02+PJLGDgQevfOOhpJaiK+/DLNwfDkkzB0KKyzjgmTpIpWbPe8McB6\ntWxbJ7ddalJmzIBLLklTh2yzTdbRSFIT8X//B2usAS++CFdfnQo/SFKFKzZpqmv4eytgZgPEIjWq\nu++GTz+FM86wwIMkFeXww2GvvWDlldM4piOP9AtUUrNQa/e8EMJCwCJ5qzqFEJausduCwL7A1yWI\nTSqZmTPhootSb5Kdd846GklqItZcEwYMSGebWhU71aMkNX11feOdBJyd+zsC/65lvwBc2JBBSaV2\n2WUwfDj8619OUC9JtZoyJSVI668Pv/sdHHNM1hFJUibqSpr+A4wlJUVXA5cCI2vsMxUYHmN8vdg7\nDCFsB/ydNN/TjTHGSwrssxcwgJSsvR1j3LfY40vFuPzytNxrr2zjkKSy9dZbsP/+8P77cOqpKWmS\npGaq1qQpxjgUGAoQQojAfTHG8bXtX4wQQkvgn8A2wGjgjRDCwzHG4Xn7rAycBmwSY/w+V7lPajBV\nVTB2LJxyir1LJGkOVVXw17+myWoXXxwefRT69Mk6KknKVLHzNF3XQPe3PjAixvgpQAjhLmAXYHje\nPocB/4wxfp+7b0uZa558+y1Mnjzn+u+/T8uuXRs3HklqEp5+OrUs9e0L114LnTplHZEkZa7o8+wh\nhFWAg4FVgbY1NscYYzFTgHYFvsi7PhrYoMY+q+Tu7yVSF74BMcbHi41TAnjjjdQFvy5LLdU4sUhS\n2YsRPvgAunVLk9Y99xz07GllPEnKKXZy2/WAF0hV8pYDPgQ6Ap2Br4DPi7y/Qt++sUBMKwO9gGWA\nF0IIv40x/lAjpsOBwwGWW265Iu9ezcU3ufbJ00+HFVaYc3vbtrDTTo0bkySVpXHj4IgjUje8d99N\n5cQ32yzrqCSprBTb0nQJ8AiwDzAN2D/G+GYIYXvgRuCUIo8zGlg27/oypKSr5j6vxhinAyNDCB+S\nkqg38neKMV4PXA/QvXv3momXBMCuu0KPHllHIUll6pFHoH//1G/5ootgxRWzjkiSylKxxZbXAm5h\n1iS2LQFijI8CF5Eq6xXjDWDlEMIKIYTWwN7AwzX2eRDYAiCE0InUXe/TIo+vZuxvf4O11koXq+JK\nUh1ihKOPhh13hM6dYcgQOOkk52CQpFoU29LUBvgpxjgzhPAdsGTetuHAmsUcJMY4I4RwDPAEKfEa\nGGN8P4RwHjAkxvhwblvvEMJwoAo4Ocb4bZFxqhm7+274+mvYaKN0ffPN0+S1kqQaQoCFF4aTT4bz\nz4c2bbKOSJLKWohx7j3bQgjDgD/HGO8KITwHjIsx7pHbNhDYIsZYYORI4+jevXscMmRIVnevMrHS\nSqn4wx13ZB2JJJWh6dPhggtgq63SmKUYLfQgqVkJIQyNMXafl9sW29L0GGlupbuAi4F/51qcZgCL\nA3+clzuX5tc776QWJoCvvrIyriQV9OGHaaLaIUNg6tSUNJkwSVLRip2n6fS8vx8PIfQE9gDaAY/n\nutVJje7yy+HWW9MktS1bwgY1C9hLUnMWI1x9deqGt+CC8H//B3vskXVUktTkFD1PU74Y46vAqw0c\ni1RvM2emkuKfWipEkuZ0772pMk6fPnDTTdClS9YRSVKTNN9lckIIvwkh3NkQwUj19fXXKXGSJOX5\nKjebR9++cN99qbS4CZMkzbM6k6aQrBlC2DGE0K3GtjVCCP8HvAPsWMogpdo8+yxMmJB1FJJUJn74\nAQ44AH77WxgzJpUQ3313xy9J0nyqNWkKISwFvAQMAx4C3gsh3BpCaBVC+Edu/Y7A1cBKjRGsVFPH\njo5jkiQABg+GNdeEO++E44+3Mo4kNaC6xjRdAqwNXAi8CawA/Al4DtgIuJs0h9LoUgcp1aZlS1h2\n2ayjkKQMVVXBKaekyjgrrQQvveTZJElqYHUlTdsA58UYL6leEUJ4jzT57LUxxt+XOjipLs8+m7rt\nO6ZJUrPWsmX6MjzySPjLX6B9+6wjkqSKU1fS1JnUPS9f9XULPyhz77+flnvvnW0cktToqqrgiitg\nhx2gWzcYNCglT5KkkqirEERLYGqNddXXfy5NOFL9rbNO1hFIUiP67DPYcss099Ltt6d1JkySVFJz\nm6epdwghv8hDCyAC24UQVsvfMcZ4R0MHJ0mScmKE226DY49N12+9NVXKkySV3NySpvNqWX9BjesR\nMGmSJKlUbr0VDj4YNtss/b388llHJEnNRl1JU7c6tkmSpMbw44+wyCJpAOf06XDIIXbHk6RGVmvS\nFGP8sDEDkepr6NCsI5CkEvr55zRu6fHH4a23UuJ02GFZRyVJzVJdhSCksnbLLWnZrl2mYUhSw3vt\ntVTl5tproW9faN0664gkqVkzaVKT1aZNGgNt0iSpYsyYAQMGwCabwNSp8Mwzae6ltm2zjkySmrW5\nFYKQykZVFUyePOt6q1awxBLZxSNJDS4EGDwY9t0XrroKOnTIOiJJEiZNakI22ST1WMnXynewpKYu\nRrjhBthpJ+jSBR57DBZcMOuoJEl5/MmpJmPkSNhww9S9H6BFC9hrr2xjkqT5MmZMqob3+OPp73PO\nMWGSpDJU76QpN9nt4sC7McZJDR+SVLu114Y//jHrKCSpAdx3Hxx+eOp3/M9/wlFHZR2RJKkWRSdN\nIYT+wLlAl9yqHsCbIYR7gadijNeWID5VsG++SfM0TpxY3P7ffVfaeCSp0Vx7bUqSevSAQYNg1VWz\njkiSVIeiqueFEPoB1wPPAAcBIW/za8DvGjwyVbx334VHH03zNrZoMfdLz56wyy5ZRy1J82HatLTc\nay+45BJ46SUTJklqAoptaToZ+HuM8cQQQkvgtrxt/wNObPDI1GxceWVKiCSpYk2dCmeeCc8/Dy++\nCB07wimnZB2VJKlIxSZNKwKP1LLtJ2CxhglHlWzsWPjXv1LpcICPP842HklqFO+8A/vvn5rXjzwy\nzcW0wAJZRyVJqodik6bvgGVr2bYKMKZhwlElu/lmOP302de1bQtLL51NPJJUUlVVcMUVcMYZsNhi\n8J//wA47ZB2VJGkeFJs0PQKcGUJ4Gvgqty6GEBYF/gA8VIrgVFlmzEjLH3+Eli3T361aQevW2cUk\nSSUzbVo6W7T99nD99c7GLUlNWLFJ0xnAq8Bw4EUgAn8FugETSVX1pFpVVcEHH6S/27WblTRJUkWJ\nEe65JyVKCy+cxjB17AghzP22kqSyVVT1vBjjN8C6wJXAEsCXQEfgVmCDGOP3JYtQFeHSS+GOO1I3\nfn87SKpI48fDnnvC3nvDNdekdYsv7peeJFWAoudpijH+QGpxOqN04ahSVc+x9MILqXy4JFWUxx9P\nE899+y38+c9w0klZRyRJakDFztN0UQhhtVIHo8rWrh1ssEHWUUhSA7vqKujTBzp1gjfegD/9yT7I\nklRhij3nfyzwfghhSAjh2BCCo1lVtBkz4K9/TdOUSFLFiDEtd9opzbn0xhuw1lrZxiRJKolik6bO\nwIHAOOBy4MsQwn9CCHuGENqULDpVhB9/TMtu3bKNQ5IaxPTpcO65sPvuKXFafnm45JI0h4IkqSIV\nWwhicozxXzHGPsAywKlAF+Bu4OsQwg0ljFEV4rDDso5AkubTRx/BppvCgAGw0EI2oUtSM1HvIfkx\nxq9jjJfHGNcDtgJ+Ag5p8MhUlmZUzeTHKdOpmhnnvu8MOPvs1GtFkpq0GFNFvLXXho8/hrvvhkGD\nbF2SpGai6Op51XLd8XYF9gd6A4E0+a0q1NQZVTz67hiuGfwJH38zkVYtAjNmRlbpvBBH9lqR7dfo\nQptWcw56Hj4czj8f2reHJZeENdbIIHhJaggTJqQvtJ49YeBA6No164gkSY0oxDj3FgOAEEIv4ACg\nL7AIMAQYBNwZYxxfqtqvNEMAACAASURBVACL0b179zhkyJAsQ6hYb33xA/0Gvs70qpn8PK1qju3t\nW7dkgZYtuPWQ9Vlr2UVn2/b22+mk7P33w267NVbEktSA/vtf2GKLNMncqFGw3HLOuyRJTVQIYWiM\nsfu83LaolqYQwudAV+AL4J/AbTHGD+flDtV0vP3FD+xz/atMnj5nslQtJVJV7H39q1yzx0ZMGdPh\nl22fftoIQUpSKUyYAMcfD7feCldfDUcdBb/6VdZRSZIyUmz3vCdJidJzpQxG5WPqjCoOGvh6nQlT\nvsnTq9h138lMGN5hjm2LLNLQ0UlSCT3/PBx4IHzxBZx1Fhx6aNYRSZIyVlTSFGPsX+pAVF4efXcM\n/8/encfZXPZ/HH9dM2PsQnbZQ7ZBRCgJ0S7dyC1+KiW5ad/uW0kpUrpTt26VFpXEfVtapBstSrJE\n9jVrlixjyzJmvX5/XGOyzIwzY875nuX9fDzm8T3ne75nzns46nzmuq7PlZyalqPnJOwuSsPmJ3hz\n1J8LowsWhLi4vE4nIuInr78ODz4INWrAvHlw+eVeJxIRkSCQZdFkjGkGrLLWHk+/nS1r7aI8TSae\nGjNnU6ZrmLJiU6JIOliII8V+4/LLNYVFREJUq1Zw773w8suupbiIiAjZjzQtAC4HFqXfzqpjhEl/\n7Oz2aRKSUtMsv+49muljSfuKsHt8K2zyGX/dFsDwR8EDpKZVJjpKC6VFJASkpcGrr8Jvv8Frr0GT\nJu5LRETkFNkVTdcBa9NvX0/WRZOEmWNJKcREGZJTz/4rP7GlNDYphmLNN2KiTn/c5EulWK29HEtK\noViBfIGKKyKSO9u2wR13wJw5cMstbnO5mBzvxCEiIhEgy/87WGtnnnL7f4GJI8GgcGwMKVlsXpu4\nqzjRxY5Tok3mzRPTjHu+iEjQshbGj4cBA9xI03vvueJJrcRFRCQLUb5cZIxZY4zJdGtSY0xdY8ya\nvI0lXoqOMtQsk/lc/sTfi5O/wqEsn1urTBFNzROR4LZnD/Tv77rUrFgBd96pgklERLLlU9EEXAIU\nzOKxQkDtvIkjweK+NjUoHHv6uqXUY7Gk/lGI/OUzL5oKx0ZzX5uLAxFPRCTnFi92o0zlysGPP7pp\nedWqeZ1KRERCgK9FE2S9pikOOJwHWSSIXN+gPPmiT397JO4qDkBsFiNN+aKjuK5BOb9nExHJkePH\n3VS8yy6DSZPcuYYNIVr9i0RExDfZtRwfCAxMv2uBycaYxDMuKwhUACb7J554JX9MNB/c1Yzuby/I\n2OA28ffiYNKILXt2jVwwn7s+f4w+hIhIEPn5Z+jZEzZsgIcecg0fREREcii7Ffu7gCXpty8G1gP7\nz7gmEVgDjMn7aOK1hpWKM7Hv5fR+bxHJqWns2VWc2DJHiMr356a3hWOjyRcdxQd3NaNhpeIephUR\nOcPrr8PDD0P58vD119CundeJREQkRGXXPW8KMAXAuAWyg6y1mwOUS4JEw0rFWTioHV8u381tL5eg\nYJ2d5Is2pKRZapUpyn1tanBdg3IaYRKR4HPxxXDbbfDGG1Bcv9QREZHc86k3tLX2r/4OIsErf0w0\nFVMqknICRg2oTJceFSgcG6MueSISXKyFt9+GI0fg0Ufh+uvdl4iIyHnKbk3T48CH1trd6bezY621\nL+dtNAkWu3bB5Ze72y1bGm1cKyLBZ/du6NMHZsxwhdLDD0NUTnodiYiIZC27kaYXgTnA7vTb2bGA\niqYwtT99JVvPnnDJJd5mERE5y9Sp0LcvHDvm1jH97W8qmEREJE9lVzQVtNae7JaX1R5NEkHUdEpE\ngs6WLdCtGzRqBB99BHXqeJ1IRETCUHaNIBIzuy3h78QJuPlm2LfP3U9I8DaPiMhZNm2CGjXc5rSz\nZ8MVV0A+TR0WERH/8Gn+gjGmujGm0Sn38xtjnjHG/NcYc7f/4okXdu50n0GMgcqVoXZt14CqZUuv\nk4lIxEtMhCefhFq13H+oAK6+WgWTiIj4lU/d84B/4/ZjWpZ+fyjwELAB6GyMibbWvuWHfOKhhx6C\nXr28TiEikm7VKre4cvlyt4apRQuvE4mISITwdaVsI+AHAOM2bboD+Ie1th6uSUQ/v6QTEREB+Pe/\noUkT+P13+OILeOstKFLE61QiIhIhfC2aigPx6bcbARcC/0m/Pxuokce5RERE/hQTA9deCytXwo03\nep1GREQijK9F016gevrta4At1tpt6fcLA6l5HUxERCKYtTB+PEyY4O7fcw98+imUKeNtLhERiUi+\nFk3TgReMMc8DjwKTT3msHrAlr4OJd9LSvE4gIhHtwAHXfaZXL1c4Wes60xjjdTIREYlQvhZNT+I2\nur0N+Bp4/pTHugHf5m0s8dI333idQEQi1syZUL++G1UaPtytX1KxJCIiHvOpe5619g8g0z5q1trL\n8jSReC4qvZS+6ipvc4hIhFm92q1bqlsXvvwSGjf2OpGIiAjge8txAIwxRYFmQElgP/CztfaIP4KJ\n92Jy9O4QEcmlffugdGmoVw8mTXK7axco4HUqERGRDL5Oz8MY8xTwOzALmISbpve7MWaQn7KJiEg4\nS0mB55+HKlXgl1/cuW7dVDCJiEjQ8WkswRjzN+A54GNgPLAbKAf0BJ4zxhyw1o7xW0oREQkvGze6\nRg8LFkCPHlCtmteJREREsuTrBKwBwL+ttQNOObccmGmMOQwMBFQ0iYjIub3zDjzwAMTGwiefQPfu\nXicSERHJlq/T86oDn2Xx2Gf8uYeThLjERPj7371OISJhbft2aNHCbVSrgklEREKAryNNB4DawOxM\nHqud/riEgTVr3BYpxsCFF3qdRkTCxqefQtGi0K4dPP20a9MZ5fOyWhEREU/5+n+sT3Gb23Y15s8N\nM4wxnYGh6Y9LGJk2DfLn9zqFiIS8I0egTx/o3BlGjXLnYmJUMImISEjxdaTpSeBSXNe8RGPMXqA0\nkB/4Of1xCQOLF3udQETCxo8/wv/9H2zbBoMGweDBXicSERHJFV83tz1sjGkJdAauxO3TdAD4HvjM\nWpvqv4gSSH37umPJkt7mEJEQt2gRtG7tuuLNnQstW3qdSEREJNd83r40vTCanP4lYez22+HKK71O\nISIh6dgxKFwYLrsMXnsN7rwTihTxOpWIiMh5yXZSuTGmuzFmgTEm3hiz0RjzgjHG50JLQk90NFSt\n6nUKEQk5aWnw6qtQvTps3eq6yQwcqIJJRETCQpZFkzGmKzABt4ntPOA4bu3S84GJJoFmrfvc82er\nDxERH2zfDtdcAw8/DJdfDoUKeZ1IREQkT2U30vQw8CVQ01rbyVobB4wABhpj1PYoDO3e7Qqn8uW9\nTiIiIWPCBGjQABYuhLFjXWvxMmW8TiUiIpKnsit+agNjrLXJp5x7HSgIVPFrKvHE5s3uWF1bFYuI\nr775BurWheXL4e67NVQtIiJhKbv1ScWB+DPO7Us/lgC2+CWReGbTJndU0SQi2Zo9240mNWwIo0dD\nvnxu7yUREZEwda5pdjaH5yWEbd7sfkmsRhAikqmEBHjgAejQAYYOdecKFlTBJCIiYe9c/6ebZzKf\narHwjPPWWps/z1KJJzZvhkqVIDbW6yQiEnSWLIGePWHdOlc4DR/udSIREZGAya5oGhGwFBIUNm/W\n1DwRycT330P79lC2rJua176914lEREQCKsuiyVr790AGEe9t3gzXXed1ChEJGikpbupdy5bwxBPw\nyCNQooTXqURERAJOrcMFgOPH4fffNdIkIri9B8aOhfr14cAB1+jh+edVMImISMRS0SQAbEnvhVij\nhrc5RMRje/bAzTdD375w0UWQmOh1IhEREc+paBLArfEGFU0iEe3TT93o0uzZMGoUzJql3a5FREQ4\nd/c8iQBpafDyy25/yiZNvE4jIp6wFt56y7XQHD/e/QdBREREABVNAnzxBaxaBR99BFEaexSJLPPm\nuUKpcmX4+GMoUkT7DoiIiJxBH5Ej0Lp1bgZOjRruq1cv1wCie3evk4lIwCQlwT/+Aa1bw9NPu3Ml\nS6pgEhERyYTPI03GmLLAA0BroCTQxVq7xhjTH1hkrV3sp4ySxxYsgNWr4dZboVAhd+6OO1xnYRGJ\nAKtXu41qly2DPn3g1Ve9TiQiIhLUfPqYbIy5BPgByAf8DLQACqQ/XBtoCfT0R0DJe/v2ueO4cVC0\nqKdRRCTQZs+Gm26CYsVc44dOnbxOJCIiEvR8nZ43EtgCVAOuB8wpj83DFVES5FJT4dln4b//dTNw\nihTxOpGIBIy17ti8OfTuDStXqmASERHxka9F01XAMGvtIcCe8dhuQD1pQ8D69TBkiJuZ06EDGHPO\np4hIOJgwAdq0gRMn3AjTW29B2bJepxIREQkZOWkEkZrF+QuBhDzIIn528hfN48a5jnkiEuYOHIC/\n/hVuvx2Sk+HgQa8TiYiIhCRfi6bFQK8sHvsLsCBv4oi/HD/uGkCISISYPRvi4mDyZHj+efjhB21U\nKyIikku+9kt7AfifMeYL4GPcFL3Wxph7gW7A1X7KJ3nk8cfhjTfc7WLFvM0iIn6WlgZPPuk6vXz2\nmXatFhEROU/G2jOXKGVxoTG3AqOAi045vQsYYK391A/ZfNa0aVO7eLE6nmfl99+hWjXo3NkVT40a\naT2TSFhauhSqVoUSJeC336B0aShY0OtUIiIiQcEYs8Ra2zQ3z/V5TZO1dipQBYgD2gONgcpeF0xy\nbq+84pYzPP88NG6sgkkk7KSkwLBh0KwZDB7szlWurIJJREQkj+RoO1PrhqVW+SmL+MFPP7mi6fbb\noUYNr9OISJ7btAn+7//cP/Zu3dy+AiIiIpKnfN3cttu5rrHW/uf840he+zR9HFCfo0TC0FdfQdeu\nEBMDH3/sOuVpKFlERCTP+TrSNDGL86cuiFLRFITmzoWWLTXKJBKW4uKgY0cYNQoqVfI6jYiISNjy\ndU1TnUy+rgBGAJvSb/vEGHOtMWa9MWajMebJbK7rYoyxxphcLdYS1zRrwQJo3drrJCKSZz7/HHr0\ncB3yKlaEKVNUMImIiPiZTyNN1tr1WTz0kzEmFbgPmH+u72OMiQbeAK4BdgA/G2M+t9auOeO6osD9\nwEJf8knmPvnEHbt39zaHiOSBI0fg4YfhnXdcC8z9+113PBEREfE7n7vnZeM74GYfr20GbLTWbrbW\nJuGm/XXK5LqhwEvAiTzIF9Fq14aGDb1OISLnZd48Vyi9+67bf2nhQhVMIiIiAZQXRVNT4LiP11YE\ntp9yf0f6uQzGmMZAJWvt9DzIFrE2bIBJk9wMHhEJYcnJ0KuX+8f8ww8wfDjExnqdSkREJKL42j3v\n8UxOxwL1gc7AWB9fL7O2ThnNJIwxUcCrwB0+ZOoL9AWoXLmyjy8fOe67zx0rVPA2h4jk0vr1bqPa\n/PndOqbKlaFYMa9TiYiIRCRfu+e9mMm5VGAnrsjxtaH1DuDUFcsXAbtOuV8UV4jNMa5tbjngc2PM\nzdbaxad+I2vt28DbAE2bNj21i58Aq1e77sPvv+91EhHJkbQ0GD0anngCHn/c7RdQv77XqURERCKa\nr0VTZtvKJ1trczr562egpjGmGq7g6g70OPmgtfYwUOrkfWPMHODRMwsmydrRo7BrF+zZA5de6n5J\nLSIhYscOuPNO+PpruPFG6N/f60QiIiKCD2uajDGxwBCgvrU28ZSvHK+WsdamAAOAmcBa4D/W2tXG\nmOeMMb42k5As7N8P5cq55g8A9ep5m0dEcuCrr6BBA/jpJ3jrLTclr2xZr1OJiIgIPow0WWuTjDEP\nAF/lxQtaa2cAM844NziLa9vkxWtGih9+gGPHYNAgVzhdc43XiUTEZxdd5DrkjR0LF1/sdRoRERE5\nha/T85YDdYEf/JhFztPcuVCgADz9tKbliYSEb76BmTPhpZfcKNN333mdSERERDLha8vxx4EnjDHt\n/RlGci8hAV59FZo3V8EkEvQSEuChh6B9ezcN7/BhrxOJiIhINnwdaXoPKA7MNMYcB3ZzSqtwwFpr\na+d1OPHdjh3uWFt/CyLBbelS6NkT1qyBgQPhxRehUCGvU4mIiEg2fC2alnB6kSRBqnVrrxOISJYS\nEuDaayEmxk3L69DB60QiIiLiA5+KJmttd38HkfMzZYrXCUQkSzt2uJ2mCxaEyZNda8uSJb1OJSIi\nIj7Kck2TMWazMaZhIMNI7g0Z4o7Vq3saQ0ROZS28+y7UqQNjxrhzV16pgklERCTEZNcIoiqglgIh\nokABeOABaNHC6yQiAsDevdC5M9x9N1x2Gdx0k9eJREREJJd87Z4nQU7Nt0SCyKxZroX4//4H//wn\nfP01VK7sdSoRERHJpXOtaVLzhxCwdq07njjhbQ4RSRcTAxUrun2Y6tf3Oo2IiIicp3MVTc8aY+J9\n+D7WWts7LwJJzm3b5o633uptDpGINn8+LFrk5sm2bQuLF0OUBvNFRETCwbmKpkZAog/fRyNSHtq3\nzx2rVfM2h0hESk6G556DYcPcP8J77nH7LqlgEhERCRvnKppusdYuCkgSyRVroXf6GF/p0t5mEYk4\na9dCr16wZAnceSeMGqWNakVERMKQr5vbSpA6etQVTnXrQvHiXqcRiSCHD7t2lfnywdSprlOeiIiI\nhCUVTSHu5NS8xx7zNodIxDh4EEqUgAsugPffd4VTuXJepxIRERE/0qT7EDd2rDuWKuVtDpGIMGkS\n1KjhRpbAjS6pYBIREQl7WRZN1toorWcKfl9+6Y6NGnmbQySsHTwIt98O3btDrVpuDyYRERGJGBpp\nCnH79sHdd8NFF3mdRCRMffcdxMW5UabnnoMff4SaNb1OJSIiIgGkNU0h6rvvYP9+iI/X1DwRv9q5\nEwoXdvswXXaZ12lERETEAyqaQtDWrW7vzJOqV/csikh4WrYM1q+H225z0/K6doX8+b1OJSIiIh7R\n9LwQlJDgji+95LaJuftub/OIhI3UVBgxApo1g0GD3Ma1xqhgEhERiXAqmkJY5cpwySXuM52InKct\nW6BNG3jySejUCRYudHswiYiISMTT9DwRkb17/2xB+dFHbkqefhshIiIi6VQ0iUjkSkx0U+/KlHHT\n8q6/3g3hioiIiJxC0/NEJDJNn+66qPz0k7vfr58KJhEREcmUiqYQ8/PPcM89XqcQCWFHj8K998JN\nN7l+/cWKeZ1IREREgpyKphAzaxbMmwft2mnLGJEcW7AAGjeGsWPh8cdh0SKoX9/rVCIiIhLktKYp\nRH31lRp7ieTYnDmujficOdC6tddpREREJERopCnEzJjhdQKRELNuHXz7rbv92GOwYoUKJhEREckR\nFU0hJDn5zzXr0dHeZhEJetbC6NFuOt7f/gZpae4fjtYwiYiISA6paAohx4+744gREKW/OZGs7doF\n114LAwfC1Ve7kSb9oxEREZFc0pqmEHKyaLrgAm9ziAS1335zG9UmJsKYMa5TnjaqFRERkfOgoimE\nHDvmjoULe5tDJCilpbnRpEqVYMAA6NkTatXyOpWIiIiEAc1XCSEnR5oKFfI2h0jQ+e47qFcPNmxw\no0rPPaeCSURERPKMiqYQoqJJ5AwnTsAjj0DbtpCa+uc/EhEREZE8pOl5IeL++2HqVHdbRZMIsGyZ\nm4K3ejX07w8vvaS5qyIiIuIXKppCxFdfQUwM9OsHTZp4nUYkCLz/Puzf7zYvu+46r9OIiIhIGNP0\nvBDSqpVrBqZfpkvE2rLFjTABDB8OK1eqYBIRERG/00hTELIWRo6EnTv/PLdvn3d5RDxnLYwb5+ap\n1q4NP//s5qlqrqqIiIgEgIqmILRrFzz+OBQsCLGxf55v2tS7TCKe2bfP7bU0bRq0bg0ffqh9l0RE\nRCSgVDQFob173XHCBLjlFm+ziHhqwwZXKB08CC+/DA89BNHRXqcSERGRCKOiKQgtWeKOpUp5m0PE\nc9Wrw003wcCBEBfndRoRERGJUGoEEYQGDXLHSpW8zSHiiYUL4Yor3JBrTAyMHauCSURERDyloikI\n5c8PzZpBlSpeJxEJoORkeOYZ1yZy+3a3uE9EREQkCGh6XhCKioI6dbxOIRJA69e7jWoXL4beveG1\n1+CCC7xOJSIiIgKoaAoqTZu6bWeSklzhJBIxhgyBzZth8mT4y1+8TiMiIiJyGhVNQWTZMjct76qr\n4K9/9TqNiJ/t2uV+Q1C1KvzrX256XvnyXqcSEREROYvGM4JM27YwfLjWvUuY++9/oUED6NPH3S9V\nSgWTiIiIBC0VTUHi0CFITfU6hYifHT4MvXpBt25QowaMGeN1IhEREZFz0vS8IPHTT+5Ypoy3OUT8\nZs0auO462LnTdckbNAjy5fM6lYiIiMg5qWgKEvv2ueP113ubQ8RvqlRxU/L+8x9o3tzrNCIiIiI+\nU9HkoS1b4Lvv3O3Zs92xdGnv8ojkuRUr4Nln4aOPoHBhmD7d60QiIiIiOaaiyUOPPALTpv15v2xZ\nKFbMuzwieSY1Ff75T3jqKShZEjZuVHcTERERCVlqBOGh33+HK6+Ebdvc18aNYIzXqUTO09atrg3k\n44/DjTe6zcdUMImIiEgI00iTh/btc/syVa7sdRKRPNSvHyxdCh984Drl6TcBIiIiEuJUNHlo/364\n8EKvU4jkgfh4VxxdeKFrI26M27RWREREJAxoep6HUlPVcVnCwIwZUL8+9O/v7lerpoJJREREwoqK\nJhHJnWPH4L774IYbXNvHQYO8TiQiIiLiF5qeF2BPPw3/+5+7ffSot1lEcm31aujc2XUvefRRGDoU\nChTwOpWIiIiIX6hoCrBJk1yx1Lix+wV9585eJxLJhdKloXhx+PZbaNPG6zQiIiIifqWiyQNt2sCE\nCV6nEMmhDRvgtdfg9dehTBlYuFCd8URERCQiaE1TgCxZAkOGuI55IiHFWvj3v6FRI5g4Edavd+dV\nMImIiEiE0EhTgAwdCp995m7XqeNtFhGf/f473HWXW4jXsSO89x5UqOB1KhEREZGAUtEUIKmpcOml\nbsRJJCRYC7feCsuXwxtvuE55Gl0SERGRCKSiKQD++AN++QXKlvU6iYgPDh92G4gVKuSm5RUqBLVr\ne51KRERExDNa0xQAPXvCrl1QpIjXSUTO4fvvIS4OnnzS3W/cWAWTiIiIRDwVTQFw6JA7jh/vbQ6R\nLCUmwmOPwdVXQ2ws9OjhdSIRERGRoKHpeQFgjGszXrmy10lEMrFmDXTvDitXQr9+MHIkFC7sdSoR\nERGRoKGiKQCs9TqBSDZiYuDYMZg+3e24LCIiIiKn0fS8APjjDyha1OsUIqfYtg2ef95V9LVqub2X\nVDCJiIiIZEpFUwDs2welS3udQgRXJH34oWv28NJLsGWLOx+jQWcRERGRrKho8jNrIT4eSpXyOolE\nvPh46NoVeveGhg3d/kvVq3udSkRERCTo6dfLfpaSAklJmp4nHrMW2rWDtWthxAh45BGIjvY6lYiI\niEhIUNEUIMZ4nUAi0vHjkD+/K5BeecXNE23Y0OtUIiIiIiFF0/NEwtWiRW5z2pEj3f327VUwiYiI\niOSCiiY/S0tzR400ScCkpMCzz0LLlpCQAM2aeZ1IREREJKRpep6fHT/ujoUKeZtDIsSvv0LPnm6U\nqWdP+Ne/oHhxr1OJiIiIhDQVTX6mokkCau9e2LwZJk2Cbt28TiMiIiISFlQ0+ZmKJvG733+HGTOg\nTx9o1Qq2boXChb1OJSIiIhI2tKbJz04WTfoMK34xZQo0aAD33++KJ9CbTURERCSPqWjyM400iV8c\nPuw2qe3SBapVgyVLoHx5r1OJiIiIhCVNz/Ozo0fdUUWT5JmUFGjRAtavh6efdl/58nmdSkRERCRs\nqWjys8mTITYWatf2OomEvORkiIlxX089BdWrw+WXe51KREREJOxpep4f7dgB48bBXXdBmTJep5GQ\ntnIlNG0Kn3zi7vfooYJJREREJEBUNPnRyJGQmgpPPOF1EglZaWnwyiuuYNqzB0qU8DqRiIiISMTR\n9Dw/mjYNOnWCqlW9TiIhads2uOMOmDMHbrkF3n4bSpf2OpWIiIhIxFHR5CfWuoGBGjW8TiIh65df\nYPFieO89VzwZ43UiERERkYikoslPjh2DxEQoVcrrJBJS9u+Hn36Cm26Czp1h82aNLomIiIh4TGua\n/GTqVHfUEhTx2f/+5zaq7dEDDh5051QwiYiIiHhORZOfJCa6Y9u23uaQEHD8OAwYANdd56rsH35Q\ntS0iIiISRDQ9z88KFPA6gQS1EydcZ7y1a+Ghh2DYML1pRERERIKMiiY/SUnxOoEENWtdY4cCBaBP\nH2jUCNq18zqViIiIiGRC0/P85KuvoFw5KFvW6yQSdH79FVq1gm+/dfcfeUQFk4iIiEgQU9HkBzt2\nwJdfwl13QYzG8uQka+Gtt9yo0rp1cPSo14lERERExAcqmvzg/vvd5+O77/Y6iQSN3bvhxhuhXz83\nyrRyJdx8s9epRERERMQHKpryWEoKTJsGVatCtWpep5GgMXWqm473+uuutXjFil4nEhEREREfafJY\nHtu/3x0fecTbHBIE/vgDVq2Cli3dCNO110L16l6nEhEREZEc0khTHouPd8dSpbzNIR6bOxcaNnRT\n8I4dg6goFUwiIiIiIUpFUx47WTSVLu1tDvFIYiI8+SRcdRVER8MXX0Dhwl6nEhEREZHzoOl5eWzf\nPnfUSFMEOnIErrwSli+Hvn3hlVegSBGvU4mIiIjIedJIUx7TSFMEK1oU2raFzz93rcVVMImIiIiE\nhYAXTcaYa40x640xG40xT2by+MPGmDXGmBXGmG+MMVUCnfF8nBxpuvBCb3NIgPz2G1x/vWv4APDP\nf8JNN3mbSURERETyVECLJmNMNPAGcB1QF/irMabuGZctBZpaa+OAycBLgcx4vuLj4YILIDbW6yTi\nV9bC+PHQoIFr+rBxo9eJRERERMRPAj3S1AzYaK3dbK1NAiYCnU69wFr7nbX2ePrdBcBFAc54Xvbt\n09S8sHfgANx2ACij4gAAIABJREFUG/Tq5Yqm5cvhllu8TiUiIiIifhLooqkisP2U+zvSz2WlD/CV\nXxPloZ9+gk8+UROIsDd6NHz6KQwfDt9/r1biIiIiImEu0N3zTCbnbKYXGtMTaApclcXjfYG+AJUr\nV86rfOdl3Dh37NLF0xjiD8ePw7ZtUKcOPPEEdO7sRplEREREJOwFeqRpB1DplPsXAbvOvMgY0x4Y\nBNxsrU3M7BtZa9+21ja11jYtHSTz4eLjoV49eOQRr5NInlq8GC69FK67zu3DlD+/CiYRERGRCBLo\noulnoKYxppoxJhboDnx+6gXGmMbAW7iCaW+A852X+HitZworKSkwdCi0aAHHjsG777qCSUREREQi\nSkCn51lrU4wxA4CZQDTwnrV2tTHmOWCxtfZz4GWgCPBfYwzAb9bamwOZM7cOHYIaNbxOIXniwAG4\n4QZYsAB69HDrmEqU8DqViIiIiHgg0GuasNbOAGaccW7wKbfbBzpTXrEWorRdcHgoXhwuush19uje\n3es0IiIiIuIhfcTPI0ePutlcEsL27IGePWHnTlf9/ve/KphEREREREVTXli3zrUZX7cO8uXzOo3k\nyqefQv36MGWKa/wgIiIiIpJORVMeGD4coqPh9ddd3wAJIX/8AX36uBbilSvDkiXQqdO5nyciIiIi\nEUNF03navBk+/hjuvRcGDoSaNb1OJDkyZIjbYGvQIJg/H+rW9TqRiIiIiASZgDeCCDe33upGmR59\n1Osk4rOkJNi3DypWhMGDoWtX11ZcRERERCQTGmk6D8eOwfLlcPnlUKGC12nEJ6tXQ7NmcNNNkJrq\nuuSpYBIRERGRbKhoOg/x8e7Yu7e3OcQHaWnw6qvQpAns2uWm5UVHe51KREREREKApufl0tixsHSp\nu12qlLdZ5Bz27XOtw7/9Fm6+2f3llSnjdSoRERERCREqmnJh927o29fdLlRIvQOCXtGikJAA77wD\nd90FxnidSERERERCiKbn5UJqqju++SYcOQIXX+xtHsnEgQPwwAOupXiBAjBvnmstroJJRERERHJI\nRVMu7NvnjlFR7kuCzOzZEBcH//43zJ3rzqlYEhEREZFc0kf+XLjuOncsVMjbHHKGhAQ3utShAxQr\nBgsXwg03eJ1KREREREKciqYcshb27oXLLnPb+0gQuf9+eP11VzgtWQKXXup1IhEREREJA2oEkUOH\nD7vu1d27Q2ys12mElBQ4etTtt/T003DbbdC+vdepRERERCSMqGjKoZN7M6nNeBDYtAl69XJT8b76\nCipXdl8iIiIiInlI0/NyKCnJHQsU8DZHRLPW7bXUsCGsXet2F1ajBxERERHxE400SWiJj3d7LX3x\nBbRrB++/D5UqeZ1KRERERMKYRpoktERFwZo1MGoUzJqlgklERERE/E4jTRL8jhxxRdITT0DJkq5o\nUhcOEREREQkQjTRJcJs3z61dGjIEvv/enVPBJCIiIiIBpJGmc9i3z31uP2n7du+yRJSkJFcojRgB\nVaq4gumKK7xOJSIiIiIRSEXTOfz97/Duu2efL1Ei8Fkiyp13woQJ0KcPvPoqFC3qdSIRERERiVAq\nms7h+HHXa+Dzz/88V7Ag1KrlXaawlZbmRpgKFIDHHoNu3aBTJ69TiYiIiEiEU9GUhdRU9xk+Lc19\nhm/UyOtEYW77drjjDqhe3e3B1KiR/tBFREREJCioEUQmli+HwoVdv4FJkyA62utEYW7CBGjQABYu\nhObNvU4jIiIiInIajTRlYuVKSEyEhx92Ha4vv9zrRGHq4EHo3x8mToQWLeDDD+Hii71OJSIiIiJy\nGhVNmYiPd8dBg1zRJH7yxx8wezY8/7zbgylGb0cRERERCT76lJqJffvclLzixb1OEoYSEtyIUt++\nrpX45s1QrJjXqUREREREsqQ1TZlYtcr1I4jSn07e+uUXaNIE+vX7c/MrFUwiIiIiEuRUFpwhLQ3m\nzoUrr/Q6SRhJSYFhw1yTh8OHYeZMbVQrIiIiIiFD0/NOcfgwjB7t+hO0bu11mjDy17/C5Mlu36Ux\nY7RQTERERERCioqmU7z6Kjz7rNuXqV07r9OEOGvdsF10tFu/1LmzK56M8TqZiIiIiEiOaHreKaZO\nhVatYP9+uOgir9OEsD17oFMnGDrU3b/mGujRQwWTiIiIiIQkFU3pVq92+zN17QqFCnmdJoR9/rnb\nqHbWLE3DExEREZGwoKIpXf/+7ti5s7c5QtaRI3D33W6EqWJFWLIE7r/f61QiIiIiIudNRROuudsv\nv0CbNlC5stdpQtTGjTB+PDz5JCxcCPXqeZ1IRERERCRPqBEEsHw5HD0K997rdZIQk5QEM2bALbdA\n48Zuo9oKFbxOJSIiIiKSpyJ6pCkpyXXJu/lmd197M+XAmjVw+eVuPuOyZe6cCiYRERERCUMRXTTt\n3w/ffgvlysHgwW4pjpxDWhq8/jo0aQLbt8O0adCokdepRERERET8RtPzcNsIaWqej7p2db3Zb7gB\n3n0Xypb1OpGIiIiIiF+paBLfWOv2WercGTp2hHvu0b5LIiIiIhIRVDRJ9g4ehAED4OqrXUvxnj29\nTiQiIiIiElARvaZJzuGbbyAuDv7zHzh0yOs0IiIiIiKeiNii6dAh+Oknr1MEqYQEeOghaN8eihSB\n+fPh0Ue9TiUiIiIi4omILZoGDIAuXdztokW9zRJ05s+H116DgQNhyRJo2tTrRCIiIiIinonYNU17\n90LdujB2LDRr5nWaIJCa6oqlK66Atm1h9WqoU8frVCIiIiIinovYkabjx93+TC1bQkzElo7pNm+G\n1q2hTRv49Vd3TgWTiIiIiAgQ4UVToUJep/CYtW6vpYYN3cjSBx/AxRd7nUpEREREJKhE7BjL8eNQ\nuLDXKTxkLXTrBpMnu3bi48ZB5cpepxIRERERCToRWzQdOxbhI03GuAYPLVvCAw9AVMQOOoqIiIiI\nZCtii6aInJ539Cg88gjcfDPccAM88YTXiUREREREgl7EFk3JyZAvn9cpAmj+fOjVyzV9qF7dFU0i\nIiIiPkhLSyM+Pp5Dhw6RmprqdRyRs0RHR1O8eHFKlSpFlB9mUEVs0QRuhlrYS06G556DYcPcmqXv\nv4crr/Q6lYiIiISQHTt2YIyhatWq5MuXDxMRH6IkVFhrSU5OZs+ePezYsYPKflinH5ELWX79FY4c\n8TpFgHz2GTz/PPTuDcuXq2ASERGRHDt27BgVK1YkNjZWBZMEHWMMsbGxVKxYkWPHjvnlNSJypKlP\nH3csU8bbHH6Tlgbr1rnde//yF/jxR2jVyutUIiIiEsL8MeVJJC/58z0ace9+a2HVqjDug7BzJ1x7\nLTRv7m4bo4JJREREROQ8RFzRtHcvHDwIHTqE4ZqmSZOgQQOYNw9GjoQKFbxOJCIiIiIS8iKuaFq6\n1B3r1PE2R55KTYWePaF7d6hVC5Ytg3vvDcOqUERERCTnli5dSnR0NK0ymX2zdetWjDEsXrz4rMfa\ntGnDgAEDTju3bNkybrvtNsqVK0eBAgW4+OKLueOOO1i5cqXf8gN8//33NGnShAIFClC9enXefPPN\ncz7n559/pn379pQoUYLixYvTrl07Fi1alPH4iRMnuOOOO4iLiyNfvny0adPGjz9BaIu4omncOChR\nAq64wuskeSg6Gi64wHXJ+/FHqFnT60QiIiIiQWPs2LH079+fVatWsXbt2lx/n+nTp9O8eXOOHj3K\nRx99xNq1a5k4cSLly5fnySefzMPEp9uyZQvXX389LVu2ZOnSpfz9739n4MCBTJkyJcvnHD16lGuv\nvZYKFSrw008/MX/+fMqXL0/Hjh05kt4RLTU1lQIFCjBgwABu0HY02YqoRhB798LUqdC/PxQs6HWa\n83TiBAwaBD16QJMmMHq0RpZEREREzpCQkMCECRP44YcfOH78OO+++y4jR47M8fc5fvw4d955Jx07\nduTzzz/POF+tWjWaNm3KoUOH8jL2ad58800qVKjAv/71LwDq1KnDwoULGTlyJH/5y18yfc66des4\ncOAAzz77LNWqVQNg6NChfPzxx6xfv56mTZtSuHDhjBGrFStW+PVnCHURNdL0wQdu26K+fb1Ocp6W\nLoWmTeGf/4Svv3bnVDCJiIiInGXy5MlUqVKFuLg4evXqxYcffkhycnKOv8/MmTOJj4/PckSpePHi\nWT532LBhFClSJNuvuXPnZvn8+fPn06FDh9POdezYkcWLF2f5s9SuXZvSpUvz7rvvkpiYSGJiImPH\njqVy5crUq1fPh59YThVRI02TJkGzZq4Td0hKTYWXX4bBg6FUKfjqK9cpT0RERCSAHnzQLaEOpEaN\nYNSonD/vnXfeoVevXgBcddVVFCpUiM8//zzLEZqs/Prrr4Ab5cmpfv360a1bt2yvqVixYpaP7d69\nm/bt2592rmzZsqSkpBAfH0/58uXPek7RokWZM2cOnTp1Yvjw4QBUrVqV2bNnUzDkp1wFXsQUTdu2\nwZIlMGKE10nOw/vvw9//Dl26wJtvwoUXep1IREREJGht3LiRefPm8cknnwBuE9Tbb7+dd955J8dF\nk7U21zlKlixJyZIlc/184KxNhU/myWqz4YSEBO666y5atGjBhAkTSE1NZeTIkXTq1InFixdTuHDh\n88oTaSKmaJo2zR07d/Y2R45ZC7t2QcWK0Lu325H3pps0HU9EREQ8k5sRHy+88847pKamUrly5Yxz\nJ4uN7du3U6lSJS644AIADh8+fNbzDx06lPF4rVq1AFi7di0tW7bMUY5hw4YxbNiwbK/56quvuPLK\nKzN9rFy5cuzevfu0c3v37iUmJoYLs/gl+oQJE9i0aRPz5s0jOjo641yJEiWYNm0aPXv2zNHPEOki\npmiaOtVtYRRSjeX27XMLsBYuhNWrXdu/m2/2OpWIiIhI0EtJSeGDDz5g+PDh3Hjjjac91qtXL95/\n/30GDx5MiRIlKFWqFEuWLKFdu3YZ1/zxxx9s3LiR2rVrA9ChQwdKlSrFiy++eFojiJMOHTqU5bqm\n852e16JFCz799NPTzs2ePZumTZuSL1++TJ9z/PhxjDFERf3ZwiAqKgpjDGlpadlmkbNFRNG0Z4/r\nxD14sNdJcmD6dOjTBw4dguHDXUtxEREREfHJl19+SXx8PPfcc89ZozHdu3dnzJgxPPXUU0RFRfHw\nww/z4osvUqFCBVq0aMH+/fsZOnQopUqVomvXrgAULlyYd955h65du3LDDTfw4IMPUrNmTQ4cOMC0\nadP45Zdf+PLLLzPNcr7T8/r168fo0aN58MEHuffee5k3bx7jxo3LmHYIMHr0aEaPHs26desAuOaa\na3jsscfo378/999/P2lpabz44otER0fTtm3bjOetWbOGpKQk4uPjOXr0KMvSF6s1atQo13nDkrU2\n5L+aNGlis/Pkk9aCtStWZHtZcEhMtLZvXxc4Li5EQouIiEg4W7NmjdcRcuymm26y11xzTaaPbdq0\nyQJ25syZ1lprU1JS7Ouvv24bNGhgCxcubCtWrGhvu+02u2XLlrOeu3jxYtulSxdbpkwZGxsba6tX\nr2579+5tV61a5c8fx86ZM8c2btzYxsbG2qpVq9oxY8ac9vgzzzxj3Uf7P82aNcu2atXKXnDBBbZ4\n8eK2TZs2dt68eaddU6VKFQuc9RWqsnuvAottLusNY89jUVuwaNq0qc1sF2eAH36Aq66CatVg06YQ\nWApkLdxyC1xyidusNn9+rxOJiIhIhFu7dm2uusaJBFp271VjzBJrbdPcfN+wn573zjvuOH16EBdM\nyckwbBj83/+56m7qVEhfsCciIiIiIt4K681tjxyBKVNcL4Wg3Ztp3Tpo0QKGDIHJk905FUwiIiIi\nIkEjrIumyZPh+HHXqTvopKXB6NHQuDFs3erCPvaY16lEREREROQMYV00jRvnWoy3aOF1kky8/joM\nHAhXXw0rV0ION1gTEREREZHACNs1TZs3uyYQL7wQZGuZDh927cP79IHixd0wWFAFFBERERGRU4Xt\nSNOHH7papFcvr5OkO3TIhWnZEhISoGhRuOMOFUwiIiIiIkEuLIumtDT44ANo1w4qVfI6DfDddxAX\nB598ArfdBlns3CwiIiIiIsEnLIumuXNdb4U77vA4SGIiPPIItG0LBQvC/PkweDDEhO2sSBERERGR\nsBNWRZO18O23MHKkm/3WubPHgaKiXAXXvz8sXQqXXeZxIBERERERyamwKpoWLHBT8qZPh9tvh0KF\nPAiRmgqvvQYHDrhpeN9/D2+84VEYERERkeCRkprGHyeSSU2zXkcBYOvWrRhjWLx4ccBfe8iQIdSv\nX/+8v48xhskn9/rMRHx8PMYY5syZc96vFcnCqmjavt0dJ0+GUaM8CLBlC7RpAw8+6DpRgJuWJyIi\nIhKhElNSmbZ0Bx1e/Z6aT31Fk6GzuXjQDDq++j3Tlu4gMSXVL69rjMn26w7P13GEjzlz5mT6Z7xu\n3brTrpsyZQp169Ylf/781K1bl2nTpp32+MiRIylbtixlypThlVdeOe2xpUuXUrt2bRISEvz+82Qm\nrBbXxMe7Y6tWkD9/AF/YWrcp1P33uyl5H34IPXsGMICIiIhI8Fm2/RB3vLeI5NQ0jiW54ig51Y0y\nrd9zlKemreLZz9fwwV3NaFipeJ6+9u+//55xe/r06dxzzz2nnStYsCAHDx7M1fdOTk4mnxp7nWX1\n6tWULFky437p0qUzbs+fP5/bbruNZ599lltvvZWpU6fStWtX5s2bR/PmzVmxYgWDBw9m+vTpWGu5\n8cYb6dChAw0aNCA1NZV77rmHN954g4IeDUiE1UjTyaLpwgsD/MIjRsBdd8Gll8KKFa61uFqJi4iI\nSARbvv0Qf317AYcSkjMKpjMdS0rlUEIy3d9ewPLth/L09cuVK5fxVbx48bPOXXDBBRnXbtu2jWuu\nuYZChQpRt25dZs+enfHYyVGUGTNm0KxZM2JjY5k5cyYAX3zxBU2aNKFAgQJUq1aNQYMGkZSUlPHc\nqVOnEhcXR8GCBSlZsiRXXXUVe/bsOS3nxIkTqVGjBkWLFuWWW24h/uQHWiAtLY2hQ4dSqVIl8ufP\nT4MGDfjss8+y/bl//vnnjEyNGzdm4cKFuf9DzKEyZcqc9mccHR2d8dioUaO4+uqrGTRoEHXq1GHQ\noEG0adOGUenTw9atW0dcXBxt27alXbt2xMXFZYxUjRo1ivr169O+ffuA/SxnCquiad8+t19swAr/\nxER37N0bXn3VdaGoUiVALy4iIiISnBJTUun93iISkn2bepeQ7K7311S9cxk0aBD3338/y5cv57LL\nLqN79+4cPXr0tGueeOIJnn/+edatW0fz5s2ZOXMmt99+OwMGDGD16tW89957TJ48mX/84x8A7N69\nm+7du9O7d2/Wrl3LDz/8QK8zNhDdunUrkyZNYtq0acyaNYulS5cyaNCgjMdfe+01Xn75ZUaMGMHK\nlSvp3Lkzt956K8uWLcv05zh27Bg33HAD1atXZ/Hixbz44os8+uij5/z5hw0bRpEiRbL9mjt37jm/\nT9OmTSlfvjzt2rXju+++O+2x+fPn06FDh9POdezYkZ9++gmABg0asGHDBn777Te2bdvGhg0bqF+/\nPlu3bmX06NFnTdcLtLCbnnfKKKD/HDsGjz4K69bBN99A+fJuHZOIiIiIMGPl7ySnpuXoOcmpaXy1\ncje3NK7op1RZe+ihh7jpppsAV0B8+OGHLFu2jCuuuCLjmiFDhpz2of+FF17gscce48477wSgRo0a\njBgxgp49e/Lyyy+za9cukpOT6dKlC1XSf6l+ZuOHlJQUxo0blzHq1bdvX95///2Mx0eOHMmjjz5K\njx49AHjuuef44YcfGDlyJOPHjz/r5/j4449JSkri/fffp0iRItSvX59BgwadVaydqV+/fnTr1i3b\naypWzPrvpXz58owZM4bLLruMpKQkPvroI9q1a8ecOXNo3bo14IrIsmXLnva8smXLsnv3bgDq1KnD\nsGHDuOaaawAYPnw4derU4dprr+WFF15g7ty5DB48mLS0NJ5//nluueWWbPPmtbAqmvbtg1Kl/Pwi\nCxe69UqbNrnCKTXVrWMSEREREQDGzNmU5ZS8rBxLSmXMnI2eFE1xcXEZtytUqADA3r17T7umadOm\np91fsmQJixYtYsSIERnn0tLSSEhIYPfu3TRs2JD27dtTv359OnToQPv27enSpctp63yqVKly2jTB\nChUqZLzuH3/8wa5du2jVqtVpr3vFFVcwY8aMTH+OtWvXEhcXR5EiRTLOtWjR4pw/f8mSJU9bi5RT\ntWvXpnbt2qe95tatWxk5cmRG0QSuOceprLWnnevXrx/9+vXLuH+yMGzfvj21atVi/vz5pKWl0apV\nKzZs2ECZMmVynTmnwurTvl9HmpKT4ZlnXJeJpCT47jt46aUAzgUUERERCX6paZZf9x4994WZ2LD3\nqCftyE9t6nDyQ3xa2ukjZYULFz7tflpaGs888wzLli3L+FqxYgW//vorpUuXJjo6mlmzZjFr1izi\n4uJ49913qVmzJsuXL8/0dU++9pmve2ahkdU5cEVIbuTV9LxTNW/enF9//TXjfrly5TJGlU7au3fv\nWaNPJ+3fv5+nn36aN998kwULFlCzZk3q1KlDvXr1qFmzZkDXakEYjjQ1aeKnb37ihOuKd/vt8Prr\ncMpvBURERETEOZaUQkyUyeiSlxMxUYZjSSkUKxD8v5S+9NJLWbduHRdffHGW1xhjaNGiBS1atGDw\n4MHUq1ePSZMm0bBhw3N+/2LFilGhQgV+/PFH2rZtm3H+xx9/pG7dupk+p27dunzwwQccO3Yso8hb\nsGDBOV/rfKfnZWbZsmWUL18+436LFi2YPXs2jz32WMa52bNn07Jly0yf//DDDzNw4ECqVq3KsmXL\nSE5OzngsKSmJ1NTArn8Lm6Jp6lRXNOXpSJO18NFH0K0bFC0KS5bAeQxdioiIiIS7wrExpORytCgl\nzVI4NjQ+ng4ePJgbb7yRKlWq0K1bN2JiYli1ahWLFi3ipZdeYsGCBXz99dd07NiRsmXLsnTpUrZv\n355lwZOZxx57jMGDB1OzZk2aNGnC+PHjmTt3LkuWLMn0+h49ejBo0CDuuusuBg8ezK5du3jhhRfO\n+TrnOz1v1KhRVK1alXr16pGUlMT48eP59NNPmTJlSsY1DzzwAK1bt2b48OF07tyZadOm8d133/Hj\njz+e9f2+/vpr1qxZw3vvvQfAZZddxvr16/niiy9IS0tj/fr1NGvWLNd5cyM03pXnkJICf/mLu12v\nXh5901274M47YdYsN8rUt68KJhEREZFziI4y1CxThA17cj5Fr1aZIkRHhca2LR07duTLL79k6NCh\njBw5kpiYGGrVqpWxae4FF1zAvHnz+Ne//sWhQ4eoVKkSTz/9ND1zsJfn/fffz5EjR3j88cfZs2cP\ntWvXZsqUKTRq1CjT64sUKcL06dO57777uPTSS7nkkksYMWIEN998c178yFlKSkri0UcfZefOnRQs\nWJB69erx5Zdfcv3112dc07JlSyZOnMhTTz3FM888Q40aNZg0aRLNmzc/7XslJCTwt7/9jYkTJ2a0\nLK9YsSJvvvkm/fr1w1rLW2+9lbH2LFBMbuc+BpOGDZvaFSsW88ILkN7l8fz897/Qrx8kJMArr7jb\n2ndJREREItTatWupU6eOz9dPW7qDp6atylEziMKx0bzQuYEnjSAkfGT3XjXGLLHWNs30wXMIq0YQ\nJUrkwTcZOtRNx6tRA5YuhfvuU8EkIiIikgPXNyhPvuicfczMFx3FdQ3K+SmRyPkJq6LpvDp/n+xU\n0qULDBkC8+bBKa0TRURERMQ3+WOi+eCuZhTMF+3T9QXzuevzx/h2vUighUXRlJjojlWr5uLJJ064\n/ZZ693b369RxrcXVSlxEREQk1xpWKs7EvpdTvGA+CsdmXgwVjo2meMF8TOx7OQ0rFQ9wQhHfhUUj\niIQEd8zBVFtnxQq3Ue3KlW4aXkoKxITFH4mIiIiI5xpWKs7CQe34auVuxszZyIa9R4mJMqSkWWqV\nKcp9bWpwXYNyGmGSoBcWFcJvv0HBglCpko9PSE2Ff/4TnnrKdcSbMQOuu86vGUVEREQiUf6YaG5p\nXJFbGlckNc1yLCmFwrExIdMlTwTCZHoeuBl2PvdriI+HF1+EG290o0wqmERERESylRcdl6OjDMUK\n5FPBJH7hz67gYVM0Pf74OS6wFr780jV8KFvWdcabPBlKlQpIPhEREZFQlS9fPhJOrocQCVIJCQnk\n81NfgrApmvLnz+bB+HjXFe/GG2HiRHeucmW1EhcRERHxQZkyZdi5cyfHjx/362/zRXLDWsvx48fZ\nuXMnZcqU8ctrhMWaJsimf8OMGXDXXXDwILz8Mtx2W0BziYiIiIS6YsWKAbBr1y6Sk5M9TiNytnz5\n8lG2bNmM92peC4uiyZgsBo2ee861D2/QAGbNgri4gGcTERERCQfFihXz2wdSkWAXFtPzspxl16YN\nPPIILFqkgklERERERHIl4EWTMeZaY8x6Y8xGY8yTmTye3xgzKf3xhcaYquf6nlEnf4rkZBgyBP7x\nD3e/dWsYORIKFMi7H0BERERERCJKQIsmY0w08AZwHVAX+Ksxpu4Zl/UBDlprLwZeBUac+/sC69dD\nq1bw7LOwa5frliciIiIiInKeAj3S1AzYaK3dbK1NAiYCnc64phPwQfrtyUA7Y7Jvc1fK7oPGjWHT\nJvjPf2DcOHXGExERERGRPBHooqkisP2U+zvSz2V6jbU2BTgMXJjdNy2fst1NxVu5Erp2zcO4IiIi\nIiIS6QLdPS+z4Z8z59H5cg3GmL5A3/S7iWbmzFVUPLP+EvGLUkC81yEkoug9J4Gk95sEkt5vEki1\nc/vEQBdNO4BKp9y/CNiVxTU7jDExwAXAgTO/kbX2beBtAGPMYmttU78kFjmD3m8SaHrPSSDp/SaB\npPebBJK58USLAAAMtUlEQVQxZnFunxvo6Xk/AzWNMdWMMbFAd+DzM675HOidfrsL8K3V1tMiIiIi\nIuKRgI40WWtTjDEDgJlANPCetXa1MeY5YLG19nPgXeAjY8xG3AhT90BmFBEREREROVWgp+dhrZ0B\nzDjj3OBTbp8ActrN4e08iCbiK73fJND0npNA0vtNAknvNwmkXL/fjGa+iYiIiIiIZC3Qa5pERERE\nRERCSkgVTcaYa40x640xG40xT2byeH5jzKT0xxcaY6oGPqWECx/ebw8bY9YYY1YYY74xxlTxIqeE\nh3O93065rosxxhpj1G1Kcs2X95sxplv6f+NWG2MmBDqjhBcf/p9a2RjznTFmafr/V6/3IqeEPmPM\ne8aYveb/27v3YLvGM47j358EUXfCaN3ChGpkFEVFta6joSqNy5RK3BLaISUodafoqLqru6FJqHtd\nztCgrjUqIa1b3DoRqfslIdFQIenTP953s7Jyzj4rJycns/f5fWb27L3Xevdaz177nX32c95nvUua\n2MZ6Sbo498XnJG1aZbsNkzRJ6gFcCuwM9AP2kdSv1GwY8FFE9AUuAM7u2iitWVTsb08Dm0XERsBt\nwO+7NkprFhX7G5KWBQ4HxndthNZMqvQ3SesBxwPfi4gNgZFdHqg1jYrfcScBt0TEJqRJwC7r2iit\niYwCBtZZvzOwXr4dAlxeZaMNkzQBWwCTImJyRHwO3AQMKrUZBIzOj28DdpDU2sVyzdrTbn+LiIcj\n4tP8dBzpumNmHVHl+w3gDFJy/llXBmdNp0p/Oxi4NCI+AoiI97s4RmsuVfpcAMvlx8sz73U8zSqJ\niL/RyjVeCwYBYyIZB6wg6evtbbeRkqbVgTcKz9/My1ptExGzgRnAyl0SnTWbKv2taBgwdqFGZM2s\n3f4maRNgzYi4uysDs6ZU5fttfWB9SY9LGiep3n9tzdpTpc+dBgyR9CZpluVfdk1o1g3N7288YBFM\nOb4AWhsxKk/9V6WNWRWV+5KkIcBmwDYLNSJrZnX7m6TFSCXHB3RVQNbUqny/9SSVrmxLGkV/TFL/\niJi+kGOz5lSlz+0DjIqI8yQNIF2zs39E/G/hh2fdTIfyhUYaaXoTWLPwfA3mHbr9so2knqTh3XrD\nc2ZtqdLfkLQjcCKwW0TM6qLYrPm019+WBfoDj0iaAmwJtHgyCOugqn9P74qILyLiNeAVUhJl1hFV\n+tww4BaAiHgC6AX07pLorLup9BuvrJGSpqeA9SStI2kJ0kmCLaU2LcD++fGewEPhC1FZx7Tb33K5\n1JWkhMn1/rYg6va3iJgREb0jok9E9CGdQ7dbRExYNOFag6vy9/ROYDsASb1J5XqTuzRKayZV+tzr\nwA4Akr5FSpo+6NIorbtoAfbLs+htCcyIiHfae1HDlOdFxGxJI4D7gB7AtRHxgqTTgQkR0QJcQxrO\nnUQaYdp70UVsjaxifzsHWAa4Nc838npE7LbIgraGVbG/mXWKiv3tPmAnSS8Cc4BjImLaoovaGlnF\nPnc0cLWkI0mlUgf4H9/WEZJuJJUW987nyJ0KLA4QEVeQzpnbBZgEfAocWGm77o9mZmZmZmZta6Ty\nPDMzMzMzsy7npMnMzMzMzKwOJ01mZmZmZmZ1OGkyMzMzMzOrw0mTmZmZmZlZHU6azMwaiKQDJEUb\ntx3nc1vD8+vWWFjxlvZ3ZinejySNl9Tpl4eQ1DPv46TCst0ljWyl7Y657dadHUed+PqWjsUcSe9I\nuk7S6h3c5qaSTpO0QmfHa2bW3TXMdZrMzGwue5Gual704qIIpAMG5PuVgZ8DN0paIiLGdNYO8nVh\nBgBvFBbvDmwNXFhq/mSO6YXO2v98OBO4B1gyx3AKsIGkARExez63tSnpeiSjgOmdGaSZWXfnpMnM\nrDE9ExGTFnUQHRER42qPJd0PvAKMBDotaSrvp512HwOV2i4ErxbifFTSksBpwMbAhEUUk5mZlbg8\nz8ysyUhaStJFkl6Q9Eku+2qR9M0Krx0q6Zn8uhmSnpM0vNRmO0kPSZqZb2Ml9etIrBHxBfAM0Lew\n/eUlXZbj/lzSK5KOKMWwnKRLJL0haZak9yT9VdL6ef1c5XmSrgf2BdYulMRNyuvmKs+TdJWktyX1\nKO2zVz4m5xaWrSrpytz+c0kvSRrWkWOR/TPfr1Xa95mSnpb0saSpkh6UtEVh/XDg6vz0tcJ7XKNw\nPE7Mx3KWpLcknZOTNDMza4dHmszMGlMPScXv8IiIOfnxUvl2OvAuqQzuMOAJSRtExPutbVDSNsBo\nUvna0UAPoB+wYqHNIODPQAvwM9I/344DHpO0UUS81YH3sg65nCwnKmOBjYCTSSVzuwEXSlo5Ik7J\nr7kIGAicCEwCepNK75ZvYx+n5jbfBgbnZZ+10XYMcDCwA3B/YfkgYDnguhzrCsDjwOKksropwC7A\n1bnc8PJK735uffL9q6Xl3wDOI5VkLgPsTzrmm0bEC8BdwLrA8aQyxHfy62qf9Y3AzsDvSKNqG5L6\nx1rATzsQp5lZt+KkycysMb1cev44KWkgIj4EDqmtyInIfcAHpB/If2hjmwOAqRFxVGHZl0mDJJGS\nlQcjYvfC8keAycCRwK/aC7yQ7K0MjCCVop2Xl/04xzE0Iq6vxSBpGeBYSRfm9zcAuC4iri1s+va2\n9hkRr0qaCsyqULb3OPAaMJS5k6ahwPMR8Wx+fiSwBtA/ImpJzgOSVgROk3RVIZFty2L5eCwJbElK\nem4u7KMW/0G1x4XP82XgIODoiPhA0uTc5OmImFJovx2wJ7BvRNxQiHM6MErSmRHxfDtxmpl1ay7P\nMzNrTIOBzQu3uUrCJO0t6UlJM4DZwEzS6FO9Er2ngFUkjZH0I0nlUZsNgLWBP+Vyr575B/9MYDzw\ng4qxf5Fv7wLHAOeTRozI25gN3FR6zfWkxOK7hViHSTpO0nckddrfs4iIvL/BkpYGkLQK8EPmPu9q\nIPB34N+l43EfsCr1j3XNNaRjMRN4gDSStH+5kaSdJD0iaRrp+HxOGlmqso+BpFG1O0px1hLC71fY\nhplZt+akycysMU2MiAmF2yu1FZIGk8qxJgL7kBKNzYEPgV5tbTAiHiSNRPUB7gSmSrpfUv/cZNV8\nP5qvEp/abSBp5KiKWqLXF1g2Io6OiFl53Uqk0a7yzHHvFtYDHEo6h+dg0oQJ70s6T9JSFWNozxhg\naVKpG6TjuBhwQ6HNqsD2zHssbszrqxyP35COxbbA5fnxXCOBkjYnzbA3gzSytGVuN5E6n2cpzl7A\np6U4356POM3MujWX55mZNZ+9gZdLJV29gHav3xMRtwC35HK47YGzgbGS1gKm5WbHAg+38vJZrSxr\nbR/1ZoX7EOgtqWcpcVot30/L2/gP6Vyq4yT1IU3BfhZpROVEFlBETJI0DhhCOodpCKks8e1Cs2mk\nKc2PamUTkGYFbM+UwvF4VNJywHBJV0REbVKIPUnva4/iMZG0EvBehX1MIyVM27Sx/u02lpuZWeak\nycys+XyNVMJVtB/zUV0QETOBFkl9SecbrUi6DtQbQL+IOKeTYi17lHSu0B7AzYXl+5ISh/GtxDoF\nOEfSUKB/eX3BLFKJYlXXARfnc4I2J53TVHQv6TpTUyJi6nxst55fk977qaSJJ+CrzzNqjSTtRJoc\n4qXCa2tJa/k93kua2GPpiHi0k+I0M+tWnDSZmTWfe4FL8tTYY0k/+A8DPq73Ikm/JZVqPUyafW0t\n0kQNE/LkC0gaAdyeR65uJY1irAZsBUyOiIsWMPa7gSdIM9CtRkoKdgUOAM6IiI9yHONJEz9MBD4B\ntiPNCHdlnW2/CBwk6RDgaeC/ETGxTvubgAtIydMnwB2l9eeSRrgek3QB8C9gWdK5X1tFxGDmU0S8\nJekKYKSkjSPiGdLnOQL4o6TRefsnMe8IUe3ixiPyFOtfAM9GxAOSbiWd03Q+6WK+kMowdyFNJFGe\nrc/MzAqcNJmZNZ8rgNVJicahpNGZXUnnxdQznvTjfBDp3KH3ST/YT641iIiWPDX5CaRJDJYiJVjj\n+Opcng6LiDmSdiaV2h1HSuJeA0aWErLHSGWI65KmRp8MHBERl9bZ/FXAFqSSwxVI03r3batxRHwo\n6S/AT0gz9X1SWj9d0gDSqNAJpJGf6aRZ7W6t/KbndRbpXK2TSSV590g6knQB4L2A50nTvZ9Riucf\nks4AhgO/II0srkmaXGIf4HDgQFLC9RlpivTarIpmZlaH0iRBZmZmZmZm1hrPnmdmZmZmZlaHkyYz\nMzMzM7M6nDSZmZmZmZnV4aTJzMzMzMysDidNZmZmZmZmdThpMjMzMzMzq8NJk5mZmZmZWR1OmszM\nzMzMzOpw0mRmZmZmZlbH/wFifRFuJSCPNQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11b3b7b70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fpr, tpr, threshold = roc_curve(y_test, probas)\n",
    "roc_auc = auc(fpr, tpr)\n",
    "\n",
    "fix, ax = plt.subplots(figsize=(14,10))\n",
    "\n",
    "plt.title('Receiver Operating Characteristic', fontsize=20)\n",
    "plt.plot(fpr, tpr, 'b', label = 'AUC = %0.2f' % roc_auc)\n",
    "plt.scatter(fpr[150], tpr[150], s=200, label='Threshold = 50%')\n",
    "plt.legend(loc = 'lower right', fontsize=14)\n",
    "plt.plot([0, 1], [0, 1],'r--')\n",
    "plt.xlim([0, 1])\n",
    "plt.ylim([0, 1])\n",
    "plt.ylabel('True Positive Rate', fontsize=16)\n",
    "plt.xlabel('False Positive Rate', fontsize=16)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA00AAAJnCAYAAABLQ7jIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x118544630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fpr, tpr, threshold = roc_curve(y_test, probas)\n",
    "roc_auc = auc(fpr, tpr)\n",
    "\n",
    "fix, ax = plt.subplots(figsize=(14,10))\n",
    "\n",
    "plt.title('Receiver Operating Characteristic', fontsize=20)\n",
    "plt.plot(fpr, tpr, 'b', label = 'AUC = %0.2f' % roc_auc)\n",
    "plt.scatter(fpr[350], tpr[350], s=200, label='Threshold = 25%')\n",
    "plt.legend(loc = 'lower right', fontsize=14)\n",
    "plt.plot([0, 1], [0, 1],'r--')\n",
    "plt.xlim([0, 1])\n",
    "plt.ylim([0, 1])\n",
    "plt.ylabel('True Positive Rate', fontsize=16)\n",
    "plt.xlabel('False Positive Rate', fontsize=16)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA00AAAJnCAYAAABLQ7jIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x132e9d518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fpr, tpr, threshold = roc_curve(y_test, probas)\n",
    "roc_auc = auc(fpr, tpr)\n",
    "\n",
    "fix, ax = plt.subplots(figsize=(14,10))\n",
    "\n",
    "plt.title('Receiver Operating Characteristic', fontsize=20)\n",
    "plt.plot(fpr, tpr, 'b', label = 'AUC = %0.2f' % roc_auc)\n",
    "plt.scatter(fpr[655], tpr[655], s=200, label='Threshold = 5%')\n",
    "plt.legend(loc = 'lower right', fontsize=14)\n",
    "plt.plot([0, 1], [0, 1],'r--')\n",
    "plt.xlim([0, 1])\n",
    "plt.ylim([0, 1])\n",
    "plt.ylabel('True Positive Rate', fontsize=16)\n",
    "plt.xlabel('False Positive Rate', fontsize=16)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.25003712822469315"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "threshold[350]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0502421412539015"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "threshold[655]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Kaggle Submission"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Xs_holdout = ss.transform(X_holdout)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "proba_pairs = rt_rand.predict_proba(Xs_holdout)\n",
    "probas = [item[1] for item in proba_pairs]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 361,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "submission = pd.DataFrame(probas)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 271,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "submission.index += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 363,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Manually entire 'Id', 'WnvPresent'\n",
    "submission.to_csv('./kaggle_submission_1.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Xs_holdout = ss.transform(X_holdout)\n",
    "proba_pairs = rt_rand.predict_proba(Xs_holdout)\n",
    "probas = [item[1] for item in proba_pairs]\n",
    "submission = pd.DataFrame(probas)\n",
    "submission.index += 1\n",
    "submission.to_csv('./kaggle_submission_2.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Balance Classes by: Changing Threshold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def threshold_predict(probas, threshold):\n",
    "    return probas[:,1] > threshold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "probas = rt_rand.predict_proba(Xs_test)\n",
    "\n",
    "new_preds = threshold_predict(probas, 0.05)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.99      0.28      0.43      2959\n",
      "          1       0.07      0.96      0.13       165\n",
      "\n",
      "avg / total       0.94      0.31      0.42      3124\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Pred -</th>\n",
       "      <th>Pred +</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Act -</th>\n",
       "      <td>817</td>\n",
       "      <td>2142</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Act +</th>\n",
       "      <td>6</td>\n",
       "      <td>159</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Pred -  Pred +\n",
       "Act -     817    2142\n",
       "Act +       6     159"
      ]
     },
     "execution_count": 159,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Way too many false positives!\n",
    "\n",
    "print(classification_report(y_test, new_preds))\n",
    "pd.DataFrame(confusion_matrix(y_test, new_preds), columns=['Pred -', 'Pred +'], index=['Act -', 'Act +'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# -------------------THE GRAVEYARD OF FAILED IDEAS -----------------------------"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Balance Classes by: Oversampling (ros)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "ros = RandomOverSampler(random_state=42)\n",
    "X_over, y_over = ros.fit_sample(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10413, 293)\n",
      "(10413,)\n"
     ]
    }
   ],
   "source": [
    "print(X.shape)\n",
    "print(y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(19724, 293)\n",
      "(19724,)\n"
     ]
    }
   ],
   "source": [
    "print(X_over.shape)\n",
    "print(y_over.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(X_over, y_over, test_size=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "ss = StandardScaler()\n",
    "Xs_train = ss.fit_transform(X_train)\n",
    "Xs_test = ss.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=10, n_jobs=1,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False)"
      ]
     },
     "execution_count": 126,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rf = RandomForestClassifier()\n",
    "rf.fit(Xs_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "             precision    recall  f1-score   support\n",
      "\n",
      "          0       0.95      1.00      0.97      9862\n",
      "          1       0.00      0.00      0.00       551\n",
      "\n",
      "avg / total       0.90      0.95      0.92     10413\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.6/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Pred -</th>\n",
       "      <th>Pred +</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Act -</th>\n",
       "      <td>9862</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Act +</th>\n",
       "      <td>551</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Pred -  Pred +\n",
       "Act -    9862       0\n",
       "Act +     551       0"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_preds = rf.predict(X)\n",
    "print(classification_report(y, y_preds))\n",
    "pd.DataFrame(confusion_matrix(y, y_preds), columns=['Pred -', 'Pred +'], index=['Act -', 'Act +'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "# Getting some Water Information (This takes too long too)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "water = pd.read_csv('./Water_Locations.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Convert a polygon made of many lat/longs to one average lat/long point\n",
    "\n",
    "def shape_to_point(shape_string):\n",
    "    coord_list = shape_string.split(' ')[3:-1]\n",
    "    lats = coord_list[1::2]\n",
    "    longs = coord_list[0::2]\n",
    "    \n",
    "    clean_lat = []\n",
    "    clean_long = []\n",
    "    \n",
    "    for i in range(len(lats)):\n",
    "        clean_lat.append( float(lats[i].strip('(').strip(',').strip(')')) )\n",
    "        clean_long.append( float(longs[i].strip('(').strip(',').strip(')')) )\n",
    "        \n",
    "    return(np.mean(clean_lat), np.mean(clean_long))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Apply function to each body of water\n",
    "\n",
    "latitudes = []\n",
    "longitudes = []\n",
    "\n",
    "for i in range(len(water)):\n",
    "    latitudes.append(shape_to_point(water['the_geom'][i])[0])\n",
    "    longitudes.append(shape_to_point(water['the_geom'][i])[1])\n",
    "    \n",
    "water['lat'] = latitudes\n",
    "water['long'] = longitudes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 rows parsed\n",
      "100 rows parsed\n",
      "200 rows parsed\n",
      "300 rows parsed\n",
      "400 rows parsed\n",
      "500 rows parsed\n",
      "600 rows parsed\n",
      "700 rows parsed\n",
      "800 rows parsed\n",
      "900 rows parsed\n",
      "1000 rows parsed\n",
      "1100 rows parsed\n",
      "1200 rows parsed\n",
      "1300 rows parsed\n",
      "1400 rows parsed\n",
      "1500 rows parsed\n",
      "1600 rows parsed\n",
      "1700 rows parsed\n",
      "1800 rows parsed\n",
      "1900 rows parsed\n",
      "2000 rows parsed\n",
      "2100 rows parsed\n",
      "2200 rows parsed\n",
      "2300 rows parsed\n",
      "2400 rows parsed\n",
      "2500 rows parsed\n",
      "2600 rows parsed\n",
      "2700 rows parsed\n",
      "2800 rows parsed\n",
      "2900 rows parsed\n",
      "3000 rows parsed\n",
      "3100 rows parsed\n",
      "3200 rows parsed\n",
      "3300 rows parsed\n",
      "3400 rows parsed\n",
      "3500 rows parsed\n",
      "3600 rows parsed\n",
      "3700 rows parsed\n",
      "3800 rows parsed\n",
      "3900 rows parsed\n",
      "4000 rows parsed\n",
      "4100 rows parsed\n",
      "4200 rows parsed\n",
      "4300 rows parsed\n",
      "4400 rows parsed\n",
      "4500 rows parsed\n",
      "4600 rows parsed\n",
      "4700 rows parsed\n",
      "4800 rows parsed\n",
      "4900 rows parsed\n",
      "5000 rows parsed\n",
      "5100 rows parsed\n",
      "5200 rows parsed\n",
      "5300 rows parsed\n",
      "5400 rows parsed\n",
      "5500 rows parsed\n",
      "5600 rows parsed\n",
      "5700 rows parsed\n",
      "5800 rows parsed\n",
      "5900 rows parsed\n",
      "6000 rows parsed\n",
      "6100 rows parsed\n",
      "6200 rows parsed\n",
      "6300 rows parsed\n",
      "6400 rows parsed\n",
      "6500 rows parsed\n",
      "6600 rows parsed\n",
      "6700 rows parsed\n",
      "6800 rows parsed\n",
      "6900 rows parsed\n",
      "7000 rows parsed\n",
      "7100 rows parsed\n",
      "7200 rows parsed\n",
      "7300 rows parsed\n",
      "7400 rows parsed\n",
      "7500 rows parsed\n",
      "7600 rows parsed\n",
      "7700 rows parsed\n",
      "7800 rows parsed\n",
      "7900 rows parsed\n",
      "8000 rows parsed\n",
      "8100 rows parsed\n",
      "8200 rows parsed\n",
      "8300 rows parsed\n",
      "8400 rows parsed\n",
      "8500 rows parsed\n",
      "8600 rows parsed\n",
      "8700 rows parsed\n",
      "8800 rows parsed\n",
      "8900 rows parsed\n",
      "9000 rows parsed\n",
      "9100 rows parsed\n",
      "9200 rows parsed\n",
      "9300 rows parsed\n",
      "9400 rows parsed\n",
      "9500 rows parsed\n",
      "9600 rows parsed\n",
      "9700 rows parsed\n",
      "9800 rows parsed\n",
      "9900 rows parsed\n",
      "10000 rows parsed\n",
      "10100 rows parsed\n",
      "10200 rows parsed\n",
      "10300 rows parsed\n",
      "10400 rows parsed\n",
      "10500 rows parsed\n"
     ]
    }
   ],
   "source": [
    "# For each observation in Training Set, find the distance to closest body of water\n",
    "\n",
    "\n",
    "min_dist = []\n",
    "\n",
    "for i in range(len(X_train)):\n",
    "    distances = []\n",
    "    for j in range(len(water)):\n",
    "        distances.append( vincenty((X_train['Latitude'][i], X_train['Longitude'][i]), (water['lat'][j], water['long'][j])).miles )\n",
    "    if i % 100 == 0:\n",
    "        print(i, 'rows parsed')\n",
    "    min_dist.append(min(distances))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Generate Features\n",
    "X_train['min_dist_water_miles'] = min_dist\n",
    "X_train['min_dist_water_feet'] = [i*5280 for i in X_train['min_dist_water_miles'] ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 rows parsed\n",
      "100 rows parsed\n",
      "200 rows parsed\n",
      "300 rows parsed\n",
      "400 rows parsed\n",
      "500 rows parsed\n",
      "600 rows parsed\n",
      "700 rows parsed\n",
      "800 rows parsed\n",
      "900 rows parsed\n",
      "1000 rows parsed\n",
      "1100 rows parsed\n",
      "1200 rows parsed\n",
      "1300 rows parsed\n",
      "1400 rows parsed\n",
      "1500 rows parsed\n",
      "1600 rows parsed\n",
      "1700 rows parsed\n",
      "1800 rows parsed\n",
      "1900 rows parsed\n",
      "2000 rows parsed\n",
      "2100 rows parsed\n",
      "2200 rows parsed\n",
      "2300 rows parsed\n",
      "2400 rows parsed\n",
      "2500 rows parsed\n",
      "2600 rows parsed\n",
      "2700 rows parsed\n",
      "2800 rows parsed\n",
      "2900 rows parsed\n",
      "3000 rows parsed\n",
      "3100 rows parsed\n",
      "3200 rows parsed\n",
      "3300 rows parsed\n",
      "3400 rows parsed\n",
      "3500 rows parsed\n",
      "3600 rows parsed\n",
      "3700 rows parsed\n",
      "3800 rows parsed\n",
      "3900 rows parsed\n",
      "4000 rows parsed\n",
      "4100 rows parsed\n",
      "4200 rows parsed\n",
      "4300 rows parsed\n",
      "4400 rows parsed\n",
      "4500 rows parsed\n",
      "4600 rows parsed\n",
      "4700 rows parsed\n",
      "4800 rows parsed\n",
      "4900 rows parsed\n",
      "5000 rows parsed\n",
      "5100 rows parsed\n",
      "5200 rows parsed\n",
      "5300 rows parsed\n",
      "5400 rows parsed\n",
      "5500 rows parsed\n",
      "5600 rows parsed\n",
      "5700 rows parsed\n",
      "5800 rows parsed\n",
      "5900 rows parsed\n",
      "6000 rows parsed\n",
      "6100 rows parsed\n",
      "6200 rows parsed\n",
      "6300 rows parsed\n",
      "6400 rows parsed\n",
      "6500 rows parsed\n",
      "6600 rows parsed\n",
      "6700 rows parsed\n",
      "6800 rows parsed\n",
      "6900 rows parsed\n",
      "7000 rows parsed\n",
      "7100 rows parsed\n",
      "7200 rows parsed\n",
      "7300 rows parsed\n",
      "7400 rows parsed\n",
      "7500 rows parsed\n",
      "7600 rows parsed\n",
      "7700 rows parsed\n",
      "7800 rows parsed\n",
      "7900 rows parsed\n",
      "8000 rows parsed\n",
      "8100 rows parsed\n",
      "8200 rows parsed\n",
      "8300 rows parsed\n",
      "8400 rows parsed\n",
      "8500 rows parsed\n",
      "8600 rows parsed\n",
      "8700 rows parsed\n",
      "8800 rows parsed\n",
      "8900 rows parsed\n",
      "9000 rows parsed\n",
      "9100 rows parsed\n",
      "9200 rows parsed\n",
      "9300 rows parsed\n",
      "9400 rows parsed\n",
      "9500 rows parsed\n",
      "9600 rows parsed\n",
      "9700 rows parsed\n",
      "9800 rows parsed\n",
      "9900 rows parsed\n",
      "10000 rows parsed\n",
      "10100 rows parsed\n",
      "10200 rows parsed\n",
      "10300 rows parsed\n",
      "10400 rows parsed\n",
      "10500 rows parsed\n",
      "10600 rows parsed\n",
      "10700 rows parsed\n",
      "10800 rows parsed\n",
      "10900 rows parsed\n",
      "11000 rows parsed\n",
      "11100 rows parsed\n",
      "11200 rows parsed\n",
      "11300 rows parsed\n",
      "11400 rows parsed\n",
      "11500 rows parsed\n",
      "11600 rows parsed\n",
      "11700 rows parsed\n",
      "11800 rows parsed\n",
      "11900 rows parsed\n",
      "12000 rows parsed\n",
      "12100 rows parsed\n",
      "12200 rows parsed\n",
      "12300 rows parsed\n",
      "12400 rows parsed\n",
      "12500 rows parsed\n",
      "12600 rows parsed\n",
      "12700 rows parsed\n",
      "12800 rows parsed\n",
      "12900 rows parsed\n",
      "13000 rows parsed\n",
      "13100 rows parsed\n",
      "13200 rows parsed\n",
      "13300 rows parsed\n",
      "13400 rows parsed\n",
      "13500 rows parsed\n",
      "13600 rows parsed\n",
      "13700 rows parsed\n",
      "13800 rows parsed\n",
      "13900 rows parsed\n",
      "14000 rows parsed\n",
      "14100 rows parsed\n",
      "14200 rows parsed\n",
      "14300 rows parsed\n",
      "14400 rows parsed\n",
      "14500 rows parsed\n",
      "14600 rows parsed\n",
      "14700 rows parsed\n",
      "14800 rows parsed\n",
      "14900 rows parsed\n",
      "15000 rows parsed\n",
      "15100 rows parsed\n",
      "15200 rows parsed\n",
      "15300 rows parsed\n",
      "15400 rows parsed\n",
      "15500 rows parsed\n",
      "15600 rows parsed\n",
      "15700 rows parsed\n",
      "15800 rows parsed\n",
      "15900 rows parsed\n",
      "16000 rows parsed\n",
      "16100 rows parsed\n",
      "16200 rows parsed\n",
      "16300 rows parsed\n",
      "16400 rows parsed\n",
      "16500 rows parsed\n",
      "16600 rows parsed\n",
      "16700 rows parsed\n",
      "16800 rows parsed\n",
      "16900 rows parsed\n",
      "17000 rows parsed\n",
      "17100 rows parsed\n",
      "17200 rows parsed\n",
      "17300 rows parsed\n",
      "17400 rows parsed\n",
      "17500 rows parsed\n",
      "17600 rows parsed\n",
      "17700 rows parsed\n",
      "17800 rows parsed\n",
      "17900 rows parsed\n",
      "18000 rows parsed\n",
      "18100 rows parsed\n",
      "18200 rows parsed\n",
      "18300 rows parsed\n",
      "18400 rows parsed\n",
      "18500 rows parsed\n",
      "18600 rows parsed\n",
      "18700 rows parsed\n",
      "18800 rows parsed\n",
      "18900 rows parsed\n",
      "19000 rows parsed\n",
      "19100 rows parsed\n",
      "19200 rows parsed\n",
      "19300 rows parsed\n",
      "19400 rows parsed\n",
      "19500 rows parsed\n",
      "19600 rows parsed\n",
      "19700 rows parsed\n",
      "19800 rows parsed\n",
      "19900 rows parsed\n",
      "20000 rows parsed\n",
      "20100 rows parsed\n",
      "20200 rows parsed\n",
      "20300 rows parsed\n",
      "20400 rows parsed\n",
      "20500 rows parsed\n",
      "20600 rows parsed\n",
      "20700 rows parsed\n",
      "20800 rows parsed\n",
      "20900 rows parsed\n",
      "21000 rows parsed\n",
      "21100 rows parsed\n",
      "21200 rows parsed\n",
      "21300 rows parsed\n",
      "21400 rows parsed\n",
      "21500 rows parsed\n",
      "21600 rows parsed\n",
      "21700 rows parsed\n",
      "21800 rows parsed\n",
      "21900 rows parsed\n",
      "22000 rows parsed\n",
      "22100 rows parsed\n",
      "22200 rows parsed\n",
      "22300 rows parsed\n",
      "22400 rows parsed\n",
      "22500 rows parsed\n",
      "22600 rows parsed\n",
      "22700 rows parsed\n",
      "22800 rows parsed\n",
      "22900 rows parsed\n",
      "23000 rows parsed\n",
      "23100 rows parsed\n",
      "23200 rows parsed\n",
      "23300 rows parsed\n",
      "23400 rows parsed\n",
      "23500 rows parsed\n",
      "23600 rows parsed\n",
      "23700 rows parsed\n",
      "23800 rows parsed\n",
      "23900 rows parsed\n",
      "24000 rows parsed\n",
      "24100 rows parsed\n",
      "24200 rows parsed\n",
      "24300 rows parsed\n",
      "24400 rows parsed\n",
      "24500 rows parsed\n",
      "24600 rows parsed\n",
      "24700 rows parsed\n",
      "24800 rows parsed\n",
      "24900 rows parsed\n",
      "25000 rows parsed\n",
      "25100 rows parsed\n",
      "25200 rows parsed\n",
      "25300 rows parsed\n",
      "25400 rows parsed\n",
      "25500 rows parsed\n",
      "25600 rows parsed\n",
      "25700 rows parsed\n",
      "25800 rows parsed\n",
      "25900 rows parsed\n",
      "26000 rows parsed\n",
      "26100 rows parsed\n",
      "26200 rows parsed\n",
      "26300 rows parsed\n",
      "26400 rows parsed\n",
      "26500 rows parsed\n",
      "26600 rows parsed\n",
      "26700 rows parsed\n",
      "26800 rows parsed\n",
      "26900 rows parsed\n",
      "27000 rows parsed\n",
      "27100 rows parsed\n",
      "27200 rows parsed\n",
      "27300 rows parsed\n",
      "27400 rows parsed\n",
      "27500 rows parsed\n",
      "27600 rows parsed\n",
      "27700 rows parsed\n",
      "27800 rows parsed\n",
      "27900 rows parsed\n",
      "28000 rows parsed\n",
      "28100 rows parsed\n",
      "28200 rows parsed\n",
      "28300 rows parsed\n",
      "28400 rows parsed\n",
      "28500 rows parsed\n",
      "28600 rows parsed\n",
      "28700 rows parsed\n",
      "28800 rows parsed\n",
      "28900 rows parsed\n",
      "29000 rows parsed\n",
      "29100 rows parsed\n",
      "29200 rows parsed\n",
      "29300 rows parsed\n",
      "29400 rows parsed\n",
      "29500 rows parsed\n",
      "29600 rows parsed\n",
      "29700 rows parsed\n",
      "29800 rows parsed\n",
      "29900 rows parsed\n",
      "30000 rows parsed\n",
      "30100 rows parsed\n",
      "30200 rows parsed\n",
      "30300 rows parsed\n",
      "30400 rows parsed\n",
      "30500 rows parsed\n",
      "30600 rows parsed\n",
      "30700 rows parsed\n",
      "30800 rows parsed\n",
      "30900 rows parsed\n",
      "31000 rows parsed\n",
      "31100 rows parsed\n",
      "31200 rows parsed\n",
      "31300 rows parsed\n",
      "31400 rows parsed\n",
      "31500 rows parsed\n",
      "31600 rows parsed\n",
      "31700 rows parsed\n",
      "31800 rows parsed\n",
      "31900 rows parsed\n",
      "32000 rows parsed\n",
      "32100 rows parsed\n",
      "32200 rows parsed\n",
      "32300 rows parsed\n",
      "32400 rows parsed\n",
      "32500 rows parsed\n",
      "32600 rows parsed\n",
      "32700 rows parsed\n",
      "32800 rows parsed\n",
      "32900 rows parsed\n",
      "33000 rows parsed\n",
      "33100 rows parsed\n",
      "33200 rows parsed\n",
      "33300 rows parsed\n",
      "33400 rows parsed\n",
      "33500 rows parsed\n",
      "33600 rows parsed\n",
      "33700 rows parsed\n",
      "33800 rows parsed\n",
      "33900 rows parsed\n",
      "34000 rows parsed\n",
      "34100 rows parsed\n",
      "34200 rows parsed\n",
      "34300 rows parsed\n",
      "34400 rows parsed\n",
      "34500 rows parsed\n",
      "34600 rows parsed\n",
      "34700 rows parsed\n",
      "34800 rows parsed\n",
      "34900 rows parsed\n",
      "35000 rows parsed\n",
      "35100 rows parsed\n",
      "35200 rows parsed\n",
      "35300 rows parsed\n",
      "35400 rows parsed\n",
      "35500 rows parsed\n",
      "35600 rows parsed\n",
      "35700 rows parsed\n",
      "35800 rows parsed\n",
      "35900 rows parsed\n",
      "36000 rows parsed\n",
      "36100 rows parsed\n",
      "36200 rows parsed\n",
      "36300 rows parsed\n",
      "36400 rows parsed\n",
      "36500 rows parsed\n",
      "36600 rows parsed\n",
      "36700 rows parsed\n",
      "36800 rows parsed\n",
      "36900 rows parsed\n",
      "37000 rows parsed\n",
      "37100 rows parsed\n",
      "37200 rows parsed\n",
      "37300 rows parsed\n",
      "37400 rows parsed\n",
      "37500 rows parsed\n",
      "37600 rows parsed\n",
      "37700 rows parsed\n",
      "37800 rows parsed\n",
      "37900 rows parsed\n",
      "38000 rows parsed\n",
      "38100 rows parsed\n",
      "38200 rows parsed\n",
      "38300 rows parsed\n",
      "38400 rows parsed\n",
      "38500 rows parsed\n",
      "38600 rows parsed\n",
      "38700 rows parsed\n",
      "38800 rows parsed\n",
      "38900 rows parsed\n",
      "39000 rows parsed\n",
      "39100 rows parsed\n",
      "39200 rows parsed\n",
      "39300 rows parsed\n",
      "39400 rows parsed\n",
      "39500 rows parsed\n",
      "39600 rows parsed\n",
      "39700 rows parsed\n",
      "39800 rows parsed\n",
      "39900 rows parsed\n",
      "40000 rows parsed\n",
      "40100 rows parsed\n",
      "40200 rows parsed\n",
      "40300 rows parsed\n",
      "40400 rows parsed\n",
      "40500 rows parsed\n",
      "40600 rows parsed\n",
      "40700 rows parsed\n",
      "40800 rows parsed\n",
      "40900 rows parsed\n",
      "41000 rows parsed\n",
      "41100 rows parsed\n",
      "41200 rows parsed\n",
      "41300 rows parsed\n",
      "41400 rows parsed\n",
      "41500 rows parsed\n",
      "41600 rows parsed\n",
      "41700 rows parsed\n",
      "41800 rows parsed\n",
      "41900 rows parsed\n",
      "42000 rows parsed\n",
      "42100 rows parsed\n",
      "42200 rows parsed\n",
      "42300 rows parsed\n",
      "42400 rows parsed\n",
      "42500 rows parsed\n",
      "42600 rows parsed\n",
      "42700 rows parsed\n",
      "42800 rows parsed\n",
      "42900 rows parsed\n",
      "43000 rows parsed\n",
      "43100 rows parsed\n",
      "43200 rows parsed\n",
      "43300 rows parsed\n",
      "43400 rows parsed\n",
      "43500 rows parsed\n",
      "43600 rows parsed\n",
      "43700 rows parsed\n",
      "43800 rows parsed\n",
      "43900 rows parsed\n",
      "44000 rows parsed\n",
      "44100 rows parsed\n",
      "44200 rows parsed\n",
      "44300 rows parsed\n",
      "44400 rows parsed\n",
      "44500 rows parsed\n",
      "44600 rows parsed\n",
      "44700 rows parsed\n",
      "44800 rows parsed\n",
      "44900 rows parsed\n",
      "45000 rows parsed\n",
      "45100 rows parsed\n",
      "45200 rows parsed\n",
      "45300 rows parsed\n",
      "45400 rows parsed\n",
      "45500 rows parsed\n",
      "45600 rows parsed\n",
      "45700 rows parsed\n",
      "45800 rows parsed\n",
      "45900 rows parsed\n",
      "46000 rows parsed\n",
      "46100 rows parsed\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "46200 rows parsed\n",
      "46300 rows parsed\n",
      "46400 rows parsed\n",
      "46500 rows parsed\n",
      "46600 rows parsed\n",
      "46700 rows parsed\n",
      "46800 rows parsed\n",
      "46900 rows parsed\n",
      "47000 rows parsed\n",
      "47100 rows parsed\n",
      "47200 rows parsed\n",
      "47300 rows parsed\n",
      "47400 rows parsed\n",
      "47500 rows parsed\n",
      "47600 rows parsed\n",
      "47700 rows parsed\n",
      "47800 rows parsed\n",
      "47900 rows parsed\n",
      "48000 rows parsed\n",
      "48100 rows parsed\n",
      "48200 rows parsed\n",
      "48300 rows parsed\n",
      "48400 rows parsed\n",
      "48500 rows parsed\n",
      "48600 rows parsed\n",
      "48700 rows parsed\n",
      "48800 rows parsed\n",
      "48900 rows parsed\n",
      "49000 rows parsed\n",
      "49100 rows parsed\n",
      "49200 rows parsed\n",
      "49300 rows parsed\n",
      "49400 rows parsed\n",
      "49500 rows parsed\n",
      "49600 rows parsed\n",
      "49700 rows parsed\n",
      "49800 rows parsed\n",
      "49900 rows parsed\n",
      "50000 rows parsed\n",
      "50100 rows parsed\n",
      "50200 rows parsed\n",
      "50300 rows parsed\n",
      "50400 rows parsed\n",
      "50500 rows parsed\n",
      "50600 rows parsed\n",
      "50700 rows parsed\n",
      "50800 rows parsed\n",
      "50900 rows parsed\n",
      "51000 rows parsed\n",
      "51100 rows parsed\n",
      "51200 rows parsed\n",
      "51300 rows parsed\n",
      "51400 rows parsed\n",
      "51500 rows parsed\n",
      "51600 rows parsed\n",
      "51700 rows parsed\n",
      "51800 rows parsed\n",
      "51900 rows parsed\n",
      "52000 rows parsed\n",
      "52100 rows parsed\n",
      "52200 rows parsed\n",
      "52300 rows parsed\n",
      "52400 rows parsed\n",
      "52500 rows parsed\n",
      "52600 rows parsed\n",
      "52700 rows parsed\n",
      "52800 rows parsed\n",
      "52900 rows parsed\n",
      "53000 rows parsed\n",
      "53100 rows parsed\n",
      "53200 rows parsed\n",
      "53300 rows parsed\n",
      "53400 rows parsed\n",
      "53500 rows parsed\n",
      "53600 rows parsed\n",
      "53700 rows parsed\n",
      "53800 rows parsed\n",
      "53900 rows parsed\n",
      "54000 rows parsed\n",
      "54100 rows parsed\n",
      "54200 rows parsed\n",
      "54300 rows parsed\n",
      "54400 rows parsed\n",
      "54500 rows parsed\n",
      "54600 rows parsed\n",
      "54700 rows parsed\n",
      "54800 rows parsed\n",
      "54900 rows parsed\n",
      "55000 rows parsed\n",
      "55100 rows parsed\n",
      "55200 rows parsed\n",
      "55300 rows parsed\n",
      "55400 rows parsed\n",
      "55500 rows parsed\n",
      "55600 rows parsed\n",
      "55700 rows parsed\n",
      "55800 rows parsed\n",
      "55900 rows parsed\n",
      "56000 rows parsed\n",
      "56100 rows parsed\n",
      "56200 rows parsed\n",
      "56300 rows parsed\n",
      "56400 rows parsed\n",
      "56500 rows parsed\n",
      "56600 rows parsed\n",
      "56700 rows parsed\n",
      "56800 rows parsed\n",
      "56900 rows parsed\n",
      "57000 rows parsed\n",
      "57100 rows parsed\n",
      "57200 rows parsed\n",
      "57300 rows parsed\n",
      "57400 rows parsed\n",
      "57500 rows parsed\n",
      "57600 rows parsed\n",
      "57700 rows parsed\n",
      "57800 rows parsed\n",
      "57900 rows parsed\n",
      "58000 rows parsed\n",
      "58100 rows parsed\n",
      "58200 rows parsed\n",
      "58300 rows parsed\n",
      "58400 rows parsed\n",
      "58500 rows parsed\n",
      "58600 rows parsed\n",
      "58700 rows parsed\n",
      "58800 rows parsed\n",
      "58900 rows parsed\n",
      "59000 rows parsed\n",
      "59100 rows parsed\n",
      "59200 rows parsed\n",
      "59300 rows parsed\n",
      "59400 rows parsed\n",
      "59500 rows parsed\n",
      "59600 rows parsed\n",
      "59700 rows parsed\n",
      "59800 rows parsed\n",
      "59900 rows parsed\n",
      "60000 rows parsed\n",
      "60100 rows parsed\n",
      "60200 rows parsed\n",
      "60300 rows parsed\n",
      "60400 rows parsed\n",
      "60500 rows parsed\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-38-e26cae7aa63a>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      6\u001b[0m     \u001b[0mdistances\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      7\u001b[0m     \u001b[0;32mfor\u001b[0m \u001b[0mj\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mwater\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 8\u001b[0;31m         \u001b[0mdistances\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m \u001b[0mvincenty\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX_test\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'Latitude'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX_test\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'Longitude'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mwater\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'lat'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mj\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwater\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'long'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mj\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmiles\u001b[0m \u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      9\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mi\u001b[0m \u001b[0;34m%\u001b[0m \u001b[0;36m100\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     10\u001b[0m         \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'rows parsed'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "# For each observation in Testing Set, find the distance to closest body of water\n",
    "\n",
    "min_dist = []\n",
    "\n",
    "for i in range(len(X_test)):\n",
    "    distances = []\n",
    "    for j in range(len(water)):\n",
    "        distances.append( vincenty((X_test['Latitude'][i], X_test['Longitude'][i]), (water['lat'][j], water['long'][j])).miles )\n",
    "    if i % 100 == 0:\n",
    "        print(i, 'rows parsed')\n",
    "    min_dist.append(min(distances))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Generate Features\n",
    "X_test['min_dist_water_miles'] = min_dist\n",
    "X_test['min_dist_water_feet'] = [i*5280 for i in X_test['min_dist_water_miles'] ]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "---\n",
    "# Getting some Park Information (This Section on Hold, takes a lot of time)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "parks = pd.read_csv('./Park_Locations.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "park_feats = ['PARK NUMBER', 'PARK NAME', 'STREET ADDRESS', 'ZIP', 'ACRES', 'WARD', 'PARK CLASS', 'LABEL', 'LOCATION']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "parks = parks[park_feats]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PARK NUMBER</th>\n",
       "      <th>PARK NAME</th>\n",
       "      <th>STREET ADDRESS</th>\n",
       "      <th>ZIP</th>\n",
       "      <th>ACRES</th>\n",
       "      <th>WARD</th>\n",
       "      <th>PARK CLASS</th>\n",
       "      <th>LABEL</th>\n",
       "      <th>LOCATION</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>288</td>\n",
       "      <td>GOLDEN GATE</td>\n",
       "      <td>500 E 130TH ST</td>\n",
       "      <td>60627.0</td>\n",
       "      <td>5.20</td>\n",
       "      <td>9</td>\n",
       "      <td>COMMUNITY PARK</td>\n",
       "      <td>Golden Gate</td>\n",
       "      <td>500 E 130TH ST\\n60627\\n</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>548</td>\n",
       "      <td>MAJOR TAYLOR BIKE TRAIL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>31.24</td>\n",
       "      <td>0</td>\n",
       "      <td>COMMUNITY PARK</td>\n",
       "      <td>Major Taylor Bike Trail</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>255</td>\n",
       "      <td>CARVER (GEORGE WASHINGTON)</td>\n",
       "      <td>939 E 132ND ST</td>\n",
       "      <td>60627.0</td>\n",
       "      <td>19.42</td>\n",
       "      <td>9</td>\n",
       "      <td>REGIONAL PARK</td>\n",
       "      <td>Carver</td>\n",
       "      <td>939 E 132ND ST\\n60627\\n</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>127</td>\n",
       "      <td>RUTHERFORD SAYRE</td>\n",
       "      <td>6871 W BELDEN AVE</td>\n",
       "      <td>60635.0</td>\n",
       "      <td>12.33</td>\n",
       "      <td>36</td>\n",
       "      <td>COMMUNITY PARK</td>\n",
       "      <td>Rutherford Sayre</td>\n",
       "      <td>6871 W BELDEN AVE\\n60635\\n</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1256</td>\n",
       "      <td>HOWARD (URE) BEACH</td>\n",
       "      <td>1300 W HOWARD ST</td>\n",
       "      <td>60626.0</td>\n",
       "      <td>0.89</td>\n",
       "      <td>49</td>\n",
       "      <td>NEIGHBORHOOD PARK</td>\n",
       "      <td>Howard Beach</td>\n",
       "      <td>1300 W HOWARD ST\\n60626\\n(42.01933233897205, -...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PARK NUMBER                   PARK NAME     STREET ADDRESS      ZIP  ACRES  \\\n",
       "0          288                 GOLDEN GATE     500 E 130TH ST  60627.0   5.20   \n",
       "1          548     MAJOR TAYLOR BIKE TRAIL                NaN      NaN  31.24   \n",
       "2          255  CARVER (GEORGE WASHINGTON)     939 E 132ND ST  60627.0  19.42   \n",
       "3          127            RUTHERFORD SAYRE  6871 W BELDEN AVE  60635.0  12.33   \n",
       "4         1256          HOWARD (URE) BEACH   1300 W HOWARD ST  60626.0   0.89   \n",
       "\n",
       "   WARD         PARK CLASS                    LABEL  \\\n",
       "0     9     COMMUNITY PARK              Golden Gate   \n",
       "1     0     COMMUNITY PARK  Major Taylor Bike Trail   \n",
       "2     9      REGIONAL PARK                   Carver   \n",
       "3    36     COMMUNITY PARK         Rutherford Sayre   \n",
       "4    49  NEIGHBORHOOD PARK             Howard Beach   \n",
       "\n",
       "                                            LOCATION  \n",
       "0                            500 E 130TH ST\\n60627\\n  \n",
       "1                                                NaN  \n",
       "2                            939 E 132ND ST\\n60627\\n  \n",
       "3                         6871 W BELDEN AVE\\n60635\\n  \n",
       "4  1300 W HOWARD ST\\n60626\\n(42.01933233897205, -...  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "parks.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(41.6590468805167, -87.6093631364668)\n"
     ]
    }
   ],
   "source": [
    "geolocator = Nominatim()\n",
    "location = geolocator.geocode(\"500 E 130TH ST CHICAGO\")\n",
    "print((location.latitude, location.longitude))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "ename": "GeocoderTimedOut",
     "evalue": "Service timed out",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mtimeout\u001b[0m                                   Traceback (most recent call last)",
      "\u001b[0;32m/anaconda3/lib/python3.6/urllib/request.py\u001b[0m in \u001b[0;36mdo_open\u001b[0;34m(self, http_class, req, **http_conn_args)\u001b[0m\n\u001b[1;32m   1317\u001b[0m                 h.request(req.get_method(), req.selector, req.data, headers,\n\u001b[0;32m-> 1318\u001b[0;31m                           encode_chunked=req.has_header('Transfer-encoding'))\n\u001b[0m\u001b[1;32m   1319\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mOSError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0merr\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;31m# timeout error\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/http/client.py\u001b[0m in \u001b[0;36mrequest\u001b[0;34m(self, method, url, body, headers, encode_chunked)\u001b[0m\n\u001b[1;32m   1238\u001b[0m         \u001b[0;34m\"\"\"Send a complete request to the server.\"\"\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1239\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_send_request\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmethod\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0murl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbody\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mheaders\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mencode_chunked\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1240\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/http/client.py\u001b[0m in \u001b[0;36m_send_request\u001b[0;34m(self, method, url, body, headers, encode_chunked)\u001b[0m\n\u001b[1;32m   1284\u001b[0m             \u001b[0mbody\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_encode\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbody\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'body'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1285\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mendheaders\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbody\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mencode_chunked\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mencode_chunked\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1286\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/http/client.py\u001b[0m in \u001b[0;36mendheaders\u001b[0;34m(self, message_body, encode_chunked)\u001b[0m\n\u001b[1;32m   1233\u001b[0m             \u001b[0;32mraise\u001b[0m \u001b[0mCannotSendHeader\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1234\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_send_output\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmessage_body\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mencode_chunked\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mencode_chunked\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1235\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/http/client.py\u001b[0m in \u001b[0;36m_send_output\u001b[0;34m(self, message_body, encode_chunked)\u001b[0m\n\u001b[1;32m   1025\u001b[0m         \u001b[0;32mdel\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_buffer\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1026\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmsg\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1027\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/http/client.py\u001b[0m in \u001b[0;36msend\u001b[0;34m(self, data)\u001b[0m\n\u001b[1;32m    963\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mauto_open\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 964\u001b[0;31m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mconnect\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    965\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/http/client.py\u001b[0m in \u001b[0;36mconnect\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m   1399\u001b[0m             self.sock = self._context.wrap_socket(self.sock,\n\u001b[0;32m-> 1400\u001b[0;31m                                                   server_hostname=server_hostname)\n\u001b[0m\u001b[1;32m   1401\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_context\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcheck_hostname\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_check_hostname\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/ssl.py\u001b[0m in \u001b[0;36mwrap_socket\u001b[0;34m(self, sock, server_side, do_handshake_on_connect, suppress_ragged_eofs, server_hostname, session)\u001b[0m\n\u001b[1;32m    406\u001b[0m                          \u001b[0mserver_hostname\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mserver_hostname\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 407\u001b[0;31m                          _context=self, _session=session)\n\u001b[0m\u001b[1;32m    408\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/ssl.py\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, sock, keyfile, certfile, server_side, cert_reqs, ssl_version, ca_certs, do_handshake_on_connect, family, type, proto, fileno, suppress_ragged_eofs, npn_protocols, ciphers, server_hostname, _context, _session)\u001b[0m\n\u001b[1;32m    813\u001b[0m                         \u001b[0;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"do_handshake_on_connect should not be specified for non-blocking sockets\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 814\u001b[0;31m                     \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdo_handshake\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    815\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/ssl.py\u001b[0m in \u001b[0;36mdo_handshake\u001b[0;34m(self, block)\u001b[0m\n\u001b[1;32m   1067\u001b[0m                 \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msettimeout\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1068\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_sslobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdo_handshake\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1069\u001b[0m         \u001b[0;32mfinally\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/ssl.py\u001b[0m in \u001b[0;36mdo_handshake\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m    688\u001b[0m         \u001b[0;34m\"\"\"Start the SSL/TLS handshake.\"\"\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 689\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_sslobj\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdo_handshake\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    690\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcontext\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcheck_hostname\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mtimeout\u001b[0m: _ssl.c:761: The handshake operation timed out",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mURLError\u001b[0m                                  Traceback (most recent call last)",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/geopy/geocoders/base.py\u001b[0m in \u001b[0;36m_call_geocoder\u001b[0;34m(self, url, timeout, raw, requester, deserializer, **kwargs)\u001b[0m\n\u001b[1;32m    142\u001b[0m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 143\u001b[0;31m             \u001b[0mpage\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mrequester\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mreq\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtimeout\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtimeout\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtimeout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    144\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0merror\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;31m# pylint: disable=W0703\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/urllib/request.py\u001b[0m in \u001b[0;36murlopen\u001b[0;34m(url, data, timeout, cafile, capath, cadefault, context)\u001b[0m\n\u001b[1;32m    222\u001b[0m         \u001b[0mopener\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_opener\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 223\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0mopener\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0murl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtimeout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    224\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/urllib/request.py\u001b[0m in \u001b[0;36mopen\u001b[0;34m(self, fullurl, data, timeout)\u001b[0m\n\u001b[1;32m    525\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 526\u001b[0;31m         \u001b[0mresponse\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_open\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mreq\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    527\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/urllib/request.py\u001b[0m in \u001b[0;36m_open\u001b[0;34m(self, req, data)\u001b[0m\n\u001b[1;32m    543\u001b[0m         result = self._call_chain(self.handle_open, protocol, protocol +\n\u001b[0;32m--> 544\u001b[0;31m                                   '_open', req)\n\u001b[0m\u001b[1;32m    545\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/urllib/request.py\u001b[0m in \u001b[0;36m_call_chain\u001b[0;34m(self, chain, kind, meth_name, *args)\u001b[0m\n\u001b[1;32m    503\u001b[0m             \u001b[0mfunc\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgetattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mhandler\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmeth_name\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 504\u001b[0;31m             \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    505\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mresult\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/urllib/request.py\u001b[0m in \u001b[0;36mhttps_open\u001b[0;34m(self, req)\u001b[0m\n\u001b[1;32m   1360\u001b[0m             return self.do_open(http.client.HTTPSConnection, req,\n\u001b[0;32m-> 1361\u001b[0;31m                 context=self._context, check_hostname=self._check_hostname)\n\u001b[0m\u001b[1;32m   1362\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/urllib/request.py\u001b[0m in \u001b[0;36mdo_open\u001b[0;34m(self, http_class, req, **http_conn_args)\u001b[0m\n\u001b[1;32m   1319\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mOSError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0merr\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;31m# timeout error\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1320\u001b[0;31m                 \u001b[0;32mraise\u001b[0m \u001b[0mURLError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0merr\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1321\u001b[0m             \u001b[0mr\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mh\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgetresponse\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mURLError\u001b[0m: <urlopen error _ssl.c:761: The handshake operation timed out>",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mGeocoderTimedOut\u001b[0m                          Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-90-e35a34f7d453>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mparks\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 6\u001b[0;31m     \u001b[0mloc\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgeolocator\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgeocode\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mparks\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'STREET ADDRESS'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      7\u001b[0m     \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      8\u001b[0m         \u001b[0mlats\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mloc\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlatitude\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/geopy/geocoders/osm.py\u001b[0m in \u001b[0;36mgeocode\u001b[0;34m(self, query, exactly_one, timeout, addressdetails, language, geometry)\u001b[0m\n\u001b[1;32m    191\u001b[0m         \u001b[0mlogger\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdebug\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"%s.geocode: %s\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__class__\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__name__\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0murl\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    192\u001b[0m         return self._parse_json(\n\u001b[0;32m--> 193\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_call_geocoder\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0murl\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtimeout\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtimeout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mexactly_one\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    194\u001b[0m         )\n\u001b[1;32m    195\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/anaconda3/lib/python3.6/site-packages/geopy/geocoders/base.py\u001b[0m in \u001b[0;36m_call_geocoder\u001b[0;34m(self, url, timeout, raw, requester, deserializer, **kwargs)\u001b[0m\n\u001b[1;32m    161\u001b[0m             \u001b[0;32melif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0merror\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mURLError\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    162\u001b[0m                 \u001b[0;32mif\u001b[0m \u001b[0;34m\"timed out\"\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mmessage\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 163\u001b[0;31m                     \u001b[0;32mraise\u001b[0m \u001b[0mGeocoderTimedOut\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'Service timed out'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    164\u001b[0m                 \u001b[0;32melif\u001b[0m \u001b[0;34m\"unreachable\"\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mmessage\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    165\u001b[0m                     \u001b[0;32mraise\u001b[0m \u001b[0mGeocoderUnavailable\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'Service not available'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mGeocoderTimedOut\u001b[0m: Service timed out"
     ]
    }
   ],
   "source": [
    "lats = []\n",
    "longs = []\n",
    "geolocator = Nominatim()\n",
    "\n",
    "for i in range(len(parks)):\n",
    "    loc = geolocator.geocode(parks['STREET ADDRESS'][i])\n",
    "    try:\n",
    "        lats.append(loc.latitude)\n",
    "    except:\n",
    "        lats.append(np.nan)\n",
    "    try:\n",
    "        longs.append(loc.longitude)\n",
    "    except:\n",
    "        longs.append(np.nan)\n",
    "        \n",
    "    time.sleep(5)"
   ]
  }
 ],
 "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}