08-01_descriptives_sample.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Descriptives Sample"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "NAME = '08-01_descriptives_sample'\n",
    "PROJECT = 'conference-calls-sentiment'\n",
    "PYTHON_VERSION = '3.7.0'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import re\n",
    "import numpy as np\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "workdir = re.sub(\"(?<={})[\\w\\W]*\".format(PROJECT), \"\", os.getcwd())\n",
    "os.chdir(workdir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "# Main Code"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = pd.read_feather(os.path.join('2_pipeline', '07-02_dataset_construction', 'out', 'dataset.feather'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "descriptive_columns = {'lm_tone': '$Tone_{LM}$',\n",
    "                       'finbert_tone': '$Tone_{FinBERT}$',\n",
    "                       'car_st': '$CAR[-1,+1]$',\n",
    "                       'surprise': '$Surprise$',\n",
    "                       'size': '$Size$',\n",
    "                       'roa': '$ROA$',\n",
    "                       'leverage': '$Leverage$',\n",
    "                       'loss': '$Loss$',\n",
    "                       'capex': '$Capex$'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Mean</th>\n      <th>Std. Dev.</th>\n      <th>Pctl(25)</th>\n      <th>Median</th>\n      <th>Pctl(75)</th>\n      <th>N</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>$Tone_{LM}$</th>\n      <td>0.154</td>\n      <td>0.242</td>\n      <td>0.000</td>\n      <td>0.172</td>\n      <td>0.326</td>\n      <td>21,966</td>\n    </tr>\n    <tr>\n      <th>$Tone_{FinBERT}$</th>\n      <td>0.302</td>\n      <td>0.272</td>\n      <td>0.126</td>\n      <td>0.322</td>\n      <td>0.500</td>\n      <td>21,966</td>\n    </tr>\n    <tr>\n      <th>$CAR[-1,+1]$</th>\n      <td>0.098</td>\n      <td>6.074</td>\n      <td>-2.893</td>\n      <td>0.110</td>\n      <td>3.231</td>\n      <td>21,966</td>\n    </tr>\n    <tr>\n      <th>$Surprise$</th>\n      <td>1.381</td>\n      <td>16.842</td>\n      <td>0.885</td>\n      <td>1.331</td>\n      <td>1.821</td>\n      <td>21,966</td>\n    </tr>\n    <tr>\n      <th>$Size$</th>\n      <td>9.792</td>\n      <td>1.118</td>\n      <td>9.037</td>\n      <td>9.669</td>\n      <td>10.442</td>\n      <td>21,966</td>\n    </tr>\n    <tr>\n      <th>$ROA$</th>\n      <td>1.609</td>\n      <td>2.453</td>\n      <td>0.610</td>\n      <td>1.417</td>\n      <td>2.539</td>\n      <td>21,966</td>\n    </tr>\n    <tr>\n      <th>$Leverage$</th>\n      <td>0.233</td>\n      <td>0.165</td>\n      <td>0.115</td>\n      <td>0.213</td>\n      <td>0.323</td>\n      <td>21,966</td>\n    </tr>\n    <tr>\n      <th>$Loss$</th>\n      <td>0.082</td>\n      <td>0.274</td>\n      <td>0.000</td>\n      <td>0.000</td>\n      <td>0.000</td>\n      <td>21,966</td>\n    </tr>\n    <tr>\n      <th>$Capex$</th>\n      <td>2.367</td>\n      <td>2.777</td>\n      <td>0.592</td>\n      <td>1.479</td>\n      <td>3.137</td>\n      <td>21,966</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
      "text/plain": "                   Mean  Std. Dev.  Pctl(25)  Median  Pctl(75)       N\n$Tone_{LM}$       0.154      0.242     0.000   0.172     0.326  21,966\n$Tone_{FinBERT}$  0.302      0.272     0.126   0.322     0.500  21,966\n$CAR[-1,+1]$      0.098      6.074    -2.893   0.110     3.231  21,966\n$Surprise$        1.381     16.842     0.885   1.331     1.821  21,966\n$Size$            9.792      1.118     9.037   9.669    10.442  21,966\n$ROA$             1.609      2.453     0.610   1.417     2.539  21,966\n$Leverage$        0.233      0.165     0.115   0.213     0.323  21,966\n$Loss$            0.082      0.274     0.000   0.000     0.000  21,966\n$Capex$           2.367      2.777     0.592   1.479     3.137  21,966"
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "descriptives = (dataset\r\n",
    "                .drop(columns=['gvkey', 'year'])\r\n",
    "                .dropna()\r\n",
    "                .filter(descriptive_columns.keys())\r\n",
    "                .rename(descriptive_columns, axis=1)\r\n",
    "                .describe().transpose()\r\n",
    "                .filter(['mean', 'std', '25%', '50%', '75%', 'count'])\r\n",
    "                .round(3)\r\n",
    "                .assign(count=lambda x: x['count'].apply(lambda x : \"{:,.0f}\".format(x)))\r\n",
    "                .rename(columns={'mean': 'Mean', 'std': 'Std. Dev.', '25%': 'Pctl(25)',\r\n",
    "                                 '50%': 'Median', '75%': 'Pctl(75)', 'count': 'N'}))\r\n",
    "descriptives"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>(1)</th>\n      <th>(2)</th>\n      <th>(3)</th>\n      <th>(4)</th>\n      <th>(5)</th>\n      <th>(6)</th>\n      <th>(7)</th>\n      <th>(8)</th>\n      <th>(9)</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>(1) $Tone_{LM}$</th>\n      <td>1.000</td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n    </tr>\n    <tr>\n      <th>(2) $Tone_{FinBERT}$</th>\n      <td>0.710</td>\n      <td>1.000</td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n    </tr>\n    <tr>\n      <th>(3) $CAR[-1,+1]$</th>\n      <td>0.114</td>\n      <td>0.163</td>\n      <td>1.000</td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n    </tr>\n    <tr>\n      <th>(4) $Surprise$</th>\n      <td>0.036</td>\n      <td>0.018</td>\n      <td>0.005</td>\n      <td>1.000</td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n    </tr>\n    <tr>\n      <th>(5) $Size$</th>\n      <td>0.142</td>\n      <td>0.173</td>\n      <td>-0.009</td>\n      <td>0.020</td>\n      <td>1.000</td>\n      <td></td>\n      <td></td>\n      <td></td>\n      <td></td>\n    </tr>\n    <tr>\n      <th>(6) $ROA$</th>\n      <td>0.106</td>\n      <td>0.057</td>\n      <td>0.046</td>\n      <td>0.073</td>\n      <td>0.174</td>\n      <td>1.000</td>\n      <td></td>\n      <td></td>\n      <td></td>\n    </tr>\n    <tr>\n      <th>(7) $Leverage$</th>\n      <td>0.027</td>\n      <td>0.024</td>\n      <td>-0.015</td>\n      <td>0.013</td>\n      <td>-0.042</td>\n      <td>-0.094</td>\n      <td>1.000</td>\n      <td></td>\n      <td></td>\n    </tr>\n    <tr>\n      <th>(8) $Loss$</th>\n      <td>-0.090</td>\n      <td>-0.026</td>\n      <td>-0.054</td>\n      <td>-0.081</td>\n      <td>-0.164</td>\n      <td>-0.486</td>\n      <td>0.081</td>\n      <td>1.000</td>\n      <td></td>\n    </tr>\n    <tr>\n      <th>(9) $Capex$</th>\n      <td>-0.032</td>\n      <td>-0.047</td>\n      <td>-0.003</td>\n      <td>0.047</td>\n      <td>-0.013</td>\n      <td>0.029</td>\n      <td>0.046</td>\n      <td>0.033</td>\n      <td>1.000</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
      "text/plain": "                         (1)     (2)     (3)     (4)     (5)     (6)    (7)  \\\n(1) $Tone_{LM}$        1.000                                                  \n(2) $Tone_{FinBERT}$   0.710   1.000                                          \n(3) $CAR[-1,+1]$       0.114   0.163   1.000                                  \n(4) $Surprise$         0.036   0.018   0.005   1.000                          \n(5) $Size$             0.142   0.173  -0.009   0.020   1.000                  \n(6) $ROA$              0.106   0.057   0.046   0.073   0.174   1.000          \n(7) $Leverage$         0.027   0.024  -0.015   0.013  -0.042  -0.094  1.000   \n(8) $Loss$            -0.090  -0.026  -0.054  -0.081  -0.164  -0.486  0.081   \n(9) $Capex$           -0.032  -0.047  -0.003   0.047  -0.013   0.029  0.046   \n\n                        (8)    (9)  \n(1) $Tone_{LM}$                     \n(2) $Tone_{FinBERT}$                \n(3) $CAR[-1,+1]$                    \n(4) $Surprise$                      \n(5) $Size$                          \n(6) $ROA$                           \n(7) $Leverage$                      \n(8) $Loss$            1.000         \n(9) $Capex$           0.033  1.000  "
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "correlation = (dataset\n",
    "               .drop(columns=['gvkey', 'year'])\n",
    "               .filter(descriptive_columns.keys())\n",
    "               .rename(descriptive_columns, axis=1)\n",
    "               .corr()\n",
    "               .apply(lambda s: s.apply(lambda x: f'{x:.3f}')))\n",
    "\n",
    "# Convert table to one-sided\n",
    "for i, _ in enumerate(correlation):\n",
    "    for j, _ in enumerate(correlation):\n",
    "        if i < j:\n",
    "            correlation.iloc[i, j] = ''\n",
    "\n",
    "# Fix index and columns\n",
    "correlation.index = [f'({i+1}) {index}' for i, index in enumerate(correlation.columns)]\n",
    "correlation.columns = [f'({i+1})' for i, _ in enumerate(correlation.columns)]\n",
    "\n",
    "correlation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## LaTeX"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lYYYYYY}\n",
      "\\toprule\n",
      "{} &   Mean &  Std. Dev. &  Pctl(25) &  Median &  Pctl(75) &       N \\\\\n",
      "\\midrule\n",
      "$Tone_{LM}$      &  0.154 &      0.242 &     0.000 &   0.172 &     0.326 &  21,966 \\\\\n",
      "$Tone_{FinBERT}$ &  0.302 &      0.272 &     0.126 &   0.322 &     0.500 &  21,966 \\\\\n",
      "$CAR[-1,+1]$     &  0.098 &      6.074 &    -2.893 &   0.110 &     3.231 &  21,966 \\\\\n",
      "$Surprise$       &  1.381 &     16.842 &     0.885 &   1.331 &     1.821 &  21,966 \\\\\n",
      "$Size$           &  9.792 &      1.118 &     9.037 &   9.669 &    10.442 &  21,966 \\\\\n",
      "$ROA$            &  1.609 &      2.453 &     0.610 &   1.417 &     2.539 &  21,966 \\\\\n",
      "$Leverage$       &  0.233 &      0.165 &     0.115 &   0.213 &     0.323 &  21,966 \\\\\n",
      "$Loss$           &  0.082 &      0.274 &     0.000 &   0.000 &     0.000 &  21,966 \\\\\n",
      "$Capex$          &  2.367 &      2.777 &     0.592 &   1.479 &     3.137 &  21,966 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(descriptives.to_latex(column_format='lYYYYYY', escape=False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\\begin{tabular}{lYYYYYYYYY}\n",
      "\\toprule\n",
      "{} &     (1) &     (2) &     (3) &     (4) &     (5) &     (6) &    (7) &    (8) &    (9) \\\\\n",
      "\\midrule\n",
      "(1) $Tone_{LM}$      &   1.000 &         &         &         &         &         &        &        &        \\\\\n",
      "(2) $Tone_{FinBERT}$ &   0.710 &   1.000 &         &         &         &         &        &        &        \\\\\n",
      "(3) $CAR[-1,+1]$     &   0.114 &   0.163 &   1.000 &         &         &         &        &        &        \\\\\n",
      "(4) $Surprise$       &   0.036 &   0.018 &   0.005 &   1.000 &         &         &        &        &        \\\\\n",
      "(5) $Size$           &   0.142 &   0.173 &  -0.009 &   0.020 &   1.000 &         &        &        &        \\\\\n",
      "(6) $ROA$            &   0.106 &   0.057 &   0.046 &   0.073 &   0.174 &   1.000 &        &        &        \\\\\n",
      "(7) $Leverage$       &   0.027 &   0.024 &  -0.015 &   0.013 &  -0.042 &  -0.094 &  1.000 &        &        \\\\\n",
      "(8) $Loss$           &  -0.090 &  -0.026 &  -0.054 &  -0.081 &  -0.164 &  -0.486 &  0.081 &  1.000 &        \\\\\n",
      "(9) $Capex$          &  -0.032 &  -0.047 &  -0.003 &   0.047 &  -0.013 &   0.029 &  0.046 &  0.033 &  1.000 \\\\\n",
      "\\bottomrule\n",
      "\\end{tabular}\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(correlation.to_latex(column_format='lYYYYYYYYY', escape=False))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Processed transcripts\r\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "cc = pd.read_feather(os.path.join('2_pipeline', '02-02_conference_calls_preprocess', 'out', 'cc_transcripts.feather'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of transcripts: 26,237\n"
     ]
    }
   ],
   "source": [
    "num_transcripts = cc['transcript_id'].nunique()\r\n",
    "print(f\"Number of transcripts: {num_transcripts:,}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of unique firms: 715\n"
     ]
    }
   ],
   "source": [
    "num_firms = cc['gvkey'].nunique()\r\n",
    "print(f\"Number of unique firms: {num_firms:,}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Sentences</th>\n    </tr>\n    <tr>\n      <th>speaker_role</th>\n      <th></th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>Analyst</th>\n      <td>1,840,233</td>\n    </tr>\n    <tr>\n      <th>Management</th>\n      <td>4,232,635</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
      "text/plain": "              Sentences\nspeaker_role           \nAnalyst       1,840,233\nManagement    4,232,635"
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(cc.groupby('speaker_role').size()\r\n",
    "   .to_frame('Sentences')\r\n",
    "   .apply(lambda s: s.apply(lambda x: f'{x:,}')))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>Words</th>\n    </tr>\n    <tr>\n      <th>speaker_role</th>\n      <th></th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>Analyst</th>\n      <td>35,095,298</td>\n    </tr>\n    <tr>\n      <th>Management</th>\n      <td>85,330,348</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
      "text/plain": "                   Words\nspeaker_role            \nAnalyst       35,095,298\nManagement    85,330,348"
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(cc\r\n",
    " .groupby('speaker_role')['num_words'].sum()\r\n",
    " .to_frame('Words')\r\n",
    " .apply(lambda s: s.apply(lambda x: f'{x:,}')))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": "False    0.991653\nTrue     0.008347\nName: num_words, dtype: float64"
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(cc['num_words'] > 64).value_counts(normalize=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": "<AxesSubplot:xlabel='year'>"
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEGCAYAAAB7DNKzAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8QVMy6AAAACXBIWXMAAAsTAAALEwEAmpwYAAA0l0lEQVR4nO3deXgUZbr38e+dPYTsCZiQQNgxrJGACKgIKuCIuIDCqGw6oKPjco6vg3PGcT+jcxwcl9ERBwQHBQXEbQQGQUSQLQiyI1uAQIQskJCQtfO8f1QldCA0AZJ0J7k/11VXVz9dVbk7qfSva3tKjDEopZRS5+Ll7gKUUkp5Ng0KpZRSLmlQKKWUckmDQimllEsaFEoppVzycXcBNS0qKsokJCS4uwyllKpXNmzYkGmMia7qtQYXFAkJCaSkpLi7DKWUqldE5MC5XtNdT0oppVzSoFBKKeWSBoVSSimXznuMQkQCgBWAvz39PGPMMyISAXwMJACpwJ3GmOP2PE8B9wEO4BFjzGK7vScwAwgEvgYeNcYYEfEHPgB6AlnAXcaYVHuescAf7XJeNMbMvNA3WVJSQlpaGoWFhRc6q6olAQEBxMXF4evr6+5SlFLnUZ2D2UXAQGNMnoj4AitFZCFwO7DUGPOyiEwGJgO/F5FEYBTQGYgFvhGRDsYYB/AOMBFYgxUUQ4CFWKFy3BjTTkRGAa8Ad9lh9AyQDBhgg4h8UR5I1ZWWlkZwcDAJCQmIyIXMqmqBMYasrCzS0tJo3bq1u8tRSp3HeXc9GUue/dTXHgwwHCj/dj8TuNUeHw7MMcYUGWP2A3uA3iISA4QYY1YbqyfCD86Yp3xZ84BBYn2iDwaWGGOy7XBYghUuF6SwsJDIyEgNCQ8hIkRGRuoWnlL1RLWOUYiIt4hsAo5hfXCvBZobY9IB7Mdm9uQtgENOs6fZbS3s8TPbK81jjCkFcoBIF8s6s76JIpIiIikZGRnneg/VeauqjujfQ6n6o1pBYYxxGGN6AHFYWwddXExe1SeAcdF+sfM41zfVGJNsjEmOjq7yehGllGqwHGWGrzYfYfa6g7Wy/As668kYcwJYjrX756i9Own78Zg9WRoQ7zRbHHDEbo+ror3SPCLiA4QC2S6WVe+ICPfee2/F89LSUqKjo7n55pvdWFXtSk1N5aOPPnJ3GUo1WMWlZXy8/iDXT/mOhz/ayNyUQ9TGPYbOGxQiEi0iYfZ4IHA9sBP4AhhrTzYW+Nwe/wIYJSL+ItIaaA+ss3dPnRSRPvbxhzFnzFO+rBHAMvs4xmLgRhEJF5Fw4Ea7rd4JCgpi69atFBQUALBkyRJatDhrL1qDokGhVO04VVzK9JX7ufb/vuX387cQ5O/NO3dfwdwH+tbKbt3qbFHEAN+KyGZgPdYxiq+Al4EbRGQ3cIP9HGPMNuATYDuwCHjIPuMJ4EHgn1gHuPdinfEEMA2IFJE9wH9hnUGFMSYbeMH+ueuB5+22emno0KH8+9//BmD27NmMHj264rV169bRt29fkpKS6Nu3L7t27QJgxowZ3H777QwZMoT27dvz5JNPVszz4IMPkpycTOfOnXnmmWcq2r/++ms6depE//79eeSRRyq2WvLz85kwYQK9evUiKSmJzz//vOJn3HrrrQwbNozWrVvz1ltvMWXKFJKSkujTpw/Z2davfO/evQwZMoSePXty9dVXs3PnTgDGjRvHI488Qt++fWnTpg3z5s0DYPLkyXz//ff06NGD1157rbZ+rUo1GjmnSnhz6W76vbyM57/aTsuIJnwwoTdfPtyfoV1j8PaqnWN/5z091hizGUiqoj0LGHSOeV4CXqqiPQU46/iGMaYQGHmOZU0Hpp+vzup67sttbD+SW1OLAyAxNoRnhnU+73SjRo3i+eef5+abb2bz5s1MmDCB77//HoBOnTqxYsUKfHx8+Oabb/jDH/7A/PnzAdi0aRMbN27E39+fjh078rvf/Y74+HheeuklIiIicDgcDBo0iM2bN9OhQwcmTZrEihUraN26daUweumllxg4cCDTp0/nxIkT9O7dm+uvvx6ArVu3snHjRgoLC2nXrh2vvPIKGzdu5PHHH+eDDz7gscceY+LEifzjH/+gffv2rF27lt/+9rcsW7YMgPT0dFauXMnOnTu55ZZbGDFiBC+//DKvvvoqX331VY3+vpVqbI6dLGTayv18uOYgeUWlDOrUjN9e15aerSLq5Oc3uE4BPVm3bt1ITU1l9uzZ3HTTTZVey8nJYezYsezevRsRoaSkpOK1QYMGERoaCkBiYiIHDhwgPj6eTz75hKlTp1JaWkp6ejrbt2+nrKyMNm3aVFyfMHr0aKZOnQrAf/7zH7744gteffVVwDpt+OBB6+DXddddR3BwMMHBwYSGhjJs2DAAunbtyubNm8nLy+OHH35g5MjTeV5UVFQxfuutt+Ll5UViYiJHjx6t6V+dUo3SoexTTF2xj49TDlHqKONX3WJ58Nq2JMaG1GkdjS4oqvPNvzbdcsstPPHEEyxfvpysrKyK9qeffprrrruOBQsWkJqayoABAype8/f3rxj39vamtLSU/fv38+qrr7J+/XrCw8MZN24chYWFLg9kGWOYP38+HTt2rNS+du3aSj/Dy8ur4rmXlxelpaWUlZURFhbGpk2bqly28/y1cTBNqcZk99GTvLN8L5//dAQvgRE945h0TVsSooLcUk+jCwp3mzBhAqGhoXTt2pXly5dXtOfk5FQc3J4xY8Z5l5Obm0tQUBChoaEcPXqUhQsXMmDAADp16sS+fftITU0lISGBjz/+uGKewYMH8+abb/Lmm28iImzcuJGkpLP2KlYpJCSE1q1bM3fuXEaOHIkxhs2bN9O9e/dzzhMcHMzJkyertXyl6oPj+cXM3XCIo7lFxIQGEBMaSExYALGhgUQH+1/yMYKfDp3g7eV7WLztKIG+3ozrm8D9V7cmJjSwht7BxdGgqGNxcXE8+uijZ7U/+eSTjB07lilTpjBw4MDzLqd79+4kJSXRuXNn2rRpQ79+/QAIDAzk7bffZsiQIURFRdG7d++KeZ5++mkee+wxunXrhjGGhISECzp+8OGHH/Lggw/y4osvUlJSwqhRo1wGRbdu3fDx8aF79+6MGzeOxx9/vNo/SylPsuuXk8z4YT8LNh6msKSMAF8vCkvKKk3j4yU0DwkgJjSAy0IDiA0LrAiT2DDrMTLID68zwsQYw+p9Wbz97V5W7skkJMCHRwa1Z1zfBCKC/OrybZ6TNLTdBMnJyebMGxft2LGDyy+/3E0V1b28vDyaNm2KMYaHHnqI9u3be+SHdGP7u6j6xVFmWLbzGO+v2s8Pe7Pw9/HitqQWjOuXQMfmweQUlHDkRCHpOQUcySkk/UQB6TnWc+uxkOLSymHi5+3FZeVBEhrAZaGBrN2fxcaDJ4gO9uf+/q25u08rmvrX/Xd4EdlgjEmu6jXdomiA3nvvPWbOnElxcTFJSUlMmjTJ3SUpVW/kFJQwN+UQH6w+wMHsU8SEBvDkkI6M7tWScKdv+GFN/Ahr4nfOA8vGGLLyi0k/UciRnAJ+ybEe0+1wWZ96nKO56cSEBfDirV0Y0TOOAF/vunqbF0SDogF6/PHHPXILQilPtjcjj5k/pDJvQxqnih30Sgjn90M6Mbhzc3y8L/zWPSJCVFN/opr60zUutMppysoMIp7f95kGhVKq0SorM3y3O4MZq1L57ucM/Ly9GNY9lnF9E8754V6Tzjxe4ak0KJRSjU5eUSnzN6Qx84dU9mXmEx3sz3/d0IHRvVsSHex//gU0MhoUSqlG42DWKWb8kMrclEOcLCqle3wYr4/qwdAuMfj56J2hz0WDQinVoBljWLUnixk/pLJ051G8Rbipawzj+yWQ1DLc3eXVCxqhdWzBggWISEWHehdj3LhxFR3vXYgZM2Zw5Ei97KVdqQuWU1DC9JX7GTTlO+6ZtpYfDx7n4evasfL3A3ljdJKGxAXQLYo6Nnv2bPr378+cOXN49tln6/Rnz5gxgy5duhAbG1unP1epurT9SC7/WnOAzzYepqDEQY/4MKbc2Z2busZ47Omnnk63KOpQXl4eq1atYtq0acyZMweA5cuXM2DAAEaMGEGnTp24++67K/pKev755+nVqxddunRh4sSJZ/WhtHTpUm677baK50uWLOH222/H4XAwbtw4unTpQteuXXnttdeYN28eKSkp3H333fTo0YOCggImT55MYmIi3bp144knnqi7X4RSNay4tIzPNx1mxDs/cNMb3/Ppj2kM6x7Dlw/357OH+nH7FZ57jUJ90Pi2KBZOhl+21OwyL+sKQ18+72SfffYZQ4YMoUOHDkRERPDjjz8CsHHjRrZt20ZsbCz9+vVj1apV9O/fn4cffpg//elPANx777189dVXFb26AgwcOJCHHnqIjIwMoqOjef/99xk/fjybNm3i8OHDbN26FYATJ04QFhbGW2+9xauvvkpycjLZ2dksWLCAnTt3IiKcOHGiZn8nStWBIycK+GjtQeasP0hmXjGtIpvwx19dzoiecYQ18YzuLxoC3aKoQ7Nnz2bUqFGAdW+K2bNnA9C7d2/i4uLw8vKiR48epKamAvDtt99y5ZVX0rVrV5YtW8a2bdsqLa/89qqzZs3ixIkTrF69mqFDh9KmTRv27dvH7373OxYtWkRIyNlXjoaEhBAQEMD999/Pp59+SpMmTWr3zStVQ4wxrNydycQPUuj/yjL+vnwPPeLDmDmhN9/+9wDuv7qNhkQNa3xbFNX45l8bsrKyWLZsGVu3bkVEcDgciAg33XRTld2IFxYW8tvf/paUlBTi4+N59tlnKSwsPGu548ePZ9iwYQQEBDBy5Eh8fHwIDw/np59+YvHixfz973/nk08+Yfr0yvd+8vHxYd26dSxdupQ5c+bw1ltvVdyESClPlFNQwvwNacxac4B9mflEBPkx6dq2/Lp3S+Ij9ItObWp8QeEm8+bNY8yYMbz77rsVbddeey0rV66scvryUIiKiiIvL4958+YxYsSIs6aLjY0lNjaWF198kSVLlgCQmZmJn58fd9xxB23btmXcuHFA5W6/8/LyOHXqFDfddBN9+vShXbt2Nfl2laox1sHpVD7beISCEgdJLfXgdF3ToKgjs2fPZvLkyZXa7rjjDt555x3atm171vRhYWH85je/oWvXriQkJNCrV69zLvvuu+8mIyODxMREAA4fPsz48eMpK7N6rvzzn/8MWKfVPvDAAwQGBrJw4UKGDx9ecbMjvae18iSFJQ4Wbf2FWWsOkHLgOAG+Xgzv3oJ7r2pFlxa137WGqky7GW8AHn74YZKSkrjvvvvcXcoFaeh/F3Xhdv6Sy5x1h1iw8TA5BSUkRDbhnj6tGNkzntAmvu4ur0HTbsYbsJ49exIUFMRf//pXd5ei6pmyMkNBiYP84lLCAv3c1oVFXlEpX/10hNnrD/HToRP4eXsxuMtljO4VT582kfWm47yGTIOintuwYYO7S1C1yBhDicNQWOqgsMRBUUkZhSUOCkvKOFVcyqli64P+VJH9WOwgv6jUGoodnCouJb/o9GO+0/NTxY6Kn9PEz5srW0fQv300V7ePon2zprXa9bUxhk2HTvDx+kN8+dMR8osdtG/WlKdvTuT2pBaV7vug3K/RBIUxxuP7fG9MGtouz/MpLHEwN+UQ+zLzKSwpo6jEYX/4l3/w2+OlDgqLHRSWnm4vu8BflY+XEOTvQ5CfN03KH/18iA3zI8jfGq/8mje7j+Wxcncm3+7aDkCzYH/6t4uif/so+reLollIQI38Hk6cKuazjYeZs/4QO385SaCvN8O6x3BXr5Zc0TJM/0c9VKMIioCAALKysoiMjNQV0QMYY8jKyiIgoGY+fDxZqaOMuRvSeP2b3fySW0hTfx8CfL3w9/EmwNeLAF9vAnytD++IIC/8fb0JqPSal/3cGve3pw/w8SLI34cmft6nH/18CPL3uaRdSIdPFLBydwbf785k+c8ZfLrxMAAdmwdXhMaVbSJo4lf9jw5jDGv2ZfPx+oN8vfUXikvL6BYXyv/e1pVh3WMIDtBjD56uURzMLikpIS0trcrrEJR7BAQEEBcXh69vw/yQKCszfL01nSn/+Zl9mfkktQzjycGduKptpLtLq7ayMsP29FxW7slk5e5M1qVmU1xahq+3cEXLcK5uH0W/dlF0iwvDu4rjCBkni5i3IY1PUg6xPzOf4AAfbktqwV294ukcq2cueRpXB7MbRVAoVVeMMXz3cwb/t3gX247k0rF5ME8M7sj1lzer91uzhSUOUlKP8/2eDFbuzmTbkVwAQgJ86NvW2k3Vr10UqVn5fLzuEN/sOEppmaF36whG9YrX6x48nAaFUnVgw4Hj/GXRTtbuzyYuPJD/uqEDw3u0qPLbdkOQlVfED3uzWLk7k5V7Mjl8oqDitcggP0b0jOPOXvG0jW7qxipVdenpsUrVop2/5PLq4l18s+MYUU39eX54Z0b1atng75gW2dSfYd1jGdY9FmMM+zPz+WFvFpFBfgy6vHmDf/+NiQaFUhfpYNYppizZxec/HaGpvw//b3BHxvdLuKADvQ2FiNAmuiltdOuhQWp8a7RSl+hYbiFvLtvD7HUH8fYSJl3Tlgeu1R5LVcOlQaFUNeWcKuHdFXuZvmo/pQ7DXb3ieWRQe5rX0DUGSnkqDQqlzqOg2MH7P+znH8v3kltYyvAesTx+fQcSooLcXZpSdUKDQikX5m1I4y+LdnLsZBEDOzXjiRs7khh79o2glGrINCiUOodpK/fzwlfb6dkqnL/ffQW9EiLcXZJSbqFBoVQVPlx7gBe+2s7QLpfx5ugkfLz1VE/VeJ137ReReBH5VkR2iMg2EXnUbn9WRA6LyCZ7uMlpnqdEZI+I7BKRwU7tPUVki/3aG2Jfqioi/iLysd2+VkQSnOYZKyK77WFsjb57parw6Y9p/PGzrQzs1IzXR2lIKFWdLYpS4L+NMT+KSDCwQUSW2K+9Zox51XliEUkERgGdgVjgGxHpYIxxAO8AE4E1wNfAEGAhcB9w3BjTTkRGAa8Ad4lIBPAMkAwY+2d/YYw5fmlvW6mqfb0lnSfm/sRVbSJ5++4r9KIxpajGFoUxJt0Y86M9fhLYAbRwMctwYI4xpsgYsx/YA/QWkRggxBiz2lj9hnwA3Oo0z0x7fB4wyN7aGAwsMcZk2+GwBCtclKpxy3Ye5ZHZG7miZTjvjUnWfomUsl3Q1yV7l1ASsNZuelhENovIdBEJt9taAIecZkuz21rY42e2V5rHGFMK5ACRLpalVI1auTuTB2b9SGJsCNPH9yLIXw/fKVWu2kEhIk2B+cBjxphcrN1IbYEeQDpQfi/OqnpAMy7aL3Ye59omikiKiKRkZGS4ehtKnWV9aja/+SCF1pFBzBzfmxC9P4JSlVQrKETEFyskPjTGfApgjDlqjHEYY8qA94De9uRpQLzT7HHAEbs9ror2SvOIiA8QCmS7WFYlxpipxphkY0xydHR0dd6SUgD8dOgE499fT0xoALPuv1JvwalUFapz1pMA04AdxpgpTu0xTpPdBmy1x78ARtlnMrUG2gPrjDHpwEkR6WMvcwzwudM85Wc0jQCW2ccxFgM3iki4vWvrRrtNqUu2Iz2XMdPXER7ky4e/uZLoYH93l6SUR6rOjth+wL3AFhHZZLf9ARgtIj2wdgWlApMAjDHbROQTYDvWGVMP2Wc8ATwIzAACsc52Wmi3TwP+JSJ7sLYkRtnLyhaRF4D19nTPG2OyL+aNKuVsz7E87p22lkBfbz66vw8xoYHuLkkpj6U3LlKNzsGsU4x89wccZYaPJ12lN9ZRCr1xkVIVjpwo4Nf/XENRaRlzJvbRkFCqGvRqItVoHDtZyD3/XEvOqRI+mNCbTpdp535KVYduUahG4Xh+Mff+cx3pOYX8677edIsLc3dJStUbGhSqwcstLGHM9HXsz8rn/XG9SNZeYJW6ILrrSTVo+UWljH9/PTt/yeXde3rSr12Uu0tSqt7RoFANVmGJg/tnprDx4HHeGJXEdZ2aubskpeol3fWkGqSiUgcPzNrAmv1ZTLmzO0O7xpx/JqVUlXSLQjU4pY4yHp29ieW7Mnjp1q7clhR3/pmUUuekWxSqQSgscbDzl5NsO5LD4m1HWfFzBn+6OZFfX9nS3aUpVe9pUKh6J6eghO1Hctl2JIdt9uPejHwcZVYvA8EBPvzxV5czoX9rN1eqVMOgQaE82rHcQrYdyWXrYTsU0nM4lF1Q8XqzYH+6tAhlcOfL6BwbQufYUOLCA7HvsquUqgEaFMojGGM4mH2qcigcySUzr6himoTIJnRrEcaoXi0rQkF7fFWq9mlQKLfKyiviw7UHmbXmAMdOWqHg4yW0a9aUaztE24EQQmJsCMF6QyGl3EKDQrnFz0dPMn3lfhZsPExRaRkDOkbzeOfL6BIbSvvmTfV+1Up5EA0KVWeMMazYncm0lftZ8XMG/j5e3NEzjgn9EmjXLNjd5SmlzkGDQtW6whIHCzYeZvrK/ew+lkezYH+euLEDv76yFRF661GlPJ4Ghao1x04WMmv1AWatPUh2fjGdY0OYcmd3bu4Wi5+PXuupVH2hQaFq3PYjuUxbuZ8vfzpCSVkZgzo1577+renTJkJPW1WqHtKgUDWirMywbOcxpq3cz+p9WTTx82Z073jG9WtN66ggd5enlLoEGhTqkpwqLmX+hjTeX5XKvsx8YkIDmDy0E6N7tSS0iZ7OqlRDoEGhLkpuYQnvfreXWWsOklNQQvf4MN4YncTQLpfh663HH5RqSDQo1AUpKzPM3XCI/1u8i6z8YoZ0voz7r27NFS3D9fiDUg2UBoWqtpTUbJ77cjtbDudwRcswpo/rpfeeVqoR0KBQ55WeU8DLC3fy+aYjNA/x52939WB4j1jdglCqkdCgUOdUWOLgvRX7eHv5XhzG8LuB7Xjg2rYE+etqo1Rjov/x6izGGBZt/YWXvt5B2vEChna5jD/cdDnxEU3cXZpSyg00KFQlO9Jzee7LbazZl02ny4L56P4r6dsuyt1lKaXcSINCAZCdX8yUJbv4aO1BQgJ9eeHWLozuFY+PnuqqVKOnQdHIlTjK+HDNAaYs+Zn8YgdjrkrgsevbE9ZEO+tTSlk0KBqxlbszee7Lbew+lkf/dlH8aVgiHZprd99Kqco0KBqhA1n5vPjvHSzZfpSWEU2Yem9Pbkhsrqe7KqWqpEHRCBhjOJpbxOa0E/ywN4uP1h7Ex1t4ckhHJvRrrXeTU0q5pEHRAGWcLGLL4RNsTsthS1oOmw/nkGHfj9rbSxjePZbfD+1E85AAN1eqlKoPNCjquez8YrYczmFLmh0Mh3NIzykEQATaRjfl6nZRdI0LpVtcKIkxoQT66RaEUqr6NCjqkZyCErYdtrYQNtvBkHa8oOL11lFB9EqIoFtcKF1bhNK5RShN9SpqpdQl0k+RWlJWZigqLaOwxEFhqYPCEnu8xB4vdVBUckZ7+fR2W5E9X35RKbuP5bE/M79i+fERgXSPC+OePq3oZodCaKDe/0EpVfPOGxQiEg98AFwGlAFTjTGvi0gE8DGQAKQCdxpjjtvzPAXcBziAR4wxi+32nsAMIBD4GnjUGGNExN/+GT2BLOAuY0yqPc9Y4I92OS8aY2Ze8ruuZT/szeSR2RvJzCu+qPn9vL3w9/UiwNebAF8vAny86dC8KSN6xtG1hbW1EB6k1zkopepGdbYoSoH/Nsb8KCLBwAYRWQKMA5YaY14WkcnAZOD3IpIIjAI6A7HANyLSwRjjAN4BJgJrsIJiCLAQK1SOG2Paicgo4BXgLjuMngGSAWP/7C/KA8kTzVl3kD9+tpXWUUHc17+N9UHv9IEf4Ot9OgR8vJ1et8b9fbzx9tLTVJVSnuO8QWGMSQfS7fGTIrIDaAEMBwbYk80ElgO/t9vnGGOKgP0isgfoLSKpQIgxZjWAiHwA3IoVFMOBZ+1lzQPeEuuk/sHAEmNMtj3PEqxwmX0J77lWOMoMryzaydQV+7imQzRv/TqJkADdFaSUqv8u6BiFiCQAScBaoLkdIhhj0kWkmT1ZC6wthnJpdluJPX5me/k8h+xllYpIDhDp3F7FPB4jv6iUR+ds4psdRxlzVSv+dHOi9pGklGowqh0UItIUmA88ZozJdXEVb1UvGBftFzuPc20TsXZp0bJly3PVVSvScwq4b0YKO3/J5blbOjO2b0Kd/nyllKpt1fraKyK+WCHxoTHmU7v5qIjE2K/HAMfs9jQg3mn2OOCI3R5XRXuleUTEBwgFsl0sqxJjzFRjTLIxJjk6Oro6b6lGbE47wfC3VnEw+xTTxvXSkFBKNUjnDQr7WME0YIcxZorTS18AY+3xscDnTu2jRMRfRFoD7YF19m6qkyLSx17mmDPmKV/WCGCZMcYAi4EbRSRcRMKBG+02t1u0NZ07312Nr7cX8x/sy3Udm51/JqWUqoeqs+upH3AvsEVENtltfwBeBj4RkfuAg8BIAGPMNhH5BNiOdcbUQ/YZTwAPcvr02IX2AFYQ/cs+8J2NddYUxphsEXkBWG9P93z5gW13Mcbwznd7+cuiXSS1DGPqvclEB/u7sySllKpVYn1xbziSk5NNSkpKrSy7uLSMPyzYwrwNadzSPZa/jOimHeoppRoEEdlgjEmu6jW9MruajucXM2nWBtbtz+ax69vz6KD22i23UqpR0KCohr0ZeUyYsZ70nEJeH9WD4T087gxdpZSqNRoU57FqTyYPztqAr7cXs3/Th56twt1dklJK1SkNChdmrzvI059tpU10ENPG9iI+oom7S1JKqTqnQVEFR5nhz1/v4J8r93Ot3R1HsHbHoZRqpDQozuDcHce4vgn88VeXa3ccSqlGTYPCyZETBdw3M4Vdv+Ty/PDOjLkqwd0lKaWU22lQ2FIz8xn57moKix1MH9eLAXqltVJKARoUFVqEB3Jth2gmXtOGDs2D3V2OUkp5DA0Km6+3F6+O7O7uMpRSyuPoUVqllFIuaVAopZRySYNCKaWUSxoUSimlXNKgUEop5ZIGhVJKKZc0KJRSSrmkQaGUUsolDQqllFIuaVAopZRySYNCKaWUSxoUSimlXNKgUEop5ZIGhVJKKZc0KJRSSrmkQaGUUsolDQqllFIuaVAopZRySYNCKaWUSxoUSimlXNKgUEop5ZIGhVJKKZc0KJRSSrmkQaGUUsolDQqllFIuaVAopZRy6bxBISLTReSYiGx1antWRA6LyCZ7uMnptadEZI+I7BKRwU7tPUVki/3aGyIidru/iHxst68VkQSnecaKyG57GFtj71oppVS1VWeLYgYwpIr214wxPezhawARSQRGAZ3ted4WEW97+neAiUB7eyhf5n3AcWNMO+A14BV7WRHAM8CVQG/gGREJv+B3qJRS6pKcNyiMMSuA7GoubzgwxxhTZIzZD+wBeotIDBBijFltjDHAB8CtTvPMtMfnAYPsrY3BwBJjTLYx5jiwhKoDSymlVC26lGMUD4vIZnvXVPk3/RbAIadp0uy2Fvb4me2V5jHGlAI5QKSLZZ1FRCaKSIqIpGRkZFzCW1JKKXWmiw2Kd4C2QA8gHfir3S5VTGtctF/sPJUbjZlqjEk2xiRHR0e7KFsppdSFuqigMMYcNcY4jDFlwHtYxxDA+tYf7zRpHHDEbo+ror3SPCLiA4Ri7eo617KUUkrVoYsKCvuYQ7nbgPIzor4ARtlnMrXGOmi9zhiTDpwUkT728YcxwOdO85Sf0TQCWGYfx1gM3Cgi4faurRvtNqWUUnXI53wTiMhsYAAQJSJpWGciDRCRHli7glKBSQDGmG0i8gmwHSgFHjLGOOxFPYh1BlUgsNAeAKYB/xKRPVhbEqPsZWWLyAvAenu6540x1T2orpRSqoaI9eW94UhOTjYpKSnuLkMppeoVEdlgjEmu6jW9MlsppZRLGhRKKaVc0qBQSinlkgaFUkoplzQolFJKuaRBoZRSyiUNCqWUUi5pUCillHJJg0IppZRLGhRKKaVc0qBQSinlkgaFUkoplzQolFJKuXTebsaVUjWorAzKSsBRYj+WOj0vdWo/83np2fP5+IF/MPiH2o/BEBACvkHgpd8BVc3RoFCqpjlK4cQByNgJGbsg82f7cTcUn6yDAsQOjpDKAVI+7h9y+rXy9tB4aHY5+AbWQX2qvtGgUOpilRRC1h7I3GUFQXkoZO0BR/Hp6YJjIKoD9BgNTSLBywe8fcHL1350fu7j1H7m8zPmKy2CopNQlGsPJ6HQfjyz/VQ2HD9wur3k1NnvR7wgoi007wzNu9iPiRDWCqSqW9irxkKDQqnzKcx12irYBRk/W1sLJw6AKbMnEghPgOiO0O566zGqI0R3gIBQd1ZfNUfp6RApyoXs/XB0GxzdCuk/wfbPTk/rF2wFRvPOp0Ok2eWe+b5UrdA73ClVlYxdsGUubP0Usveebvf2g8h21hZCdMfTj5HtGtZum6I8OLbDCo6j26zh2DYozDk9TWhLp/CwAySijbUVpOodV3e407+oUuVyDsPW+bDlE/hli7UrpvW1kHSPFQbRnazdMI3hg9C/KcT3soZyxkDu4dNbHuUBsvs/YBzWND4BENQMfPydhgArYH0CXLQHWAfnz5w2oT80beae34Gq0AjWeKVcKDgO2z+HLfMgdSVgoEVPGPIydL4dgpu7u0LPIQKhcdbQYfDp9pJCa5dceXCcyoLSQusYSvlQnAenMiu3OcrHC5124Z0hrBU88L3u5nIzDQrV+JQUwM+LYPNc69twWYm162jAU9B1BES2dXeF9YtvAMR0t4aL5Si1AsNRbIdMobX7b87d8OWjMOJ9PaDuRhoUqnFwlML+76wthx1fWqepNr0Mek+EbiMhpod+ELmTtw94N63cFtEGBv4Rlj4Hba6DnmPdU5vSoFANmDFweMPpg9L5x6zrBzoPh64jIeFq8PJ2d5XKlX6PWQG/8PcQ39s620rVOQ0K1bCUFlm7LHZ8aQXE8f3g7W/tU+86EtrfaO0qUfWDlxfcNhX+0Q/mjoeJ3zass8vqCQ0KVT8V55++tqFi2GkFgykDBFpfA9c8AZ1uhsAwd1esLlZwc7jtHzDrDlj0FAz7m7sranQ0KJRnKzhuXeDmfPVzxi7IOXh6Gi+f01cUd7ndutAtoT+ExLivblWz2l0P/R6FVa9Dm2uh823urqhR0aBQ7lfmgLxjZ3SHsdMKiLxfTk/nEwBR7a191VeMsa56ju5kX+Tl6776Vd0Y+DSkroIvHoXYKyC8lbsrajQ0KFTtKi22Puxzj5wxHIaT6db4yXSrp9RyfsF2VxiDnLrC6AhhLfXgc2Pm7QsjpsE/rob598H4hfoFoY5oUKiLV5wPJ3+xPvTPFQR5x4AzuonxbQIhLaxdQwn9ISTW6jgvsq21hRAco6eqqqqFJ8Atb8DccfDtS3D9s24uqHHQoFCnlTmsYwL5GdYHfH4G5Gdap5WWjzu3l+SfvYyAMDsEYiGm2+nxkFgIth8DQjUI1MXrfBvsWw4rX7NOcW43yN0VNXgaFA2FMdZVrcX51pXHJafsocCprcD6cD+VZQdAeSDY46cyq+5KQbwhKAqCoq0hoo3V/06TyNMhENLC2hLwa1L37101PoP/DAfXwoJJ8MAq7WqllmlQ1Jbs/ZB31PrwdhRbVwY7ik/fpcxRbD+WnNHu9FqZ03jFB70dAMVOQVDedq7+cqri1/T0h394AsQlWx/+QdF2e7PTwRAYrndMU57FrwmMmA7vXWeFxT2f6jpaizQoalpZGaz8Kyx7ibP2zVeHeFu9Z3r7Vr5JjW+gtW/ft4m9eyfWfh5o3frSN9Aa/IIqT1s+7uf0PDBCv/mr+q95otV541ePwQ+vQ//H3V1Rg6VBUZMKc2DBg7Dr39D1TuuOZt5+1uDlU0UAOD339rPa9FuRUtXXc5x1vGLZi9Cqf+Vu0VWN0aCoKcd2wsd3w/FUGPoXq7M5PWCrVO0SgWGvw5EfYf4EmPS9XoVfC/Tra03Y9hm8N9C6ZeaYL+DKSRoSStWVwDC4Y7p1WvaXj1ondqgadd6gEJHpInJMRLY6tUWIyBIR2W0/hju99pSI7BGRXSIy2Km9p4hssV97Q8T6JBURfxH52G5fKyIJTvOMtX/GbhHxvD6GHaWw5E8wd6y1v3TSd5DQz91VKdX4xPeyrtze/hlsmOHuahqc6mxRzACGnNE2GVhqjGkPLLWfIyKJwCigsz3P2yJSfintO8BEoL09lC/zPuC4MaYd8Brwir2sCOAZ4EqgN/CMcyC5XX4mzLrd6nsm+T4Y97V1gFkp5R59H4G2A2HRZDi63d3VNCjnDQpjzAog+4zm4cBMe3wmcKtT+xxjTJExZj+wB+gtIjFAiDFmtTHGAB+cMU/5suYBg+ytjcHAEmNMtjHmOLCEswPLPQ7/CFMHwME1MPxtuHmKdb9fpZT7eHnBbe9a9xyZN946hVzViIs9RtHcGJMOYD+W3/28BXDIabo0u62FPX5me6V5jDGlQA4Q6WJZZxGRiSKSIiIpGRkZF/mWqmnjLJhu59V9iyHp7tr9eUqp6mvaDG5/1+pUctFkd1fTYNT0weyqjuAaF+0XO0/lRmOmGmOSjTHJ0dHR1Sr0gpUWwVePw+cPQaurYOJ3EJtUOz9LKXXx2g60rqn4caZ1Z0N1yS42KI7au5OwH4/Z7WlAvNN0ccARuz2uivZK84iIDxCKtavrXMuqe7lHYMavIGW6dWvGez6FoEi3lKKUqobr/gfiellnQR1PdXc19d7FBsUXQPlZSGOBz53aR9lnMrXGOmi9zt49dVJE+tjHH8acMU/5skYAy+zjGIuBG0Uk3D6IfaPdVrdSV8G718CxHXDnB3DDc9rVtVKeztsX7pgGCMybYHWDoy5adU6PnQ2sBjqKSJqI3Ae8DNwgIruBG+znGGO2AZ8A24FFwEPGGIe9qAeBf2Id4N4LLLTbpwGRIrIH+C/sM6iMMdnAC8B6e3jebqsbxsCad2DmMKu30/uXQuLwOvvxSqlLFN7K6pL88AZY9oK7q6nXxDSwi1OSk5NNSkrKpS2kON/aZN0yFzr+yrpfb0BIzRSolKpbXz1u7Ta+81+QeIu7q/FYIrLBGJNc1Wt6ZfaZsvfBtBthyzzrAp67ZmlIKFWfDf5faN4VPrkX3v8V7FmqV29fIA0KZ7uXWNdH5KTBPfPgmie0kz6l6jvfQOtU9sF/tr4Izrrd6p58x5dWb8/qvPRTsFzGz/DRndZ9mSd9B+2ud3dFSqma4hcEV/0WHt0Ew96AghPw8T3wdh/4aY4e7D4PPUbhbOun0GGI3qtBqYbOUWr1C/X9FDi2zfqC2O9R6HEP+Aa4uzq3cHWMQoNCKdV4lZXB7sWw4lU4nAJNm8NVD0PyePAPdnd1dUoPZiulVFW8vKDjULj/Gxj7JTS7HJY8Da91gW//DKfq7ox8T6ZBoZRSItD6GhjzOdy/DBL6w3cvW4Gx+H/g5C/urtCtNCiUUspZXE8Y9SE8uBo6/QrWvA1/62pdj9FIuwPRYxRKKeVK9j5Y9QZs+hDKHNB1BHQfDZHtIKRFgzmFXg9mK6XUpco9Aqv/bl3lXWLf68LbH8ITIKKNPbQ+PR4aD94+bi35QmhQKKVUTSk4DumbrS2NimG/9VhacHo6Lx8Ia+UUIk5DWEuPu9mZq6CoP3GnlFKeIDAc2lxrDc6MsQ56VwoQezi4BopPnp5WvKwtjog2ENnW2o0V2c4aD23pcVsinlWNUkrVVyIQEmMNCf0qv2YM5GdWESJ7YfNcKMo5Pa2Xr7U7qzw4nEMkOMb6OXVMg0IppWqbCDSNtoaWV1Z+zRg4lQVZe84Y9sG+b6G08PS0vkEQ2cYpPNpBRFsrRJpE1Fr5GhRKKeVOIhAUZQ0t+1R+rawMcg87hcdeaysk/SfY/gVU3O4HCIyAttfBiOk1XqIGhVJKeSovLwiLt4a211V+rbQYThyovBUSWDtbFRoUSilVH/n4QVR7a6hlDeNKEaWUUrVGg0IppZRLGhRKKaVc0qBQSinlkgaFUkoplzQolFJKuaRBoZRSyiUNCqWUUi41uG7GRSQDOHAJi4gCMmuonJqkdV0YrevCaF0XpiHW1coYE13VCw0uKC6ViKScq092d9K6LozWdWG0rgvT2OrSXU9KKaVc0qBQSinlkgbF2aa6u4Bz0LoujNZ1YbSuC9Oo6tJjFEoppVzSLQqllFIuaVAopZRyqcEHhYjEi8i3IrJDRLaJyKN2e4SILBGR3fZjuNM8T4nIHhHZJSKDq1jmFyKy1VPqEhE/EZkqIj+LyE4RucND6hotIltEZLOILBKRqLqqS0Qi7enzROStM5bV065rj4i8IXLxd6uvqbpEpImI/Nv++20TkZcvtqaarOuMZdb5en+ev6Pb1vvz1OXO9f4GEdlg//wNIjLQaVkXv94bYxr0AMQAV9jjwcDPQCLwF2Cy3T4ZeMUeTwR+AvyB1sBewNtpebcDHwFbPaUu4DngRXvcC4hyd11Yd088Vl6LPf+zdVhXENAfeAB464xlrQOuAgRYCAx1d11AE+A6e9wP+N4T6vKA9d7V39Gd6/25/o7uXu+TgFh7vAtwuCbW+4v+g9fXAfgcuAHYBcQ4/TF22eNPAU85Tb8YuMoebwqstP9Ql/QPU8N1HQKCPOn3BfgCGUAre8X8BzCxrupymm7cGf/IMcBOp+ejgXfdXVcVy3kd+I0n1OXO9f48dbltvXexfnnEem+3C5CF9SXuktb7Br/ryZmIJGAl7lqguTEmHcB+bGZP1gJrBSyXZrcBvAD8FTjlKXWJSFh5bSLyo4jMFZHm7q7LGFMCPAhsAY5gfchMq8O6zqWFXWOlej2gLuflhAHDgKUeUpc71/tzzRtWXpub1vsqedh6fwew0RhTxCWu940mKESkKTAfeMwYk+tq0irajIj0ANoZYxZ4Ul1Ym7pxwCpjzBXAauBVd9clIr5Y/zBJQCywGWvro67qOuciqmi75HPEa6Cu8uX4ALOBN4wx+9xdlwes9+fi7vX+XPN7xHovIp2BV4BJ5U1VTFbt9b5RBIX9x5sPfGiM+dRuPioiMfbrMVj7FcFK2nin2eOwvhlcBfQUkVSszfAOIrLcA+rKwvqmV/6PPBe4wgPq6gFgjNlrrG3dT4C+dVjXuaTZNZ5Zr7vrKjcV2G2M+dul1FSDdbl7vT8Xd6/359ID3Lvei0gc1u9ljDFmr918Set9gw8K+8j+NGCHMWaK00tfAGPt8bFY+/7K20eJiL+ItAbaA+uMMe8YY2KNMQlYB7F+NsYM8IC6DPAlUF7LIGC7u+sCDgOJIlLeG+UNwI46rKtK9mb6SRHpYy9zzPnmqYu67GW9CIQCj11sPTVdlwes9+eqy93r/bm4db23d8n9G+u44aryiS95va+pgyyeOmCt3AZrE3CTPdwERGLtA95tP0Y4zfM/WGfv7KKKMwOABC797I8aqwvrwNkKe1lLgZYeUtcDWP8km7H+qSPruK5UIBvIw/pGlWi3JwNb7Zrfwu6hwJ11YX3DM/bvq3w597u7Lg9Z78/1d3T3en+uuty23gN/BPKdpt0ENLvU9V678FBKKeVSg9/1pJRS6tJoUCillHJJg0IppZRLGhRKKaVc0qBQSinlkgaFUkoplzQolPJAIuLt7hqUKqdBodQlEpEXyu8TYD9/SUQeEZH/JyLrxbovwXNOr39m3ytgm4hMdGrPE5HnRWQtVtcZSnkEDQqlLt007O4URMQLGAUcxerOpDdW/z89ReQae/oJxpieWFfKPiIikXZ7ENaVz1caY1bWYf1KueTj7gKUqu+MMakikiUiSUBzYCPQC7jRHgfrng7tsbqceEREbrPb4+32LMCB1fmbUh5Fg0KpmvFPrJvYXAZMx+qk7s/GmHedJxKRAcD1WDedOmX3xBpgv1xojHHUUb1KVZvuelKqZiwAhmBtSSy2hwn2fQQQkRYi0gyrd9jjdkh0Avq4q2Clqku3KJSqAcaYYhH5FjhhbxX8R0QuB1ZbvTqTB9wDLAIeEJHNWL3trnFXzUpVl/Yeq1QNsA9i/wiMNMbsdnc9StUk3fWk1CUSkURgD7BUQ0I1RLpFoZRSyiXdolBKKeWSBoVSSimXNCiUUkq5pEGhlFLKJQ0KpZRSLv1/myAbwHE4QqcAAAAASUVORK5CYII=\n",
      "text/plain": "<Figure size 432x288 with 1 Axes>"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cc[cc['speaker_role'] == 'Management'].groupby('year').size().plot(label='Management', legend=True)\r\n",
    "cc[cc['speaker_role'] == 'Analyst'].groupby('year').size().plot(label='Analysts', legend=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": "<AxesSubplot:xlabel='year'>"
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": "<Figure size 432x288 with 1 Axes>"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cc[cc['speaker_role'] == 'Management'].groupby('year')['num_words'].sum().plot(label='Management', legend=True)\r\n",
    "cc[cc['speaker_role'] == 'Analyst'].groupby('year')['num_words'].sum().plot(label='Analysts', legend=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>year</th>\n      <th>Frequency</th>\n      <th>Percent</th>\n      <th>Cumulative</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>2004</td>\n      <td>1370</td>\n      <td>5.221634</td>\n      <td>5.221634</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>2005</td>\n      <td>1451</td>\n      <td>5.530358</td>\n      <td>10.751991</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2006</td>\n      <td>1455</td>\n      <td>5.545604</td>\n      <td>16.297595</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>2007</td>\n      <td>1497</td>\n      <td>5.705683</td>\n      <td>22.003278</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>2008</td>\n      <td>1537</td>\n      <td>5.858139</td>\n      <td>27.861417</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>2009</td>\n      <td>1550</td>\n      <td>5.907688</td>\n      <td>33.769105</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>2010</td>\n      <td>1580</td>\n      <td>6.022030</td>\n      <td>39.791135</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>2011</td>\n      <td>1575</td>\n      <td>6.002973</td>\n      <td>45.794108</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>2012</td>\n      <td>1591</td>\n      <td>6.063955</td>\n      <td>51.858063</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>2013</td>\n      <td>1605</td>\n      <td>6.117315</td>\n      <td>57.975378</td>\n    </tr>\n    <tr>\n      <th>10</th>\n      <td>2014</td>\n      <td>1584</td>\n      <td>6.037276</td>\n      <td>64.012654</td>\n    </tr>\n    <tr>\n      <th>11</th>\n      <td>2015</td>\n      <td>1561</td>\n      <td>5.949613</td>\n      <td>69.962267</td>\n    </tr>\n    <tr>\n      <th>12</th>\n      <td>2016</td>\n      <td>1566</td>\n      <td>5.968670</td>\n      <td>75.930937</td>\n    </tr>\n    <tr>\n      <th>13</th>\n      <td>2017</td>\n      <td>1587</td>\n      <td>6.048710</td>\n      <td>81.979647</td>\n    </tr>\n    <tr>\n      <th>14</th>\n      <td>2018</td>\n      <td>1592</td>\n      <td>6.067767</td>\n      <td>88.047414</td>\n    </tr>\n    <tr>\n      <th>15</th>\n      <td>2019</td>\n      <td>1599</td>\n      <td>6.094447</td>\n      <td>94.141861</td>\n    </tr>\n    <tr>\n      <th>16</th>\n      <td>2020</td>\n      <td>1537</td>\n      <td>5.858139</td>\n      <td>100.000000</td>\n    </tr>\n  </tbody>\n</table>\n</div>",
      "text/plain": "    year  Frequency   Percent  Cumulative\n0   2004       1370  5.221634    5.221634\n1   2005       1451  5.530358   10.751991\n2   2006       1455  5.545604   16.297595\n3   2007       1497  5.705683   22.003278\n4   2008       1537  5.858139   27.861417\n5   2009       1550  5.907688   33.769105\n6   2010       1580  6.022030   39.791135\n7   2011       1575  6.002973   45.794108\n8   2012       1591  6.063955   51.858063\n9   2013       1605  6.117315   57.975378\n10  2014       1584  6.037276   64.012654\n11  2015       1561  5.949613   69.962267\n12  2016       1566  5.968670   75.930937\n13  2017       1587  6.048710   81.979647\n14  2018       1592  6.067767   88.047414\n15  2019       1599  6.094447   94.141861\n16  2020       1537  5.858139  100.000000"
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distribution = (cc.groupby('year')['transcript_id']\r\n",
    "                  .nunique()\r\n",
    "                  .sort_index()\r\n",
    "                  .to_frame('Frequency')\r\n",
    "                  .assign(Percent=lambda x: x['Frequency'] / x['Frequency'].sum() * 100,\r\n",
    "                          Cumulative=lambda x: x['Percent'].cumsum())\r\n",
    "                  .reset_index())\r\n",
    "distribution"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "4ad8bff83ca7a203c8f5c1f2f3a00576dbcaaca98e752b3367803050219b2dc1"
  },
  "kernelspec": {
   "display_name": "Python 3.7.0 64-bit ('ccs': conda)",
   "name": "python3"
  },
  "language_info": {
   "name": "python",
   "version": ""
  },
  "orig_nbformat": 2
 },
 "nbformat": 4,
 "nbformat_minor": 2
}