analyses.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Analyses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "from utils.plot_utils import compare_parts_doggos, qual_helper, pearsonr_ci, get_questionnaire_dataframe\n",
    "from utils.plot_utils import regplot_helper, readibility_helper, helper_k, lines, bars_plotter\n",
    "from utils.plot_utils import big_plot_helper, table_helper, group_by_count, compare_parts\n",
    "from utils.categories import SetTokens\n",
    "from utils.hopper import Hopper\n",
    "from itertools import product, combinations\n",
    "from matplotlib.lines import Line2D\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import patches\n",
    "import networkx as nx\n",
    "import seaborn as sns\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import textstat\n",
    "import warnings\n",
    "\n",
    "warnings.filterwarnings('ignore')\n",
    "sns.set_style(style=\"whitegrid\")\n",
    "sns.set_context(\"paper\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "num_hops = 5\n",
    "\n",
    "topics = [\"breast\", \"cardio\", \"diet\", \"immunization\"]\n",
    "\n",
    "qualifications = [\"breast:3KSRR0ANT0SQYOX6MT2LB8HJ8EFCHJ\", \"cardio:3RJTH7YUGU2T60P6IF1ZMICQMWRYTM\",\n",
    "                  \"nutrition:36FDFKUBEQYUVKNYWYMXNPALOD93NJ\", \"immunization:3RVVHO1V7S1LCQ80JEK422CLLUJWGC\"]\n",
    "\n",
    "# =====================================================================\n",
    "# This normalizes things such that each sub-categorie counts the same\n",
    "# =====================================================================\n",
    "\n",
    "weights_categories = {\"Participants\": 1 / 5, \"Intervention\": 1 / 4, \"Outcomes\": 1 / 3, \"Conclusion\": 1}\n",
    "weights_categories_tag = {\"Participants\": 1 / 5, \"Intervention\": 1 / 4, \"Outcomes\": 1 / 3, \"Conclusion\": 1}\n",
    "\n",
    "# ===================================================================\n",
    "# This relates to the questions in the qualification questionnaires\n",
    "# ===================================================================\n",
    "\n",
    "true_or_false = {\n",
    "    \"breast\": [\"1\", \"2\", \"3\", \"4\", \"5\", \"6\", \"7\", \"8\", \"9\", \"10\", \"11\", \"12\", \"13\"],\n",
    "    \"cardio\": [\"1\", \"2\", \"3\", \"4\", \"5\", \"6\", \"7\", \"8\", \"9\", \"10\", \"11\", \"12\", \"13\"],\n",
    "    \"immunization\": [\"1\", \"2\", \"3\", \"4\", \"5\", \"6\", \"7\", \"8\"],\n",
    "    \"diet\": ['1_1', '1_2', '1_3', '1_4', '1_5', '1_6', '1_7', '1_8', '2', '3_1',\n",
    "             '3_2', '3_3', '4', '5', '6', '7', '8', '9']\n",
    "}\n",
    "# ==================================\n",
    "# Maps values to a nice phrase :)\n",
    "# ==================================\n",
    "\n",
    "mapping_vals = {\"A\": \"A: Fact captured\", \"B\": \"B: Fact partially captured\", \"C\": \"C: Fact not captured\",\n",
    "                \"D\": \"D: Fact contradicts the text\", \"Keyphrases\": \"Keyphrases\"}\n",
    "\n",
    "# =========================\n",
    "# Color map for the plots\n",
    "# ========================= \n",
    "\n",
    "colors_map = {\"A\": \"#226666\", \"B\": \"#552700\", \"C\": \"#2D882D\", \"D\": \"#AA3939\", \"Participants\": \"#276666\",\n",
    "              \"Intervention\": \"#d29460\", \"Outcomes\": \"#ef88bf\", \"Conclusion\": \"#a83c3b\", \"Keyphrases\": \"#4C72B0\",\n",
    "              None: \"#4C72B0\"}\n",
    "\n",
    "# ===========================\n",
    "# Maps branches into papers\n",
    "# ===========================\n",
    "\n",
    "branch_to_topic = {\n",
    "    \"NEJMoa1000727\": \"breast\", \"NEJMoa1103507\": \"breast\", \"NEJMoa1206809\": \"breast\",\n",
    "    \"NEJMoa1404037\": \"breast\", \"NEJMoa072761\": \"cardio\", \"NEJMoa1003114\": \"cardio\",\n",
    "    \"NEJMoa1003833\": \"cardio\", \"NEJMoa1304127\": \"cardio\", \"NEJMoa0804748\": \"diet\",\n",
    "    \"NEJMoa1014296\": \"diet\", \"NEJMoa1112010\": \"diet\", \"NEJMoa1200303\": \"diet\",\n",
    "    \"NEJMoa021134\": \"immunization\", \"NEJMoa0907413\": \"immunization\",\n",
    "    \"NEJMoa1102287\": \"immunization\", \"NEJMoa1200850\": \"immunization\"\n",
    "}\n",
    "\n",
    "# =================================================\n",
    "# These helpers are useful to the doggos analyses\n",
    "# =================================================\n",
    "\n",
    "category_helper_tag = {\n",
    "    \"Coarse\": [\"Participants\", \"Intervention\", \"Outcomes\"],\n",
    "    \"Fine\": [\"Sex\", \"Age\", \"Condition\", \"Sample-size\", \"Location\",\n",
    "             \"Outcomes-general\", \"Duration\", \"Intensity\", \"Control\",\n",
    "             \"Intervention-general\", \"Effect-strength\", \"Adverse-effects\"]\n",
    "}\n",
    "\n",
    "hierarchy_helper_tag = {\n",
    "    \"Hierarchy\": {\"Coarse\": \"Fine\"},\n",
    "    \"Hierarchy_Mapping\": {\n",
    "        \"Fine-Coarse\": {\"Sex\": \"Participants\", \"Age\": \"Participants\",\n",
    "                        \"Condition\": \"Participants\", \"Sample-size\": \"Participnts\",\n",
    "                        \"Location\": \"Participants\", \"Duration\": \"Intervention\",\n",
    "                        \"Intervention-general\": \"Intervention\", \"Intensity\": \"Intervention\",\n",
    "                        \"Control\": \"Intervention\", \"Outcomes-general\": \"Outcomes\",\n",
    "                        \"Effect-strength\": \"Outcomes\", \"Adverse-effects\": \"Outcomes\"\n",
    "                        }\n",
    "    }\n",
    "}\n",
    "\n",
    "category_helper = {\n",
    "    \"Coarse\": [\"Participants\", \"Intervention\", \"Outcomes\", \"Conclusion\"],\n",
    "    \"Fine\": [\"Sex\", \"Age\", \"Condition\", \"Sample-size\", \"Location\",\n",
    "             \"Intervention-general\", \"Duration\", \"Intensity\", \"Control\",\n",
    "             \"Outcomes-general\", \"Effect-strength\", \"Adverse-effects\",\n",
    "             \"Conclusion-general\"\n",
    "             ]\n",
    "}\n",
    "\n",
    "hierarchy_helper = {\n",
    "    \"Hierarchy\": {\"Coarse\": \"Fine\"},\n",
    "    \"Hierarchy_Mapping\": {\n",
    "        \"Fine-Coarse\": {\"Sex\": \"Participants\", \"Age\": \"Participants\", \"Condition\": \"Participants\",\n",
    "                        \"Sample-size\": \"Participants\", \"Location\": \"Participants\", \"Duration\":\n",
    "                            \"Intervention\", \"Intensity\": \"Intervention\", \"Control\": \"Intervention\",\n",
    "                        \"Effect-strength\": \"Outcomes\", \"Adverse-effects\": \"Outcomes\",\n",
    "                        \"Outcomes-general\": \"Outcomes\", \"Intervention-general\": \"Intervention\",\n",
    "                        \"Conclusion-general\": \"Conclusion\"\n",
    "                        },\n",
    "        \"Coarse-Fine\": {\"Participants\": [\"Sex\", \"Age\", \"Condition\", \"Sample-size\", \"Location\"],\n",
    "                        \"Intervention\": [\"Intervention-general\", \"Duration\",\n",
    "                                         \"Intensity\", \"Control\"],\n",
    "                        \"Outcomes-general\": [\"Conclusion-general\", \"Intensity\", \"Control\",\n",
    "                                             \"Effect-strength\", \"Adverse-effects\"],\n",
    "                        \"Conclusion\": [\"Conclusion-general\"]\n",
    "                        }\n",
    "    }\n",
    "}\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Loads the graphs :)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "hopper = Hopper()\n",
    "hopper.g = nx.read_gpickle(\"./data/iterative_cascades.gpickle\")\n",
    "hopper.num_hops = num_hops\n",
    "hopper_control = Hopper()\n",
    "hopper_control.g = nx.read_gpickle(\"./data/control_cascades.gpickle\")\n",
    "hopper_control.num_hops = num_hops\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Telephone Effect\n",
    "\n",
    "In our first analysis, we compare the distortion of information in the cascading and control conditions.\n",
    "\n",
    "## Coarse\n",
    "\n",
    "We begin obtaining two dataframes: `df_plots` and `df_props`. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tmp_root(root):\n",
    "    return root[\"Doggos_crowd\"][\"Fine\"]\n",
    "\n",
    "\n",
    "def tmp_node(node, root):\n",
    "    return node[\"Doggos_crowd\"][\"Fine\"]\n",
    "\n",
    "\n",
    "df_1 = hopper.get_tagging_dataframe_list(tmp_node, hierarchy_helper=hierarchy_helper, f_root=tmp_root)\n",
    "df_1[\"control\"] = 0\n",
    "df_2 = hopper_control.get_tagging_dataframe_list(tmp_node, hierarchy_helper=hierarchy_helper, f_root=tmp_root)\n",
    "df_2[\"control\"] = 1\n",
    "df = pd.concat([df_1, df_2])\n",
    "df_props = df.groupby([\"hop\", \"control\", \"key_coarse\", \"val\"]).sum().reset_index()\n",
    "df_norm = df.groupby([\"hop\", \"val\", \"control\"]).sum().reset_index()\n",
    "\n",
    "df_props[\"count\"] = df_props.apply(lambda x: dict(x)[\"count\"] /\n",
    "                                             df_norm[(df_norm.hop == dict(x)[\"hop\"]) &\n",
    "                                                     (df_norm.val == dict(x)[\"val\"]) &\n",
    "                                                     (df_norm.control == dict(x)[\"control\"])][\"count\"].values[0],\n",
    "                                   axis=1)\n",
    "\n",
    "df_plots = group_by_count(df, [\"hop\", \"control\", \"branch\", \"key_coarse\"], \"val\", \"count\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The first containts the percentage of facts retained for each hop, branch and category:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>hop</th>\n",
       "      <th>control</th>\n",
       "      <th>branch</th>\n",
       "      <th>key_coarse</th>\n",
       "      <th>val</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Conclusion</td>\n",
       "      <td>A</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Conclusion</td>\n",
       "      <td>B</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Conclusion</td>\n",
       "      <td>C</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Conclusion</td>\n",
       "      <td>D</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   hop  control        branch  key_coarse val  count\n",
       "0    0        0  NEJMoa021134  Conclusion   A    1.0\n",
       "1    0        0  NEJMoa021134  Conclusion   B    0.0\n",
       "2    0        0  NEJMoa021134  Conclusion   C    0.0\n",
       "3    0        0  NEJMoa021134  Conclusion   D    0.0"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_plots.head(4)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The latter contains the distribution of values at each hop for each category:"
   ]
  },
  {
   "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>hop</th>\n",
       "      <th>control</th>\n",
       "      <th>key_coarse</th>\n",
       "      <th>val</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>130</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>Conclusion</td>\n",
       "      <td>A</td>\n",
       "      <td>0.084746</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>131</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>Conclusion</td>\n",
       "      <td>B</td>\n",
       "      <td>0.236948</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>132</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>Conclusion</td>\n",
       "      <td>C</td>\n",
       "      <td>0.047833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>133</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>Conclusion</td>\n",
       "      <td>D</td>\n",
       "      <td>0.447368</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     hop  control  key_coarse val     count\n",
       "130    4        0  Conclusion   A  0.084746\n",
       "131    4        0  Conclusion   B  0.236948\n",
       "132    4        0  Conclusion   C  0.047833\n",
       "133    4        0  Conclusion   D  0.447368"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_props[(df_props.hop == 4) & (df_props.key_coarse == \"Conclusion\") & (df_props.control == 0)].head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We do the same for the keyphrases:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tmp_root(root):\n",
    "    return root[\"Tagging\"].get_occurrences(\"Coarse\", relative=False)\n",
    "\n",
    "\n",
    "def tmp_nodes(x, root):\n",
    "    return x[\"Tagging\"].get_occurrences(\"Coarse\", relative=False)\n",
    "\n",
    "\n",
    "df_1_token = hopper.get_tagging_dataframe(tmp_nodes, tmp_root)\n",
    "df_2_token = hopper_control.get_tagging_dataframe(tmp_nodes, tmp_root)\n",
    "df_1_token[\"control\"] = 0\n",
    "df_2_token[\"control\"] = 1\n",
    "df_tag = pd.concat([df_1_token, df_2_token])\n",
    "df_diff_treatment_tag = df_tag[df_tag.control == 0].sort_values([\"hop\", \"control\", \"branch\", \"key\"])\n",
    "df_plots_tag = df_tag[df_tag.control == 1].sort_values([\"hop\", \"control\", \"branch\", \"key\"])\n",
    "df_plots_tag[\"diff\"] = (df_diff_treatment_tag[\"val_norm\"].values - df_plots_tag[\"val_norm\"].values)\n",
    "df_plots_tag[\"reduction\"] = df_plots_tag[\"hop\"]\n",
    "df_props_tag = df_tag.groupby([\"hop\", \"control\", \"key\"]).sum().reset_index()\n",
    "df_norm_tag = df_tag.groupby([\"hop\", \"control\"]).sum().reset_index()\n",
    "df_props_tag[\"count\"] = df_props_tag.apply(\n",
    "    lambda x: dict(x)[\"val_norm\"] / df_norm_tag[(df_norm_tag.hop == dict(x)[\"hop\"]) &\n",
    "                                                (df_norm_tag.control == dict(x)[\"control\"])][\"val_norm\"].values[0],\n",
    "    axis=1)\n"
   ]
  },
  {
   "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>branch</th>\n",
       "      <th>hop</th>\n",
       "      <th>key</th>\n",
       "      <th>node</th>\n",
       "      <th>val</th>\n",
       "      <th>val_norm</th>\n",
       "      <th>control</th>\n",
       "      <th>diff</th>\n",
       "      <th>reduction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1843</th>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>0</td>\n",
       "      <td>Intervention</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1844</th>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>0</td>\n",
       "      <td>Outcomes</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>6</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1842</th>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>0</td>\n",
       "      <td>Participants</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>982</th>\n",
       "      <td>NEJMoa072761</td>\n",
       "      <td>0</td>\n",
       "      <td>Intervention</td>\n",
       "      <td>NEJMoa072761</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            branch  hop           key          node  val  val_norm  control  \\\n",
       "1843  NEJMoa021134    0  Intervention  NEJMoa021134    3       1.0        1   \n",
       "1844  NEJMoa021134    0      Outcomes  NEJMoa021134    6       1.0        1   \n",
       "1842  NEJMoa021134    0  Participants  NEJMoa021134    3       1.0        1   \n",
       "982   NEJMoa072761    0  Intervention  NEJMoa072761    3       1.0        1   \n",
       "\n",
       "      diff  reduction  \n",
       "1843   0.0          0  \n",
       "1844   0.0          0  \n",
       "1842   0.0          0  \n",
       "982    0.0          0  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_plots_tag.head(4)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>hop</th>\n",
       "      <th>control</th>\n",
       "      <th>key</th>\n",
       "      <th>val</th>\n",
       "      <th>val_norm</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>Intervention</td>\n",
       "      <td>98</td>\n",
       "      <td>24.971212</td>\n",
       "      <td>0.398721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>Outcomes</td>\n",
       "      <td>91</td>\n",
       "      <td>19.909524</td>\n",
       "      <td>0.317900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>Participants</td>\n",
       "      <td>76</td>\n",
       "      <td>17.747619</td>\n",
       "      <td>0.283380</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    hop  control           key  val   val_norm     count\n",
       "24    4        0  Intervention   98  24.971212  0.398721\n",
       "25    4        0      Outcomes   91  19.909524  0.317900\n",
       "26    4        0  Participants   76  17.747619  0.283380"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_props_tag[(df_props_tag.hop == 4) & (df_props_tag.control == 0)].head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We use these dataframes to calculate the percentage of keyphrases in the summaries of a given hop and the percentage of facts with a given value for all papers in the cascading and control settings.\n",
    "We depict the difference between these for different target lengths in **percentage points** in the line plots below.\n",
    "Values above $0\\%$ mean that the percentage of keyphrases or facts of a given value is higher in the cascading summaries. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "plots = [df_plots[\"key_coarse\"] != \"Bla\", df_plots[\"key_coarse\"] == \"Participants\",\n",
    "         df_plots[\"key_coarse\"] == \"Intervention\", df_plots[\"key_coarse\"] == \"Outcomes\",\n",
    "         df_plots[\"key_coarse\"] == \"Conclusion\"]\n",
    "\n",
    "plots_tags = [df_plots_tag.key != \"Bla\", df_plots_tag.key == \"Participants\",\n",
    "              df_plots_tag.key == \"Intervention\", df_plots_tag.key == \"Outcomes\", None]\n",
    "\n",
    "super_categories = [None, \"Participants\", \"Intervention\", \"Outcomes\", \"Conclusion\"]\n",
    "\n",
    "y_ticks_helper = [([-0.15, -0.075, 0, 0.075, 0.15], [\"-15%\", \"-7.5%\", \"0%\", \"7.5%\", \"15%\"], [-0.15, 0.15]),\n",
    "                  ([-0.15, -0.075, 0, 0.075, 0.15], [\"-15%\", \"-7.5%\", \"0%\", \"7.5%\", \"15%\"], [-0.15, 0.15]),\n",
    "                  ([-0.15, -0.075, 0, 0.075, 0.15], [\"-15%\", \"-7.5%\", \"0%\", \"7.5%\", \"15%\"], [-0.15, 0.15]),\n",
    "                  ([-0.15, -0.075, 0, 0.075, 0.15], [\"-15%\", \"-7.5%\", \"0%\", \"7.5%\", \"15%\"], [-0.15, 0.15]),\n",
    "                  ([-0.30, -0.15, 0, 0.15, 0.30], [\"-30%\", \"-15%\", \"0%\", \"15%\", \"30%\"], [-0.30, 0.30])]\n",
    "\n",
    "x_ticks_helper = [([0, 1, 2, 3, 4, 5], [\"\", \"\", \"500\", \"250\", \"125\", \"64\"], [2, 5])]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 25 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axszes = plt.subplots(5, 5, figsize=(15, 6))\n",
    "\n",
    "axszes_tokens = axszes[:, 0]\n",
    "axszes = axszes[:, 1:]\n",
    "for val_idx in range(5):\n",
    "    axs = axszes[val_idx]\n",
    "\n",
    "    df_plot = df_plots[plots[val_idx]]\n",
    "    df_diff_treatment = df_plot[df_plot.control == 0].sort_values([\"hop\", \"control\", \"branch\", \"key_coarse\"])\n",
    "    df_diff_control = df_plot[df_plot.control == 1].sort_values([\"hop\", \"control\", \"branch\", \"key_coarse\"])\n",
    "    df_diff_control[\"diff\"] = (df_diff_treatment[\"count\"].values - df_diff_control[\"count\"].values)\n",
    "    df_diff_control[\"reduction\"] = df_diff_control[\"hop\"]\n",
    "\n",
    "    t_stack, c_stack = [], []\n",
    "    for idx, val in enumerate([\"A\", \"B\", \"C\", \"D\"]):\n",
    "        big_plot_helper(df_diff_treatment, t_stack, df_diff_control, c_stack, val, axs[idx],\n",
    "                        colors_map[super_categories[val_idx]], \"\", mapping_vals[val], val_idx == 0,\n",
    "                        y_ticks_helper[val_idx], False, x_ticks_helper[0], val_idx == 4,\n",
    "                        super_categories[val_idx], False)\n",
    "\n",
    "    if plots_tags[val_idx] is None:\n",
    "        plt.sca(axszes_tokens[val_idx])\n",
    "        axszes_tokens[val_idx].set_frame_on(False)\n",
    "        axszes_tokens[val_idx].grid(False)\n",
    "        plt.xticks([], [])\n",
    "        plt.yticks(y_ticks_helper[val_idx][0], y_ticks_helper[val_idx][1])\n",
    "        plt.ylim(y_ticks_helper[val_idx][2])\n",
    "        plt.ylabel(super_categories[val_idx])\n",
    "    else:\n",
    "        df_plot_tag = df_plots_tag[plots_tags[val_idx]]\n",
    "        big_plot_helper(None, None, None, None, None, axszes_tokens[val_idx],\n",
    "                        colors_map[super_categories[val_idx]], \"\", \"Keyphrases\", val_idx == 0,\n",
    "                        y_ticks_helper[val_idx], True, x_ticks_helper[0], val_idx == 3,\n",
    "                        super_categories[val_idx] if super_categories[val_idx] is not None else \"All\",\n",
    "                        True, df_tag=df_plot_tag)\n",
    "\n",
    "fig.tight_layout()\n",
    "plt.savefig(\"./images/telephone_effect_coarse.pdf\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We also observe how the distribution of categories differ for the distinct settings, which we visualize as tables below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x504 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axszes = plt.subplots(5, 1, figsize=(4, 7))\n",
    "all_stack = []\n",
    "\n",
    "for idx, val in enumerate([\"A\", \"B\", \"C\", \"D\"]):\n",
    "    t_stack, c_stack = [], []\n",
    "    for idy, val_super_cat in enumerate([\"Participants\", \"Intervention\", \"Outcomes\", \"Conclusion\"]):\n",
    "        t_stack.append(df_props[(df_props.val == val) & (df_props.key_coarse == val_super_cat) &\n",
    "                                (df_props.control == 0)][\"count\"])\n",
    "        c_stack.append(df_props[(df_props.val == val) & (df_props.key_coarse == val_super_cat) &\n",
    "                                (df_props.control == 1)][\"count\"])\n",
    "\n",
    "    table_helper(t_stack, c_stack, row_labels=[\"P\", \"I\", \"O\", \"C\"],\n",
    "                 col_labels=[\"500\", \"250\", \"125\", \"64\"], number_cols=4, ax=axszes[idx])\n",
    "\n",
    "    axszes[idx].set_title(mapping_vals[val])\n",
    "\n",
    "t_stack, c_stack = [], []\n",
    "for idy, val_super_cat in enumerate([\"Participants\", \"Intervention\", \"Outcomes\"]):\n",
    "    t_stack.append(df_props_tag[(df_props_tag.key == val_super_cat) & (df_props_tag.control == 0)][\"count\"])\n",
    "    c_stack.append(df_props_tag[(df_props_tag.key == val_super_cat) & (df_props_tag.control == 1)][\"count\"])\n",
    "\n",
    "table_helper(t_stack, c_stack, row_labels=[\"P\", \"I\", \"O\"],\n",
    "             col_labels=[\"500\", \"250\", \"125\", \"64\"], number_cols=3, ax=axszes[4])\n",
    "axszes[4].set_title(mapping_vals[\"Keyphrases\"])\n",
    "\n",
    "fig.tight_layout()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fine\n",
    "\n",
    "Analyzing the coarse categories, we find evidence above that more is lost in the cascading summaries and that the messages that get passed differ. We zoom into the fine categories to drill deeper in this regard."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# DOGGOS\n",
    "\n",
    "\n",
    "def tmp_root(root):\n",
    "    return root[\"Doggos_crowd\"][\"Fine\"]\n",
    "\n",
    "\n",
    "def tmp_node(node, root):\n",
    "    return node[\"Doggos_crowd\"][\"Fine\"]\n",
    "\n",
    "\n",
    "df_1 = hopper.get_tagging_dataframe_list(tmp_node, hierarchy_helper=hierarchy_helper, f_root=tmp_root,\n",
    "                                         output_nodes=True)\n",
    "df_1[\"control\"] = 0\n",
    "df_2 = hopper_control.get_tagging_dataframe_list(tmp_node, hierarchy_helper=hierarchy_helper, f_root=tmp_root,\n",
    "                                                 output_nodes=True)\n",
    "df_2[\"control\"] = 1\n",
    "df = pd.concat([df_1, df_2])\n",
    "# Average\n",
    "df_plot_fine = group_by_count(df, [\"hop\", \"control\", \"branch\", \"key\"], \"val\", \"count\")\n",
    "df_diff_treatment_fine = df_plot_fine[df_plot_fine.control == 0].sort_values([\"hop\", \"control\", \"branch\", \"key\"])\n",
    "df_diff_control_fine = df_plot_fine[df_plot_fine.control == 1].sort_values([\"hop\", \"control\", \"branch\", \"key\"])\n",
    "df_diff_control_fine[\"diff\"] = (df_diff_treatment_fine[\"count\"].values - df_diff_control_fine[\"count\"].values)\n",
    "df_diff_control_fine[\"reduction\"] = df_diff_control_fine[\"hop\"]\n",
    "# Best\n",
    "df_custom = df.groupby([\"branch\", \"node\", \"key\", \"control\", \"hop\"]) \\\n",
    "    .apply(lambda x: x[x.val == \"A\"][\"count\"].sum()).to_frame().reset_index()\n",
    "df_custom[\"father\"] = df_custom.node.apply(lambda x: \"_\".join(x.split(\"_\")[:-1]))\n",
    "df_custom = df_custom.rename({0: \"val\"}, axis='columns')\n",
    "df_t = helper_k(df_custom, 0, category_helper, branch_to_topic, df, num_hops)\n",
    "df_plot_fine_top = group_by_count(df_t, [\"hop\", \"control\", \"branch\", \"key\"], \"val\", \"count\")\n",
    "df_diff_treatment_fine_top = df_plot_fine_top[df_plot_fine_top.control == 0].sort_values([\"hop\", \"control\",\n",
    "                                                                                          \"branch\", \"key\"])\n",
    "df_diff_control_fine_top = df_plot_fine_top[df_plot_fine_top.control == 1].sort_values([\"hop\", \"control\",\n",
    "                                                                                        \"branch\", \"key\"])\n",
    "df_diff_control_fine_top[\"diff\"] = df_diff_treatment_fine_top[\"count\"].values - \\\n",
    "                                   df_diff_control_fine_top[\"count\"].values\n",
    "df_diff_control_fine_top[\"reduction\"] = df_diff_control_fine_top[\"hop\"]\n",
    "# Worse\n",
    "df_custom = df.groupby([\"branch\", \"node\", \"key\", \"control\", \"hop\"]) \\\n",
    "    .apply(lambda x: x[x.val == \"A\"][\"count\"].sum()).to_frame().reset_index()\n",
    "df_custom[\"father\"] = df_custom.node.apply(lambda x: \"_\".join(x.split(\"_\")[:-1]))\n",
    "df_custom = df_custom.rename({0: \"val\"}, axis='columns')\n",
    "df_t = helper_k(df_custom, -1, category_helper, branch_to_topic, df, num_hops)\n",
    "df_plot_fine_bot = group_by_count(df_t, [\"hop\", \"control\", \"branch\", \"key\"], \"val\", \"count\")\n",
    "df_diff_treatment_fine_bot = df_plot_fine_bot[df_plot_fine_bot.control == 0] \\\n",
    "    .sort_values([\"hop\", \"control\", \"branch\", \"key\"])\n",
    "df_diff_control_fine_bot = df_plot_fine_bot[df_plot_fine_bot.control == 1] \\\n",
    "    .sort_values([\"hop\", \"control\", \"branch\", \"key\"])\n",
    "df_diff_control_fine_bot[\"diff\"] = df_diff_treatment_fine_bot[\"count\"].values - \\\n",
    "                                   df_diff_control_fine_bot[\"count\"].values\n",
    "df_diff_control_fine_bot[\"reduction\"] = df_diff_control_fine_bot[\"hop\"]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The above cell creates the percentage of facts retained per each hop. It also considers the \"best ancestor\" and \"worse ancestor\" scenarios!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "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>hop</th>\n",
       "      <th>control</th>\n",
       "      <th>branch</th>\n",
       "      <th>key</th>\n",
       "      <th>val</th>\n",
       "      <th>count</th>\n",
       "      <th>diff</th>\n",
       "      <th>reduction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>840</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Conclusion-general</td>\n",
       "      <td>A</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>841</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Conclusion-general</td>\n",
       "      <td>B</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>842</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Conclusion-general</td>\n",
       "      <td>C</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>843</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Conclusion-general</td>\n",
       "      <td>D</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2265</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Conclusion-general</td>\n",
       "      <td>A</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2266</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Conclusion-general</td>\n",
       "      <td>B</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2267</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Conclusion-general</td>\n",
       "      <td>C</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2268</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Conclusion-general</td>\n",
       "      <td>D</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      hop  control        branch                 key val  count  diff  \\\n",
       "840     0        1  NEJMoa021134  Conclusion-general   A    1.0   0.0   \n",
       "841     0        1  NEJMoa021134  Conclusion-general   B    0.0   0.0   \n",
       "842     0        1  NEJMoa021134  Conclusion-general   C    0.0   0.0   \n",
       "843     0        1  NEJMoa021134  Conclusion-general   D    0.0   0.0   \n",
       "2265    1        1  NEJMoa021134  Conclusion-general   A    1.0   0.0   \n",
       "2266    1        1  NEJMoa021134  Conclusion-general   B    0.0   0.0   \n",
       "2267    1        1  NEJMoa021134  Conclusion-general   C    0.0   0.0   \n",
       "2268    1        1  NEJMoa021134  Conclusion-general   D    0.0   0.0   \n",
       "\n",
       "      reduction  \n",
       "840           0  \n",
       "841           0  \n",
       "842           0  \n",
       "843           0  \n",
       "2265          1  \n",
       "2266          1  \n",
       "2267          1  \n",
       "2268          1  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_diff_control_fine[(df_diff_control_fine.branch == \"NEJMoa021134\") &\n",
    "                     (df_diff_control_fine.key == \"Conclusion-general\")].head(8)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We do the same for keyphrases!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# TOKENS \n",
    "\n",
    "\n",
    "def tagging(x, root):\n",
    "    return x[\"Tagging\"].get_occurrences(\"Fine\", relative=False)\n",
    "\n",
    "\n",
    "def tagging_root(root):\n",
    "    return root[\"Tagging\"].get_occurrences(\"Fine\", relative=False)\n",
    "\n",
    "\n",
    "df_1 = hopper.get_tagging_dataframe(tagging, tagging_root)\n",
    "df_2 = hopper_control.get_tagging_dataframe(tagging, tagging_root)\n",
    "df_1[\"control\"] = 0\n",
    "df_2[\"control\"] = 1\n",
    "df_tag_fine = pd.concat([df_1, df_2])\n",
    "# Average\n",
    "df_diff_treatment_fine_tag = df_tag_fine[df_tag_fine.control == 0] \\\n",
    "    .sort_values([\"hop\", \"control\", \"branch\", \"key\"])\n",
    "df_diff_control_fine_tag = df_tag_fine[df_tag_fine.control == 1].sort_values([\"hop\", \"control\", \"branch\", \"key\"])\n",
    "df_diff_control_fine_tag[\"diff\"] = (df_diff_treatment_fine_tag[\"val_norm\"].values \\\n",
    "                                    - df_diff_control_fine_tag[\"val_norm\"].values)\n",
    "df_diff_control_fine_tag[\"reduction\"] = df_diff_control_fine_tag[\"hop\"]\n",
    "# Best\n",
    "df_custom = df_tag_fine.groupby([\"branch\", \"node\", \"key\", \"control\", \"hop\"]) \\\n",
    "    .apply(lambda x: x[\"val_norm\"].sum()).to_frame().reset_index()\n",
    "df_custom[\"father\"] = df_custom.node.apply(lambda x: \"_\".join(x.split(\"_\")[:-1]))\n",
    "df_custom = df_custom.rename({0: \"val\"}, axis='columns')\n",
    "df_t = helper_k(df_custom, 0, category_helper_tag, branch_to_topic, df_tag_fine, num_hops)\n",
    "df_diff_treatment_fine_tag_top = df_t[df_t.control == 0].sort_values([\"hop\", \"control\", \"branch\", \"key\"])\n",
    "df_diff_control_fine_tag_top = df_t[df_t.control == 1].sort_values([\"hop\", \"control\", \"branch\", \"key\"])\n",
    "df_diff_control_fine_tag_top[\"diff\"] = (df_diff_treatment_fine_tag_top[\"val_norm\"].values \\\n",
    "                                        - df_diff_control_fine_tag_top[\"val_norm\"].values)\n",
    "df_diff_control_fine_tag_top[\"reduction\"] = df_diff_control_fine_tag_top[\"hop\"]\n",
    "# Worst\n",
    "df_custom = df_tag_fine.groupby([\"branch\", \"node\", \"key\", \"control\", \"hop\"]) \\\n",
    "    .apply(lambda x: x[\"val_norm\"].sum()).to_frame().reset_index()\n",
    "df_custom[\"father\"] = df_custom.node.apply(lambda x: \"_\".join(x.split(\"_\")[:-1]))\n",
    "df_custom = df_custom.rename({0: \"val\"}, axis='columns')\n",
    "df_t = helper_k(df_custom, -1, category_helper_tag, branch_to_topic, df_tag_fine, num_hops)\n",
    "df_diff_treatment_fine_tag_bot = df_t[df_t.control == 0].sort_values([\"hop\", \"control\", \"branch\", \"key\"])\n",
    "df_diff_control_fine_tag_bot = df_t[df_t.control == 1].sort_values([\"hop\", \"control\", \"branch\", \"key\"])\n",
    "df_diff_control_fine_tag_bot[\"diff\"] = (df_diff_treatment_fine_tag_bot[\"val_norm\"].values \\\n",
    "                                        - df_diff_control_fine_tag_bot[\"val_norm\"].values)\n",
    "df_diff_control_fine_tag_bot[\"reduction\"] = df_diff_control_fine_tag_bot[\"hop\"]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>node</th>\n",
       "      <th>key</th>\n",
       "      <th>branch</th>\n",
       "      <th>hop</th>\n",
       "      <th>val</th>\n",
       "      <th>val_norm</th>\n",
       "      <th>control</th>\n",
       "      <th>diff</th>\n",
       "      <th>reduction</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9873</th>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>Intervention-general</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9393</th>\n",
       "      <td>NEJMoa021134_3PS7W85Z80NO5VSK9TX20BURQDGT9I</td>\n",
       "      <td>Intervention-general</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9405</th>\n",
       "      <td>NEJMoa021134_3L6L49WXW1ILNNBIR2JDGNDXOJ545X</td>\n",
       "      <td>Intervention-general</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9417</th>\n",
       "      <td>NEJMoa021134_32ZKVD547G82U79NSY5794LJKFIB37</td>\n",
       "      <td>Intervention-general</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9429</th>\n",
       "      <td>NEJMoa021134_30OG32W0SVW75EEHK8RCJQPASUSENY</td>\n",
       "      <td>Intervention-general</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9441</th>\n",
       "      <td>NEJMoa021134_36AHBNMV1SXWP74VVFP3UO5WW20YDK</td>\n",
       "      <td>Intervention-general</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9453</th>\n",
       "      <td>NEJMoa021134_3TEM0PF1Q6IZSC8AN7EU2WBCKPWD09</td>\n",
       "      <td>Intervention-general</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9465</th>\n",
       "      <td>NEJMoa021134_3HYA4D452S43M6PKT0DY9K7CT0PF2Z</td>\n",
       "      <td>Intervention-general</td>\n",
       "      <td>NEJMoa021134</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             node                   key  \\\n",
       "9873                                 NEJMoa021134  Intervention-general   \n",
       "9393  NEJMoa021134_3PS7W85Z80NO5VSK9TX20BURQDGT9I  Intervention-general   \n",
       "9405  NEJMoa021134_3L6L49WXW1ILNNBIR2JDGNDXOJ545X  Intervention-general   \n",
       "9417  NEJMoa021134_32ZKVD547G82U79NSY5794LJKFIB37  Intervention-general   \n",
       "9429  NEJMoa021134_30OG32W0SVW75EEHK8RCJQPASUSENY  Intervention-general   \n",
       "9441  NEJMoa021134_36AHBNMV1SXWP74VVFP3UO5WW20YDK  Intervention-general   \n",
       "9453  NEJMoa021134_3TEM0PF1Q6IZSC8AN7EU2WBCKPWD09  Intervention-general   \n",
       "9465  NEJMoa021134_3HYA4D452S43M6PKT0DY9K7CT0PF2Z  Intervention-general   \n",
       "\n",
       "            branch  hop  val  val_norm  control  diff  reduction  \n",
       "9873  NEJMoa021134    0    1       1.0        1   0.0          0  \n",
       "9393  NEJMoa021134    1    1       1.0        1   0.0          1  \n",
       "9405  NEJMoa021134    1    1       1.0        1   0.0          1  \n",
       "9417  NEJMoa021134    1    1       1.0        1   0.0          1  \n",
       "9429  NEJMoa021134    1    1       1.0        1   0.0          1  \n",
       "9441  NEJMoa021134    1    1       1.0        1   0.0          1  \n",
       "9453  NEJMoa021134    1    1       1.0        1   0.0          1  \n",
       "9465  NEJMoa021134    1    1       1.0        1   0.0          1  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_diff_control_fine_tag_bot[(df_diff_control_fine_tag_bot.branch == \"NEJMoa021134\") &\n",
    "                             (df_diff_control_fine_tag_bot.key == \"Intervention-general\")].head(8)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lastly, in the plot below, we show the differences in the percentage of facts and keyphrases between cascading and control summaries for fine-grained categories. The positive side of the $x$-axis (in gray) means that percentage of facts\\slash keyphrases for that category was higher in the cascading scenario. Each category is associated with three values (with bootstrapped 95\\% CIs) with different markers. The circle $\\circ$ represents the actual difference between cascading and control settings. The upward $\\triangle$ (downward $\\triangledown$) triangle represents the difference between the successor of the best (worst) cascading summary from the previous hop and control. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def lines(axs):\n",
    "    plt.sca(axs)\n",
    "    axs.axhline(y=3 * 0.5 + 1, color=\"black\", alpha=0.8)\n",
    "\n",
    "    axs.axhline(y=3 * 3.5 + 1, color=\"black\", alpha=0.8)\n",
    "\n",
    "    axs.axhline(y=3 * 7.5 + 1, color=\"black\", alpha=0.8)\n",
    "    axs.axhline(y=3 * 13.5 + 1, color=\"black\", alpha=0.8)\n",
    "\n",
    "    axs.fill_between([0, 0.5], [-.5, -.5], [3 * 12.5 + 1, 3 * 12.5 + 1], facecolor=\"black\", alpha=0.15)\n",
    "    for v in np.arange(-0.5, 3 * 12.5 + 1, 3):\n",
    "        axs.axhline(y=v, color=\"black\", alpha=0.4)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axszes = plt.subplots(1, 5, figsize=(15, 6), sharey=True)\n",
    "\n",
    "y_ticks_helper = [([-0.50, -0.25, 0, 0.25, 0.50], [\"-50%\", \"-25%\", \"0%\", \"25%\", \"50%\"], [-0.5, 0.5]),\n",
    "                  ([-0.50, -0.25, 0, 0.25, 0.50], [\"-50%\", \"-25%\", \"0%\", \"25%\", \"50%\"], [-0.5, 0.5])]\n",
    "\n",
    "for idx, val in enumerate([\"Keyphrases\", \"A\", \"B\", \"C\", \"D\"]):\n",
    "    # ORIGINAL\n",
    "    if val is not \"Keyphrases\":\n",
    "        df_tmp_fine = df_diff_control_fine[df_diff_control_fine.val == val]\n",
    "        df_tmp_fine[\"key_coarse\"] = df_tmp_fine.key.apply(\n",
    "            lambda x: hierarchy_helper[\"Hierarchy_Mapping\"][\"Fine-Coarse\"][x])\n",
    "    else:\n",
    "        df_tmp_fine = df_diff_control_fine_tag\n",
    "    bars_plotter(df_tmp_fine, category_helper, val, colors_map, hierarchy_helper, axszes[idx],\n",
    "                 helper_idy=lambda idy: idy * 3 + 1, alpha=1)\n",
    "    # TOP\n",
    "    if val is not \"Keyphrases\":\n",
    "        df_tmp_fine = df_diff_control_fine_top[df_diff_control_fine_top.val == val]\n",
    "        df_tmp_fine[\"key_coarse\"] = df_tmp_fine.key.apply(\n",
    "            lambda x: hierarchy_helper[\"Hierarchy_Mapping\"][\"Fine-Coarse\"][x])\n",
    "    else:\n",
    "        df_tmp_fine = df_diff_control_fine_tag_top\n",
    "    bars_plotter(df_tmp_fine, category_helper, val, colors_map, hierarchy_helper, axszes[idx],\n",
    "                 helper_idy=lambda idy: idy * 3 + 2, alpha=0.8, marker=\"^\", is_err=False)\n",
    "    # BOT\n",
    "    if val is not \"Keyphrases\":\n",
    "        df_tmp_fine = df_diff_control_fine_bot[df_diff_control_fine_bot.val == val]\n",
    "        df_tmp_fine[\"key_coarse\"] = df_tmp_fine.key.apply(\n",
    "            lambda x: hierarchy_helper[\"Hierarchy_Mapping\"][\"Fine-Coarse\"][x])\n",
    "    else:\n",
    "        df_tmp_fine = df_diff_control_fine_tag_bot\n",
    "    bars_plotter(df_tmp_fine, category_helper, val, colors_map, hierarchy_helper, axszes[idx],\n",
    "                 helper_idy=lambda idy: idy * 3, alpha=0.8, marker=\"v\", is_err=False)\n",
    "    axszes[idx].set_xlabel(\"\")\n",
    "    axszes[idx].set_ylabel(\"\")\n",
    "    lines(axszes[idx])\n",
    "    plt.sca(axszes[idx])\n",
    "    plt.xlim(y_ticks_helper[0][2])\n",
    "    plt.ylim(-0.5, 3 * 12.5 + 1)\n",
    "    plt.xticks(y_ticks_helper[0][0], y_ticks_helper[0][1])\n",
    "    plt.yticks(np.array(list(range(len(category_helper[\"Fine\"])))) * 3 + 1, \n",
    "               [\"Sex\", \"Age\", \"Condition\", \"Sample size\", \"Location\",\n",
    "               \"General\", \"Duration\", \"Intensity\", \"Control\",\n",
    "               \"General\", \"Effect strength\", \"Adverse effects\",\n",
    "               \"General\"][::-1])\n",
    "    plt.title(mapping_vals[val], fontsize=11)\n",
    "    plt.grid(False)\n",
    "    axszes[idx].set_facecolor((1, 1, 1))\n",
    "\n",
    "fig.subplots_adjust(left=0.15)\n",
    "trans = axszes[0].get_xaxis_transform() # x in data untis, y in axes fraction\n",
    "ann = axszes[0].annotate('Conclusion', xy=(-1.63,0.027), xycoords=trans, fontweight=\"bold\")\n",
    "ann = axszes[0].annotate('Outcomes', xy=(-1.63,0.183), xycoords=trans, fontweight=\"bold\")\n",
    "ann = axszes[0].annotate('Intervention', xy=(-1.63,0.483), xycoords=trans, fontweight=\"bold\")\n",
    "ann = axszes[0].annotate('Participants', xy=(-1.63,0.783), xycoords=trans, fontweight=\"bold\")\n",
    "axszes[0].plot([-1.1,-1.1 ], [0,0.08 ], transform=trans, clip_on=False, lw=3, color=colors_map[\"Conclusion\"])\n",
    "axszes[0].plot([-1.1,-1.1 ], [0.08,0.31 ], transform=trans, clip_on=False, lw=3, color=colors_map[\"Outcomes\"])\n",
    "axszes[0].plot([-1.1,-1.1 ], [0.31,0.617 ], transform=trans, clip_on=False, lw=3,\n",
    "               color=colors_map[\"Intervention\"])\n",
    "axszes[0].plot([-1.1,-1.1 ], [0.617,1 ], transform=trans, clip_on=False, lw=3, \n",
    "               color=colors_map[\"Participants\"])\n",
    "\n",
    "\n",
    "# fig.tight_layout(h_pad=3)\n",
    "plt.savefig(\"./images/telephone_effect_fine.pdf\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conditionals Message Distortion\n",
    "\n",
    "Next, we calculate the **conditional chance of survival** for keyphrases and facts across hops. That is, given that a fact or a keyphrase is present in a given hop $k$, what is the chance that it will also be present in hop $k+1$? In the case of keyphrases, for a given summarization process and category, we define the survival probability as the percentage of keyphrases that continue to exist in the summary, out of those that already exist in the reference text. For instance, if there are three keyphrases in the **Outcomes** category in hop $1$, and two remain at hop $2$, the probability of survival for this category from hop 1 to hop 2 is $2/3$.  We also consider the influence of the number of difficult words in the original abstract on the survival of facts and keyphrases, and the influence of the number of difficult words in the original abstract on the survival of facts and keyphrases."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 540x612 with 12 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(4, 3, figsize=(7.5, 8.5), sharey=True, sharex=True)\n",
    "\n",
    "for idx, h in enumerate([hopper, hopper_control]):\n",
    "    df_doggos = h.get_tagging_hop_dataframe(\n",
    "        lambda root, src, dst: compare_parts_doggos(root, src, dst, weights_categories))\n",
    "\n",
    "    df_tag = h.get_tagging_hop_dataframe(lambda root, src, dst: compare_parts(root, src, dst, weights_categories))\n",
    "    df_tag[\"Topic\"] = df_tag.branch.apply(lambda x: branch_to_topic[x])\n",
    "\n",
    "    for cat in category_helper[\"Coarse\"]:\n",
    "        df_doggos_tmp = df_doggos[df_doggos.key == cat]\n",
    "        regplot_helper(color=[colors_map[cat]], df=df_doggos_tmp, axis=[axs[1 + 2 * idx][0]])\n",
    "\n",
    "    for cat in category_helper_tag[\"Coarse\"]:\n",
    "        df_tag_tmp = df_tag[df_tag.key == cat]\n",
    "        regplot_helper(color=[colors_map[cat]], df=df_tag_tmp, axis=[axs[0 + 2 * idx][0]])\n",
    "    read_fun = textstat.difficult_words\n",
    "\n",
    "    df_tag[\"Qualification\"] = df_tag.dst.apply(lambda x: qual_helper(x, hopper, hopper_control))\n",
    "    df_doggos[\"Qualification\"] = df_doggos.dst.apply(lambda x: qual_helper(x, hopper, hopper_control))\n",
    "    df_tag[\"Readibility\"] = df_tag.branch.apply(\n",
    "        lambda x: readibility_helper(x, read_fun, hopper, hopper_control))\n",
    "    df_doggos[\"Readibility\"] = df_doggos.branch.apply(\n",
    "        lambda x: readibility_helper(x, read_fun, hopper, hopper_control))\n",
    "\n",
    "    df_tag = df_tag[df_tag.key != \"All\"]\n",
    "    df_doggos = df_doggos[df_doggos.key != \"All\"]\n",
    "\n",
    "    for i in range(1, 6):\n",
    "        df_tag_tmp = df_tag.loc[(df_tag.hop == i)]\n",
    "        df_doggos_tmp = df_doggos.loc[(df_doggos.hop == i)]\n",
    "\n",
    "        for col, new_col in [(\"Qualification\", \"Qualification_acc_p\"), (\"Readibility\", \"Readibility_p\")]:\n",
    "            quals = df_tag_tmp[col].values\n",
    "            df_tag_tmp.loc[df_tag_tmp[col] < np.percentile(quals, 25), new_col] = \"1st Quartile\"\n",
    "            df_tag_tmp.loc[df_tag_tmp[col] > np.percentile(quals, 75), new_col] = \"4th Quartile\"\n",
    "            df_tag.loc[df_tag.hop == i, new_col] = df_tag_tmp[new_col]\n",
    "\n",
    "            quals = df_doggos_tmp[col].values\n",
    "            df_doggos_tmp.loc[df_doggos_tmp[col] < np.percentile(quals, 25), new_col] = \"1st Quartile\"\n",
    "            df_doggos_tmp.loc[df_doggos_tmp[col] > np.percentile(quals, 75), new_col] = \"4th Quartile\"\n",
    "            df_doggos.loc[df_doggos.hop == i, new_col] = df_doggos_tmp[new_col]\n",
    "\n",
    "    regplot_helper(color=[\"#D68910\", \"#8E44AD\"], df=df_tag, axis=[axs[0 + 2 * idx][1], axs[0 + 2 * idx][1]],\n",
    "                   row=\"Qualification_acc_p\", value=[\"1st Quartile\", \"4th Quartile\"])\n",
    "    regplot_helper(color=[\"#D68910\", \"#8E44AD\"], df=df_doggos, axis=[axs[1 + 2 * idx][1], axs[1 + 2 * idx][1]],\n",
    "                   row=\"Qualification_acc_p\", value=[\"1st Quartile\", \"4th Quartile\"])\n",
    "    regplot_helper(color=[\"#D68910\", \"#8E44AD\"], df=df_tag, axis=[axs[0 + 2 * idx][2], axs[0 + 2 * idx][2]],\n",
    "                   row=\"Readibility_p\", value=[\"1st Quartile\", \"4th Quartile\"])\n",
    "    regplot_helper(color=[\"#D68910\", \"#8E44AD\"], df=df_doggos, axis=[axs[1 + 2 * idx][2], axs[1 + 2 * idx][2]],\n",
    "                   row=\"Readibility_p\", value=[\"1st Quartile\", \"4th Quartile\"])\n",
    "\n",
    "custom_lines = [Line2D([0], [0], color=colors_map[cat], lw=4) for cat in category_helper[\"Coarse\"]]\n",
    "axs[0][0].legend(custom_lines, category_helper[\"Coarse\"], loc='upper center', bbox_to_anchor=(0.5, 1.8),\n",
    "                 fancybox=False, shadow=False, ncol=1)\n",
    "\n",
    "custom_lines = [Line2D([0], [0], color=col, lw=4) for col in [\"#8E44AD\", \"#D68910\"]]\n",
    "axs[0][1].legend(custom_lines, [\"1st quartile \\n (best scores)\", \"4th quartile \\n (worst scores)\"],\n",
    "                 loc='upper center', bbox_to_anchor=(0.5, 1.8),\n",
    "                 fancybox=False, shadow=False, ncol=1)\n",
    "\n",
    "custom_lines = [Line2D([0], [0], color=col, lw=4) for col in [\"#8E44AD\", \"#D68910\"]]\n",
    "axs[0][2].legend(custom_lines, [\"1st quartile \\n (harder texts)\", \"4th quartile \\n (easier texts)\"],\n",
    "                 loc='upper center', bbox_to_anchor=(0.5, 1.8),\n",
    "                 fancybox=True, shadow=False, ncol=1)\n",
    "\n",
    "plt.yticks([0.0, 0.25, 0.50, 0.75, 1], [\"0%\", \"25%\", \"50%\", \"75%\", \"100%\"])\n",
    "plt.xlim([0.5, 5.5])\n",
    "plt.ylim([0, 1])\n",
    "\n",
    "for idy in range(3):\n",
    "    for idx in range(4):\n",
    "        axs[idx][idy].set_xlabel(\"\")\n",
    "        axs[idx][idy].set_ylabel(\"\")\n",
    "    plt.sca(axs[3][idy])\n",
    "    plt.xticks([1, 2, 3, 4, 5], [\"2000 to 1000\", \"1000 to 500\", \"500 to 250\", \"250 to 125\", \"125 to 64\"], \n",
    "               rotation=90)\n",
    "\n",
    "axs[0][0].set_ylabel(\"Keyphrases\")\n",
    "axs[1][0].set_ylabel(\"Facts\")\n",
    "axs[2][0].set_ylabel(\"Keyphrases\")\n",
    "axs[3][0].set_ylabel(\"Facts\")\n",
    "\n",
    "# axs[3][0].set_xlabel(\"Hop\")\n",
    "axs[0][0].set_title(\"Categories\")\n",
    "axs[0][1].set_title(\"Knowl. scores\")\n",
    "axs[0][2].set_title(\"Readibility\")\n",
    "\n",
    "for ax, row in zip([axs[0,0], axs[2,0]] , [\"Cascading\", \"Control\"]):\n",
    "    ax.annotate(row, xy=(0, -0.25), xytext=(-ax.yaxis.labelpad - 5, 0),\n",
    "    xycoords=ax.yaxis.label, textcoords='offset points',\n",
    "    size='large', ha='right', va='center', rotation=90)\n",
    "fig.subplots_adjust(left=0.15)\n",
    "\n",
    "plt.savefig(\"./images/conditional_effect.pdf\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Strategies"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Another aspect that may influence distortion effects in information cascades is the summarization strategy employed. Here we compare abstractive and extractive summarization strategies. To be able to do so, we have to distinguish abstractive and extractive summmaries, as well as successful and unsuccessful ones.\n",
    "\n",
    "To this end, we define **keyphrase scores** and **fact scores** for each of the summaries. The keyphrase score is defined as the percentage of keyphrases that the fact retained from the original abstract, whereas for fact scores we take a weighted average of the percentages of $A$- and $B$-valued facts retained from the original abstract (assigning weight $1$ to $A$-valued facts and weight $0.5$ to $B$-valued facts). \n",
    "\n",
    "With these newly defined scores, for each hop, we can plot the summary as a point in a two-dimensional plane. We depict the kernel density estimate of these points in the columns of the plot below. We consider that the points with high fact scores ($y$-axis) are better summaries, and that the points with high keyphrase scores ($x$-axis) are more extractive. These abstractions allows us to measure the quality and extractiveness of summaries."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def node_func(x, root):\n",
    "    tmp = dict()\n",
    "    root_aux = root[\"Tagging\"].get_occurrences(\"Coarse\", relative=False)\n",
    "    top_token, bot_token, top_doggo, bot_doggo = 0, 0, 0, 0\n",
    "    for key, item in x[\"Tagging\"].get_occurrences(\"Coarse\", relative=False).items():\n",
    "        top_token += item * weights_categories_tag[key]\n",
    "        bot_token += root_aux[key] * weights_categories_tag[key]\n",
    "    for key in category_helper[\"Coarse\"]:\n",
    "        top_doggo += (x[\"Doggos_crowd\"][\"Coarse\"][key].count('A') + 0.5 *\n",
    "                      x[\"Doggos_crowd\"][\"Coarse\"][key].count('B')) * weights_categories[key]\n",
    "        bot_doggo += (root[\"Doggos_crowd\"][\"Coarse\"][key].count('A') + 0.5 *\n",
    "                        root[\"Doggos_crowd\"][\"Coarse\"][key].count('B')) * weights_categories[key]\n",
    "    tmp[\"All\"] = (top_token / bot_token, top_doggo / bot_doggo)\n",
    "    return tmp\n",
    "\n",
    "\n",
    "def root_func(root):\n",
    "    return {key: (1, 1) for key in root[\"Tagging\"].get_occurrences(\"Coarse\", relative=False).keys()}\n",
    "\n",
    "\n",
    "df_tag_1 = hopper.get_tagging_dataframe(node_func, root_func, norm_enabled=False)\n",
    "df_tag_1[\"control\"] = 0\n",
    "df_tag_2 = hopper_control.get_tagging_dataframe(node_func, root_func, norm_enabled=False)\n",
    "df_tag_2[\"control\"] = 1\n",
    "df_tag = pd.concat([df_tag_1, df_tag_2])\n",
    "df_tag[\"val_tokens\"] = df_tag.val.apply(lambda x: x[0])\n",
    "df_tag[\"val_doggos\"] = df_tag.val.apply(lambda x: x[1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axs = plt.subplots(2, 5, figsize=(10, 4))\n",
    "titles = [\"1000\", \"500\", \"250\", \"125\", \"64\"]\n",
    "rows = [\"Cascading\", \"Control\"]\n",
    "pad = 5\n",
    "\n",
    "for control in [0, 1]:\n",
    "    x_means, y_means = [], []\n",
    "    for i in range(1, 6):\n",
    "        tmp = df_tag[(df_tag.hop == i) & (df_tag.control == control)]\n",
    "        x_mean, y_mean = np.mean(tmp.val_tokens), np.mean(tmp.val_doggos)\n",
    "        ci_x = sns.utils.ci(sns.algorithms.bootstrap(tmp.val_tokens))\n",
    "        ci_y = sns.utils.ci(sns.algorithms.bootstrap(tmp.val_doggos))\n",
    "        e1 = patches.Ellipse(((ci_x[0] + ci_x[1]) / 2, (ci_y[0] + ci_y[1]) / 2), (ci_x[1] - ci_x[0]) / 2,\n",
    "                             (ci_y[1] - ci_y[0]) / 2, linewidth=1, fill=True, color=\"black\", zorder=2)\n",
    "        axs[control][i - 1].add_patch(e1)\n",
    "        if i != 1:\n",
    "            e2 = patches.Ellipse(((ci_x[0] + ci_x[1]) / 2, (ci_y[0] + ci_y[1]) / 2), (ci_x[1] - ci_x[0]) / 2,\n",
    "                                 (ci_y[1] - ci_y[0]) / 2, linewidth=1, fill=True, color=\"red\", zorder=2)\n",
    "            axs[abs(control - 1)][i - 1].add_patch(e2)\n",
    "        r, p, lo, hi = pearsonr_ci(tmp.val_tokens, tmp.val_doggos)\n",
    "        # print(lo, hi)\n",
    "        x_means.append(x_mean)\n",
    "        y_means.append(y_mean)\n",
    "        sns.kdeplot(tmp.val_tokens, tmp.val_doggos, ax=axs[control][i - 1], color=colors_map[\"Keyphrases\"],\n",
    "                    shade=True, shade_lowest=False)\n",
    "        props = dict(boxstyle='round', facecolor='wheat', alpha=0.5)\n",
    "        axs[control][i - 1].text(0.18, 0.125, 'r: {0}'.format(round(r, 3)), fontsize=8, verticalalignment='top',\n",
    "                                 bbox=props)\n",
    "        axs[control][i - 1].set_xlim([0, 0.6])\n",
    "        axs[control][i - 1].set_ylim([0, 1])\n",
    "        axs[control][i - 1].set_title(\"{0} characters\".format(titles[i - 1])) \\\n",
    "            if control == 0 else axs[control][i - 1].set_title(\"\")\n",
    "        axs[control][i - 1].set_xlabel(\"\")\n",
    "        axs[control][i - 1].set_ylabel(\"Fact score\") if i == 1 else axs[control][i - 1].set_ylabel(\"\")\n",
    "        axs[control][i - 1].set_xlabel(\"Keyphrase score\") if control == 1 else axs[control][i - 1].set_xlabel(\"\")\n",
    "        axs[control][i - 1].plot(x_means, y_means, color=\"black\", linestyle=\":\")\n",
    "\n",
    "for ax, row in zip(axs[:, 0], rows):\n",
    "    ax.annotate(row, xy=(0, 0.5), xytext=(-ax.yaxis.labelpad - pad, 0), xycoords=ax.yaxis.label,\n",
    "                textcoords='offset points', size='large', ha='right', va='center', rotation=90)\n",
    "fig.tight_layout()\n",
    "fig.subplots_adjust(left=0.10)\n",
    "plt.savefig(\"./images/summarization_strategies.pdf\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our last analysis is as follows: for each hop and abstract in the cascading setting, we match summaries in pairs. Summaries are only matched if \n",
    "**(i)** the relative difference in their keyphrase score is **greater** than $\\alpha$ and \n",
    "**(ii)** the relative difference in their fact score is *less* than $\\beta$. \n",
    "\n",
    "We only match each summary once, and we do so greedily. Within each pair, we consider the summary with the higher keyphrase score to be more extractive (while of similar quality). \n",
    "We then compare how the fact score of each summary in the pair decreases with the subsequent summarization. The idea is that if extractive texts lose more of their fact score than abstractive texts, this is evidence that extractive summarization is less effective (and vice versa).\n",
    "\n",
    "We show the results for each of summarization step in the figure bellow for several values of $\\alpha$ and $\\beta$."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tagging(x, root):\n",
    "    tmp = dict()\n",
    "    root_aux = root[\"Tagging\"].get_occurrences(\"Coarse\", relative=False)\n",
    "    top_token, bot_token = 0, 0\n",
    "    top_doggo, bot_doggo = 0, 0\n",
    "    for key, item in x[\"Tagging\"].get_occurrences(\"Coarse\", relative=False).items():\n",
    "        top_token += item * weights_categories_tag[key]\n",
    "        bot_token += root_aux[key] * weights_categories_tag[key]\n",
    "    for key in category_helper[\"Coarse\"]:\n",
    "        top_doggo += (x[\"Doggos_crowd\"][\"Coarse\"][key].count('A') +\n",
    "                      0.5 * x[\"Doggos_crowd\"][\"Coarse\"][key].count('B')) * weights_categories[key]\n",
    "        bot_doggo += (root[\"Doggos_crowd\"][\"Coarse\"][key].count('A') +\n",
    "                      0.5 * root[\"Doggos_crowd\"][\"Coarse\"][key].count('B')) * weights_categories[key]\n",
    "    tmp[\"All\"] = (top_token / bot_token, top_doggo / bot_doggo)\n",
    "    return tmp\n",
    "\n",
    "\n",
    "def tagging_root(root):\n",
    "    return {key: (1, 1) for key in root[\"Tagging\"].get_occurrences(\"Coarse\", relative=False).keys()}\n",
    "\n",
    "\n",
    "other_lines = [(alpha, beta, None) for alpha, beta in\n",
    "               list(product(np.arange(0.3, 0.71, 0.1), np.arange(0.05, 0.151, 0.025)))]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAATwAAAEuCAYAAAATG66WAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvOIA7rQAAIABJREFUeJzsvXl0XHd5//+6984+o9GMZrRLlmTZGu97bGcxZGUJhJAmNQmkIQkQ2tKclJZf037P9/D7fU+BlpZvSw+FNgFK2KGEpUkbIJDEkH1R4i22x7Ysydo1mk2z37n3fn5/XHmXE9nWZuu+ztGxNcu9n7maec/zPJ9nkYQQWFhYWCwE5LlegIWFhcVsYQmehYXFgsESPAsLiwWDJXgWFhYLBkvwLCwsFgyW4FlYWCwYbHO9gJliz549P1IUpXWu12FhYTE5uq73rF69+vbZPOclK3iKorSuWLHikn19FhYXO/v27Wud7XNaLq2FhcWCwRI8CwuLBYMleNNEqVSSbrvttshtt90WWb9+/boPfvCDy2677bbIb37zG/9cr2226O7udmzatGndsevw8Y9/fPFkj3vjjTfczz//vPdCz/fwww9XH/v/d77zndCOHTt8F3rM+cx99923+Ng13bFjR8Vf/dVfLZrK806+TufLjh07Kt58803Xsd/vv//+1gs95lxgxbimCafTKR599NEowG233Rb5t3/7t67q6mptrtc12yxfvjz73e9+9/BbPWbPnj2eZDJpu/LKK3Mn365pGjbb1N+SP/nJT2ruu+++GMBdd90VP68FXyRkMhk5lUrZDMOQ0un0ORkqJ1+nkzmX6/3iiy9WdHR0FFauXFkE+MpXvtJzLmuYL1iCN4Ps3LnT/fDDD9d97Wtf6wa48847l3zhC1/o/fu///smr9er9/f3O3Vdl772ta91hcNh/Yc//GHVT37yk2rDMKRbbrkl9tGPfnTaPsSJnKrkVe2CLHqPw2ZUeR36VB9vGAZ33XXXkk996lNDHR0dxbvvvrvjW9/61sEf/vCHtfl8Xn722WcDDz300OFbb711+fr16zO5XE554IEHBv/2b/+2uVwuyx6PR3/44Ye73G63+NGPflT1gx/8oMZut4t3v/vdiXK5LI2MjDgmrMlYd3e3q6OjozA8PGx3u93GvffeO6aqqnTLLbcse/zxx/fv2LGj4qtf/WqDEILNmzeP//Vf//XQhVyLU8jHFdTchXlLDq+BJ3TWa/vYY48FrrvuuiTAf/3XfwUXLVqkDg4OOu655572oaEh56c+9amBm266Kf3Nb34z/Itf/CLsdDrF9ddfn6iqqtKOXad3vvOdqWw2qwwODjpSqZRt+/btsc7OTt++ffu82WxW+eQnPzl08803p+LxuPKZz3ymdXx83CZJEt/4xjcOP/HEE6Gnn37aePTRR8M//OEPD91www0rn3jiiTdvvvnm5Y8//vh+RVH4yle+Uuv3+7Xbb7898ZnPfKZlZGTEAfD5z3++d+nSpaULuj7ThCV4M8i6desKw8PDjlQqpWSzWRlg0aJFZYDW1tbiP/7jPx595JFHwg8//HDtxz/+8dGf/exn4f/8z/+MSpLErbfeuuwP/uAPkhUVFcaFrqOsG1z1xadX51VduZDjeByKvuv/fddOu3L2z/b+/ft9t912WwSgvb09/w//8A+9f/zHf9xeVVVVfuCBB/rD4bB+xx13jCSTSdsDDzwwAhCPxx0PPvjgQF1dnZbP56Xvf//7hxRF4f/8n//T+NhjjwWuvfbazHe+853an/70pwfcbrc4Zpn84he/qD5mVf/d3/1dA8Btt92WuP/++xffe++9Y7/+9a/9W7duHQf4p3/6p6Yf/ehHUZ/PZ3zsYx9bfODAAeeyZcsu/EOol+HLq1ej5i7o2uLw6jzYuxPFPundTz75ZPBLX/pSryzL4tOf/nTbvffeOxKPxx2PPfbYvmw2q2zfvn3ZTTfdlP7lL38Z+o//+I/D1dXVmq7rKIrC17/+9YaTr1NFRYV+zEK75pprMj6fz0gkEsqdd97ZcfPNN6e+/OUv11933XWpO++8M24Y5tvvxhtvjHd0dBRuvfXW5LE12e12Vq9enf3d735Xce2112aefvrpwCOPPHLo29/+dnj58uX5r3zlKz27d+92f+lLX2p46KGHui/o+kwTluDNMO95z3sSjz76aDCdTttuuumm4xbbxo0bcxP/Zl988cX6w4cPOwcGBly33357BCCXyymDg4P2SCRywR9KuyLz3IPX7pkOC++txA4md2lXrVqV27dvn/f666/PTPacmpqaUl1dnQZw9OhRx+c///nmfD6vJJNJWzAY1I4cOeJYunRp3u12C+At3bDa2lrN4XAYPT09jscffzz053/+54NjY2O24eFh5913370UIJvNKn19fdMjeIod/nzPnmmx8M4idrFYzHbw4EHvAw880AbQ3d3tHh0dtS1btixnt9sJBoO62+3Ws9ms/Dd/8zd9n/vc5xo1TZNuv/32sW3btmVPP966deuO3/bQQw/VPP/885WyLBOLxRwTx3d94hOfGAGQ5bd+WbfcckviBz/4QXV9fX25urq6XFlZaRw6dMj95ptvep9++ukggMvlmrJXMNNYgjfDbN++PfGJT3yivVQqyT/+8Y+jx25/4403vFu3bs29/vrr3paWllJ7e3uppaWl8L3vfe+Qoiioqio5HI5pa1ZY5XXo5+KOThednZ2e7u5u19KlS/M//vGPqz70oQ8lHA6H0DRNOvaYkz9UjzzySM327dtjN910U/qzn/1skxCC9vb20uHDhz2FQkFyu93imOUiSdKk1+d973tf4tvf/nb12NiYfcWKFUXDMGhsbCx+5zvfOejxeISu60xrH0hPSH8rd/RC+fnPfx785Cc/OXj33XePAXz3u98NxeNxezQa9WiaRjabVQqFguLz+Yzly5cX/uVf/qW3v7/f/md/9mft27ZtOyDL8ikvVlFMYzQejytPPfVU8L//+7/3x+Nx2/ve976VAIsXLy4+++yzFR/5yEcSxyw8u91uaNqZIemtW7fmPve5zy36/ve/Hz72hb5kyZJCJBLJf/zjHx8Dc0Nvpq7NuWIJ3gwTCAT06urqstfr1Y9ZKABdXV2uO+64Y+nJMbybb745vn379ogsy8LpdBrf+9733jL4Px852aW12WxGoVBQ/vVf/7UrFAppH/7whyNbtmzJbtmyJfvpT3968b59+zxf+tKXek9+/nXXXZf+4he/2Pzoo4+GvV6vEQqFyuFwWL/zzjtHtm/fHnG5XMYNN9yQvO+++2JLly7N33PPPe0f/OAHT4l1fuADH0ht27Zt0T333DMIpqA+8MADA3fddVeHJEnCZrOJL3/5y921tbUXxabSr371q6p//ud/Pu4SXn/99ek/+ZM/WVJVVVX+2Mc+tmRkZMRx//33DwA8+OCDLcPDw45SqSRv3749BnDFFVek77zzziXXXHNN6uTjBoNBvba2Vr311lsjHR0deZ/PpwM88MADQ3/xF3/R+rOf/axaURTxrW996/BVV12V+eIXv9j0zDPPBP793//9FPf0qquuSv385z+v+exnPzsAcPfdd489+OCDi7Zv314FsG3bttT9998/OrNXaWpIl2rH43379r00Xyot/uzP/qz17rvvHt20aVMe4E//9E/bPvGJTwyvX7++MNdrs7CYK/bt26etWLFi62ye08rDm2Huu+++xYZhSMfEzsLCYu6wLDwLC4s5wbLwLCwsLGYQS/AsLCwWDJesy6fres9ctJ+xsLCYGrqu98z2OS/ZGJ6FhYXF6VgurYWFxYLBEjwLC4sFgyV4FhYWCwZL8CwsLBYMluBZWFgsGCzBs7CwWDBYgmdhYbFgsATPwsJiwWAJnoWFxYLBEjwLC4sFgyV4FhYWCwZL8CwsLBYMluBZWFgsGCzBs7CwWDBYgmdhYbFgsATPwsJiwWAJnoWFxYLBEjwLC4sFgyV4FhYWCwZL8CwsLBYMluBZWFgsGCzBs7CwWDBYgmdhYbFguCQHcXd2dlrDdi0sLhI2btwozda5LknBA9i4ceNcL+GiY//+/Sxfvnyul3FRYl2786Ozs3NWz2e5tBYWFgsGS/AsLCwWDJbgWVhYLBgswbOwsFgwWIJnYWGxYLAEz8LCYsEw52kpkUjkk8BHgTLwsWg0evik+5YA3wTswLej0ehDc7NKCwuLS4E5tfAikUgVcC+wDfhL4O9Oe8jfA38BvAO4d+LxFhYWFufFXLu0W4BnotGoHo1GXwM6Trt/aTQa7YxGoxqwA9g82wtcEKh5OPgk4de/DIYx16uxsJgx5tqlDQKpk34/vcTk5N+TwJQtvP3791/AshYIhkbF4LPYciMo6jjusd30/forZFvfNdcru+goFovWe+4iYK4FLwmsPun3082Lk38PADunemCrzOcslIswshdKGWi/Buo8YPfAoScZlh009/0c3vvAXK/yosMqLTs/Flpp2cvA1ZFIRIlEIhuAQ6fdfygSiWyIRCIK8E7glVlf4aWAEObP6AF49RtQTEHdGvM+Twj2/AQUG8llfwSKA7qfm9v1WljMEHNq4UWj0UQkEvk28CwTu7SRSOQ9QGU0Gv0x8Dec2KV9JBqNJuZutRchmgqjb8LgG7D4Ggi2wpY/BmXizy4EHPgfWHId1K6E/fth7Yfh1a9D21VzunQLi5lgrl1aotHovwP/ftJNh0+67zCmZWdxLpSy4PTB3kfBE4ZlN4Gv+szHqVnw1UHvCxCe2C9a9xHoehpih6F6yeyu28JihplzwbOYJvQyxA6Y1pxsh7W3w9o7QDpLq7FCEiQZQm1QTIJiN2+32cHfAL/933DHj2Zv/RYWs8Bcx/AsLpRc3PzJxyHdD0tugHUTQnc2sTMM2PcYZEdheM+JeN4xtvwJJHshG5v59VtYzCKWhXexMnYI+l4BrQDt10LVYoi8d2rP7XsJvGEILwWn3/z/yVS1mML5zOfhpi9P/9otLOYIy8K7mMgnoPtZ00IzNFj8Ttj0MVPszgVdhfbrINENsmL+nM6WPzZTV/LWPpHFpYMleBcDag52/Qj2/hRsLhAG1CyHyqazu62TYegQ74LFV4PdBd2/N2+bDH89yDZ48V+n4xVYWMwLLJd2vlJMw9Auc2Oh5UpouQIqm89N4E6n5znQShBqh9yYaSVW1E7+WEkyrcBXv26mt9gc539eC4t5gmXhzScMw8yN63vVtOhkOzSsN8UnsOjCxC49ALGoad2BKahNl731c1Z+0IzvdT5y/ue1sJhHWBbefKCUgaHd5o7p6tugfg00bbowgTudcgGW3WhaaoZhJiFPFrs7GZsDrvpzeOGrsP4j4PBO33osLOYAy8KbK4SAQsoUn70/BYQpKt4w2JzTK3aj+00LsbLJ/D3ZDdFfTu254Qi4KuHwU9O3HguLOcISvNlGzcPRl+GVr5sVDrIMG++G1qvAWTH950v2Qu/zZizwGMO7oToytee7A+a6Oh8xRdrC4iLGErzZQAhIHTWrG3KjoBVhzXbTxZxJtBIc/BUse/+J+tlyAcYHzy2VZeNd5rF6X5yZdVpYzBJWDG+mGXgdBjrB4TPbMQVbzZ/ZQFfNHd6KuhO32Vyw7sNvH787mXAEVt1qpqg0X3aiDM3C4iLDsvCmGyHMHdEjO8z/21ymWKy741ThmWmSvab7XLfq1NuPvmhWV5wLkgRLbzBz/wZ3Td8aLSxmGUvwppN8Al77DzjyDHgnupPUrgDPLI/iKBcg+sSZVlw2ZqamnIt1d4yKerMy49VvTM8aLSzmAMulvVAyI2aHEk8IGtbB8g9M3oppNjn0JDRuPLNGdnj3mY0Cpopig453mc1Cx4fBP4vWqoXFNGFZeOeDrpn/Hn4KDv7SLMNqWGfGtuZa7ISAYBs0bjrzPtlmWpzny7L3Q1U7/Pb/s4b9WFyUWBbeuZAbM625sYNmKknbO+ZXAL+UNdfWuOHM+8pFc70Xkt/n9MGGj8KbP4OxqBnTs7C4iLAsvLdD18zYnFYy26F7QnDZx82qg/kkdkKYKShnI/o/ZsLxheIJmgnTu39s5eVZXHRYFt7ZKCRNa250P9SvNRODN350rld1dkb2mp1PGtafeZ+aM2ONgdYLP4+/AexOSB41j+v0XfgxLSxmCcvCOxlDNyd7lbKm++qogE33mmI333FXmQ1AJ3NZRw+Y7qc8TX/uNbeb53n5IcvKs7iosCw8MD+0Pc/C8F6obDRbJoWXzvWqpoYQZl+7RZefvYVT/VqzFdR0Ub/GjGHu/ZlZQRJsmb5jW1jMIAvXwjMMGDsMR35nWiveatNlXXEzuINzvbqpM/g65MfOHk/Mxkx31+6a3vMqdnO04+Gnp/e4FhYzyMIUvHQ/vPIQDO+CQLN5W83yi6/9UT5hzrXoeM/Zd1+HdpkdkqebutVmQ9H+V0ArT//xLSxmgIUheEJA4ojZhml4jzmrdf2dZsnXuc6DmE/YPaZFejahNvSJ9JELyL07Gw6vmeYiDDj82+k/voXFDHBpx/COtSbf9wszvaRhvSlwsgxMs4s32wx0mkO0KxvP/phyAZq3TL87e4x1H4FyHvb82GwqcHplh4XFPOPStfD2/Rd0fssUuuUfgDV/COEl07dTOZdkY6Yr+3Y1ukI3OyfPFHan2SZ+zYfNpgQWFvOcS+DTfxaql5kJwort/Irl5yuGAQf+G5a+C+zusz+ulIWdP5j5tBF3wCyvO/zUiZI7C4t5yiUseJFLS+iOIUnQ8W5z8thbMbrPjN1NZ6v4yQgtNWt07Z7pqeSwsJhBLl3BuxQZHzLbwvsb3v6xo/vNndSZRpZh0z2msP7+H82BRBYW8xRL8C4WdM10ZY+l0bwd6z48e334aleYnVTarzVjixYW8xRL8C4Wep8zd5gDi97+sUdfMmuBZ5Ni2qytPfyUuTtuYTEPsQTvYqF+HbS98+0fp2vQ/5o5WnE2adpkWneKA9Ts7J7bwmKKWII339FU2POoOSpRmULaZPywWdtqc8782k7GG4bmzWZM78Wvmu20LCzmGZbgzXeOPAO+mqnvONvdZm7cXLDpXrO1vM1pttaysJhnWII3n0n1wfiAOWpxKpQLZl7cbE5HOxlJMmOHy24yh39beXkW8wxL8OYz/kZY/YdTt+6GdsHQ7pld09sR6oCXvmY2Nijn53YtFvOWYlknnp39sIclePOV6C8hM2TG7qaCEGZjhNPn0M42DWvN+mVvtdkS39Dndj0Wc45hCLIljcFknr0DaV7sGuO1ngT7h8ZnfS1z2jwgEoksAb4J2IFvR6PRh067f9PE/RGgKRqNjs3+KueA2EHIxcxZsFOlkASHb+57+dmccNVfmJ1pjr5k9syrP8/RkBYXJapmkFc1knmVRK5MQdVAAgUZp13G5zzRu1GdZc2b624pfw/8BbALeD4SifwkGo0mTrr/IHAV8PhcLG5O0DXoehrWbD+3RgeeKljzoZlb17ngbzDL2lZ/CA48ZlZ8zHSJm8WcIISgUNbJFDQSeZV0XkUzBAJwKDIuu0Kl+yyduOeAuRa8pdFotBMgEonsADYDx0dvRaPR8Yn75mRxs44QZrxuw13g8Ez9eXoZ9j8OK2+ZubWdC5JkdmHe+yhkBs2kZHdgrldlMQ2UdYO8qpPOqyRyKjlVBwQSEk6bgtc59Ul+ybzKbLfcnWvBO/lrPwlMWy3U/v37p+tQs4YzdRillCRfe25pJc7UYezZUbLKgQs6f7FYnLbrZlfDVHX9ilLlYoqdT1GsWn5JW3nTee3mC0IIVF1QKBuMF3XSJQPNMLvvKLKE0yYhn8ffVDcEiiyxe7jAe9tm1/qbccGLRCL1wM8nuesLwMm9xwPAzuk67/LlF9mQ6OI4vPEMbP3IuVdJ7NoFy2+cWtnZW7B///7pu25iGTiTkI/DyDOwfD2ELuLu0m/DtF67OUI3BHlVI10ok8ipZEsahhBIbomqoEy9TUa6gC8t3RAcGslwcDTLe1fV0dYqo44cnsZX8PbMuOBFo9EhYOtk90UikT+KRCIbMGN47wT+YabXM2/pedYcB3k+JWE1K6Byik0FZgtJMlvo971slsX1Pn9JC97FSEnTyZd0EnmVZE6lVDYQmNabyyZTcQ7u6duhG4In9w0T8jp514paZEni4EiG1mk7w9SYa5f2bzixS/vIsQ2LSCTyrWg0ek8kEmkFvgGsBX4aiUS+GY1GvzNnq50pDAOW3HD2yWNvxfgQhDvmp7sodHPIT8s7zF552Rj4qud6VQsSw5jYXChqJHIl0oUyugAQOBUFp03GaZv+/pHJvEpXLMvGRUGuidTgsiuUdYMn9gxR63fCOYSqp4M5FbxoNHoY07I7/fZ7Jv7tAa6f5WXNLoWk2Y5+/V3nLlpCmJsVaz90/nMr9LLpTudG8Yx0QlieaOo5DSmaih1qVsKeH0KgFXSri8psoWoGBVUnkS+RzJUplDUQIEkSLrtySmrITFDWDXb2pRgZL7K2OYAQglimxGimxMaWINcvr8XtUFBHZrerz1xbeAsbIczk3OYt5ycw6X4zMflc3GDDMLuZ5BOQHTHL0RCUZSdF2QuZYSiNQ93asw/2PheaLzPddXcQXv0GbPnk3JW+XaIIISiWDTJFM/aWKqhougAk7Iq5e+p3zc7mgG4IUnmVgMdBpdvOhkVBsiWN3+wfxWmTWdsUwBCCsm7gErNf92AJ3lwSP2wmC9ecZ7A73T+1rsblAhRSkBlGFFOMlwBFwevx8Ot+B2N5g7Kho2ckXs7JrA0VWFt+yZzyNtVKj7PhqoR3/D/Q/XsItpodm1f9wYUdc4Gj6Qb5sk4qXyaZU8kd21yQJJw2Ga9jZq23yRBCMJAqsLMvRWPATcjnpMrrYDRTpMrrYP2iAC6bwsHRDP3JArUVTi5vn/0pd5bgzRW6BqElFzYXt+XyyYf0TLip6vgI8fgYuZLGkqDCM4MKh1I2fA6JNTV2VlUorKm1E3bLuO0yR46MMWy38e19BRb5JW5NvkTzklXgP4eKj8kQBmhFM5n65X+HXBy8oQs75gKiWNbJlbQJ662MWjYQksAmmYm9Fa7ZF7iTEUKQzJeJDme4ckkYWZL4/cEYOVWjo6aCWCZDc5WHgqqRK2p4HApjWRVNn4EB8W+DJXhzgWHAzu9D5MbzD+KP7DOL85s2IXSd8fEUsbExYmOjrAmW6RkXvDwiE/I5qPM5WOJysqXZ4J1tp+ZONftPvAUkCa5odrG10UHfuE40rvHC73fTVDfE5evWINvO8+3irDDrgl/6GmhlMKwuKmdDn9hcGM+rxPMq2aKGZggkwGkzNxdcM7C5cD4Uyzp7BtLIksTGliBb2kLYFIlYpkRLlYdDoxkODI/jcdqodNuprnDicdpYHvJQ7XMiSRKzHdW1BG8uOPqiWVx/nmI3lC4wuvt1xtyLaUq8QpMzxy8Pa4S9CuEKN7Lbw4oKiZVNp26CeOxTi5nIskxLQGZRpY1oXOE7b8T4n/2/465r1hFprDr3XCxJgrZ3mGMjm7eYYx1X3jL7XZnnISVNNzcXcirJvFl3KgCbdGbd6XxiOF3k5e44HbUVNAfdvHhkjOhwhiqvA5/TxrXLaonlSvQnCrhsMl6nDY/DxmWtszRn5SxYgjfbaCUYO2gO2XkLDEMgSTCaKdE1miU2nkfRiry/VefImz1IsV5aOhqo83nxugLcvn76lypJEsvCDr5wQw19Y1kOd+1kT38zTdVVbGoJYj8XS6N6uTnoJ3UUEt3Q+yJE3jP9i57HHEsNyRY14rkS44UymjBdQqei4LTL86ru9HSEEAymisgyBD0Oru6oJlvS2dmXwhCCer8bRYZcSUPTDTpqKljXFMCmzJ+mTJbgzSa6BpICGz56yq5ssWy2UFJ1g5ePJBjLlsgUVO5cH0KNx/ClRmm1lwj5JCi6ubI9YHYW9s1eZ5TmsI/mKo2BWC/f2p3iey/Z+OgVLaxrDuKwTeENrdjMzimuSjMh+ZWHzWHhTt/ML36OOFZ3mppI7M2pOkII5InUkHOpO51r0oUynb1JDCFY0+jniT1D7BsaZ21TgLXNAQJuO6/1JmgJeWkOerApMn735O8LQwj29KcZTBe4bpbTMi3Bm0WM7t+TLNspN1xGXaWLX+4ZYiBVwCZLbOuoptELba48m50J/IwjxfrxKArN9R6QJsqshYBUr7njOdvINhprQvzvihR9VNNV0nhy3zB1fherGvy4HG/zdpJtZp+/XNxMfSkkLxnBE0JQUHVypTLxnEoqX0YzDIQAuyJPpIZcPAJ3jGJZR5ElBlIFdGHgUmRe601RKuusqK/AYZOxyRJ+t51rl9VOeoyCqjOYLjCULmKXJSJ1FdgUibVNASjlZvX1WII3Q+RVjbGMSjxXYl1zgJf3Rtm9a5hg5AoWpwrUVbrY3OKnosWGozgGuZ2Q0ViCBE4PyGfpLpIdMfvlzYXggRmP8wRpLiZo9hUZ9S7jp7tG+OZzR3jfmgauXVaD52zC5/SZO9NdT5lTzpK95ryO2R44NAWEEGiGQDcm/tUFmmFQKuuUdANVMyhpBuWJ248MF0nYk0gSOG0ybruCJF28Hy/dEOwZSHEklqMl5GFXX4psSWNJjY/rl9eRyqt4HDZ8rjNfoyEEY9kSQ+kiK+v9dMezFFQzEbo3U5zY6ICR8SJ/ufESax5wqSOEYCyrMpYtMZYt0VFbgaoZPBMdJexzEvY5MQRslA6y5X1XIbkDZkH90YOEjrVAt7vAMcVGOfFDEF4ycy9oqrj8UC5Qk3qDP7l8PQOFBrpjOXZEYzQGXLSEvAQ8k7yZF201X7/iMCtMENC2bUaXeq7iZUw81kRCkk5k/yiyhCJJyDIokoRdlnAoNrwOmUr3xWfBTUb3WJan9o+QVXWu6ajh0EiGTEnjstYqltf7sSsSNf5TK3vyqoYhzPZHT+4bxq7IpPJldkRHyZU0dAMOjWRIFsqnPO8vN85uEroleOfIULrAQLLAWLZE0ONgc1sVL3SNEfA4CPsceJ02av027rq81XxCuQDJIyiBJkgegYQw41l2D9jOsUecEObxLrAryrRhd5spJn2v0liznMal9SRzKk/uG+HfdnRxxZIwH1j/4lT6AAAgAElEQVTbcKrw+eth7e2QOAKXfwoO/Le5czvFqo7JxKus66ZoTUG8TuZs4rXQ0A3BULpAf7JAoaxTUjXcdgUhoMJl4/oVtVS67afszouJROfocIbDsQz9iQJlw2BkvIiqGRxN5MmWzmzv77YrtIU91PpdEy7+7ObiLby/7hTQdANJkkjmVd4cHGcsUyKvaty5tYV41swcWlbnp7rCzCW6eV3jiSfrZTNGlRs1LZn0AAzvhhU3T08axtJ3XfgxphPZZjb3HN0PpTTBUAcfuqyZ65ZXs6d/nN8djNEa8tIS8pwQvvFBiB1AtFyBoWtoo4cohyLnJV5iovnkMfFSZAlZYsGK11QRwuxKPDJe5KWuOCPjRcaLGu9dVceb6QJOu8INK+uoqThhyRXLOn2JPHsG0kRHMiAEfckCo+NFSvqZCfAVLhvt1T5aQm7CXiceh426Shcr6v283pdkLKMCVgxv1hBCTPT8Mr95frt/hLFsCVUz+MDaBmyKTE2FkxX1fqq8DiRJYlXjaaJ1vDY1DtnRE5O67C5TDEb2QPs159cJ5XQO/9Z0CS+03GsCQwhSRYPRnEEsr1OrloglUuRkH3U+BZdtivl2koRwB9HSIxi5cco1q7DLdlY3+knlVV7qTvB/n4zSGHRzdUc1AXeAuq596D1HKDtCZIbyFLNJhMQZ4qXIC9fymglSeZXeeJ6eeI61jZUkCyqGEIR8Dq5YEqapykO4wkmd34WqG7zanWBnf5J4VqUnnmNkvHS8CejJVHkdtFd7aa/20RT0UOm2oRuCdc0BDgxn6I3nKJR1Do1m0HTB5tYQvfEcaJbgzQhl3SCeVcmrGourfTx9YISu0Rwep8K65gAr6v2saqgkXOE4Jehe5Z3E1TqpNpVS2nQ1J3NTk70QXgqeaSijKmVAzZm1t+fJaE5nNKcTyxu0BmwEnDK/PVKiylGmJfUyNleA0v432O9/B7/V/FS5JN7T7uJQvIxDEQQcpgNS1iesLl2gCQNhCI5VCclaHKnvKbKhVRj2ChRZoqPGz6Kgh70D4+wZGKc15MFTu5XqkWfJhTaw6OgvSLe9l2J4CnXBFudMXtWI51Sagx5e60mQL5uJzk8fjNEa8hD0OhgvaGSKGs8cGOXQaIaj8Tx9yQKTFC5S63fSXu07/tMYcKHqAr/LFLkd0RixTBGA/mSBm9Y2EPQ4MIQg7HPispv5my0hL+rILF4ILmHBOzyaZSxbYm1TgK5YlpeOxAn7nDQE3Cyuhi1tIa7uqEGWT1gxi0Jnac51Ugsl8vGJ0ijJnDtxNje1lDFjbdPVpy5+GELtUz7eWF6nb9wUOIcicU2ri90jZZwKVDglqj0y3SnNTJ3ofYFUZRUZtZJ+fS01h54kXXEFdQ0hdvWrvDKsU9IERQ3W1Nhoq1Toywqq3RIBt4x8iiXoA0PDN76XUrADzW0Gpd0OhW0d1RhCsLMvxVcPBWmWL+eadB6pegPe4VcohlaCNH+SVC924tkSO/tSJHMqXpeNpoCbcIUTwzB3UW2yxDPRUQaSBWLZyYu8mgIu2sI+ltaa4rY47MOmSORKGiGfk6f2D7NvMI19ouRtU2uQ61fU0p/MU13hJOR1IksSDQH3LL/6yblkBa97LEfY50CSYGWD/wxX1Ot8i5f+Vm7qVHZTtSIc/DV0vHva3E+CbZP2vBNCkCsL9sXKxPKma3rHSi/JgkFJE7QGbFR7ZOJ5nVzZoCdlmM0Dqh34FMF6Ww8ur0q6tgP3wJtcHlmL0lbBIskDEhQMG0E3JIuCsjBYUqUwkBFEEzp7dUHZgA8sdZArC7KqIOyR8dkVdGclzuQhZDWDGmg/LmSyJLFhUZAV9X72vZnl9awNv6uJrYxhT0Qphy7uNulziW4IBlMFeuI52kJeHHaZoXSRTLFMf6rAU/tHGEgVJt1MAGip8tAUdBOpq2BZnZ/WkBe3Q6FY1nHYZPqTBZ7cP8x4vozbodBRW8G2pTXsGUgfr5X1u2xIkkSle36WDV6ygnfDismTIM/KVN3UqXD0JbPl03SJXT6O0HVK9kpcwIv9JUayOmMFg3cuclLnU1BkidXVNvwuJw4F+jM6AxmdYkxwQ5uLGq/MhjoHXrvZr2zfUBqh67gqm8Ch4LMLJHUM79CL5Oq3YEfG3/trCqGVXN5oto83JqoEfA6DpgqZZFHgVARuG7zQrxEvCDQhCLll3rPYQbTgw50bpSqfQa5bhVBOhAdcdoXNHS1UHfg+h5MtfLfXSWpY5aoVaSJ1fhR5mizjSxwhBJmiRoXLxk9f7zMzCHIqpbLO6HjpjDQQMLd7av1OFlV5aK7ysL45yPL6ChwTpYJCCEqawa7+FD1jOTJFjbawl82tVSyp9qHpgnCFk2qfE4dNZmPLHM9CPgcuWcF7W87HTZ0K5YJpzdSsOO9DHNtMqHIrvBkrk97/Ov1KI9WNfq5pdVHllmmttFHhkPA5ZQ7Fy/SlNV4fMlhaZePqVheLAzY21TvwOWTyZY2xTJFEXmVMgFOR8Sll/P2/oeypxl5MkqwPEPctp1gcJdTzHKJ1G9nGbVT0PUNRVykF2o93WanxytR4T3U9V1XbSBQNEkVBc4XEeEnw4kAZQ9jRtCKXJ16huX01h9ISIa/DjI16qlG9dSzJdHHH5mvYfeggvX0Ko5katrSFcNrl85qKdaljCMGR0RwvHBnjwNA4ibxqxuBKZ3ahsUkQ8Nhx2BQag242LgqwuS084f2cuLaDqQK98RzdYzkqXHY2tQZx2xQCHjsrGyqprnDiddlY03Rxj9tcOIJ3oW7qVCjngYnOIFN9ii4YyxtUuiQyJcHTPUWyqiDolrkl4qbGpdHiibNl3dUI2Ua6aHA0rTOcVSkbgj9a7cXvlNnS6KTaa5b5ANR4JZK5Ej2xEpphljcd34wRBr6+FwCBLrvZWbGR5w4abPJJPFdYzsFYHr2/wJUNMptqruPA0WFcBY0Gn0SNV57U+qrzydT5Toigbgjev8RBoiBIFm20hwwORnfxcqoKw+4BJO7Y3MyYbzPOghN3PsO2wBjCsZN9tX/AE3sHGU6XuLy9ilUNganV616CaLpBX7LAkViWA8PjHBjOMJwuUtTOzF/zOWTqK914HDbieRWPQ6Ep4GZ9c5DNi6uO//1Lms7RRJ6jiTxD6SIbW4IIAUcTeeorXTQGPfhdduor3azh4ha407m0BW863dS3Qwizm2+gxdyZnYSiJojldCQJGisUvr83T1EThN0ylzc5Cbhkblzixu+UMISZGNs3miSjtXFwd4lrWqGhQqEtoHB5k2m9AdT6TFekpBnEs+bcAE0X2BQJl912ar6t0PH1PwvCIL34Jp7s1hjIGjT6ZEJOuGOxC3Ch5ZLY+18mLV3FuKuRg30Jni+WuDpSh2rAK0MaYbfMIr/M2hobNllgk6XjVoMiSwRdEsGTwo6rGm10BJKMygp5bxMeh40dCTfp0WrKY058wfdzj+tFxoe6WNXQQr1f5dBIlv5kgWsiNfictgsaEzjfUTWDnniOrliWIzHz3+6xLJNoGyGvnTq/m4aAC0WRGU0XaQ17aQq6qPG56E8VuHF1PSXNrOcdHi/QM5bH77ZR53fzak+CYtmgLezBNWH9Lam5NOqa34pLV/B6nps+N3UqxA+bohdaYub3qYLRvEG6aLCh3sHT3UWOjmtUexSWVpkf3NuWe3DZpOOxsd6UxoF4meGsQa1P5j3tbryVVTTW1nKV54RltaTqhLVT1gXpQplYpmQWeksSLofCZHXqcilNRd/TDCnNvMpKrhQSVS4JSZJpD8rs7FNoaBCM5AR5zU9FYBXVA7/n/a1XoLUG8Q8+jy6GiVVtwG2zM5wzSJXMYc0/3q9yJKVT4ZRYEbJxbYuNo+MGXodEjVvCYZORZBlnRSUtahLDKFLSV3Dj6noc4X70sSMUwj4YijGQq+HNQRdlQ3DtsmrCXidf//0RZFlidWMlWxdXXVSdRiYjr2p0j+Xoih0TuCxHE3kmSXGjvtKFz2Uj7LHTFq6grOvkVJ2GgJu6Shc1PifJgsrBkSyDqRKyJNMYcLGzP0VLlZffHRzlaDxPXaULv7uC6gont25ouqS/PM7GpSt40+Wmvg1CCFJFQXYkRr97E5GiQbIoeLG/RLVHpsarIITg6lYnsuQ6/pysarBrpMxQVidZMLhrjRdFhiVVdrYtks1mncU0kfgLEHnvKefUDEGmUCaWNecZSBK47ZMXck+cEBAoyW52jPoYCC1jXb0DYRgE3BJ7xzTyZUGjy+C/D5fJlQ3yGtR6aqixX0Z1vMQLSReGcRkb1NcJSHGWhqspG+C2SaRKgjtWODEMg5G8AElCNWDXqE7PuEG2ZHB9m53WSoXXhjQqnQ4a3SXaiq8i166hXL2a8NALGIk8xboNfEArcH1NNeP4cDsUhtMFGqvcRIcz/Oz1fuJZlUqPDbsi01LlpTHoMi3Zecp4ocyRsRxHYlm6Ylm6YjkGU2fmuElAQ8C02gxDEHTbqfZP1GMbAlUzqJyoVumoraC+0oUh4I3eJHuH0ryjo5qjY3l2DaSQkGgLe2modPPBdY3HPYaFzvx9l8xT4nmdoayZ/hFyy6wI23juQB++4AaqPTIum0R7UKE9aF7aYzWHe0dVBsZ1hrI6W5ucLPIrVDgkli5yEvaYwfkm/2l/jvjh43WzhgHZkmnJjRc1kMBtU84uchPI5TxK/yt029pYUR4kuGgdisPJE0fKuBWocenUqgNcmX+JJeOvEV99L/HgGgoaVDolhAgTTeh0GAdJ20LULbmSo+MaL73RR9IWxmWXWRJUSJcEiaLBoYSOU5EIuGRuXmrn6LhBomDgtklkSgZVbonetEFPSsbjdLL3lZ3szvnZLDeyTBkgGK7E170DV24MY/l2ABqDHj60ybwOx8r+vv9yD91jOX5THmFJjY97rmzjZ6/3E/A4WFTlpqXKi9M++63QEzl1QtROuKWjmdIZj7PJEotCHtqrfbjtCpqu43XYiNRWoNgkDgxn8NhtZEoauZLOZa1BGoMeZEkiVyozlC7yn6/18frRJCXNoCngIZUvs62jmq3tIZw2+RQLTrG0DgBJTDYE5iKns7NTbKy/8CB3umhwJKURy+kUNMHNEQ8vD5RQdUGNV6HOp1CZjpotmxZfffx5w1mdQ4kyQ1kDj03i/R1udo+oVDhl6qdasiUMxN6fkVv8bsaKZucJIcBlU7BNseRLGh/gSNch3qCDsM/NDVXD7CgsYe9QFn9xgD/kN3jjb9Ioho/nMwsk4ss/QrJj+ylJwLKapaL/GQqhlaj+Npyx3ejZMRJ1V+H3OOnP6KSKgoIGQSc0+2We7NZIFA1KOnRUyVS5ZHJlQVdSp6CB3wHtlRJuo8jvx9xUO3Vagk72xGHxyK950XkFi1sXc2V7mCNjWXxOG7V+FwGPmV/ZPZbj+UMxHHaFjpoKBIK+RIGhdIHNbVWEfU6eiY7SUOmmOeiho86HMh3zdjlRcN89Zu5svnk0xmDOIJU/Mw3EYZNZHDbLrhZXe03LzIB4TmVLW5DHdg2SzJVx2WUGUkVWNfq5vD2M12EO6FE1A0MI3jia4qXuOMOpPG6Hnesi1QS9DpbU+ua1hftWqCOH2bhx46zJsSV4mNv8mgG7RtTjdaU3LnFTNgR9aZ1qr0KNR6bCedoxi2nUA0/yZujdDBXtDOd0PrzKy0hWJ6Ma1PkUQm75nGMleVUnni0xnoyh2itx2ORz2qUUWpGSqpMvqTxxVKFaxAm5wD7ey2gsRj6f46PKk1TL4+bLCCwlW7+FXCrOoqFfISHI1W5ieONfYjhO5BJKWgHvSCfZ+stBVnAl9mPLj5JtOmOW+ikcK/jXBYzkDAqauYGzPKTQnzHYN6aRypVYVtxJ7dJNHBV1hPt+SbRcz3BwHS67glrWOTSaZbxYZsOiICGvg5EJy6m2woFNVlANHd0wq2hqKpwUyjrR4Qx9iTyjmRIf3rKI3x2MMZgq0FDpZmmtjyU1FcdjqGf/e2j0xPPHxa1nLEdPPEdpkt0Er0NhcbXveF3p4mofjQH38Wagw+kiv9w7iCzLHI3nqHTb2bo4RNjnJOBxUFNh9gZM5FWeOzTGkVgWWZZ4/5oG+pJ5YuMl7IrMVUvD1PrPc/j6PMISvGngbIInhKCgCTx2mZ3DKr1pjVjeYF2tnfV1DnaOlI/H3c5mhR1OlBnI6Axndda7RmjyK+wu1VHvU6g9l4L70yiWDZJ5lbFsEU0HX7YLqbIFMUl1xdkQQtAzMMzvDsWpC1fx3vAYOw8ewZbs4mbxDG+IdtoYptKmUqheZ4pc7WXobrPWd3BwkCXKEHWvfQmlnKHsrmFoy/+iFDi1/55czuNMd1EIrUIyNBAakjAw7BcWN3Wku/H1P0e89goKwsa4s4FyeAXhChev9iTIFs1cs/ZqHyVN49WeJCPjRQxDsHFRkDf6khxNFihpBmGvk3evrKGlyovDrpBTdbPWU4ecqtGbyCFLEu9bXc+//a4LuyxTV+miLewlXShzJJalL2lacMPjxUnXG/Y5aA15aQt78Yk8V6xcTK3fecoXXM9Yjq6xLPsGUmg6+D3mZkvI62BJjY9FVR5cdoX9Qxl2D6QYSBVY3VjJZS1VDKQKNATclMo6ewbHSeVVVjdWsrKh8pKJx81rwYtEIiuBpdFo9BeRSMQPOKPRaGzGVneedHZ2ivV1EomCQcAl0zeu0zmkkiiYVtcHOtz0pjQcikTYI2M/S4AjUdDpTukMZXRkCW5c6ubVwRIeu0yjGKEyVId0AV08VN0gnVOJZVVKmoEiS2anXEOlsvsJUu03mTMw3gbdEMgSPLPnKDv7ElznPMja/PP8To3QKo2wRu4ib6/iCfdNXLukElvjWoTtTCEdHBykoaEBW36U+lf+DlfqEIZsI7bmjxlvefeJOl5Dxzf0AkJSzKqM3AiekU4yTe/AcF7AbrgQVHb9AlspTXzZnfj7n6EQXkOm5YZJHipQdYNi2aBQ1qn2OYllihwaGWcoXSJbKuN12nmha4yiaiAQLK/3s6LBT1+iQKZYJl0om6MCyzqpYpmBZIFJuhwhYdZZLwp66Kg1rbbWkBf/SQ0/u3u6aWttwxCC4VSRzt44sqywayCFWtbxOW0sq/dz+eIQdptEZ0+KnngOQwhuXFXPwZEMLrtZrhWcaFgRz5boS+Z542iKd62opTF4llrvixRZzVJMDs9PwYtEIg8C1wJt0Wi0IxKJLAK+F41Gp55lO0t0dnaKzqE8QZfMu9tdCEDVBVWuyZNmAUqaoCetMThhvX2gw81wziCe16n3KdR4FZzHrLd8Ao48Y07hOsf25OWJHdaR8Yk0EhmcNtvJM31wJg+hqOPkazee9ThCCGJ5wf6BBG/0jPIhfkNt5gAHjCaalDHWSYfJeRdhVEc4WvcufjneyrUtDgKuSVxjIbBlB4mNDBNs3wCShKSXCe/9OoHuJwAYb76O0bV/ckIohYF35DV0ewXF0HJsuWF8wy+Taboa/QJET1YzONJHAAXD5sFRGCax7COnuNZTQQjB8ESvt95EDoHA77IzkCxwYDg7aVUCmMH9xoCLKo8Dm01hvKDSUVtBTtXRDYEQAo/Txta2EOPFMqsaKs1NqcPd5BU/u/tT6MLcvLhuWQ1bF4eIZYr0JgoMjxdZ1xygoGoMpoq0hb0srfHhOy2HqKTp7O5P89SBUd6xNMymlqpLJ/FaCJRSCnvqCEYpQ9bVNG8FbyewAeiMRqPrJ27bFY1G187g+s6Lzs5OsaFOesvY2WBGZzCjMZg1aA/aaK1UeGVQpc6nUO9TqHSe5fmGDgf+x5zJ4G+Y0np0AzLFMrFsidyxHVb7qSJ3Mq7EAcqeWnTXmTWKOdXASPRQG3+ZH3Y5KRaLvF95ibioYFyqYFkVhBo70D1hc4PBac4BzZUF/tNjkELgSuzHmepCMkqMj2eo9LrINF6FVmHWz1YcfZqaXV9F1kuU/K0Mbf5flH0Nx58PAlt+FN0VRC7n0Y+1r5LPP0/ONbaXUjCCpBVxj+1GDSwl03L92z7vWFXC8Vhb3EwFGS+eKW4S0FTloS1kbiIE3HYcdpmA205HrY8fv9bPULoIQtBeXcGNq2t5fNcQo5kiqibY0BKgN55nNFOkoOrYRZkb1ixiMJlHls0vVo/T3HX9/aExgh47S6p9rGjwn/V9qekGPfE8T745zDs6wnTU+i8JodN0A7VcRuRiuDO9KKKM7PThdbsYU12zKnjn4o+p0WjUiEQiAiASiczPdggTnPymypQM+jOma6oLuGGxi+6UhlOBDXV2ar0K9okWSm+LVjLbNL2N2BmGGSuKZc35o0KYBfPet0kjwShTDEZOaQMl6WUSffs42DtAPjlMQBujUj7IcqOFPrmG+oZGaho2UajbhOGsJJ8dxDP6Bgkq2NFV5r2L7SfETghs+RHshRiF0CpsuRHKnhry9VsYHBjApQxR2ftbcnWbKYZWkFl0LaXKxdS/8gWc4z007/hzRjZ8mlzD5RNrlFDUDN6RTjLN14CkUNnzK/LVayn7Gid/jW+DrJfwHX0KNbgUpZRELiaQS+MYTv/xx4wXynRP1H4e++lL5CdtTul1KLSGvdT5XeiGgWZAa8jD1sUhWkKTxx3v27aYkmaQzqt4nDZyqo7PZSOn2nHaDBZVeWkMuHkmGiPgFsiaoKO2goFUkXhOpSXkYVNLkLDPyeLqt69g6Ipl+WlnP1sXh/ijy1uouAgnnB1rOqBqBgKQJIFNaPjLMerLQ/jsMvamMA67/fjbe2xodlu8n4uF93HgeuBy4N+A24H/G41Gvztzyzs/Ojs7hW6YSb3XtbkYnOgc0lAxkUpyuqUzVfJxM6bmnrwsTQjIlzUSWZVEXsUwzAlW9nP4lvaMvoFu96G6a8j17eZofx9XZ58grdk4pDeyXyxisSNBnyvC4valNLR0YLOfcKslrUhlz68ZqL2aJ4dcbG2wUz9R46oUxvANvojh8FGqWIQz04eQJHJ1WxA21/EYnqSrVHY9hu4MkG16B0JxIJdz1L7+L/iGXgAgsfRW4svvAtmMMTrGe3HHdjPe+h4kvTSRwrIatbL1nC+zrGbx9zyBsHnJhNai9b1G1GjiedtmutIS3WM54rnJ+7fVT2w8HP8JeY+34j9GPFvCaZeJDmfxOBRUTSdS5z/emPLtOLarO5Qu0DOWI5YpIQppbty8DEMInFMcUC6EoD9V4LlDY4S8diK1flrCs5Mwf6FouoE68SNNdKp22sxBRn6XHa9SxpUdwJYbMj8zTt+kvRw7h4z56dLC8U2LazA9gqej0eibM7WwC6Gzs1OAZu6cnqXY/ZwxNNj/OLRcAb5TW08VyjrJnDm5TBfgUM4tjeQY9kwfVXsfATXLz2ON+KQCy+lBwSDm7eBJ2ztY5RtnW5OdXN3mMzcehFnlIKtZDmTdeGzQqsRwpQ6j+hoo+5qQdBXD5kTWStjzw5QqTzQVPSZ45rEMXIn9+Ht/y3jzNRSr14AQBA7/nPC+R5CEQT68muFNf3Xc9ZbLOQybB7mcQ0gyjkwfparIlF9/rizoTul0pQwY2Yec7uOJfIR14gBVUoZv6++mhBnQd9nNKouTxa0l5Dn7iMhJSORUdvUlORzL4bbLrG40h0qfD8c2LabK8HiB3+wboTHgZlGVmYA8H0u9DGFWeKiagYFARkKRweuyT4ibDZddOZHoXMqYA5ryCXMyneMsGy3CgFQfncXm+Sd4kUhEAV47Frub70xX4vEp9L1iJuI2bQLMQv1kXiV2cqG+TTl9MNZbI3RciSiOoVcY7e9hMCeok5KslY+Qw8P+yqvodF/OqGsx7SEXq0KCyvIYqr9l0m9Lz/CrjDvr6NJriISd+AaeQzLKFANLTPdSGHhG30AyyuQarjjj+acI3gT2zFE8I51o7mpytZeBzYlrbC/1r34RWymJ5qpi6LIHzW7FgFTOU3n0N+RqL6Psa8CR7kEuZye6GU8M3hGC4ZzgSErnSMqgK6VzJKkzlDv1vdjMMKNU4fc4+aBnL2G/C3v1EtqqXIRaViJNstN8PiTzKoYhGBkv4bJL5FSdtrAP31s1iT2NqQpeLFvkaDzP4dEsDZVurlwanjctsMoTg5PKhmm1yUg47Qp+l41Ktx2P04bLJmNTzowFU0hCvMsUPIfnrTfzCkno/j24AnS6t80/wQOIRCL/AfxDNBo9MLNLunCmXfCEgNF9lKsipEvGGYX65/J+lbQintE38A69THroELXlAaKimRERZJl9FH9tG28ErmUXEfan7LynRbBZ64T/n733Do8rQct8fydUzsrZtmz5OOfYcaa7p3t6QjMZJgLDXJaFS7gsLMsCy3IfYNkF9u4u4QL3mYdhlx1gIhN6mJ7QaTq31c7hOMqWlWNJleuE+8dXsoIlWdEt23qfp5+2qkpVp0qnvvOF930/zUu69uCMz+sdvozWfYxvjhps9/VSs600PC+VnGoxTaT9eQqRerIV26elu0wX8OT920Svfh/FLpCqPYQdrELLDVHz5n8mOHAKV1Hp3/pZhtf/WCnDHCXS/jzZyu2k/HUULr5AWyHG89Y2LiddLg3bTCNIwKPCmpjK+rhGc0zhXfnvE6xqxlPWQLjjZfLxDeTLNqJYOQnk5dsn9fUWi9FckRPXhznXPYqmKhjVEXY3JeZU6t4q4CWzRdknUbAJejUOry+fVza6lLiRtdlC11FcydrCfg/RgIeY34PPo94kT7v5iWyxWhu8LP6SvpAsrpr1xS04+zQ4RajdRWuxecUGvFZgO3AW2a2mAK5pmjenCm8zljLgWcU8mZ7LdOuNZIoy7ZttwjodtOwAoZ43CXe9jq/3OK12M+ecJsJKlodC19Hrd9BXtpeOjEZ3eCvHey2eavFQraeJtj9HNrGJfGLjtFmdYuVQimki7S/ydLKJ6upaWhqrxwOa66Llh7B9MfTcEFagYsbjnJFvUvQAACAASURBVDHglaCnuyg//T8ZbXiITO0hcB3Kz/wdZRe/BsBA9X38oOEXOD/qp2MwRU8yzclUFNd1eVw9wpvOJvqRWVfCr9AcLwW3uEZzXKUxOu7nB+Abvoh/4DRWsAbLn0DP9JNPbMAKVoFjoRVS5BMtWKHauf8x5oCRbJGhtPgNOq6LX1cJ+z3TL3QqYaaAl85bFCyHZ05349VVPri7/raWrmNZmwxzpPfo82g3ytEZs7bZYBch2QnDbfKzN3Tjwjoj0n3QfQqaH4aOozB8FZoO0ZquWbFT2g8t21GsMNgOpEtCfeXaq9h6AKe6YfY9GBPhunhHrhLqfp1w12toQ5e56NbRRoiDagEtWsPD9Q0o9Xvp8DVypMfG6TnPtYyPXVGXz2xRCeZ7KHpqGGl8ZFoOmp7txz94Dj3dy0j5dgbXP8VuRyXkGT931GKGUNerOHqQdO2hWYPdXGCFaund+fP4u17HOvk1XvU/yOniJ6gMruMX0n9Jec8r7Ou6wueLv8J5txGI85B6ksqwylDiPj4QVzmgXyRWv4my0K2nkPnoWvwDZ/GkO8mUbyXYexxPuovhlg+Aqpf2ZlxELaQm7c1YLKIByXQKlsOx9iHMnlFsBxoTAXY3JWYNfGPIFW1OdSa51CtrCA+vL6epLLhswW66rE3XFUJevbRrYo5Z22woZmH4muwVVjRZYXCr53Is6DwKyevQdFh+HmoT/fnbgPkOLYLA1tKPp03TzCzLUS0SC8nwXBdSBYv+0TzJEo0kUughMniSkTXvurXiwbEIDJwi1PU64e430NM9uChcdmt51t1LXSJIU30D4TW7yHniXE46eBRZgnN52OFj6g8Zqj6MnyKh7tfIR9eRq9g26SUUK4dv5Aq5+Aa8qS5cQE9e5pnR9dTW1NFSNn6Mil0g2vZdshXbKETXzWnb2XQZXjLvcGnImdRvuzZic4BTbFKuy/TU3cZapZu/8f43Nirt5BUfzzf8HPa6R1gTcSnreRlX85Gu3k+w7xhaPslow4Nz4+q5Dt5kG7Y/gat6CPSfxApWkY+vv/EQtZDC1QLkKrbgavMjgs8F6bzFpd5RysI+RnMWNTE/2YJNbcx/I3iMZXi246Io8KUj7YS8Ou/eWn2Dl7dUKNqyoNyekLX5PVqpHC1lbR5t6V5zroOIqSikwROAvnNQsRFyScgNw7F/gJod0HSQ1tHKFVvSvg/4M6AVUIGdwC+apvmdhb64YRgbgM8DHuDvTNP86yn3fxT4NSAPdAKfMU1zej7CBMwn4GUKNgPpAkPpPLbDJKG+nu7B1f0zKgfUQopgbyvh7jcIdh9Bs9Ik3SDn3CZOq5s4VKcSqN9GvnI7mlfW1JkDNqf6LeI+haGcS2VQZXelQnXqNLl4C9Hrz5OqOYA9MRtzLMJdr6FnB8jH1pAr24KrefH2n+VH1wtQuYUDdSU3YKeIf/gSuYSB4li4c1wA7roub13qZtRbwaWh8eA2kL35/FCA2rDKzliKny38b0Yqd6M3P0iVz6L6xF8SbX8OgOF176V/2+dwVY1A/ylyZQau6sU/cBo9P0yq/oE5HJhNpP0F0jUHcPQAoa7XQPWQrj0w+ZiWqa83FcevD3OmawTbdkiEvOxoiNOYCHLx8mUKgQpOXk8S9GqUh73saowvqk83MWsTkrdkbRGfTjTgIeLz4Peqc6bBzAvzHURMhF2A60ckQG56j2Te+RSc+SZEayFcA+Eq+O5v0nrfX63Ykvb3gcOmaXYDGIZRA3wXWHDAA/4I+FXgOPCyYRhfNk1zcML9R4D7TdO0DMP4I+BjwN8v4vWAm4X6uqYQmHJi+gZNCrHmmwKGnu6RUrX7dQL9p1Bcm4Kr0UuMRKScbwc+RKxmHXsa1uDx6RRcl65RB7OzyKF6Hb/ucqhOJ1GySqoMqujpLhQrj6sHSK59NyiqZHPJKyh2nmzlTvLx9aTqDk/ozTnY2WH81bvYUSPBTs/K1rFCdC3g3jLYDeccWrttjnRbtHZbDOVCQHbSY/w6NMc0mhMqzXGN9XGVdTGNgEcBwmD/EtV9R/G3/T0jax6jZ8+vki3bQuXJvyZ+5Wn8wxfo2v/vyFbuQLHzhDt+RLr2IHlU1GIGF3A9s2QMioareYlcf55c2WZcTxBH86HYxUnvz9X9uI6HQN+xZenrjWFnQ5yNVRHOdY8QC3gYTBcoC3r54aVRNjaGeWxzFQDl4XlKDidmbYpI3Py6RlnISyygE/AucdY2E6YbRATnsZXMseDst4Wc33hQgl26D175H+Avg6YD0Po/ZR2CO728bzkxn4CnARONAvpLty0GLaZptgIYhvE8cAAJogCYpnllwmMtYPqFmnPAJKG+7aArpX0P08QE78g1fKNXySdawHXwDV8slaqv4xtpu/G4AaK8EnqMS4Gd1Nc3sGtdNWN5h+u62I7Lty4WifsV1sZUjnYXGcjBgVqdoEchqIN/4DSJi18XEi8ACt5kG8G+4+Rja8SpRFEoTvwCF9Kc6i2wrvoQO71jRnY2gb4TpGsPz9irK9ouZwZsjnRJgLswNJnlHvc6bKrwsj6u0pyQQUJd+BabwzThBOaiawn1tBLsPcrI2ifJxzeIAcHQeZqe/xW69/0amao9FIPVRK/9kNGGd6DlBgj2HWO04Z043pnVCLmEQajzFfxDJiP1DxO//G18ycsk1713cqmuahP6eqMU4huWZbF3wKuxu0mCwJX+NN8/24OuKhg1kVsGOrvkXFx0xrM2j64S9mlUR32E/R78nmXK2mY9sGkGEXMtXUF2Mfedh5rtktWNUYZ6z8C1N8AbAc0H//IbQuCHRW32WyjmU9L+NvAe4GuAiwwxvmOa5h8s9MUNwzhhmuaO0r//HXDNNM0vTvO4TcD/BB40TfNm+9gpaG1tdcty17Bcl1TOZjhnk7PEUcSrK7N+gVU7R13vC+S9CeIjJuXDJ/AVkzfuH1QSvBF8iGBZHenYRgbcCLUBhzGe8XBB4fKoiuXAgUqbVBH8GnRl5bY1YQdVAVwHxXWo7n8FF5VsoJZw5ioD8Z0UPFFcRZv+y+o6dF09R5dWz66GGF47TSJ5mv7EbtxpemI9WYVTQxonh3TODmvk7PH37lFcjLjN9oTNtoRNlZ7D67t1Q37Gz87K0NT1DOlAPQPx7aiuzebLn6cseRoXhav17+dq3XsIZrvwWqMMRzcRyPZQljxFT8UhLH0GlYHr4s/3EU1dYSi6Ca81Qjh9jZHwBrKB6fcPa8U0ju5jJNKCoy78Pc0FluNSzOcJBCbzAou2S9GWSS8oYsWvK4S8KmGvil8Xp5630+pJsXJ40l3o2T5QVBw9NKd+7w24Lp50B4GB0+Rj628Q2bXcEI4nhL//BHqmh2D/SYLDwmhzVC+DGz7MaNPjDAbWrMyS1jTN3zcM4xngPqSN8yumab55q98zDKMW+Po0d/0hMDHFiAPHpvn9OuDvgI/PJdiNoRiqIJOzUCJQV3ZrGomWTxLsfpNw1ysEe4+hOuOtwoK/kkztAb6tPUK3p4nmhIe6hEatR2ED4waX10Yc2rI2mxs0GiMKF4cdzo3a3Ffv4WDDuJbVO3IF/6BJqu4+gkoDeqqLfFmE/LoPUD5LpgPQc/kko8FG3r29hWCqjWDfGTJr9uCJigV6puhyrNe6kcV1piZf0NZEVfbV6uyt0dlROcEBhlvTUuaCQmMzkYEz1HY9R6Z6DwMP/T7uha9Qdu6LrO34JpVWR8lYNEpstB0n0Qh1TVT54oAz4yBDsavwpGLoMaF++AaixNOdjNbNzIVXrBz1Tv+y9/Vc18W8dJnymgbAxXXBq2uEfRrxgIfQ25W1zYYbg4hhiMTAu4AWgOvIf1evw8ZPyNTWLspU1umAggaXviy7n3Hlb7vjx1EbD1BRu4MKYHAFa2l/G/hz0zSHSz8ngJ9fZIb3ZeA/IT28l4D3TuzhGYYRR3qEv2Sa5pG5Pm9ra6sbJI1+CyN/z+h1wt2vEep6A//gWZQJa1Wy0fVcqngHx3376dcqeXK9l0wRgp5xY4JM0eXCoM2lYZuDdR5qwwoKcv/L14voKuyo1Ev9LmRFYvvzeFNdOJ4gVrCKVO1B6cvNofRK5x1i6Ytkw2vweKT5n4uu40LKz5FuCXJnBuxJa/0iXthdrbOvRoLc1AXaE7EUAe8GbItw50v4hi8yuOmTBAbPTDYWPfCbWP4yIp0vka45QDFQRaztu6Sr92KFam5+PscmfvnbFAPl5Cq2Eeg/SSFUJ3SU2eDYaIVR8vEWrPDS9fVc1yVTsLEcB11VGOntYO/2TQS8Gn5dQ12JBp2LGURMfZ7+89Bnwub3TT53r74CqhcGr8AbfwVuqQtVfwBCFXDo58A/fvFZsVpawzCOTpWWTXfbfDDTlNYwjL81TfOnS4OKzwDnS7/yBdM0v3Cr521tbXWj2jSMGcfGP3iWcPcbhLpfx5vqGL9L0SiGarne8F4ytYcY1so51W+xPq5NIsS6rstIwSXmU/lRe5GYT6ElIU38rpTD0R6Lhxo9BD2Ml86ui2/oAlaoCk+qE8XKk0+0oGf7cXT/9F/wKbjUk+Ry7yjvXudB63id72sP8mqfn9Zui+H8+N9QVWBTmVbK4jSMsrk3upc04CHUmMi1H+KqHnLxFlzdR+2R/4J/6LwYi27/V6Tq7iPS8SKjjY+AYxG5/gKZqt0US/ZUExHoO4En3YXti5OLbyDc+TJWoIp03aHZD8R10fIjFEPVi+rrOa5LJm9juQ5eTRySK8I+Qj6ds2fPsnnz5gU977Jj4iDCKoA/fGtFxIzPZcHFH0i21nRYgqaVl6lspQGdx+DNz0OqSx5f1gy7PinUlFgDbH7/pM9/JQe8E8iUNl36OQy8Zprmttl/8/ZjYsBTrCzB3qOEu14j1HMErTBy43G2J0K6Zh/p6gMMjKQ46TbT4V3H/hqdusjkL4XluFwZdjg3aOPX4NG1nhsBzXVdXmi3yBbdm7Io/8A5Itd+gOJaDBo/MSm4xa58h9G6+2/pEnxxIMdbJ07ToA8xnBzh71P76JuwEb4yqLCvRmdfrc7uap2Id2Hnz1IHPEBK+OHLBPuP4niipKv3U3b+H4lfeRqAkcZH6N3+s7ieEHqmB8cbwZPqEGXJFChWlkD/CbypbpJrnyDY04qe6ydVd/9kGs8MWAhfz3ZcskULy3EJeDRqYn7KQz4C3snl6YoMeAtRRMwE1xHCcaxBSMOhKun1DV+F9jfl51Nfhe7j8njNB4d/ATa9V4Lh4GXY9qGbMsrbHfDmE+b/HPiRYRj/WPr5x4H/sfSHtDSIXfkOoe7XCfQdR3XGx9+FUC3p2kOMVh/kqt+gI62wp0plUO1kTayWvcHJTPTRgiya9msKA1mHBxqEUgLi7nFh0GZnlcbOKo24T0paPdOLd+Qqmao9BPqOkq7ZT6Zm/6Qrm5YbwlW0aYOd67p0pByOdNm0dltEet8gaft52m2mlziaprO/SrtRpjZFF8GeX24oCoXEegqJ9YQ6XqLM/CJ92z5LrmwTVcf+nGj7s/iSl+na9+sEhkyKwSqyFTtLpqRFcmWbbjyVqwfI1BykmBLLoXT9/XhG2ggMnCHVcGvjbccbFlPRnlZy5dtm7OvZjku6YOG6LgGvzpqyIGVh35zto952FLMwdA1G56GImA3ZIbj6Mvhi4gMZroZiRoJaPi0B8MjfQlFUJdTtgXf9HhRS0HMKGvfL9HYh5fMSYz5Di78xDONV4B3I0OKnTNM8uVwHtlhUHf9LQNYOZss2k645SKr2IIVQPckCPHetSMTrYoTS+AevU1s3OVHtSTuc7rMZKbjsrtYoj6kcqpeGuuW4nOi1aUvabK3QcYGEXy2tMnweR/PhKjq4NsMtH5mBD+eKiL+EdNHlWI91oxfXnRYWfTOd7NfaeCn8KPfVVbKvVmd7pYb3Dlw0mq47jKv7iV17FssX5/oDf0hN638VY9EXf52eXb8kw6O+o2TLtxG5/jyKXSiZHcj71fJJAv3HycU3UoiuwT98mWKwas7HIHw9W/h6E/p6RdshU5ALY9ins6EyTDzkWVmDhlshNyK9s2xJETGDb+O84Fhw5UdQv0eyO9eF/gtSuroOnPzSOM0kVAnrH4WDPyvcu0vPA66oLOZDcVlGzDngGYaxEThvmuZJwzAeAd5pGEbHFKLwikGq9hDpmoOka/aT02NcHXG41GfTmHcwyjSeWOclqLtEr71BtkxKkYItTh7r4xoF22VTuSaDiNKXzXZchvIuCZ+CT4f3rfcQLPTj67qI4wmTrdhGpmInwYGTFIMJUFTc6Xolro2t6JwpVHPkdJ7Wbosz/fakBTIhDxzyXuXT0bcI7P4oT0XvDGPIWaFoZKr34Ul1oqc6CPSfpnvPr5G4+GUina9Qd+SPGNzwIYY2fBCAkfqHCHe/gVocxfFKNmb7YqCo+JKXUHDJxzfgTV7Bm2ybu9loia+nDZ4nkxokF91AOOjFqI4QD3nxzEdI/3ZjukHEUgS6dD90nxSx/+b3j2eI7a/DcDt0HJF/AwTKYOcnoWEvlK2VIHn1FdB0yew8gcUfzxJhPiXtl4C9JU7cnwNfBv4ReHw5Dmyx6DzwWwxkXSp8Km91ihfR3hqd8oAEsKAq/TXbEyYXauDNjiIdKYc1UQ3Hhcbo+JXddV3akg7He20aIiqVVTZby70E+o6j5wbIxzdQiDSAouJNtZOu2jPtIGIg63Cky6Ljehv5/mt8rTAuj1IV2FohZer9sT5OJf28W72Gf+17sYN3QbCbgGK4jmK4DtWxiXQ8T8+uXyKX2ETFmS9QdvFr+IfOM7jxY3hyA4w2PIyravgHzpIr2wiKRi5h4Bu6gH/IJLnmCQJ9Rwn19GIFKmYlMAMULVE0KAqEgwk2eLOEfG3oNTtXRMk1Z0w3iJiPImLG57Ukexu+Jma3iiqZXO9Z8Cfg/Pfg8vMIhUiH7R+VUtfOQ6hcgq83DPX7YLQLKjfd6hVvK+YztHjLNM09hmH8FjBgmuZfjd22vIc4f7S2tro/vDRE1KvycJM+yXJoDI7r0t0/xIDlY3ttiGsjNnVh9abHjmV1J3osDsSGqc5exJPpZWTN4zi6HxQpZUPdr5OuOTjpC1ewXU712TfK1CtJ4Ys8pb7CMXc9hWANe8eGDVU6Id3B6jxBJN8Dmgcn1iRqj9uEruvXqK1vWBZ1wrRwXXxD5/EPnEHBoRiopurU36DnBrF8CQaMj6MoMNLwMMGB06jFDKP1D4j0zrEJDJwkW7YFcPGOtKMXhknXHLjpZcYce1VFnFDKQ15CPg83EjkrJ0Gjbhf4F7aq5bYNLewiJDtkWACLG0RMxVSxv6pDdhjaXoKBC3Dhh1BMyWNrd8EDvwIdb8FoD2z7gJTRl54VY4CK6e3MJsHK0drnXbFDi4xhGL8LfBp4yDAMFVheCvsi8MQ6L0HP9J9j12iRPvM1klX7MMrlqt4UnXzSDOUcjnbbxL0W+xNpHmmKE+48SyG6Vow4S5pWmUCeJFO1C9sTon1EpFtHum1O9FrkJ4jh/BrsrXJ5ypfjg5vXUh/VJw0bit0X+F5PlAMtm6gM66Asv0GkYudRrRyggKKiFUZx9ACudhv+tIpCvszAClYJhaXjRfq2fY7Y5acJDp6m6uRfSXmr6KSr9hIYOEmo+03SdYdRC0PYpcXfrh5AtUZFtF5SNeQth6LtoKoQD3ooCwUJzeRjqPvly329VagVsYUtH1pWLPUgYiKmiv2rtkgG6bpw8Ydw7tswVFJ5ljXD4f9Tenr9FyCxVkpZ3S90lfKNkg2WrZMAOBNyI29LX28+36iPAJ8APmuaZmdpL+1/WZ7DWjwmBjvXdenPupwbsFkTU1mbOUtDjRe7YfrewpVhm5OdoxzyXKKl2E7eZ2AHKkg1PHzjMYqVw9V8ZCx4QX8nr53zcKQ7RW9mcsbcHFdvUEa2VpSGDe6EBduui3/IZNTx8XRyHYfqB2hOvsZo9J1L/6GUXk+xsqh2EVdVsf3lFGIbsL1RhovXKKuuxT9wCrWQxvHenlJ6bCdGLtqMf8ik68C/p+z8P5G4/E3KLnwVT6qTbMV2UnX343ijQi9RdQIDZ/ENXSJdfx+uGiCvQCY9iuINkgj6SIQ9hDz63OKCqkvvq9+UXljFRubl8rpcGBtEZAZBX6JBxETMJPa/8H0x6uwoiak0H2z9IOz/Gfm5/wLkR0oDibBkybW7oOs4tDw+c7BzHcgmId4IZeuh5+jSvp9bYF5+eHcKxnh4ruuiKAovthfJFF2MMo113iSxrpcZWfvuSdPTgu1ypjdHrdVBeUUVHiePx85QCDdMKhlsx6Wz4yq5q638feYQrw1FmLgZMOZT2Fsjvbg9NTrlgQlfGtcl1PWqlF2qjlLMEO56DVfz0F+xn8F0kS2DzzLSNL3p54Lh2qjFDIrr4qo6VrBa+l2eyZukbrj2Ojbe5OUSyTd6+0pcZJFRqOt1sdN2XcrOfwnNzlIIVJNqeIDk2idR7RyB/tNYngiWZVPUfGSrdtPQ+QxBj4pnx4dQPIvYd5EfBT0AtTvm3Ndb0pJ24iCikJIyc6n7ixPF/nZ+XOzf8Rac+QZcf1OoJwBrH4TDPw+R2nHicTEHnqBQTq6+ChufEK87uwiRGYj0Vl4y1ertEBbeZGtr64otae8ovNVtcX3U4b3rPRyq08dpHE6M0caHJwW7y0MWZy5extC6qK+LYut1uJ4oY2ra/oxzow+n9xynwu7lGfsQw0TQFNheqd3I4jYkZnYX0XKDaMWUZBOujVYYYSS0ludGG3hQ97DePk+6es/SBDuniFbMAi6OHqIQXScmmvocJmaqRiHRgu2N4R82b1+JCxQjjSRLNJPY5acZ2PRxYu3P4htpI37pW7hodGz4JJmoS6L3DQJN+6kZNfHWxyD6IIx0Q99ZmEVje0v4IhIQrr0OdTsX3NebN6YbRCx1Rue6Up52tJbcSlwJdiNdIhd7+X9AtkQziTXAI78jZf4Yrrwo2V4hI2XtpWeh4YCUxA17Z1Zw5EYlaDcdeluntndtwPPp8MQ6D5qq3PCwCvSdoBiqkZ0IVo6uzg4a6SZcfpjHjQTe8HosVSNvuZzssm74xLUlHerpY5gwGmsIB7fwYJ2XfTU6O6v1Sbbqsx5T8jL5SBOhrtdxdB/Jsp38oLtIc1zDXxwiW7ljUe9ZsXKoVgEUoW/kEk04vtiCg5UdqiLji+Dvv70lrqv5wHXJJTbiHzLp2PNrlJ/5O2K9b1J+4UsE8z2o2z6Eb81TUjoVq+T/8TXguND5FtTuXFxmOtbXaz8CVZuWt69nFUTFMHEQsRz9rTGxf7IDNr67FNjzUr6e/Mq4IsMXgU3vhz2fgbFM2bEBVz5Xf0Kywp7TEuyGr8rnPy0Fy5XBR6QWKjcu3YBlgZgPD+/fmKb5p7e6baVga8Xkt6ZnevGOXiNbZtDf38upC1fwhMqIrd9OZVDh2kgZRy5YHOnKcaLPplAaNmjYPOo5zX2RPnKND7C1oZL6yAK1mLof/6BJIbaObMVWzD6bpqjKTu0K3t6roiedlzWPg2plxdlY0bACFeQTVTieyJKdWK4eIFu1B2/yEp50D7YvcltKXNuBgcgmdK2civ4jKGsfwIpXoV94hsC1F2DQhIP/B8TXwYkvQfVmkTGluiRALcUxqrpQPfpN6aVVGkvb15s4iFD1pR1ETITrypS195yI/dc9KLdbOfjB78kKUteWIdmeT8HmpyBYNuH3HcnsQuUwcBnWHJZA2bBPennFrFBYpsIuSklevWXmMvc2Yz4Z3ieBqcHt09PctuKgWBlil79FSo3hTXVwtdhA48bdXE8r/L+nLVq7UvRNsTLfkJBhw3sCp1kTCVCofv/CBdeug57uphBuxArWkPdXMJBx2VapoueHCHScJrnm8bmd7I6NamVQXAdH81EM1WL7K3A88/Qxmw9UjUJiI7Y3jn/oHI4nNGfr+PnAsl1yRRsFMdlsTASJ1m3As3GD9I5OZWHvT8rUMNUNL/wJPPB/SfbVcxqC5ZKBnP8eDF2Flnct/G82BkWBQAIyfdCRmldfb0Ys9yBiIhxLJq2qDhsekwtBIQNv/A1c+qH0KwHKW2Dnx6H5ocmf2Viw0zyiy403QdvLUL9XnjuxFhJrbr7AFEqLDRsPSsa6QnDLs8EwjJ8BPgdsNAzjlQl3RYAzy3Vgi4brohWS2L44btcpnrV380ZuHXqbn6O9RcxBe9KwIe5TSpw4jb3VKrUZE9uXoBjaQ2ERgUQtpgl3voIndZ3k2icpBKt5qd0i6FGoCumoxTSp2sOz9tYUu4hqZXFxcfUwheh6bH8cd4kWUc8Vk0vc1C1JvnPBWJADCPt11kaDhAMePFO5k6ouGUXHW3D/L0PrF6Tn9Pwfwpr7ZTLYc0b+HW+QL3L/Rcn2lgI3+nqvLYyvN90gIriMgW6i2L9u17jY/9Jz8Mb/J1klQGIdrH9EsuOJWd0YrDz4ohCth8FLciGp2y2fx7mnb3I/kfeZFCuoqs2itlhBmMvRfAX4IfB7wH+YcHvKNM2BZTmqJUCs7V8Ytjy8VNjAmTabpzObyNoAorrQ1cnDhua4DBvUwijhrlexPRHy8ZZFZU1qYZTotWfJlhkoVpZioIpXOyw8KuyrVvENnZftW1M3orkuip1DtYuAguWPU4iulYnpXDZ9LSOWosSdqHaI+HTqYiFC/ukJ4jegKFC+Qb64PSfhwL8SHWf76yJsH+2Ew79YUgTEhC5x8YdQ0bJ0faOF9PVuxyBiKrLDcPWlUqAqif1HOuHIF+Di9+QxnhA0HYSH/51kmVPhumLkWblJnsMfFTuo/Khke+a/jCsxbrxXS4YTFS1CO1mBuGXAM00zCSQNw7gAJCcagBqG8VuLMQBdTvzESVls/XHtWZ63D5MF6sIq+2olyO2qmmDMCfIHdm303ADZss3T+rHNFYpdwJPuohBpIrn2cTzpR2d5EwAAIABJREFUHvLRtaAorItp1IQVAqXNXfl4SUnhOkIdcWxcTccKVJIPVMrE9jbSQuaEGyVuDP+QOacSd2z7loqoHRoSgclqh7nCH5UsrpiB/l1Swh77BykRn/ltuP9XwClIIKzdvfRl/lz7elMHEb7w7SHaOpaUoHW7JejYBfGnO/V1+VxAytHmd0hmN1Owa39NsrsxpxRFk95fuEouKjXbJvflihlwHJnU3q6p9gIwn3zzw6Zp/v7YD6ZpDhmG8RFgRQa8y5kAT+hHUco28ImmavbV6NSGp/92KVaWcNfrFML1i5Zy6ZleQl2vlzI3BVcPkI+u4Wi3RWTIoaVMQ0934xu5ykjjI6iFURRcHM1PMdyAFaiYfYvXCoIdqibji+LvP4layE8ucV1uqB00FeJBL4mQd2a1w3zhCYqw/eorcOBn4ej/EivxF/4T7PqUfOkGLspja7Yt7UVjur7eGG7XIGIq0v3QfUIC2ZjYv+MtePYPxmkmFZvg4V+XDDNQNvNxDVyUPl/dHrjwjNxWt0vcUEa7pFSdiFxSPu+qrdMH0BWEeW0tMwwjNMUA9O2tr2bBf31ngC2JA2ge36wnu57uJtz1OtmKrbKAZBFQ7DzB3rdI1d2HHSgHQMv209HexoBnBzurVRQ7j+I4ZCp34uo+isE1wo9bhgXStwNS4u7FN3wRLd1DRgtjOaCpkAj5KAt5CM5V7TBfROuk9zTaBU/+EXzvd2CkA479L1j7kHxJr74i2VVZ89K//gS+npZVRXSfGVr+QcRETBL7H5ZzfaQLnv196D0tjwmWQ9N9UpIm1s5c4ruuqCfKmkUaVkgLMVnVpUXQfVK878I1EiwdW7LcsubS4GLlW5bdtQag9+eeZ8TzyIzBTrELKHYexxNetLJBLYwSGDhDumY/I2uemPSHV/ov0VMI8WhDFr2YxT94npHm92EFyhc/QVwBcFyXTNElGViHXw1TX7hCLJYg4A/cnvPfE5AvXGYANjwuu06PfRHaXpQm+8Z3w6mvyaBjGSbLY309X/IMVBjLO4iYijGxvzcIW56SAPTyn8GZfy7tktBh76cgWCEDn1tRQ7qOyefY/E5pCTiO9OlUTQLq4GUwnpTz28pJyVu/W7LdOwRzzvNN0/wb4CeRTc05xAD0b5brwBaLfLTp5mFACXq6m2jbd9EzfTje8MKDneviG75I9NqzFMJ1ElwVRaRchRSdvQMEU21s37EXq2YXql2kEG/GClXf0cHOdlxSOYvhbIGC7VCfCLBnTYJ927ZQu+0dBHUFpZC6vQcVLIct75fp5J6flD2oIx2y9DmfHvduWw6ouvj13S57Kbsgmeul5+Tnys1w6QX44sfh9Fcl2DUchIOfE/Lw5vfNIdidgNFuKYnbX5fBjD8+ng0qqvT8NK9kdYouGeMdFOxg/kqLM8AI4AMxBTVN8/zsv/L2IB+/eScCgCfVQbDvOKn6B2+I1hcE1wFctNwQyTXvwtU8aPlRcB0cTxDTruF0ziGx5adRyjbgTbbhyfYy1PLRhb/m2wjbcckULBzXxe/RaCwPUDGd7bk3BI0HxGIo1SNfmttV6gTLpKxLdkjGc/U1GLoEV54DXwiqt0rv707yvZuKMbF/WTMYB2GoDV74Y5HTgahN1r9TptlTp6izwRsSnt5Ih8jE1j0EtdtlGNF9Chr2A66U7GPC/5VgrjBPzEdp8VHgd4EmZIvYDuAIsqd25WHKl0zLDaLlRyhEG0kGaxZFVdDT3YR632Kk/kHy8Y2ojoWtB29IuTpTNsd7hniyKY1Sth5ch0KkkWKo+m2X1swHlu2QKdq4rkvQq9NcGSYe9Nx6t4OqSXAJlMkX0RtennJyOlQYwscLV4nWs2Y7nP1nISt3n4KNj8PWD0jgu5Ng5WQtYs0OsXCyCvDyf4ez35T79SDs+riQh0c7INY0twtN71kJipWGlKjeKOz/rPzNHEuyyIqNklUWs/L64VsvTFqpmE+G9zvAYeBHpmnuMwxjF/Bry3NYSwjXwT9wBl/yyriP3UITDsch1P06nnQn6ZpD2P5K8qHJUi7bcUkENR5dF6Dq2jMMxhtJXPgKwxs+KGqIFY6x3Q4KCmG/TktVmHjQi1dfwNU8Wis0ks5jJQLr4onKt4TmEaqKJyBBwRcG70/Aya+JVvToF6VPte9nVsyehVnhupLFdRwp+dRZcP4ZeO2vwMrKY1qekIAVSEjWFV8zt+fuM2Ui2/K4cPde+0vpedaXPH073pILR7hUDr/Nwv+lwHwCXt40zVHDMDAMw2Oa5jHDMLYs25EtEcYcSqbaQc0Zjo1qZUvKgij5hMHQxh+XbWNTrqC9ozmOtQ/zrs3VhNLnySU2Ebn2A3KJlhUZ7GzHJW/ZFGwhAacLDrqmLO1uB29I5EX9phBw/Td/bkuOSkMIspveI19QxxZjypf/O+STsk4wWAHbP3L7Ms+FYEzsP3JdAtHAJfjq58a5fZVb4IFfliytfvfcAx1ItjZwSeR3rgtv/LVcKMZcZhwLqrdJZhmqXBHC/6XAfAJet2EYceCbwLcMwxgEepbnsBYP39AF9PwQ6ZoDpOews3QiFLuIWszgKuBqIdRiBv/IFQY3fXJGOdVAKs+rlwZ4aGMliqKgZ/rIJTaiZ/vIVN9sO3474bouRdulYDnYOOCCpiqEvDo1MT+xgJSpZYU+tjQsw5RR06XE9SekxPVFlj/QXH8TYo3CEWvYJ1l98zsky+s6Ll/wC9+Hx/6DUDVWEqaK/Ss2wot/AldekPv9cdj70xAoKUoqWubHMxzplMzNeFJ+HrwMNTvF025s12zvOenbVW9dMcL/pcB81jS+v/TP/2AYxjuAKPDd5TiopYCUnXMPNIqVQ7ULgCLWSiUpV/Tasyh2noEtn5lV73p9KMsDGypIBL3guqQaHsLxRCgsUp42Xziue0PVMGZ37tFVIn6durifkF8n4NGmzd6WfbdtrCRR6jouPaHlFJVXbYbO42JjlF0vDfzRHvlbRBvhwvdg6DJ88xfhkd+WLHQlYKLYf92D8Nb/grf+HtyitGN2flyyreyQlLiKwrx6NENtIhnb+IRQas58SwY9xrvl/syAmAOsf0xK2BUk/F8KLIgbYZrm80t8HEuOVP1DsweaMWsl28ZVVZFyBatuSLm8ySs4nhCpusPSo5vhuUZzRYYzRXY2jmdG0bZ/wT94loGtn13cJPgWsGwJbGMlqYqCz6ORCHmJB3QCXh2/R0ObTaN6u+ELS3DpOydW4stV4oZrwDki4vieM7D2fgmC5RukTKvbBa/9hQSO7/4WPPTrYDyx9McxV0wU+9fulDL1a6X9riBysEP/Wt5P/3mR1823xExel77cWLA78WUhbW96b+kYXLlIbHhcJr0rTPi/FJiLW8obpmkeKP37T03T/DfLf1hLgOm+RI4lelXA0bwUQ/XYgXIcPTj+eKdIpP059Ew3yXXvk17dDEjnLZ43+9i7ZjyoKVaO6JXv0L/tc0sW7MZK0rxl4+CiAKqiEPLp1Eb8RP1Skvp0dfmztKWApovcK9khgW85SlxFAeM9gCu9KpAgV8zB2W/B2gdEgnXxWemRvfCfhKz82O/OvnxmOTBR7G/l4JU/FxIwiPD/4M8DtuiFy9bfLO2aC1xHpuYt75KsrfuUkIkP/bz8Pay8OKHs+JgE3TvhPFoA5hLCJ56Jy7RZZvmg2AVUKwso2N4whfgGsYyawVop0vEjHM3H0MafmPUKajsuz5/vY2djnLr4eKnr7ztOPt5CrmrXgo53aknqAj6PRtinU58IEPRpM5akdxxi9ZLhLVeJq2qyQ7V+rwxMwlXihlxpSF9qz08KfaPtJemPXX0ZvvxZeO+f3L6+1ZjYv6JFyuzvlwyJVI+Qhrd9SJxJaneIZ91CAtFIp+hsW0oZ7OBlsW/a+1PCTyyk5TOo2bZiXU6WCnMJeHfklh8tNwKKguVPUIg2l6yVZni7rkuw5wjFcB2jDe+4ZQO4aDvoqsLDLZWE/ePP6RltpxhpZKDMmOW3x2HZYpNUdJwbWZvfo1EW8hIL6AR9On5dQ11JJelSY6zE7T0r/SN/dGmzC19UssjBNtj2QbE7avuRZDyVG6HlMdjwqFgnnf6aZHtf+Rl4529KFrhcGBP7r31QjuU7/xYKJTPOxoNif5UbEr7glh9beAY82iWqjJbH5Oe2l+Tz2PfT8n0YMyONr4WN71mSt7aSMZeAt90wjE6kM1pe+jeln13TNOuW7egWgWzVjtJWrtmDl1oYIdr2DI4nRLZi+5yC3bPnetnVGKc6Op4larkhYpe+yWjDw1jhyR/JdCWppigEfTp1kQBRv35nlaRLDU0XVn+yQ7hhviUkKlcakiFFa+WLXdEimWRuSHp7DfskwG55v/S4Bs5Dsh2+99uylvC+X1xapxXHlqHB8DUxGPjnnxsvucvWw4P/RgYtXUdlX4SiLO6zGL4mAd0fFwVF13EZRqCIaqKsWWR5FS13pHJivpiLH94KJirNDMcbvfWDXBfvyFVy5dvIld+6L2LZDi+YfTSVBScFO5wisSvfphiuQy2myRZsio5Iz1xXStKIX6c+ECDk0/DfLSXpUmNSiVtcGmKwNyRfcN0vmU1Fi8imXBfOfFOUGLpPSryHfw2OfxGyI3D2G3D66xIk3/UfJWAsFmNifysv73GMZuINw5YPwM5PgKaJK8nm9y+ul5juk9dpPCjv1bEkk6zaLFSTQkpcjEe7ZEp7j+DuG8PMAYqdJ9L+PIVwLbmKuW8KS+UtauN+NtdKMB0rSe1inmJsF4ne10jXP0JFxEvUf4+UpEuNGyXuGUgPCNdssVlvtE4C19oH5EuveSWolK+XQEBJW+sJwO7PyGMcWxQNXcfgSz8FD//GwgODXZC1iCPdwgs89sWSGacGzQ/Cnp8WOdi1l2Wo0LBvce83MyC9y+Z3SrDraJX3VLsT8inpD1YYcPLLwvO7h3DPBTw93U2s7btkyzaTK982p9+xHYdj7cM0V4RoSAQZyRXQFIWQ30NN7gKheBXehq34oiMoLU3L/A7uAWi6NOmHrwsFwx9ZnLuMokqPEGQn64bHJGPrPy9Zl+YdLxvH/t90UIYe116DdA8881uw/cNw8OfmRwcZE/vnR+X/mRLNJNoAj/6OZLXmd4VAvBRcQCsn+td1D0vW2nlUDEk1TUrmeJNQc078k1B1Yg2Lf807CAs+iwzD8AJPmqb5jSU8nuWD66BYOVBUkmufwArVTvsw25Ep6XhJ6nLsepKIT6elOkK41G/zaCWjxfa3YN1npGTadPc3fW8r4g2S4XUdBwqLK3GrNosTSCEt3LtYg0xse04J9aN66+THr31AjDxjjTJcaHtRMqLMoGxKu5UueEzs7w3D5Rehu0Qz8cdF4bD5KRlSeEJC+l2ChUgUs3IeGu+Rz8qxZChh56F2v2R4oQq5fctTdx2peC6YV8ArBbl3Ax8DHgUSwO1dnbUAqPlhYm3fJR/fQKZ6vFwo2g4Fy8FyXMBFUxX8ukZFxEss4CHg1Th5Pcn6ijBPbK2ZXJpaBXGq2Fxa33jk87D/c3ctf+ltgy8yXuJmBhdOVE6slaFE5eZxInLTYUCRXlqlMTmLVBQZGsSaYOtT8IP/G669IqsNu47Bfb8kvcCpxzIm9m97SeRhV36EaPm8wnHb+kFRfFx9CWp3ye8vRbDLJUUq1/K4TLp7z0gpr/sk0G1+n5TsbS9LAFz/yOJf8w7EXIjHXuAJxOH4HcCLyCaznwVOL+fBLQU8qU6iV77DYO0DjITX42SLMiVVIez3UBHxEfN78HtVfPp4qeK6LpmCzc7GOB5NnRzsXFdKn20flitm1wkpFVaD3fJA84gtUbJj4SWuqskOjGJOJG4gvLyRLohUC/k3NEVzrSgy3c0Oi4NIpSHuyZkBeO4PJPM79K/Hj8V1wLaE3nLxOSiWTFDr9shrb3hMHpMdgE3vWzrnkfwoXPyBDGf8UdnRO3BZCGXrHi4Fc030sf2m7Py4RzGXs6YPuAj8R+BnTNPMj91hGMaK5eiNjo7gzQ8wGqwivf4jhKMJ1ge8N4i7+i2mpK9fGWQwXeA926cpfdteEiuemlIPsPvEPXvFvG1QlCklbn7+JZldEF+85ndIoKtoEXqK6pW/p+tMT0EJxGWQcP1NePI/wwt/JOTdU1+VQLL1Q3hHrsJr3xEbp8HL8nuxJpFxNe6X0rL9dTlP1j64yA9jCorZUjbaICTj/ovyORWzUsorilyk+87KRXqFL9pZTswl4P174KPAnwBfNwzjq6ZpvrkUL24Yxgbg84ia4+9M0/zrKfe/D/htZJnsCPATpmmOzuW5d/Z9A63pIN6m2nlPSVuvDnFtIMMH90yzd3Tgklwld39m/LYNj0kfaBXLj8WUuJpXsrlUr2RBZetkWHDq69JzizXI5HY6xJvk910bHvs9yfD6zsGJL8HVVynHC8kSny5YIXSX+35JBjAXfyAUkKUOdIV0iVd3WD6D/Cj4YjKUyCdhz6fHby9kpJy+x3FLMphpmn9hmuY7gIeAduCPDcO4ZBjGnwKLbT78EfCrpef+rGEYU1eff880zUOmaT4ItCJl9ZwQ2v1R/Gv3zzvYua6L47o8tavuZq6cY8Pl52DLB8evkn2mNIpXy9nbh7ESt3y9lJuONfffrdoik9pYvVBV1JKuN1onVBV3lqIlUiOBa6QTDv+CWDT5IpBsx5+8JG4mjYfEZ2/3p8DKiFJi3UNie7+UeuFiRqRokVo59/rOw7nviFSsbhfs+oT07+yiZKKZ/qV77TsY87GH6gH+AvgLwzBqgA8jOy4WgxbTNFsBDMN4HjjABMsp0zQLEx7rB07N+Zkj88+4znWPoKsK+9dOjbtIbwYX9vzUuIuE68KlZ+XkXsXthaJI1uWPS5ajqHOb4gbLpay085IhgpR9dlGGDSMdt6ZqVG+RrG3NfRJ0T/wT2fQogUd/QwJgIS10kOotEiCD5Yt+uzeh/Q0J+mXr5KLb8ZZYvOeSUrKDnJ/nnpaBzdQp9D2KhdpDdVMKfot8/Ylp0RBwU6QxDOMTwG8gm9L+eK5PfPbs2XkdSHuywNHOLI9tiHB24ObEN9z5ErY3SnYCUdmT6iQwnGfk8vV5vdZKRS6Xm/fntiLgxGUvbmG0JCecPdtWrCy+kTZy8Q2o/adxPCH8A6dxFZW8VhTd7S2g+LYQvHCUTNlmwvFdWOEivcMOOAMEBk+TrdiJkw7ClVs/13yg2AUUp4CjN8GIijJ4Dm+qnVT1k4SPfl+MK/Klv6Hr4kuq5JUquBP/rsuAZSceG4ZRC3x9mrv+EHAm/BwHjk19kGmaXwS+aBjGLwO/Dvzbubzu5s1zt9BJ5y3eaL3Ozz25hVhgmrKj6wTkArDzI5NJpz0ONK+lfqa+zx2Gs2fPzutzW1Fwt4kGtv/irae4rgOnz0BMgaEOWPc41FcIATiUlQFGdHqe5iRsMKQ3Fi7S2dPPGl9Ssj52LtnbmgQrL9STSgMqmqHrpPQU7/8VcVyJ7Rk3Ceg9K1PaLU8tz7EsEVpbW2/r6y17wDNNsws4NN19hmF82jCMPcBx4GHgv0y53zdhKjzINBngYjGULhAPevjEwabp9a2uK1O3LT82Odg5tnC67gHB9R2BGyVuTC5Qs5W4iirDitywTDKzQxLkItXSG0tem1vAA+mTFbNoxVGof2jp3s9U2EUpo8ubRfDfdQKSnXDfz0twbzwwzucb6RJp2a5PLN/x3KF4u7+tvwn8P8CPgC+YpjkIYBjG35bu/1nDMJ4v9fc+DPy3pXzxnpEcXzvawUjOmj7YFXMiwt76AenNTET3yXHx9ypWDvyxcWvy7PDMj6vYKHSTqi2SGYJMO+v2yBAk1Tu311M1WHs/maq9MrxaDriuBOma7RLsxgL14X8tNlNnvi78O1WVXvPZb4mSwz+zee29irdVS2ua5kUks5t6+0+X/v9nwJ8tx2v3p/I8faKL926vnb6MdV3hbMUahaA6Fd0nRei9ipWHsVWNw+2idvBP44Wo++Rvm09BqFz+3qoudJdY48pZ1u1YMhhrPCADmvSAyN7WPyoE6Avfgx0l8oJtSWDc/ak7YwXl24C3O8N72+C4Lk9sq6EmNsNVuf11KZMap1kElBmUKd8q927lQlEg0SRux8WslKpTUcyIxGvgspDHQTh0yXaRDuZmyRBvB8aCXaAMnNLKxuFrJYK0Iv52G98NwbLSBfpb0Ht6NdjNgnsu4CWzRb5zsovKsI/6+CzSHkUV+c90Ez9/HLZ/dJV7dycgEBdunCcklI1J9yWkDFUVkaw5lmTziiZcvc6bZmi3FyOd0pcrWy+7YfvOSTbnj8Fot6g4ytbJY9tekqBXPTcHoHsV91TAS+Ut/vloBxurIzM7C+dH5URvPDB9WeM4svvANweD0VWsDOheKXHj68TldyJRuWqz9PqiJSIyiJKjapNk8lOD5O2AY4uix5+QDLVhv8jgtn9Myu+LPygNZkrnsJWHdO/MF+hV3MA9FfBevtjPgXVlbKiaQSDi2HD6n2enNAy3Qapn9cS606AoULZGTAAmlrjxJqjbLSTesclssEzIw2XNtz/guY44rAxekuOI1sOpr4Dmk2DX0Sq9uw2PyuNHeyTg3eMa2bningh4ecumP5XnXZurb7gVT4vLz0lJUzNLWdB1Qr4cq7gzMVbi6oHxYDZ8TcrFfGp8OlvMSg8vWi8B5Xah7RXR9e78lCg0zn5L/PPG6E+KKppYVRNDgtNfE7v2VcwJd33AK1gO3zjaSftg5ta62srNMv2aDb7IzALzVdwZ0L2S1cXXSokbqpBy1iqM74NNrJWStuvY7enlua4EsPL1otON18PARaFERWrkWNrflOP2BORYT31V7J+iK3KP1orEXR3wbMfl2yc6aSwLsrtplqXY6QFh2cfqZ9+2XsiIDnO+G99XsfKgKFC2Vkpc24JIlZSxY0RkRRURfrBKMsBidvmOxXHg4g9Fy9vyuGSUPadh3YMSzIo5CW7BCbx7RRHeYPWW5TuuuxB3bcBzXRdVgR0NcQ41zyLQsApSFlRuvPWTnvyyED1XcfcgEC8RjvdBMCFBrpiT+8qaIVotk9CxVYpLDbsIl58VruCuT0oZfeork4nu574t7shjlcXVV6T0rtq0PMd0F+OuDXjPnO6hYzjLhqrw7Lter7wok7qy5tmfMN0vusVw5dIe6CrefuhesXz3RaGQF9rHWN+u75wMEpbDbaSYEVZAxUbY9zOAK5mc8aQMU0Cyvw2Pjm8y6zkt7ijTkeFXcUvcxVvL3Nl5diClzNoH5saq7zm9Oqy4mzFm595vSt9O0WRzWYUh+2uD5TJMWKrAl02KZrflCdHw5lMih9v9qXEn585jcqEdMwTIjcgEd9cnltZb7x7CXZvhPb6lZvbMLnkdTvyjBLu5UEzW3C9crlXcvUisE2rSxidg+CqkB+X8KFsnwa7nzPzMRqeDY8ugJN0nQTVYJjzAY/9bytSxYDd8TdQ+Y7twrYKUvXt/Uv6/igXhrg14s05kC2nZEbr+0bkFu6Groslcvare3VAUsWHXvCU7dkWCU80OyeziTaLIWCisvFBIfFEJoNs/KmXtiX+SYdiYaa3risHn1g9KALQKcOzvxQVlqRb/3KO4i0vaWdB9UsqVuVoAdb4F1duX95hWsTJQtUkmss3vBG9AJrZdJ6Rv5glCZIEUkPyocP8qNkn/rXaH9A5zSXmtihZ5nJWXC/K2D48v3zn3LShvmfv5uooZcddmeDMiPSDSodpdc3t8MSuaxjHN4iruflx7FXpOwtVXJeA1HRLKUk/JcDM5D4dr15HnCFdLCXvhGQlkdmG8NzjGEHAcIRr3XxivPLJDEmjXPrD07/MexL0V8PovwNlvlPzF5igNs3KyhGWVe3fvoG6P7KQIVwvZV/dJyVm2HnpNuPa6BLJbwS7IoKF6u5xHXUfHy9jj/yDMgIkGsldekHJ6zKFnqE1KWuPJVSnjEuHeCXjZIbHH3vKB+bsU16yWs/cUAnFxJ1E1KIxCqk8Czs4fh83vE/ulPnP258in5MLaeFCoTKoOO39CfvfyC3L7RAmj60p/zniPvFayQ8jwdmHm11jFvHHvBLx0vxh2BufhEp/qFTOBVdx72PwUJNaA8d7xQUEgIcGo5XH5OTdy8++5rkxdg2XicjJ8Da69JkMPRZVzauO7pYc3hmSH8P2aDonSJ5eEs9+c3ml7FYvCvRHwuk8J5WCsMTzn3zsx+cRcxb0DTRfzTVWD0S4ZJIDQSvovwLaPCEE5MyS3gagmskPC3aveKuqMKy+KjXwxJ9PYkY7JFcZYcAtMkD6qugTFSM3te7/3CO7+gNd5TAKXMs+36rrihFu1qlW8ZxEsF1fhVI/sfQUxFbByQkTvOi7Dr0IKtTAi5WfjQYjVCZXk2iuw42OSIZ74J6jaKuL/MTgOnPqa7JGN1Mg5Z35XAujqkGxZcHcHvNFuKSc2PzX/vp2iwL7PrvKe7mVUbRXn44pNctG0LTkv1j00viejMApNhymE6iTY+cKlKa4Le35SskBVl99p3D/+3GODsy0/JtJGkKGFnV8tY5cRd3fA07yyl9M3g+HnbLj0LBTTS39Mq7hzoOkyRPCFRGUztuOifD3EGsRooPcs6D6sSIMQ04euCrWkkBZVxvF/kL7d1Iyt7Udi5jnWUx66Kv+NDS1WsSy4ewPehR/IvoKFeIUV0tB7DryrV9p7HmXrJBA1HZb9Eq4rt/eelds3PjF+20gnmN8R6okvAie/IhvQIlOWPfWckf7emFTRyouKY8ePr6p5lhl3b8ArjC68HO05I2XG6pLtVYAYCvSfB/NpKXFBluUMtcmS9jPfEF6ePy7DjFCFlLXRupsJw44tmd22D0twyw7Dkb8VgrtnmfbaruIG7t5v9GJKg0Bi1ShgFeOo3wfX3xRCcvubcpumi3loyQggfvmbUlG7VcF6AAAMn0lEQVQEy+B6K8TXQPPDk8/B/KgMPHZ/SnqAY67F69+5ulrxNuHu1dIudJFybkRKkNXG8SrGEKuXi2C4WiqH7JD83HBAqoBoLSn9MuDCmX8GX+zmi61dlODWeGjcuHOkQ+grlcZtf0v3Ku7egLdQtL8hV+n6PW/3kaxiJWHz+2VKu/vT4xdTVRV6Sn4UK1AhSh5vWAw7JwY815XeXmLteLDrOy/Dj1X6yW3F3VvSLgSOLf2aVe7dKqbCdaH1byVTO/+9cSJyfA1cP4JSzMCa+0SFMTW7cywIVcnCHRD3lWuvjg87VnHbsBrwJmKkU6Zlq83jVUyFosiFsPOo9NvGiMi+MCTWEOw/Jn25qcGu/4LwQdcclvuG22UnxbYPz74wahXLgtWANxHxRtnevopVTIe6XeKlWLNjnIgMsOl9pGvvu/nxoz1w8QeT+8GeIGz70MK4oatYNFYD3hjyKSGMrpI+VzETvCHY82nJ5DY8Jt54MP05Y+VlgLHpfeK+YuWFvuKPrS7geRuxGvDG0HtGrLdXsYrZoAcka6vYKE4o0/XhXFdUPts+LFXDmLFnqGq1jH2bsRrwQE7Q7hOrvneruDU0XXpyyevSzxsjIo/BdeHc06KkCFXIbZ1HZbLbdOj2H+8qJmE14IGcpE2H5+eVt4p7Fw37oeOI/H+MiDyG9jfE0Xhsz3ExJyT2jauuxSsBqwEPhABaPk+vvFXcuyhvgWi9UFI0z/jS7kJGjDzH3HmG22XbmKKulrIrBKsBz7akmcz/397dx0hVnXEc/8K6iqKivDUovtAiD8jGiljtC8iGaI22kdRqtBiNb21TWmsLjVVjAW1qKRVs01hjrQFr02h8gZpSSVOrFqPVZqkWdP0VCgRUUIFd7PK+7PSPcwYuw+yyszu7dzL3+SQbZs4c7j33mZnnnnPP3Hv9N1Guk/r2hZPHh3vL1l0OfWrC9fBqauGc68LPmnY2wTt/6tqlyVyP8Xdiy6pwKe+unormsmnfHlj5VDgftmEBA9YtDRcKzQ9bt64NM7k+I1tRvJ/98Xs+WeFKV3t0vIvZ2zD8XHa0rg/XyGtrC2frnDTOj9lVoFQTnpmNBB4BaoFHJT3UTr27gKsl1RV7vVtGXlj2RbqMGH5uuGdF3eXs3hbPzlnzQpi0GDI63ba5otIe0s4BpgMXADea2SHTpGY2COiZk1s3vB5O/nauK/oPhrFfOfBbvI1vhtMTfUa2YqWd8M6Q1CCpFXgROK9InduB+8q+5lwufEC7ckVk5/K2bw53HQM4emCYxPAZ2YqV9juT3A02AQf18MxsODBM0nKz0q4Z1tjY2OHrR+z4kP6bm9m2bhOwqaRlV6tdu3YdNm6uQC7Hiave4Kja9TRyIdCSdotcB3o84ZnZMGBRkZfuBdoSz08A3iioMzPWK9mYMWM6rrB7OIwazUmF9xvIsMbGxsPHzR1q6DWsXrvWY9cFDQ0Nvbq+Hk94kjYCRc+pMbNrzewc4E1gEjC3oMoIYH7s3Z1mZnMl3dbtRu3bCx9vhCGjur0o5xg8kr0f7U27Fa4T0h7S3sGBWdqFkrYCmNkCSTdIuihf0cxWliXZQbghS9M6T3jOZUyqCU/SakLPrrD8hiJl5ftJyqYV4eq0zrlMSXuWtvfta4U+NeF+oc65TEl7SNv7+tbAWVem3QrnXAqy1cPL5aBhYbhkj3Muc7KV8La9G26W7DfpcS6TspXwNq3wCwU4l2HZOoY3/Nxwx3jnXCZlp4fXvD78W1Obbjucc6nJTsJb93K4aKNzLrOykfB2Nof7zh5/ctotcc6lKCMJrykcv/NrlDmXadU/aZHLwQmnwsARabfEOZey6u/hNa8/cIFG51ymVX/C27QCBpd28VDnXHWq7oTXuidcBmqwXwbKOVftCa+mFs6e6vcYcM4B1Z7wNrwORx2XdiuccxWiehPejq2w6d/Q13t3zrmgehPeByvDhQL8t3fOuah6Ex594BNj026Ec66CVO947/QJ3rtzzh2kent4W1an3QLnXIWp3oR3op9K5pw7WPUmPP/tnXOuQPUmPOecK+AJzzmXGZ7wnHOZ4QnPOZcZnvCcc5nhCc85lxme8JxzmeEJzzmXGZ7wnHOZ4QnPOZcZnvCcc5nRJ5fLpd2GsmtoaKi+jXKuSo0fP77XruNWlQnPOeeK8SGtcy4zPOE55zLDE55zLjM84TnnMsMTnnMuMzzhOecyo9du/GBmRwEvAGcCN0t6KpYfAywEhgFvA9+S1GZm5wP3A32An0p6Nta/B5gM/A+4TtJHBeuZJunXnWzT6UADsCIW3SHpVTMbAvwOOA74m6SZsf5lwB1ADvi+pNe6EovuMLPtwD/j019KWlRqDBPLqgc+lPR2J9bbF/gzcCzhczNb0tJSY5gWMxsDPAy0AfuAmyStMbPrgR8BG2LVSyTtLGfsYv0HgSnAM5K+E8suAH4G7AF2ANdI2mpms4ErgY+AbZKmdHnDy6iD7/AtwLXAXmC5pFti+YtAbSx/SdKsguXVU1oMhwIPAIOADyR9LfHaXcDVkuo6WkZv9vD2Al8FflFQfiPwmqSJsc4lsXw+cAVwETDbzI4wszpgnKQJwCPAbUXWM63Edr0qqT7+vRrLfgj8Nq5nnJnVmdkRwN2xPVfE9qVhbaK9i2JZp2NYsKx6woe3M3LAt2NMvgTMTbzWqRiWsI09YTPwZUn5JHNn4rUHE+3fGcvKGTuAHwPXFJStASZLmgT8kYM/u7Nieyoi2UXtfYefA86X9AVgiJlNSLw2JW7HLA5VT2kxnAfcLmlyQbIb1Nnl9FoPT1IbsNHMCl+aCOSDsQSYZGbPAzWS3gcws1XAyFh3SaLujOSCzOwqYETcszwL/AF4FOhH2FteJ2lHwfo/Y2bLgLeAGZK2AxMIe30IvZoJhF6BJLUALWZWY2b9JO3qSjy64RQzewl4D7g19nBLieE78fkA4Hpgu5l9Q9IXzewnhA9hH0Is8skLSTngv/HpbkICzOtsDFeWLQolKhgJtBLez7ybzWwKsFjSPDPrRxljF9f/vpmNKih7t4M23WVmtwIPSfp9V7e7nNr7DktK3gQ6uR05YJGZ7ST0/BvylUqNYdzhjCbsfE4BHpD0ZHz5duA+woiiQ5VwDO9EoDk+bgIGxr/mRJ18+f66cU/cP7kgSU9woAc0nzD8/E3cg74OfL1g3RuBT8We0RpCrwTgmMSe/pB1R82xvLd9Mm7PYsIeD0qLIQCSthGGwbPih20ccFbcS1/FoXvxpHmE4R6UFsPUxeH/PRzYvsXAWMJhkolmNpmejV2xNg0l9O4ejkW/knQ2oac+zczOKGkjUxKH6EMTieqK+LmYBiw0s/2nkHUhhkOAs4F7CSOMO81soJkNB4ZJWt6ZNlZCwmsCBsTHJwBb49+ARJ18+f66cS+8/TDLHgX8Iz5+BTho1yRpd+yxATwOjI+Pd8TlF113QXmvkrQlPnwS+HR8XEoM27M/VpI2AMcUq2RmdwItkhbGuqXEMFWxl/A4MFdSI4CkZkn7JO0FniG0v0di106bjgOeIhx33RyXsSX+2wIsJXzRK5qZjSUcKpiaL0tsx2rCCGtwB4s4XAybgHclNcYRRAOh1z2TkAQ7pRIS3t8JGRvgUmBZHCbuM7NhZtafEIzVse6libovF1leW+LxKuCz8fHngf8kK5rZ8Ymn9XEdAMsK1rMsLsvMrL+ZDQNae3s4G9ddE59OBNbFx6XEMGkPBw5r7I9VHDIUDv0xs5uAOuAHibJSYpia2Lt4BHhO0uJEeTKxTQJW90Ts2mnTkYQd18+VmADLtym+1xOKrLuimNmphN7a1HzSjuXHx38HAScDWwr+a6djGN+T9fE9qSF8DtcBI4D5ZrYUOM3MkseWD9Frx/AAzOxpYBzhGNjnJM0AFhC6u8sIM4zPxeozgKcJSXm2pFbgLTN708xeBloIM0OFlpjZEsJB4DnAo2b2XUKwC+tfYGZ3x2W1ADfE8rnAY2Y2nTDD+FZs/2zgr8RZ2u5Fo0tGAw+bWQvhWMk3Y3kpMUx6HphnZpdJmmpmK83sFcIxlOnJimZ2LPAQ4dDAC2aGpHpKjGGKLibMfJ4Wj/W+Iel7wAwzu5hw3Gk5YYgLZYwdgJnNJMzSDjWz0ZIuJEw2nQf0M7MZwFJJc+JyzwRqgGcl/aucgeiOdr7Dcwkzpwvi8b05wF8In5OdhJna6fEYYFJJMYxlTwBHAo9J+pAwqZRv20pJxSYy9/OrpTjnMqMShrTOOdcrPOE55zLDE55zLjM84TnnMsMTnnMuMzzhOecywxOecy4zPOE55zLj/yw8g9Qy7m/OAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 324x309.6 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "weights_categories_list = [({\"Participants\": 1/5, \"Intervention\": 1/4, \"Outcomes\": 1/3, \"Conclusion\": 1},\n",
    "                            {\"Participants\": 1, \"Intervention\": 1/4, \"Outcomes\": 1/5})]\n",
    "\n",
    "fig, axs = plt.subplots(1, 1, figsize=(4.5, 4.3))\n",
    "lines_to_plot = [(0.3, 0.15, \"band\")] + \\\n",
    "                [(alpha, beta, None) for alpha, beta in\n",
    "                 list(product(np.arange(0.3, 0.7, 0.1), np.arange(0.05, 0.15, 0.05)))]\n",
    "\n",
    "for alpha, beta, err_style in lines_to_plot:\n",
    "    for weights_categories, weights_categories_tag in weights_categories_list:\n",
    "        df_tag_1 = hopper.get_tagging_dataframe(tagging, tagging_root, norm_enabled=False)\n",
    "        df_tag_1[\"control\"] = 0\n",
    "        df_df_tag_1tag_2 = hopper_control.get_tagging_dataframe(tagging, tagging_root, norm_enabled=False)\n",
    "        df_tag_2[\"control\"] = 1\n",
    "        df_tag = pd.concat([df_tag_1, df_tag_2])\n",
    "        df_tag[\"val_tokens\"] = df_tag.val.apply(lambda x: x[0])\n",
    "        df_tag[\"val_doggos\"] = df_tag.val.apply(lambda x: x[1])\n",
    "        df_tag[\"node_aux\"] = df_tag[\"node\"]\n",
    "        df_tag.set_index([\"node\"], inplace=True)\n",
    "        count = 0\n",
    "        tmp_df = []\n",
    "        for i in range(1, 5):\n",
    "            tmp = df_tag[(df_tag.hop == i) & (df_tag.control == 0)]\n",
    "            for branch in branch_to_topic.keys():\n",
    "                flag = []\n",
    "\n",
    "                for idx, idy in combinations(range(8), 2):\n",
    "                    if idx or idy in flag:\n",
    "                        continue\n",
    "                    extractive = dict(tmp[tmp.branch == branch].sort_values(\"val_doggos\",\n",
    "                                                                            ascending=False).iloc[idx])\n",
    "                    abstractive = dict(tmp[tmp.branch == branch].sort_values(\"val_doggos\",\n",
    "                                                                             ascending=False).iloc[idy])\n",
    "                    if extractive[\"val_tokens\"] < abstractive[\"val_tokens\"]:\n",
    "                        tmp_ = extractive\n",
    "                        extractive = abstractive\n",
    "                        abstractive = tmp_\n",
    "                    if extractive[\"val_tokens\"] - abstractive[\"val_tokens\"] < alpha* extractive[\"val_tokens\"] \\\n",
    "                            or extractive[\"val_doggos\"] - abstractive[\"val_doggos\"] > \\\n",
    "                            max(beta * extractive[\"val_doggos\"], beta * abstractive[\"val_doggos\"]):\n",
    "                        continue\n",
    "                    succ_extractive = dict(df_tag.loc[list(hopper.g.successors(extractive[\"node_aux\"]))[0]])\n",
    "                    succ_abstractive = dict(df_tag.loc[list(hopper.g.successors(abstractive[\"node_aux\"]))[0]])\n",
    "                    tmp_df.append({\"hop\": i, \"Type\": \"Extractive\",\n",
    "                                   \"$\\Delta$ Fact score\": (succ_extractive[\"val_doggos\"] -\n",
    "                                                           extractive[\"val_doggos\"])})\n",
    "                    tmp_df.append({\"hop\": i, \"Type\": \"Abstractive\",\n",
    "                                   \"$\\Delta$ Fact score\": (succ_abstractive[\"val_doggos\"] -\n",
    "                                                           abstractive[\"val_doggos\"])})\n",
    "                    flag += [idx, idy]\n",
    "\n",
    "        df_ana = pd.DataFrame(tmp_df)\n",
    "\n",
    "        if err_style == \"band\":\n",
    "            sns.lineplot(x=\"hop\", y=\"$\\Delta$ Fact score\", hue=\"Type\", data=df_ana, ax=axs,\n",
    "                         color=[\"#8E44AD\", \"#D68910\"], lw=2,\n",
    "                         err_style=err_style)\n",
    "        else:\n",
    "            ax = sns.lineplot(x=\"hop\", y=\"$\\Delta$ Fact score\", hue=\"Type\", data=df_ana, ax=axs,\n",
    "                              color=[\"#8E44AD\", \"#D68910\"], lw=0.9,\n",
    "                              err_style=None, legend=False, alpha=0.5)\n",
    "            ax.lines[-1].set_linestyle(\"--\")\n",
    "            ax.lines[-2].set_linestyle(\"--\")            \n",
    "\n",
    "    plt.sca(axs)\n",
    "    plt.legend(loc='upper center', bbox_to_anchor=(0.5, 1.2), fancybox=False, shadow=False, ncol=3)\n",
    "    plt.xticks([1, 2, 3, 4], [\"1000 to 500\", \"500 to 250\", \"250 to 125\", \"125 to 64\"])\n",
    "    plt.xlim([1, 4])\n",
    "    plt.xlabel(\"\")\n",
    "    fig.tight_layout()\n",
    "    plt.savefig(\"./images/summarization_strategies2.pdf\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "telephone_game",
   "language": "python",
   "name": "telephone_game"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}