analyze_simple_facts.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "9984390d",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "from matplotlib import pyplot as plt\n",
    "from utils.task_config import task_config\n",
    "from analysis_utils.load_scores import *\n",
    "from utils.load_task import load_test_split, format_data"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc478aed",
   "metadata": {},
   "source": [
    "## accuracies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "00fd2049",
   "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>arithmetics_addition</th>\n",
       "      <th>elements_number_from_element</th>\n",
       "      <th>elements_number_from_position</th>\n",
       "      <th>olympics_hundred_meters</th>\n",
       "      <th>olympics_downhill</th>\n",
       "      <th>writers_english</th>\n",
       "      <th>writers_german</th>\n",
       "      <th>writers_italian</th>\n",
       "      <th>writers_dutch</th>\n",
       "      <th>writers_swedish</th>\n",
       "      <th>companies_us</th>\n",
       "      <th>companies_de</th>\n",
       "      <th>companies_it</th>\n",
       "      <th>companies_nl</th>\n",
       "      <th>companies_sv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>en</th>\n",
       "      <td>99.4</td>\n",
       "      <td>100.00</td>\n",
       "      <td>37.78</td>\n",
       "      <td>55.41</td>\n",
       "      <td>37.61</td>\n",
       "      <td>90.32</td>\n",
       "      <td>78.49</td>\n",
       "      <td>67.74</td>\n",
       "      <td>76.88</td>\n",
       "      <td>67.74</td>\n",
       "      <td>89.0</td>\n",
       "      <td>86.0</td>\n",
       "      <td>61.0</td>\n",
       "      <td>75.0</td>\n",
       "      <td>80.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>en-paraphrase</th>\n",
       "      <td>98.6</td>\n",
       "      <td>100.00</td>\n",
       "      <td>42.22</td>\n",
       "      <td>54.05</td>\n",
       "      <td>31.62</td>\n",
       "      <td>90.32</td>\n",
       "      <td>79.57</td>\n",
       "      <td>67.20</td>\n",
       "      <td>75.27</td>\n",
       "      <td>68.82</td>\n",
       "      <td>87.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>62.0</td>\n",
       "      <td>65.0</td>\n",
       "      <td>82.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>de_from_en</th>\n",
       "      <td>45.2</td>\n",
       "      <td>98.89</td>\n",
       "      <td>40.00</td>\n",
       "      <td>50.68</td>\n",
       "      <td>35.04</td>\n",
       "      <td>89.78</td>\n",
       "      <td>80.65</td>\n",
       "      <td>67.20</td>\n",
       "      <td>79.57</td>\n",
       "      <td>66.67</td>\n",
       "      <td>84.0</td>\n",
       "      <td>85.0</td>\n",
       "      <td>60.0</td>\n",
       "      <td>71.0</td>\n",
       "      <td>77.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>it_from_en</th>\n",
       "      <td>44.0</td>\n",
       "      <td>100.00</td>\n",
       "      <td>36.67</td>\n",
       "      <td>51.35</td>\n",
       "      <td>35.04</td>\n",
       "      <td>90.32</td>\n",
       "      <td>79.03</td>\n",
       "      <td>65.59</td>\n",
       "      <td>76.88</td>\n",
       "      <td>64.52</td>\n",
       "      <td>80.0</td>\n",
       "      <td>79.0</td>\n",
       "      <td>64.0</td>\n",
       "      <td>69.0</td>\n",
       "      <td>76.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nl_from_en</th>\n",
       "      <td>42.4</td>\n",
       "      <td>100.00</td>\n",
       "      <td>36.67</td>\n",
       "      <td>52.03</td>\n",
       "      <td>35.04</td>\n",
       "      <td>90.86</td>\n",
       "      <td>78.49</td>\n",
       "      <td>68.28</td>\n",
       "      <td>81.18</td>\n",
       "      <td>65.05</td>\n",
       "      <td>82.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>59.0</td>\n",
       "      <td>71.0</td>\n",
       "      <td>76.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sv_from_en</th>\n",
       "      <td>19.2</td>\n",
       "      <td>100.00</td>\n",
       "      <td>41.11</td>\n",
       "      <td>50.68</td>\n",
       "      <td>33.33</td>\n",
       "      <td>86.02</td>\n",
       "      <td>77.96</td>\n",
       "      <td>62.90</td>\n",
       "      <td>75.27</td>\n",
       "      <td>69.35</td>\n",
       "      <td>83.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>56.0</td>\n",
       "      <td>65.0</td>\n",
       "      <td>77.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               arithmetics_addition  elements_number_from_element  \\\n",
       "en                             99.4                        100.00   \n",
       "en-paraphrase                  98.6                        100.00   \n",
       "de_from_en                     45.2                         98.89   \n",
       "it_from_en                     44.0                        100.00   \n",
       "nl_from_en                     42.4                        100.00   \n",
       "sv_from_en                     19.2                        100.00   \n",
       "\n",
       "               elements_number_from_position  olympics_hundred_meters  \\\n",
       "en                                     37.78                    55.41   \n",
       "en-paraphrase                          42.22                    54.05   \n",
       "de_from_en                             40.00                    50.68   \n",
       "it_from_en                             36.67                    51.35   \n",
       "nl_from_en                             36.67                    52.03   \n",
       "sv_from_en                             41.11                    50.68   \n",
       "\n",
       "               olympics_downhill  writers_english  writers_german  \\\n",
       "en                         37.61            90.32           78.49   \n",
       "en-paraphrase              31.62            90.32           79.57   \n",
       "de_from_en                 35.04            89.78           80.65   \n",
       "it_from_en                 35.04            90.32           79.03   \n",
       "nl_from_en                 35.04            90.86           78.49   \n",
       "sv_from_en                 33.33            86.02           77.96   \n",
       "\n",
       "               writers_italian  writers_dutch  writers_swedish  companies_us  \\\n",
       "en                       67.74          76.88            67.74          89.0   \n",
       "en-paraphrase            67.20          75.27            68.82          87.0   \n",
       "de_from_en               67.20          79.57            66.67          84.0   \n",
       "it_from_en               65.59          76.88            64.52          80.0   \n",
       "nl_from_en               68.28          81.18            65.05          82.0   \n",
       "sv_from_en               62.90          75.27            69.35          83.0   \n",
       "\n",
       "               companies_de  companies_it  companies_nl  companies_sv  \n",
       "en                     86.0          61.0          75.0          80.0  \n",
       "en-paraphrase          84.0          62.0          65.0          82.0  \n",
       "de_from_en             85.0          60.0          71.0          77.0  \n",
       "it_from_en             79.0          64.0          69.0          76.0  \n",
       "nl_from_en             78.0          59.0          71.0          76.0  \n",
       "sv_from_en             78.0          56.0          65.0          77.0  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "all_scores, em = accuracy_facts()\n",
    "em_df = pd.DataFrame(em)\n",
    "display(em_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c94e6a7f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "en               73.491333\n",
       "en-paraphrase    72.511333\n",
       "de_from_en       68.712000\n",
       "it_from_en       67.426667\n",
       "nl_from_en       67.733333\n",
       "sv_from_en       64.988000\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(em_df.aggregate(\"mean\", axis=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "4337e00b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "en               76.234\n",
       "en-paraphrase    76.236\n",
       "de_from_en       76.774\n",
       "it_from_en       75.268\n",
       "nl_from_en       76.772\n",
       "sv_from_en       74.300\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "en               78.2\n",
       "en-paraphrase    76.0\n",
       "de_from_en       75.4\n",
       "it_from_en       73.6\n",
       "nl_from_en       73.2\n",
       "sv_from_en       71.8\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "writers_df = em_df[[\"writers_english\", \"writers_german\", \"writers_italian\", \"writers_dutch\", \"writers_swedish\"]]\n",
    "companies_df = em_df[[\"companies_us\", \"companies_de\", \"companies_it\", \"companies_nl\", \"companies_sv\"]]\n",
    "display(writers_df.aggregate(\"mean\", axis=1), companies_df.aggregate(\"mean\", axis=1))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c735cc79",
   "metadata": {},
   "source": [
    "#### accuracies en 2nd run"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "a74e230c",
   "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>arithmetics_addition</th>\n",
       "      <th>elements_number_from_element</th>\n",
       "      <th>elements_number_from_position</th>\n",
       "      <th>olympics_hundred_meters</th>\n",
       "      <th>olympics_downhill</th>\n",
       "      <th>writers_english</th>\n",
       "      <th>writers_german</th>\n",
       "      <th>writers_italian</th>\n",
       "      <th>writers_dutch</th>\n",
       "      <th>writers_swedish</th>\n",
       "      <th>companies_us</th>\n",
       "      <th>companies_de</th>\n",
       "      <th>companies_it</th>\n",
       "      <th>companies_nl</th>\n",
       "      <th>companies_sv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>en_2nd_run</th>\n",
       "      <td>99.4</td>\n",
       "      <td>100.0</td>\n",
       "      <td>37.78</td>\n",
       "      <td>54.73</td>\n",
       "      <td>34.19</td>\n",
       "      <td>90.32</td>\n",
       "      <td>78.49</td>\n",
       "      <td>67.2</td>\n",
       "      <td>75.27</td>\n",
       "      <td>66.67</td>\n",
       "      <td>88.0</td>\n",
       "      <td>87.0</td>\n",
       "      <td>62.0</td>\n",
       "      <td>73.0</td>\n",
       "      <td>80.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            arithmetics_addition  elements_number_from_element  \\\n",
       "en_2nd_run                  99.4                         100.0   \n",
       "\n",
       "            elements_number_from_position  olympics_hundred_meters  \\\n",
       "en_2nd_run                          37.78                    54.73   \n",
       "\n",
       "            olympics_downhill  writers_english  writers_german  \\\n",
       "en_2nd_run              34.19            90.32           78.49   \n",
       "\n",
       "            writers_italian  writers_dutch  writers_swedish  companies_us  \\\n",
       "en_2nd_run             67.2          75.27            66.67          88.0   \n",
       "\n",
       "            companies_de  companies_it  companies_nl  companies_sv  \n",
       "en_2nd_run          87.0          62.0          73.0          80.0  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "_, em_en_2nd = accuracy_facts(languages=[\"en_2nd_run\"])\n",
    "em_en_2nd_df = pd.DataFrame(em_en_2nd)\n",
    "display(em_en_2nd_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d0f87c63",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "en_2nd_run    75.59\n",
      "dtype: float64 \n",
      " en_2nd_run    78.0\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "writers_2nd_run_df = em_en_2nd_df[[\"writers_english\", \"writers_german\", \"writers_italian\", \"writers_dutch\", \"writers_swedish\"]]\n",
    "companies_2nd_run_df = em_en_2nd_df[[\"companies_us\", \"companies_de\", \"companies_it\", \"companies_nl\", \"companies_sv\"]]\n",
    "print(writers_2nd_run_df.aggregate(\"mean\", axis=1), \"\\n\", companies_2nd_run_df.aggregate(\"mean\", axis=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "34b739dc",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot\n",
    "\n",
    "acc_df = make_aggregate_df_from_table(em, score=\"accuracy\")\n",
    "\n",
    "sns.set(font_scale=1.2)\n",
    "sns.set_style(\"whitegrid\")\n",
    "plt.figure(figsize=(10,4))\n",
    "p = sns.color_palette('mako') \n",
    "ax = sns.barplot(data=acc_df, y=\"accuracy\", x=\"task\", hue=\"language\", palette=p) #\"muted\")#, alpha=0.5, size=10, marker=\"o\")\n",
    "plt.xticks(ticks=np.array([0, 1, 2, 3, 4, 5, 6]), \n",
    "           labels=[\"arithmetics\\naddition\", \"elements\\nfrom-elem\", \"elements\\nfrom-pos\", \n",
    "                   \"olympics\\n100m\", \"olympics\\ndownhill\", \"writers\\naverage\", \"companies\\naverage\"], \n",
    "           rotation=30)\n",
    "\n",
    "sns.despine(left=True, right=True, top=True, bottom=False)\n",
    "\n",
    "# Update the hatches\n",
    "for i, bars in enumerate(ax.containers):\n",
    "    for bar in bars:\n",
    "        if i%6 == 0:\n",
    "            bar.set_hatch('////')\n",
    "            bar.set_edgecolor('b')\n",
    "        elif i%6 == 1:\n",
    "            bar.set_hatch(5*'x')\n",
    "            bar.set_edgecolor('b')\n",
    "            \n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "labels = [\"$en$\", \"$en^P$\", \"$de^T$\", \"$it^T$\", \"$nl^T$\", \"$sv^T$\"]\n",
    "\n",
    "legend = ax.legend(handles=handles, labels=labels, title=\"sense\", ncol=3, \n",
    "                   loc='upper right', bbox_to_anchor=(1,1.15), fontsize=12)\n",
    "plt.setp(legend.get_title(), fontsize=14)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bae3583f",
   "metadata": {},
   "source": [
    "#### differences containment versus em"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "b5761564",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 300x200 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"white\")\n",
    "differences = []\n",
    "for c, t in collection_task_all[\"facts\"]:\n",
    "    for language in [\"en\", \"en-paraphrase\", \"de_from_en\", \"it_from_en\", \"nl_from_en\", \"sv_from_en\"]:\n",
    "        delta = all_scores[c + \"_\" + t][language][\"contained\"] - all_scores[c + \"_\" + t][language][\"exact_match\"]\n",
    "        differences.append(delta)\n",
    "        \n",
    "counts, bins = np.histogram(differences, bins=np.arange(-0.5, 8, 1))\n",
    "plt.figure(figsize=(3,2))\n",
    "plt.stairs(counts, bins, fill=False, color=\"k\")\n",
    "plt.ylabel(\"counts\")\n",
    "plt.xlabel(\"contained - exact match (%)\")\n",
    "plt.xticks(ticks=[0, 2, 4, 6, 8])\n",
    "plt.yticks(ticks=[0, 20, 40, 60])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "51254fdd",
   "metadata": {},
   "source": [
    "## consistency"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "967ff503",
   "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>arithmetics_addition</th>\n",
       "      <th>elements_number_from_element</th>\n",
       "      <th>elements_number_from_position</th>\n",
       "      <th>olympics_hundred_meters</th>\n",
       "      <th>olympics_downhill</th>\n",
       "      <th>writers_english</th>\n",
       "      <th>writers_german</th>\n",
       "      <th>writers_italian</th>\n",
       "      <th>writers_dutch</th>\n",
       "      <th>writers_swedish</th>\n",
       "      <th>companies_us</th>\n",
       "      <th>companies_de</th>\n",
       "      <th>companies_it</th>\n",
       "      <th>companies_nl</th>\n",
       "      <th>companies_sv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>en_2nd_run</th>\n",
       "      <td>100.0</td>\n",
       "      <td>100.00</td>\n",
       "      <td>91.11</td>\n",
       "      <td>90.54</td>\n",
       "      <td>88.89</td>\n",
       "      <td>98.39</td>\n",
       "      <td>90.32</td>\n",
       "      <td>81.18</td>\n",
       "      <td>82.80</td>\n",
       "      <td>82.80</td>\n",
       "      <td>98.0</td>\n",
       "      <td>98.0</td>\n",
       "      <td>97.0</td>\n",
       "      <td>95.0</td>\n",
       "      <td>99.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>en-paraphrase</th>\n",
       "      <td>99.0</td>\n",
       "      <td>100.00</td>\n",
       "      <td>77.78</td>\n",
       "      <td>82.43</td>\n",
       "      <td>61.54</td>\n",
       "      <td>95.16</td>\n",
       "      <td>88.17</td>\n",
       "      <td>76.34</td>\n",
       "      <td>78.49</td>\n",
       "      <td>75.27</td>\n",
       "      <td>97.0</td>\n",
       "      <td>95.0</td>\n",
       "      <td>87.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>93.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>de_from_en</th>\n",
       "      <td>45.0</td>\n",
       "      <td>98.89</td>\n",
       "      <td>71.11</td>\n",
       "      <td>80.41</td>\n",
       "      <td>76.92</td>\n",
       "      <td>94.09</td>\n",
       "      <td>85.48</td>\n",
       "      <td>72.04</td>\n",
       "      <td>81.72</td>\n",
       "      <td>72.58</td>\n",
       "      <td>94.0</td>\n",
       "      <td>91.0</td>\n",
       "      <td>82.0</td>\n",
       "      <td>88.0</td>\n",
       "      <td>92.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>it_from_en</th>\n",
       "      <td>44.2</td>\n",
       "      <td>100.00</td>\n",
       "      <td>70.00</td>\n",
       "      <td>75.68</td>\n",
       "      <td>70.09</td>\n",
       "      <td>96.24</td>\n",
       "      <td>87.63</td>\n",
       "      <td>70.43</td>\n",
       "      <td>78.49</td>\n",
       "      <td>73.12</td>\n",
       "      <td>88.0</td>\n",
       "      <td>87.0</td>\n",
       "      <td>84.0</td>\n",
       "      <td>82.0</td>\n",
       "      <td>91.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nl_from_en</th>\n",
       "      <td>42.2</td>\n",
       "      <td>100.00</td>\n",
       "      <td>67.78</td>\n",
       "      <td>83.78</td>\n",
       "      <td>64.10</td>\n",
       "      <td>93.55</td>\n",
       "      <td>82.26</td>\n",
       "      <td>72.58</td>\n",
       "      <td>77.42</td>\n",
       "      <td>69.89</td>\n",
       "      <td>92.0</td>\n",
       "      <td>86.0</td>\n",
       "      <td>88.0</td>\n",
       "      <td>85.0</td>\n",
       "      <td>93.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sv_from_en</th>\n",
       "      <td>19.4</td>\n",
       "      <td>100.00</td>\n",
       "      <td>70.00</td>\n",
       "      <td>83.78</td>\n",
       "      <td>59.83</td>\n",
       "      <td>90.32</td>\n",
       "      <td>83.33</td>\n",
       "      <td>66.13</td>\n",
       "      <td>75.27</td>\n",
       "      <td>71.51</td>\n",
       "      <td>89.0</td>\n",
       "      <td>86.0</td>\n",
       "      <td>80.0</td>\n",
       "      <td>78.0</td>\n",
       "      <td>91.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               arithmetics_addition  elements_number_from_element  \\\n",
       "en_2nd_run                    100.0                        100.00   \n",
       "en-paraphrase                  99.0                        100.00   \n",
       "de_from_en                     45.0                         98.89   \n",
       "it_from_en                     44.2                        100.00   \n",
       "nl_from_en                     42.2                        100.00   \n",
       "sv_from_en                     19.4                        100.00   \n",
       "\n",
       "               elements_number_from_position  olympics_hundred_meters  \\\n",
       "en_2nd_run                             91.11                    90.54   \n",
       "en-paraphrase                          77.78                    82.43   \n",
       "de_from_en                             71.11                    80.41   \n",
       "it_from_en                             70.00                    75.68   \n",
       "nl_from_en                             67.78                    83.78   \n",
       "sv_from_en                             70.00                    83.78   \n",
       "\n",
       "               olympics_downhill  writers_english  writers_german  \\\n",
       "en_2nd_run                 88.89            98.39           90.32   \n",
       "en-paraphrase              61.54            95.16           88.17   \n",
       "de_from_en                 76.92            94.09           85.48   \n",
       "it_from_en                 70.09            96.24           87.63   \n",
       "nl_from_en                 64.10            93.55           82.26   \n",
       "sv_from_en                 59.83            90.32           83.33   \n",
       "\n",
       "               writers_italian  writers_dutch  writers_swedish  companies_us  \\\n",
       "en_2nd_run               81.18          82.80            82.80          98.0   \n",
       "en-paraphrase            76.34          78.49            75.27          97.0   \n",
       "de_from_en               72.04          81.72            72.58          94.0   \n",
       "it_from_en               70.43          78.49            73.12          88.0   \n",
       "nl_from_en               72.58          77.42            69.89          92.0   \n",
       "sv_from_en               66.13          75.27            71.51          89.0   \n",
       "\n",
       "               companies_de  companies_it  companies_nl  companies_sv  \n",
       "en_2nd_run             98.0          97.0          95.0          99.0  \n",
       "en-paraphrase          95.0          87.0          84.0          93.0  \n",
       "de_from_en             91.0          82.0          88.0          92.0  \n",
       "it_from_en             87.0          84.0          82.0          91.0  \n",
       "nl_from_en             86.0          88.0          85.0          93.0  \n",
       "sv_from_en             86.0          80.0          78.0          91.0  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "consistency = consistency_facts()\n",
    "consistency_df = pd.DataFrame(consistency)\n",
    "display(consistency_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b302f95c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "en_2nd_run       92.868667\n",
       "en-paraphrase    86.012000\n",
       "de_from_en       81.682667\n",
       "it_from_en       79.858667\n",
       "nl_from_en       79.837333\n",
       "sv_from_en       76.238000\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(consistency_df.aggregate(\"mean\", axis=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "f623d6ef",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "en_2nd_run       87.098\n",
       "en-paraphrase    82.686\n",
       "de_from_en       81.182\n",
       "it_from_en       81.182\n",
       "nl_from_en       79.140\n",
       "sv_from_en       77.312\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "en_2nd_run       97.4\n",
       "en-paraphrase    91.2\n",
       "de_from_en       89.4\n",
       "it_from_en       86.4\n",
       "nl_from_en       88.8\n",
       "sv_from_en       84.8\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "companies_df = consistency_df[[\"companies_us\", \"companies_de\", \"companies_it\", \"companies_nl\", \"companies_sv\"]]\n",
    "writers_df = consistency_df[[\"writers_english\", \"writers_german\", \"writers_italian\", \"writers_dutch\", \"writers_swedish\"]]\n",
    "display(writers_df.aggregate(\"mean\", axis=1), companies_df.aggregate(\"mean\", axis=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "28979e31",
   "metadata": {},
   "outputs": [],
   "source": [
    "# plot\n",
    "consistency_df = make_aggregate_df_from_table(consistency, score=\"consistency\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "ef63e569",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_tasks = acc_df['task'].unique()[2:]\n",
    "filtered_df = consistency_df[consistency_df['task'].isin(plot_tasks)]\n",
    "\n",
    "sns.set(font_scale=1.2)\n",
    "figsize=(10,4)\n",
    "n_tasks = len(plot_tasks)\n",
    "sns.set_style(\"whitegrid\")\n",
    "plt.figure(figsize=figsize)\n",
    "p = sns.color_palette('rocket_r') \n",
    "ax = sns.barplot(data=filtered_df, y=\"consistency\", x=\"task\", hue=\"language\", palette=p) #\"muted\")#, alpha=0.5, size=10, marker=\"o\")\n",
    "plt.xticks(ticks=range(n_tasks), \n",
    "           labels=[\"elements\\nfrom-pos\", \"olympics\\n100m\", \"olympics\\ndownhill\", \"writers\\naverage\", \n",
    "                   \"companies\\naverage\"], \n",
    "           rotation=30)\n",
    "\n",
    "sns.despine(left=True, right=True, top=True, bottom=False)\n",
    "\n",
    "# Loop over df to add upper- and lower accuracy lines\n",
    "linecolor='white'   # Find correct RGB, beige may be nicer\n",
    "bar_width = ax.patches[0].get_width()/n_tasks\n",
    "\n",
    "# get bar x coordinates\n",
    "bar_coords = []\n",
    "for bar in ax.patches:\n",
    "    bar_coords.append(bar.get_x()/figsize[1])\n",
    "\n",
    "# print(bar_coords)\n",
    "\n",
    "for i, task in enumerate(plot_tasks):\n",
    "    # Get base accuracy\n",
    "    acc_base = acc_df[(acc_df['task']==task) & (acc_df['language'] == 'en')]['accuracy'].iloc[0]\n",
    "    for j, language in enumerate(consistency_df['language'].unique()):\n",
    "        \n",
    "        if language == 'en_2nd_run':\n",
    "            acc2 = {'elements_number_from_position': 37.78, \n",
    "                    'olympics_hundred_meters': 54.05,\n",
    "                    'olympics_downhill': 34.19,\n",
    "                    'writers': 75.59,\n",
    "                    'companies': 78.0}[task]\n",
    "        else: \n",
    "            acc2 = acc_df[(acc_df['task']==task) & (acc_df['language'] == language)]['accuracy'].iloc[0]\n",
    "\n",
    "        # Compute values and plot them\n",
    "        # NB this is an overestimation which assumes that \n",
    "        # Answers that are both wrong are inconsistent\n",
    "        max_consistency = fully_consistent(acc_base, acc2)\n",
    "        xmin = 0.02+i/n_tasks+j*bar_width*1.025\n",
    "        plt.axhline(y=max_consistency, xmin=xmin, xmax=xmin+bar_width*0.85, linewidth=2)\n",
    "\n",
    "# plt.axhline(y=0.2, xmin=-0.1, xmax=xmin+bar_width, linewidth=2)\n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "labels = [\"$en_{id}$\", \"$en^P$\", \"$de^T$\", \"$it^T$\", \"$nl^T$\", \"$sv^T$\"]\n",
    "\n",
    "legend = ax.legend(handles=handles, labels=labels, title=\"sense\", ncol=3, \n",
    "                   loc='upper right', bbox_to_anchor=(1,1.4), fontsize=12)\n",
    "plt.setp(legend.get_title(), fontsize=14)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f87d170c",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "source": [
    "## advantage for language-specific information"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "9aa7e850",
   "metadata": {},
   "outputs": [],
   "source": [
    "em_writers_same = []\n",
    "em_writers_different = []\n",
    "em_companies_same = []\n",
    "em_companies_different = []\n",
    "\n",
    "for language in [\"en\", \"de_from_en\", \"nl_from_en\", \"it_from_en\", \"sv_from_en\"]:\n",
    "    \n",
    "    writers_mapping = {\"english\": \"en\", \"german\": \"de\", \"italian\": \"it\", \"swedish\": \"sv\", \"dutch\": \"nl\"}\n",
    "    companies_mapping = {\"us\": \"en\", \"de\": \"de\", \"it\": \"it\", \"sv\": \"sv\", \"nl\": \"nl\"}\n",
    "    \n",
    "    for subtask in [\"english\", \"german\", \"italian\", \"dutch\", \"swedish\"]: \n",
    "        \n",
    "        if writers_mapping[subtask] == language.split(\"_\")[0]:\n",
    "            em_writers_same.append(em[\"writers_\" + subtask][language])\n",
    "        else:\n",
    "            em_writers_different.append(em[\"writers_\" + subtask][language])\n",
    "    \n",
    "    for subtask in [\"us\", \"de\", \"it\", \"nl\", \"sv\"]:\n",
    "        if companies_mapping[subtask] == language.split(\"_\")[0]:\n",
    "            em_companies_same.append(em[\"companies_\" + subtask][language])\n",
    "        else:\n",
    "            em_companies_different.append(em[\"companies_\" + subtask][language])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "42c431e2",
   "metadata": {},
   "outputs": [],
   "source": [
    "writers_deviation_from_mean = {\"language\": [], \"subtask\": [], \"deviation from mean\": []}\n",
    "companies_deviation_from_mean = {\"language\": [], \"subtask\": [], \"deviation from mean\": []}\n",
    "\n",
    "for language in [\"en\", \"de_from_en\", \"nl_from_en\", \"it_from_en\", \"sv_from_en\"]:\n",
    "    \n",
    "    writers_mapping = {\"english\": \"en\", \"german\": \"de\", \"italian\": \"it\", \"swedish\": \"sv\", \"dutch\": \"nl\"}\n",
    "    companies_mapping = {\"us\": \"american\", \"de\": \"german\", \"it\": \"italian\", \"sv\": \"swedish\", \"nl\": \"dutch\"}\n",
    "    \n",
    "    for subtask in [\"english\", \"german\", \"italian\", \"dutch\", \"swedish\"]:\n",
    "        \n",
    "        mean = np.mean(list(em[\"writers_\" + subtask].values()))\n",
    "        writers_deviation_from_mean[\"language\"].append(language.split(\"_\")[0])\n",
    "        writers_deviation_from_mean[\"subtask\"].append(subtask)\n",
    "        writers_deviation_from_mean[\"deviation from mean\"].append(em[\"writers_\" + subtask][language] - mean)\n",
    "        \n",
    "    \n",
    "    for subtask in [\"us\", \"de\", \"it\", \"nl\", \"sv\"]:\n",
    "        subtask_name = {\"us\": \"american\", \"de\": \"german\", \"it\": \"italian\", \"nl\": \"dutch\", \"sv\": \"swedish\"}[subtask]\n",
    "        \n",
    "        mean = np.mean(list(em[\"companies_\" + subtask].values()))\n",
    "        companies_deviation_from_mean[\"language\"].append(language.split(\"_\")[0])\n",
    "        companies_deviation_from_mean[\"subtask\"].append(subtask_name)\n",
    "        companies_deviation_from_mean[\"deviation from mean\"].append(em[\"companies_\" + subtask][language] - mean)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "07cac15c",
   "metadata": {},
   "outputs": [],
   "source": [
    "collapsed_writers = {\"language\": [], \"subtask\": [], \"deviation from mean\": []}\n",
    "mapping = {\"english\": \"en\", \"german\": \"de\", \"dutch\": \"nl\", \"italian\": \"it\", \"swedish\": \"sv\"}\n",
    "\n",
    "for i in range(25):\n",
    "    subtask = writers_deviation_from_mean[\"subtask\"][i]\n",
    "    language = writers_deviation_from_mean[\"language\"][i]\n",
    "    collapsed_writers[\"language\"].append(language)\n",
    "    if mapping[subtask] == language:\n",
    "        collapsed_writers[\"subtask\"].append(\"same\")\n",
    "    else:\n",
    "        collapsed_writers[\"subtask\"].append(\"different\")\n",
    "    collapsed_writers[\"deviation from mean\"].append(writers_deviation_from_mean[\"deviation from mean\"][i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "d9f07cb2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(5,4))\n",
    "sns.set(font_scale=1.3)\n",
    "sns.set_style(\"whitegrid\")\n",
    "p = sns.color_palette('mako') \n",
    "plot = sns.barplot(data=pd.DataFrame(collapsed_writers), x=\"language\", y=\"deviation from mean\", hue=\"subtask\", \n",
    "                   palette=p[2:])\n",
    "plt.title(\"writers\")\n",
    "hatches = [\"\", \"//\"]\n",
    "# Loop over the bars\n",
    "for bars, hatch in zip(plot.containers, hatches):\n",
    "    # Set a different hatch for each group of bars\n",
    "    for bar in bars:\n",
    "        bar.set_hatch(hatch)\n",
    "plt.ylabel(\"acc $-$ mean(acc)\")\n",
    "plt.legend(title=\"information\", fontsize=\"small\", bbox_to_anchor=(0.35,0.35))\n",
    "plt.yticks([-6, -4, -2, 0, 2, 4, 6])\n",
    "sns.despine(left=True, right=True, top=True, bottom=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "94ae0ab0",
   "metadata": {},
   "outputs": [],
   "source": [
    "collapsed_companies = {\"language\": [], \"subtask\": [], \"deviation from mean\": []}\n",
    "mapping = {\"american\": \"en\", \"german\": \"de\", \"dutch\": \"nl\", \"italian\": \"it\", \"swedish\": \"sv\"}\n",
    "\n",
    "for i in range(25):\n",
    "    subtask = companies_deviation_from_mean[\"subtask\"][i]\n",
    "    language = companies_deviation_from_mean[\"language\"][i]\n",
    "    collapsed_companies[\"language\"].append(language)\n",
    "    if mapping[subtask] == language:\n",
    "        collapsed_companies[\"subtask\"].append(\"same\")\n",
    "    else:\n",
    "        collapsed_companies[\"subtask\"].append(\"different\")\n",
    "    collapsed_companies[\"deviation from mean\"].append(companies_deviation_from_mean[\"deviation from mean\"][i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "708512e9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(5,4))\n",
    "sns.set(font_scale=1.3)\n",
    "sns.set_style(\"whitegrid\")\n",
    "plot = sns.barplot(data=pd.DataFrame(collapsed_companies), x=\"language\", y=\"deviation from mean\", hue=\"subtask\", \n",
    "                   palette=p[2:])\n",
    "plt.title(\"companies\")\n",
    "hatches = [\"\", \"//\"]\n",
    "for bars, hatch in zip(plot.containers, hatches):\n",
    "    for bar in bars:\n",
    "        bar.set_hatch(hatch)\n",
    "plt.yticks([-6, -4, -2, 0, 2, 4, 6])\n",
    "plt.ylabel(\"acc $-$ mean(acc)\")\n",
    "plt.legend(title=\"information\", fontsize=\"small\")\n",
    "sns.despine(top=True, bottom=True, left=True, right=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "6904b304",
   "metadata": {},
   "outputs": [],
   "source": [
    "same = []\n",
    "different = []\n",
    "\n",
    "d1 = pd.DataFrame(writers_deviation_from_mean)\n",
    "d2 = pd.DataFrame(companies_deviation_from_mean)\n",
    "\n",
    "for i, language in enumerate([\"en\", \"de\", \"it\", \"nl\", \"sv\"]):\n",
    "    for j, information in enumerate([\"english\", \"german\", \"italian\", \"dutch\", \"swedish\"]):\n",
    "        value_writers = d1[(d1[\"language\"] == language) & (d1[\"subtask\"] == information)][\"deviation from mean\"].iloc[0]\n",
    "        mapping = \"american\" if information == \"english\" else information\n",
    "        value_companies = d2[(d2[\"language\"] == language) & (d2[\"subtask\"] == mapping)][\"deviation from mean\"].iloc[0]\n",
    "        if i == j:\n",
    "            same.append(value_writers)\n",
    "            same.append(value_companies)\n",
    "        else:\n",
    "            different.append(value_writers)\n",
    "            different.append(value_companies)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "eb8e3f5a",
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.stats import ttest_ind as ttest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "f35850bd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=3.4417426174022334, pvalue=0.0012070463833898407)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ttest(same, different)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2741d06d",
   "metadata": {},
   "source": [
    "## relationship consistent correct, consistent incorrect"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "4e350187",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_conditional_consistency(consistency, correct_source):\n",
    "    \n",
    "    conditional_consistency = {\"task\": [], \"language\": [], \"consistency\": [], \"correct\": []}\n",
    "\n",
    "    \n",
    "    for collection, task in collection_task_all[\"facts\"]:\n",
    "        \n",
    "        for language in [\"en_2nd_run\", \"en-paraphrase\", \"de_from_en\", \"it_from_en\", \"nl_from_en\", \"sv_from_en\"]:\n",
    "            \n",
    "            consistency_tmp = consistency[collection + \"_\" + task][language]\n",
    "            correct_src_tmp = correct_source[collection + \"_\" + task][language]\n",
    "            \n",
    "            n = len(consistency_tmp)\n",
    "            \n",
    "            consistent_correct = np.mean([consistency_tmp[i] for i in range(n) if correct_src_tmp[i]])\n",
    "            consistent_false = np.mean([consistency_tmp[i] for i in range(n) if not correct_src_tmp[i]])\n",
    "            \n",
    "            conditional_consistency[\"task\"] += [collection+ \"_\" + task, collection+ \"_\" + task]\n",
    "            conditional_consistency[\"language\"] += [language, language]\n",
    "            conditional_consistency[\"consistency\"] += [consistent_correct, consistent_false]\n",
    "            conditional_consistency[\"correct\"] += [\"correct\", \"incorrect\"]\n",
    "\n",
    "    return pd.DataFrame(conditional_consistency)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "0fdb3868",
   "metadata": {},
   "outputs": [],
   "source": [
    "_, consistent, correct_src, correct_tgt = consistency_facts(get_details=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "7bd46ff9",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Xenia\\Anaconda3\\envs\\chatgpt\\lib\\site-packages\\numpy\\core\\fromnumeric.py:3432: RuntimeWarning: Mean of empty slice.\n",
      "  return _methods._mean(a, axis=axis, dtype=dtype,\n",
      "C:\\Users\\Xenia\\Anaconda3\\envs\\chatgpt\\lib\\site-packages\\numpy\\core\\_methods.py:190: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  ret = ret.dtype.type(ret / rcount)\n"
     ]
    }
   ],
   "source": [
    "conditional_consistency = get_conditional_consistency(consistent, correct_src)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "263e9340",
   "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>task</th>\n",
       "      <th>language</th>\n",
       "      <th>consistency</th>\n",
       "      <th>correct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>en_2nd_run</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>en_2nd_run</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>0.991952</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>0.450704</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>0.440644</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>0.424547</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>0.193159</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    task       language  consistency    correct\n",
       "0   arithmetics_addition     en_2nd_run     1.000000    correct\n",
       "1   arithmetics_addition     en_2nd_run     1.000000  incorrect\n",
       "2   arithmetics_addition  en-paraphrase     0.991952    correct\n",
       "3   arithmetics_addition  en-paraphrase     0.666667  incorrect\n",
       "4   arithmetics_addition     de_from_en     0.450704    correct\n",
       "5   arithmetics_addition     de_from_en     0.333333  incorrect\n",
       "6   arithmetics_addition     it_from_en     0.440644    correct\n",
       "7   arithmetics_addition     it_from_en     0.666667  incorrect\n",
       "8   arithmetics_addition     nl_from_en     0.424547    correct\n",
       "9   arithmetics_addition     nl_from_en     0.000000  incorrect\n",
       "10  arithmetics_addition     sv_from_en     0.193159    correct\n",
       "11  arithmetics_addition     sv_from_en     0.333333  incorrect"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conditional_consistency[conditional_consistency.task == \"arithmetics_addition\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "ba322b84",
   "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>task</th>\n",
       "      <th>language</th>\n",
       "      <th>consistency</th>\n",
       "      <th>correct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>en_2nd_run</td>\n",
       "      <td>0.963415</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>en_2nd_run</td>\n",
       "      <td>0.833333</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>0.951220</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>0.890244</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>0.696970</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>0.865854</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>0.621212</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>0.902439</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>0.757576</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>0.890244</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>0.772727</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       task       language  consistency    correct\n",
       "36  olympics_hundred_meters     en_2nd_run     0.963415    correct\n",
       "37  olympics_hundred_meters     en_2nd_run     0.833333  incorrect\n",
       "38  olympics_hundred_meters  en-paraphrase     0.951220    correct\n",
       "39  olympics_hundred_meters  en-paraphrase     0.666667  incorrect\n",
       "40  olympics_hundred_meters     de_from_en     0.890244    correct\n",
       "41  olympics_hundred_meters     de_from_en     0.696970  incorrect\n",
       "42  olympics_hundred_meters     it_from_en     0.865854    correct\n",
       "43  olympics_hundred_meters     it_from_en     0.621212  incorrect\n",
       "44  olympics_hundred_meters     nl_from_en     0.902439    correct\n",
       "45  olympics_hundred_meters     nl_from_en     0.757576  incorrect\n",
       "46  olympics_hundred_meters     sv_from_en     0.890244    correct\n",
       "47  olympics_hundred_meters     sv_from_en     0.772727  incorrect"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conditional_consistency[conditional_consistency[\"task\"] == \"olympics_hundred_meters\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "f9645ad4",
   "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>task</th>\n",
       "      <th>language</th>\n",
       "      <th>consistency</th>\n",
       "      <th>correct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>en_2nd_run</td>\n",
       "      <td>0.909091</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>en_2nd_run</td>\n",
       "      <td>0.876712</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>0.795455</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>0.506849</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>0.886364</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>0.698630</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>0.840909</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>0.616438</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>0.772727</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>0.561644</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>0.772727</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>0.493151</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 task       language  consistency    correct\n",
       "48  olympics_downhill     en_2nd_run     0.909091    correct\n",
       "49  olympics_downhill     en_2nd_run     0.876712  incorrect\n",
       "50  olympics_downhill  en-paraphrase     0.795455    correct\n",
       "51  olympics_downhill  en-paraphrase     0.506849  incorrect\n",
       "52  olympics_downhill     de_from_en     0.886364    correct\n",
       "53  olympics_downhill     de_from_en     0.698630  incorrect\n",
       "54  olympics_downhill     it_from_en     0.840909    correct\n",
       "55  olympics_downhill     it_from_en     0.616438  incorrect\n",
       "56  olympics_downhill     nl_from_en     0.772727    correct\n",
       "57  olympics_downhill     nl_from_en     0.561644  incorrect\n",
       "58  olympics_downhill     sv_from_en     0.772727    correct\n",
       "59  olympics_downhill     sv_from_en     0.493151  incorrect"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conditional_consistency[conditional_consistency[\"task\"] == \"olympics_downhill\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "d1acc41b",
   "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>task</th>\n",
       "      <th>language</th>\n",
       "      <th>consistency</th>\n",
       "      <th>correct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>elements_number_from_element</td>\n",
       "      <td>en_2nd_run</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>elements_number_from_element</td>\n",
       "      <td>en_2nd_run</td>\n",
       "      <td>NaN</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>elements_number_from_element</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>elements_number_from_element</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>NaN</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>elements_number_from_element</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>0.988889</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            task       language  consistency    correct\n",
       "12  elements_number_from_element     en_2nd_run     1.000000    correct\n",
       "13  elements_number_from_element     en_2nd_run          NaN  incorrect\n",
       "14  elements_number_from_element  en-paraphrase     1.000000    correct\n",
       "15  elements_number_from_element  en-paraphrase          NaN  incorrect\n",
       "16  elements_number_from_element     de_from_en     0.988889    correct"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conditional_consistency[conditional_consistency[\"task\"] == \"elements_number_from_element\"].head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "f217f144",
   "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>task</th>\n",
       "      <th>language</th>\n",
       "      <th>consistency</th>\n",
       "      <th>correct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>en_2nd_run</td>\n",
       "      <td>0.970588</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>en_2nd_run</td>\n",
       "      <td>0.875000</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>0.970588</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>0.660714</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>0.911765</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>0.589286</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>0.852941</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>0.607143</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>0.852941</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>0.911765</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>0.571429</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             task       language  consistency    correct\n",
       "24  elements_number_from_position     en_2nd_run     0.970588    correct\n",
       "25  elements_number_from_position     en_2nd_run     0.875000  incorrect\n",
       "26  elements_number_from_position  en-paraphrase     0.970588    correct\n",
       "27  elements_number_from_position  en-paraphrase     0.660714  incorrect\n",
       "28  elements_number_from_position     de_from_en     0.911765    correct\n",
       "29  elements_number_from_position     de_from_en     0.589286  incorrect\n",
       "30  elements_number_from_position     it_from_en     0.852941    correct\n",
       "31  elements_number_from_position     it_from_en     0.607143  incorrect\n",
       "32  elements_number_from_position     nl_from_en     0.852941    correct\n",
       "33  elements_number_from_position     nl_from_en     0.571429  incorrect\n",
       "34  elements_number_from_position     sv_from_en     0.911765    correct\n",
       "35  elements_number_from_position     sv_from_en     0.571429  incorrect"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "conditional_consistency[conditional_consistency[\"task\"] == \"elements_number_from_position\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "9b49a723",
   "metadata": {},
   "outputs": [],
   "source": [
    "companies_df = conditional_consistency[conditional_consistency[\"task\"].isin([\"companies_us\", \"companies_de\", \"companies_it\", \"companies_nl\", \"companies_sv\"])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "f19f68c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "writers_df = conditional_consistency[conditional_consistency[\"task\"].isin([\"writers_english\", \"writers_german\", \"writers_italian\", \"writers_dutch\", \"writers_swedish\"])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "4722bbe0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                         consistency\n",
      "language      correct               \n",
      "en_2nd_run    correct       0.989753\n",
      "              incorrect     0.911862\n",
      "en-paraphrase correct       0.941693\n",
      "              incorrect     0.819965\n",
      "de_from_en    correct       0.933615\n",
      "              incorrect     0.756565\n",
      "it_from_en    correct       0.914660\n",
      "              incorrect     0.691074\n",
      "nl_from_en    correct       0.921384\n",
      "              incorrect     0.771078\n",
      "sv_from_en    correct       0.888737\n",
      "              incorrect     0.664966\n"
     ]
    }
   ],
   "source": [
    "companies_avg = companies_df.groupby(['language', 'correct'], sort=False).mean(numeric_only=True)\n",
    "print(companies_avg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "0e037eff",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                         consistency\n",
      "language      correct               \n",
      "en_2nd_run    correct       0.985496\n",
      "              incorrect     0.552545\n",
      "en-paraphrase correct       0.976538\n",
      "              incorrect     0.375956\n",
      "de_from_en    correct       0.973214\n",
      "              incorrect     0.328941\n",
      "it_from_en    correct       0.962470\n",
      "              incorrect     0.374845\n",
      "nl_from_en    correct       0.970483\n",
      "              incorrect     0.242494\n",
      "sv_from_en    correct       0.952142\n",
      "              incorrect     0.245891\n"
     ]
    }
   ],
   "source": [
    "writers_avg = writers_df.groupby(['language', 'correct'], sort=False).mean(numeric_only=True)\n",
    "print(writers_avg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "c3cf77d5-b85c-4a6e-9098-7dce89fb2a6e",
   "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>task</th>\n",
       "      <th>language</th>\n",
       "      <th>accuracy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>en</td>\n",
       "      <td>99.400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>98.600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>45.200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>44.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>42.400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>19.200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>elements_number_from_element</td>\n",
       "      <td>en</td>\n",
       "      <td>100.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>elements_number_from_element</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>100.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>elements_number_from_element</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>98.890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>elements_number_from_element</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>100.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>elements_number_from_element</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>100.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>elements_number_from_element</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>100.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>en</td>\n",
       "      <td>37.780</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>42.220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>40.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>36.670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>36.670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>elements_number_from_position</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>41.110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>en</td>\n",
       "      <td>55.410</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>54.050</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>50.680</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>51.350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>52.030</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>olympics_hundred_meters</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>50.680</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>en</td>\n",
       "      <td>37.610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>31.620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>35.040</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>35.040</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>35.040</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>olympics_downhill</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>33.330</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>writers</td>\n",
       "      <td>en</td>\n",
       "      <td>76.234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>writers</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>76.236</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>writers</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>76.774</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>writers</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>75.268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>writers</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>76.772</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>writers</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>74.300</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>companies</td>\n",
       "      <td>en</td>\n",
       "      <td>78.200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>companies</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>76.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>companies</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>75.400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>companies</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>73.600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>companies</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>73.200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>companies</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>71.800</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             task       language  accuracy\n",
       "0            arithmetics_addition             en    99.400\n",
       "1            arithmetics_addition  en-paraphrase    98.600\n",
       "2            arithmetics_addition     de_from_en    45.200\n",
       "3            arithmetics_addition     it_from_en    44.000\n",
       "4            arithmetics_addition     nl_from_en    42.400\n",
       "5            arithmetics_addition     sv_from_en    19.200\n",
       "6    elements_number_from_element             en   100.000\n",
       "7    elements_number_from_element  en-paraphrase   100.000\n",
       "8    elements_number_from_element     de_from_en    98.890\n",
       "9    elements_number_from_element     it_from_en   100.000\n",
       "10   elements_number_from_element     nl_from_en   100.000\n",
       "11   elements_number_from_element     sv_from_en   100.000\n",
       "12  elements_number_from_position             en    37.780\n",
       "13  elements_number_from_position  en-paraphrase    42.220\n",
       "14  elements_number_from_position     de_from_en    40.000\n",
       "15  elements_number_from_position     it_from_en    36.670\n",
       "16  elements_number_from_position     nl_from_en    36.670\n",
       "17  elements_number_from_position     sv_from_en    41.110\n",
       "18        olympics_hundred_meters             en    55.410\n",
       "19        olympics_hundred_meters  en-paraphrase    54.050\n",
       "20        olympics_hundred_meters     de_from_en    50.680\n",
       "21        olympics_hundred_meters     it_from_en    51.350\n",
       "22        olympics_hundred_meters     nl_from_en    52.030\n",
       "23        olympics_hundred_meters     sv_from_en    50.680\n",
       "24              olympics_downhill             en    37.610\n",
       "25              olympics_downhill  en-paraphrase    31.620\n",
       "26              olympics_downhill     de_from_en    35.040\n",
       "27              olympics_downhill     it_from_en    35.040\n",
       "28              olympics_downhill     nl_from_en    35.040\n",
       "29              olympics_downhill     sv_from_en    33.330\n",
       "30                        writers             en    76.234\n",
       "31                        writers  en-paraphrase    76.236\n",
       "32                        writers     de_from_en    76.774\n",
       "33                        writers     it_from_en    75.268\n",
       "34                        writers     nl_from_en    76.772\n",
       "35                        writers     sv_from_en    74.300\n",
       "36                      companies             en    78.200\n",
       "37                      companies  en-paraphrase    76.000\n",
       "38                      companies     de_from_en    75.400\n",
       "39                      companies     it_from_en    73.600\n",
       "40                      companies     nl_from_en    73.200\n",
       "41                      companies     sv_from_en    71.800"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Compute incorrect and correct \n",
    "display(acc_df)\n",
    "\n",
    "def max_consistencies(acc1, acc2):\n",
    "    # Compute the maximal consistencies for\n",
    "    # correct and incorrect examples, given\n",
    "    # the accuracies\n",
    "    max_correct = (acc1 - max(acc1 - acc2, 0)) / acc1\n",
    "\n",
    "    incor_en, incor_other = 100-acc1, 100-acc2\n",
    "    \n",
    "    max_incorrect = (incor_en - max(incor_en - incor_other, 0)) / incor_en\n",
    "    return max_correct, max_incorrect\n",
    "\n",
    "# Compute this per task\n",
    "upper_bounds = {}\n",
    "for task in acc_df['task'].unique()[2:]:\n",
    "    upper_bounds[task] = {}\n",
    "    incorr = []\n",
    "    corr = []\n",
    "    results_en = acc_df[(acc_df['task'] == task) & (acc_df['language'] == 'en')]['accuracy'].iloc[0]\n",
    "    for condition in ['en-paraphrase', 'de_from_en', 'it_from_en', 'nl_from_en', 'sv_from_en']:\n",
    "        results_condition = acc_df[(acc_df['task'] == task) & (acc_df['language'] == condition)]['accuracy'].iloc[0]\n",
    "        max_corr, max_incorr = max_consistencies(results_en, results_condition)\n",
    "        incorr.append(max_incorr)\n",
    "        corr.append(max_corr)\n",
    "    upper_bounds[task] = {'incorrect': sum(incorr)/len(incorr),\n",
    "                               'correct': sum(corr)/len(corr)\n",
    "                              }      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "e4702297",
   "metadata": {},
   "outputs": [],
   "source": [
    "CC_new = conditional_consistency[conditional_consistency[\"task\"].isin([\"arithmetics_addition\", \n",
    "                                                                       \"elements_number_from_element\", \n",
    "                                                                       \"elements_number_from_position\", \n",
    "                                                                       \"olympics_hundred_meters\",\n",
    "                                                                       \"olympics_downhill\"]\n",
    "                                                                     )].to_dict('list')\n",
    "\n",
    "companies = companies_df.to_dict('list')\n",
    "writers = writers_df.to_dict('list')\n",
    "\n",
    "CC_new[\"task\"] += [\"writers\"]*12 + [\"companies\"]*12\n",
    "CC_new[\"language\"] += CC_new[\"language\"][0:24]\n",
    "CC_new[\"correct\"] += CC_new[\"correct\"][0:24]\n",
    "CC_new[\"consistency\"] += list(writers_avg[\"consistency\"].values) + list(companies_avg[\"consistency\"].values) \n",
    "\n",
    "CC_new = pd.DataFrame(CC_new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "424ad7d4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAk0AAAIUCAYAAAAHexhnAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAA9hAAAPYQGoP6dpAACw7klEQVR4nOzdd3xT1d8H8M/N6Eo3ndCyS9lU9pSNgMhWkKG4wIELfQRFBXGBKCIyHYgCiiAyKvBDGUW2FBCojDI6oXu3aZsmOc8fpZFKadMmbZL28369fs9T770593sPafrNued+jySEECAiIiKicsksHQARERGRLWDSRERERGQEJk1ERERERmDSRERERGQEJk1ERERERmDSRERERGQEJk1ERERERmDSRERERGQEJk1ERERERmDSREREVm9P4ll8ejUUMeqUCo/NKlLj06uh+CnuiFnbtRYR2XH49GoojqReNmyzxeuwRQpLB0BERFSR5s5+cFU6wU3pZOlQrBL7p2YwaSIiIqsX5OyPIGd/S4dhtdg/NYO354iIiIiMwJEmIiKyensSz+KfnHg83KA7Gjl5G7bHqlNxIv0qEgszIYOEZipfhLg3qXT7RXodDqZE4FLOLWj0RfCyc0VH9yZo7Rpw17GRuQk4nxWDpMIsFOqKoJQp4GPvik7uTdHc2a/UsXH5aTiZfhUphdko0BfBWe6AJiofdPcMgrPCodSxOUX5OJ4eiSh1MvK0hXBS2KOpkw961GsBF4Vjpfvn06uhaK7yRa96LXE47TJu5qdBJwT8HNzQw7NFqX4EAL0Q+DsrGhHZsUjX5EImyeBn744uHs3QROVT6T6tjZg0ERGRTbqccxO7Es9CIckQ5OwPpUyO67lJiK7CZOg/ks9DJ/QIdqkPvRC4mpuA3UlnkaVVo4dnC8NxR9Ou4Hh6JNyVTmjpXB8KmRxphTm4oU5GXH4axvh3QbPbidPN/HT8cvME7GVKBDn7wV6mRFJBJv7Oika0OhmPN+wHpUwOAEgtzMHPN48hX6dBU5UvWro4I1OThwvZsbiWl4SJAT3haedc6etKL8rDj3FHUM/OBe1cGyJbm4/I3ARsvXkSUxveD297VwDFCdOOhFO4npeEenbOaO/WyNAPW2+dRH+vNujk0bTS569tmDQREZHNKdQVYV/yBShlcjwa0Mvwx7+XZyE23zyOPF1hpdrTCz2mNrzfMJG6u2cQNsUfxbG0SLRybgB3OxXytIU4mX4V9excMCWwN5Syf/+E/p0ZjX0pF/BPTrwhaQrPvAGd0GNSQC+426kMx+5NOnc7GUpEK5cGAIDdSWdQoNNgXP1upUZ1buQl4ddbf+F/SX9jUmDvSvdTuiYXnd2bop93G8O2Y2mROJZ+BeeyYjDIpx0A4GxmFK7nJaG1SwCG+naATCqevdO7XjB+jD+KsNSLaKzyRj07l0rHUJtwThMREdmcKHUyCvRFaOfa0JAwAYCTwh596rWqdHtdPJqXevLMTemELh7NISBwKfcmAEAmSRjudx8e8GlfKmECgIZOXgAAtU7z70YhAADxBemlju3r1RrPNhmMls71AQAJBZlILsxGkLP/XbfBmqp80cjJC7cKMpCmyan0dQFAN8+gUv/dTOULAMgsyjNsO58dAwkSBnq3NSRMAOAgt0MPjyAICERkx1Xp/LUJR5qIiMjmJBVmAQD8HTzu2hfg6Fnp9sp6Tf3bbScXFJ/LUW6HlrdHhjI0uUjT5CKrSI10TS7iC9IAAOJ2ogQAHdwa41peIv6X9DeOpV1BYydvNHLyRmMn71LzmRILMgEAal0hjqZduSuOIr2u+JoLsio90uMoU8JRbldqm71cCQDQCT0AQKPXIk2TCzuZAuGZN+5qQ3171K6kH+oyJk1ERGRzCnVFAAB72d1/xuzlSsggVao9p/9MygYAu9tta4TWsO16XhKOpF5GiiYbACCDhHp2LvCzd0eaJhfAv0lTY5U3Jgb0QnjGdUSrU3A+Oxbns2Mhl2Ro5dIAA7zbwk6mQKG+eHQqPj8d8fmlR6XuVKAvqtQ1AYD89pypO5X0TEmkhbfb1ei1OJ4eadbz1zZMmoiIyOY43B49KSzjD7lWr4P+juTFGIW6IkBZeluutqD4XLLicyUWZGL7rVNwkCsxxKc96jt4wsNOBbkkQ5omB//kxN/VbgNHTzRw9IRO6JFQkIFodQr+yY5DRHYcJEh4wLeDITm7v14rdPVsXqm4zcFOKj6/p50znmzUv8bPb0s4p4mIiGyOn4M7AJQ5MnOrIKPS7SUWZt61raTtknNdyomHgMD99VqhvVsjeNm7QH57/k9aYfF8o5JUTQiBk+nXDEudyCUZAhzroXe9lng0oNft9otv6fnYuwEAEu4R99nMaBxLu4KsInWlr8sY9nIl3JROyNTkIf/OOVm3JRVkIizlIq7nJlXL+W0JkyYiIrI5TZx84KxwQER2LG7l/5tsFOqKcCTtcjmvLFt4xvVSCUO6JhenM29AIcnR2qW4VpNCKr7Vpf7Pk3nZRfk4fPuc+ttzmiRJwvW8RJzIuGpIjkpk3k5+SiaeN3DwhKfSGVfzEnE1N6HUsTfz03EwJQKnM2/cNTfJnNq5NoQeAvuSLxjmOgHFt+z+SL6A8MzrZY7q1TW8PUdERDZHKZNjqE8Itif8hZ9vHkOQyh+OciWu5yWVevqrMr6POYQWLv4o1GlxNS8BRXothvneB5XCHgDQ0qUBwjNv4EjaFSQVZsFdqUJ2UT6u5yVCLskgl2TI1/+beN3v1Qpbbp7Alpsn0FzlBzelE3Jv10lSSDL0qhcMoDjBGu53H7bcPIEdCeFo7OQNb3tXZBfl41peIgBgqG+I4TZedeji0Qyx6lRcyb2F5JgsNHLyhgQJ1/ISkKMtQAtnf8Mk+LqMSRMREdmkkonWx9IjEa1Ohh4CjRy90N+7Lb6K3lepth7y74SzmdG4mB0PrdDBz8EDPT1bGEoJAIC3vSsebtAdR9OuIFadiiiRDBelI1q7BqC7RxD2JP2N2PxUpGlyUM/OBQGO9TAxoBf+Sr+KWwUZuJaXCHuZEs1Vfuju2QJe9v8+Cefn4I6pDe/HyfSriFYnI1adaqgI3tUzCP63bxFWF7kkw7gG3XA2MwoXc+IRcXvCurtShe6eLdDONRAyqXKT62sjSdz5fCQRERERlYlzmoiIiIiMwKSJiIiIyAg2mTRFR0cjJCQECxYsqPRrk5KSMG/ePAwePBjt2rVD//798f777yM9/d4FxYiIiIhsLmlKTU3F888/j/z8/Eq/Ni4uDuPGjcOmTZvg4OCA/v37Qy6XY8OGDRgzZgwSExOrIWIiIiKqDWwqabp06RImTZqE69evV+n1c+bMQUpKCl544QWEhoZi2bJl2Lt3LyZOnIjExETMmzfPzBETERFRbWETSVNWVhYWL16MRx55BDExMQgICKh0G+Hh4QgPD0fjxo0xc+ZMw3a5XI63334b9evXR1hYGK5du2bO0ImIiKiWsImk6YcffsA333wDT09PrFq1CqNHj650GwcOHAAADBw4EDJZ6ctWKpUYMGAAAGD//v0mx0tERES1j00kTX5+fpg9ezb27t1rSG4qKzKyeOXmFi1alLm/efPiRRIvX658+X0iIiKq/WyiIvjDDz9schvJyckAAF9f3zL3+/j4lDqOiIjIluiFQHx+GvJ0hVDJ7RHgWI9VvM3MJpImc1CrixdIdHR0LHO/g4NDqeOqqrCwEDqdzqQ2iEylP3MEut9+BAor/5RptbB3hHzEZMg69rJ0JNXG6vocqBP9TsWu5yfjSGYk8vT/Liasktmjt3sLNHP0Meu5aut73cnJqcJj6kzSJJcXr04tVZB1m7qqTEREhEmvJzKHgNAf4ZAcb+kwSsn77UfEO3paOoxqY419DtRMvwsIZMo10Eg62Ak53HV2kMARjpqSoshHhGNG8X/c0e15ukLsTbuAtvke8NaWPWBQFbX1vd6pU6cKj6kzSZNKpQKAe9Z3KigoAHDvkShjtW3bliNNZHH6hyZD99tG6/kmaO8I1YhJaNWqlaUjqTZW1+dApfu9KiMIWqGHRl8E/zu2yQDYyZRQSGaYNsvRsnLphcBfiUcBfRk7bydQ0S5q9Pa9z2y36mrDe72q6kzS5OPjg3/++QcpKSll7i+Zy1Qyt6mq7O3tTXo9kVn0Hlz8P6o5taDPcw/sAJIqN4KgQA38ITm4A0736NuiU4dQsO17iALTplaYi+TgBIcx06Dscn+NnC9WnVrqllxZcnWFSJfy0dDJyzwnrQXv9aqqM0lTcHAwDh48eM86TCXbg4ODazKsSrG2Dweg5j8giKj62A+bgIJt64z+jMnTFpQ5wFFCBkClcDApJsnBCfZDH7nn/sI9m6FPiDXpHOYkABT+b3ONfSbm6cpPmCp7HJWvziRNffv2xerVq/HHH3/g5ZdfLlWrqaioyFCfqX///pYKsULW9uEA1PwHBBFVH2WX+43+XY5Vp2LzzeMVHvdIgx7mG+EoQ2UTvepWUZJnbiq5cXc3jD2OylfrkqaioiLExhYnFg0bNoRSqQQAdOzYEe3atcOFCxewZMkSvPbaa5AkCTqdDh9++CESEhJw//33W/WcC1M/HHRCDyEEJEmC3BxzDVDzHxBEZB2sZYSjMolebRTgWA/OCgfkagvueYyLwgEBjvVqMKraq9YlTUlJSRg+fDiA4uredy658vHHH2PKlCn4+uuvsX//fgQFBeHSpUuIjY1FgwYN8MEHH1gqbKNU9cMhMjcBB1IiSv1SOSscMMC7LVo4+5fzSiKisnGEwzrIJAkDvNtiZ0L4PY/p790WMklChiYX7kpVhU+R073ZREVwcwkKCsKvv/6KsWPHIicnBwcPHgQATJ06FZs3b75n4UtbFpmbgJ0J4Xd9C8nVFmBnQjgicxMsFBkR2aqb+enIKlLDuZz5SkIIOELJEY4a0MLZHyP9O9/17+GicMBI/85o4eyPNE0Ovo89hNDE0yjSay0Uqe2ThKmFichq6YXAV9H7Khy2fabxIFaNrUas0ku1SUphNjbEHYYQAr3qBeNw2t1LT5X8Wck9cAXvPft/NR1inVXeZ01Edhx+TzqHhk5eGFe/G0ebqqjW3Z6jf8Xnp5WbMAFAjrYA8flp1TpRsy7jrVGqbbzsXBCk8odO6HCfexN42Dnf9R4vUhcgYv0+ICEXeNaCwdYR57Ji4KZwREMnr3t+lrd1DYSXnQtUCntDwqQXemiFHnYypgLGYk/VYtYwUdPayiTUZImEkluj/1Vya7Rk2JzI2hXoNLCTKSCTZJAkCUN9O0B+++cWzv5orvJDfH4aovNSsPXHn/HPrmMQOl2tnPJgbYr0WoSl/IMiocPkwN7wd/C457F+Du6l/vuvjGuIyI7HcN/7UN/x3q+jfzFpqsWsYaKmtZVJqKkSCXohcCCl/CV1DqZEoLnKj7fqyGro9fpS5ViA4ttx2xNOIUjlh37ebQAACpm81DEySTKMcqw+GQlh5lURyoqLihXpdWjtGoDkwiz42bsb/TqtXocL2XHIKlIjoyiXSZORmDTVYsY8iuosd4CXnUu1xVBXa6jw1ijZIplMhk8+3464+DTDNqem9vAe4objsVfx65KTEJryp8Gmp+eW+vnF1741KabAgHp449XRJrVRmzkp7DHYp32lX6eQyTE18H5czIlHa5d/nzIvKUtDZWPSVIsZ8yiqi9IBP8YfwSj/LvC2dzV7DHW1hkp2kXFrMrFKL1mbuPg0XL+R+O+GG4BfhgdSL2VDq654BEnpoULgfa1QkJGDW6diSrdFVsVBrkRH9yaG/9bqddgUfwytXBqgo3sTJk9l4HhnLVfeo6gP+LRHnrYQ2UX5fATVjAp0RTicdsmoY1nDhqyNpJDQ7EF/yO3//fOQeDrDqIQJAOq1bILgMb0R2Kc9ZHJVdYVJAJILs5BZZL5R/Is58UgszMSJjKvI12vM1m5twpGmOuDOiZr/fRS1ubM/buVnoL6jp6XDrDUc5Eo0cPDE1bwElHcjg1V6yRp5D3GFY0N7OLgp8c+PlZ+PmHnDDknn45ASEQ+5ffNqiJBKHEq5iJj8VDzg0wHt3Bqa3F4714bQCwEXhQOc+IWuTBxpqiNkkgQpLR/rPliGFe8txs344pXMHeV2aOb87xMuudoC/JZwGmotbxsZSy/0OJsZDc0do3WDfNthmO995b7Oy84VKDetIqp5meF5yE/XIO5oapVer8lW4p8NGUj+W8WRpmqkF3pIkgQJQKCZvnxJkoQQ98Zo5uxn2HYrPwOhCaeRr+PIE8CkqU5ZsWIFwsPDER4ejpUrV5Z5zP+S/sbl3Fv4X9LfNRucDduZEI79KRdwPD3SsM1Jbo/WrgFl3hp1vv0tLkqdjCNpV2o6XKJShBDIu+OhBU2yFicWX0F2jHU8vEFlk0kyjG/QHc81HQJ3u+pJToUQ2Jv8N67k3sJRflYB4O25OqVkIWMAiImJKfOYfl5t8L+kv9Hfu21NhWXz2rk2Qnx+OjyVznftu9et0Rt5STiadgUhbo1rPmCi23RCj33JF3AjLwlTGvaBi8IRACB0po+Ayu1kcGvshPTI3IoPruNMrWeXbeZ47qxnN8z3PhxJu4ze9Vqa+Sy2iUkTleJl74LJgb1LPTWRWJAJb3tXyCUOTAoh8E9OPJzlDmis8gYANHP2xTOOA2EvV5b5mpIaNndq7uyHpiofyNinZEE6oUdCQQbydIWIU6ehtWtAxS8ygiSX0OvtVlA4yHFi8RWoU3i7vzzWXM/Oz8Ed4xt0L7X/ZPo1BDp61sm5sEya6C53JkwphdnYFH8MvvZuGFO/CxzkdhaMzPLOZkXjQEoE3BROmNaoL5S3lx+4V8JUnjsTplv5GdDoi9BY5WO2WIkqYidTYHT9rsjQ5KKJGd97QieQHaeGo6cd7F0VTJoqUNl6dhq9Fhq9FnYyRbUsgVJePbtYdSoOp12CBAlPNepfbbcGrRWTJipXnrYAMkmCQiaDnazyiUFt09Y1EH9nRqGta0OzjRIlFGRi881jkCBhUmDvaqmXRVTiUs5NyCAh2KU+AMBd6QR3pZPZz3PhhxjoCvVmb7c2qmw9u19unkC0OgVDfNqjvVsjAEBcXByWL18OAJg5cyYCAwOrJVYfeze0dmkApUxR5xImgEkTVaCxygdTAvvAUW5nWO6jZAXzulD47FpuInJP7kOz/QcM3wLH396XZ6ZzqAA8evvJFDv5DxXOT6jJ9fOodrmRl4RdiWegkGTwtneFp93d8/DMhQlT9RlXvxtuFqTffgK3WMmDPgCwcuVKfPzxx9Vybge5EsP9OkIv/p33VqgrwqWcm+jg1qjW/11g0kQV+u8H66mM68goysNA77Z3rUFVm2Rq8rAj4RRG7t8DfUJStZ7rzufrKpqCW1Pr51Ht09jJB42dvOFj7wZ3Zc2NEkgyQDCHMhtJku6q8WbMgz7mdOeamQdS/sE/OXFIKMjEML+Qaj+3JTFpokrJLlLjSNpl6CHQyMkbLW8P8ddG7nYqdHJvCvWgByHt+wMoMG5plP/S6XTIzS1+gsjZ2RlyecWJplbooJDKPq6m1s+j2iFHmw9nuQMkSYJMkjC2ftcaewDBN8QdjQf5IO1SDq7tSqiRc9Zm1rouXH1Hd1zNS0A7t+q5JWhNmDRRpbgqnTC2fldEqVMQ7Oxv6XDMKj4/DUdSL+Mh/85QKYqr4fbzbgMMaAMMmFTldufMmYNTEacAAF27Nq9w2PxMZhQOpEQgSOWHh/w7l/pGR1QZUXnJ+C3xNLp6BKGbZ3F17pp8YlMIAZWPA/RaFnE1h6TCLPwv6W+0dQ1EZ49mlg7HoINbYwQ7N4DDHQ/EJBRkwE3pVOsqi/N5Z6q0xiof9PduY/jGoxN6XMqON8x1skVCCBxM+QfxBek4lm7eIm6VHTb3tnOBXJLBVekEpktkiqwiNQr1WkTlJUFvgftj6ZG5uPBDNM6uvl7j566NLufcRKomBwkFmZYO5S53JkxqbSG23TqFdTGHkFL47yxNvRCIVafiUs5NxKpTS82LshUcaSKTHUiJwLmsGNwsyMAgn3aWDsdod05olyQJg3za4UJWrMWLuAU6eeHxhn2rdZJubaUXosw1FuuqEPfGsJMp0MLZ3yI1wbT5OqREmLv0Yt3V3TMI9excrP6zoUBfBMfbSZTH7blzkbkJOJASgdw7qs87KxwwwLstWtjQXQsmTWQSIQQ8lc6QQUJTG6oxlFKYjT+Sz6O9WyO0dS2+D+/v4AF/Bw8LR1bszg9FIQRSNNnwsXezYETWr7Z8KJsipygfJzKuor9XG8NDGuYqWEnVR6/XQyarOKl1kNuZZWFeYxkb13952jljSuD9UOsKoZDJEZmbgJ0J4Xcdl6stwM6EcIz072wzv6NMmsgkkiShk0dTBDn7w1XpaNiu1eus+sm6KHUybhVkIE9biNYuDay2MrdO6PF70jlcyrmJ8Q2631VZnIrVpg/lqtILgS03TyC9KBcKSY7+3m0sHRIAQKaQ4NvRA+6NnXBpSzzXqC6DTCbDJ59vR1x8WpXbSE/PLfXzi699a1JMgQH18Maro6v8eqVMDjeZE/RC4EBKRLnHHkyJQHOVn02MCjNpIrO4M2Eq0GmwMe4I2ro2RFePZlbxtIcQAkVCZ6ie28m9KfK0hejs3tRqEyYAkEFCkdBBAMjWVu3pvdqutn0oV5VMktDfuw0Op13Cfe5NLB2OgdALNH/QH0pHOW4eT0N2HN/HZYmLT8P1G4ll7pNkQMuHA5F6MRup/2SVWb6hSKsr9fO92qpp8flppUZ/y5KjLUB8fppNfClk0kRmdynnJjKK8nAuKxohbo2qtMSIOWUXqfFH8gVohQ6PNOgBSZIgl2RW8028PJIkYbjvfUh0z7yrLgsVq20fypWhF3rkaQvhcvtLSxOVDxo5eVtVcij0QPzRVAi9QGGO1tLh2CTPFi7w7+SBei2ckfpPlqXDqZQ8nXFL6Bh7nKUxaSKzC3FrDJkkg5+9u8UTJgDQQyAuv/hJjVRNjs0tU6KQyUslTFq9Djqht4q+tQa17UPZWPk6DUITTiNbq8bk21X7AVhVwlQi6vfqLQ5b26mTCxFzMBm6Qr3NFQlVGVlywNjjLI1JE5mdJEnocHs9pBIx6hRo9Xo0c/atkRhyivIN377dlSoM9Q2Bt70r6tm51Mj5q0u+ToPtt05BJkkYV7+bVc8bM1XRqUMo2PZ9hYuY+gm9YRma8jjJ1yNbkkFAIF+ngVySwb4S6yla2/I1AgKZRXnI12mQWpiNwFo2ikb/yk/X4Poe67jdVlkBjvXgrHAodzTYReFgMyPpTJqo2mUX5SM04TQK9EUY498FzZz9qu1cWr0OvyefR2TuLTzesB88bi8o2dKlQbWdsyblaguQosmGBCCjKM/mRs0qo3DPZugTYis8TgbA2AewS+YgO/7nv419rTUtX+Mkt8eY+l0BwDbeBxLgGugEbYEO6uTaNepnDSSZE4SuODGRya1nIV2ZJGGAd9syH9Qo0d+7rVWOkJaFSZONqOqjn9XNmLhUCnu0dGmAxIJMNHLyrtZ45JIMedoCaIUeMeoUQ9JUVdbW7972rhjt3wUOktI2/lCawH7YBBRsW1fmSFO+TgOt0MNOpoC9TAGt0CFfV3TPthzlSsOyNAICOn3xPY47R+rytIXQQ/zn2OL/K0Gy+PI1QggcT7+K+g4eaKwq/j2ypfdAs6F+aNTfBzdPpuHK1puWDsdm+N7njtyEAuQllj9vT+EQBG3BVQCA3L55TYRmtBbO/hjp3/mukiAuCgf0t7GSIEyabIQtP5Iql2QY5NMORXptqT9SudoCOCscynmlcVIKs+Fp5wy5JIMkSRjs0x75eo1Zai7ZQr9be3mHqlJ2ud8wqnNnIVIASMlNRFjqRfT1ao3mt0cuy6rTVPKh3KCCD2W90ONE8gXE56dhUmAfQ3XjvzOjsT8lAh3cGlm8cOu5rBgcS78CB5kSTzbub3PLU2TcyEP9blroi1hzwFhyBxlajg+AXCnDiU+vlDtCJ5OrYKcKqbngKqmFsz+aq/xsvvgskyYbUt4jqcaw9COpStm/b7cLWbE4kBKBYb4haGHCor/H0iJxPP0KetdriW6eQQCKF9p1h/mGp62539M1udh66yT61mtlUj9as5v56QhLvYjO7k0RfPsam6p80VjlA/kd5SJM+VCWSTIM8e1w1/ZUTQ4ERKkERSf0+OXmCdR38EA3zyBDGYvq1tY1EJdzb6KNS6DNJUwAkHE1B0cWXLS5icyWpHSUI/1KDhw87GrFLU2ZJNn8E6xMmqjGCSEQmZuAIqFDmia34heUw03pCIHi+T110YXsWGQVqXEsPRLNnf2suuZUVUWrU5BQkIET6VfRwtnfsOyNvIyV+cz9oTzIpx26ejSH/I7EK6kgC3H5aUgpzC615E50XgokCajv4FHqC4IpMjS58LhdHV4hk2NCg55WUfesKpgsVV5BRhEu/BADLkJpPZg0UY2TJAlj6nfBP9nxhiVM7lTe+mEZmjwUCa1hSZHWLgHwUKpQ39GzRq/BWvSp1woySOhk5UU6KyNXWwC90MNV6QQA6OLRDBq9Fl0sVCj1zsKtAOBhV/w0pkZfVCqeo+lXkFCQgaG+IYb3tVZfPMpY0e3Tst7z57KicSDlHwz2aYf2t59GtdWE6b9kCgl6LW/TGY1dZTWYNJFFyCRZqTWUhBDYl3IB7koVTmfeKHP9MDkk7Ew8DU+lM6Y27APZ7TlMdTVhAopHVvp4tSq1TQhhs39cL+fcxN6kc2js5I1R9bsAAOxkCqsqROoot7sr2RdCoJ6dM3K0+Qi849Hp63lJ2J10Fm1cAsq8/Qfce828AAdPCAjcKsgwJE22TuEoR/tpjeHs74AjCy4ycSqHk7c9CnOKoCvgEJ01YdJEViE88wbOZcWUua9k/bAHfDtAKcnhpLBDgb7IJud1VLfovBScyLiKsfW71thcG3PysnOFVuiQpytEkV5rtttc1U2SJAz1DbkrYU0szIRO6EuNNAkhEJp4Gt72rnBVOGJP0t93tZerLcDl3Fvo4t4M9/8nKbZl2nwdHDyUUDjI4drQCZk36uZtdWO0HNcArg2dELEhFqkXsy0dDt1mG59IVOu1dgnAkbTL0JUz8eFY2hVMDuwDd6WTzY6kVCetXof/Jf+NXG0BTqZfvWsEytoIIXA1LxGFuiLDqKOXvQsmB/aBr72bTf4b/zfm++u1QjvXhqVunWYW5SEyNwHXchMrTPwv595EH69WtWpKy8VNcSjI0KAg494lIuo6mUKCwkkBSSYhO7784q5Us5g0kVVI0+SUmzABxeuH5WjzTa69VNNqquicQibHKP8uOJcVjR6eLartPOYSpU7GzoRw2MkUaKbyhZOiOIHwc3C3bGBmJEkSPO1Kl950kNthkHc73CrIwMWc+HJfXxvXzOPoUsX0WoG/lkTCsZ4dNNlcr8+aMGkiq1Cb1w+ryaJz/g7u8HcIqdZzmEIn9IYyAU2cfBDg6IkAx3q1ss7UvTjK7RDi3hj2OcoKkybANt/zZB75aRUvD0Q1i0kTWYXatqjjnSxZdO5sZjRkZawFWNPydRocSbuMxIJMTA7sA9ntsgG2/Ai9qWrze74i7k1U8L3PHemROUiJ4HydO/HJQutWO55RJptXsqhjeWxpUUdrEKNOwf6UC9iXfB7JhVmWDgeXc24hqTALsepUw7a6mjABdfs979HcGQ2614NPB3dLh2J1GvbzRo85LeHXyfQVDcj8mDSRVShZ1LE8trSoozVo6OiFdq4N0dMzGN52NbtGme722n8lHOV2GOTTDhMCehrWTavr6vJ7PuWfLMQdTcWtk+mWDsXqeLVyhaOnHSA42mSNeHuuDrHWVbBL1KZFHa2BJEkY4tO+xkdzNHotNsQeRnpRLh5reL+hEGkrlwY1GoctqKvv+dxbBbi645alw7BKZ1ZdR71WrkiPzLF0KFQGJk11iDWvgl2itizqaC3uTJj0QuBE+lWEuDUyPKlWHexkCvjYuyJfr0F2Ub4haaKy8T1Pd9JrBVIuWP52OpWNSVMdYu2rYJeoDYs6WqODKRE4mxWNKHUyHg3oZbY/yumaXJxMv4oB3m1hL1cCKL6tpJBkhv+2hLi4OCxfvhwAMHPmTAQG3r1kj7Woq+95x3p2cGngiOTzTBLINnBOE1EdcZ97EzgrHNDZvanZEiYhBEITTuOfnHj8lXHNsF2lsLdowgQAK1asQHh4OMLDw7Fy5UqLxkJ3s3NVoMfslmgzqSEUjnWn5MS92PsqEPJ0E/iGuFs6FCoHR5qI6ghPO2c81WgAlCbWRNLotVBKcki3ywb0qheM89mxaOUSYKZIzSM2Ntbwc0xM2Uv0kOVosrXIuZkPbYEOds4KaPN1lg7JolRBDnBp4YTC7CIk/Z1p6XDoHpg0EdUhdyZMhboi3FAnV2qCdkR2HA6nXsJA77Zo4VIfANDc2Q/Nnf3MHivVfqeWXQX4kBgAIPucGslxWayYbuWYNBHVQUV6LTbFH0OKJht6oUcrlwCjJiJnFamRpyvE+exYQ9JEVGVMmAy0OXrEnEup+ECyKCZNRHWQUqZAE5UP8nQFyNMW4qvofaUeeXdWOGCAd1s4yx3gJLeD++31/rp4NIOT3A7tLVxhnGoXSQZIcgn6ImZRZN2YNBHVUX3qtYS70gm/J5+/a1+utgA7E8IBAEEqP4yq3wVAcTmB+9yb1GicVLs1HuiDhn29EfVHEuIOp1b8glpGq9fhcNol2PspgRuWjoYqwqfniOooAeBYemSFx9nJFNALffUHRHWSTqOHwkEOt0ZOlg7FIqLUyTidGQWvQa4AS3NZPY40EdVR8flppW7J3Usb10DIJH6/ouqRdDYTmVF5yLmZb+lQLMJV4YjWLgE4duoy53jZAH4SEtVRebpCsx5HVBWaXC1y4vPrbMLg6+CO4X73Ieu02tKhkBGYNBHVUSq5cUupGHscEVFtx6SJqI4KcKwHZ4VDuce4KBwQ4FivhiKiukqpkqPpA75oM7mhpUOpUddyE5FnxC1ysh5MmojqKJkkYYB323KP6e/dlgvHUrUTeqBRfx/4dnCHvbtll9+pKfk6DXYmhGN11B/ILqqb87lsESeCE9VhLZz9MdK/Mw6kRJSaFO6icEB/77Zo4exvweiortDm6xB9IBn5aZo6s5xKrrYAvg7u0AkdXJWOlg6HjGQzSVNUVBRWrFiB06dPIy0tDX5+fhg2bBhmzJgBJ6fKPar6999/Y82aNThz5gzy8vJQr1499OzZE8899xwaNqxbw8NELZz90VzlZ1RFcKLqEvV7kqVDqFHe9q6YHNgbRXqtpUOhSrCJ23Pnz5/H2LFjERoaCi8vL/Tr1w9qtRqrV6/GxIkTkZuba3Rbe/bswaRJk3DgwAE0aNAA/fr1g52dHX799VeMHj0a58/fXeiPqLaTSRIaOnmhlUsDNHTyYsJEVEOUMpsZuyDYQNKk1Woxa9YsqNVqfPjhh9iyZQuWLVuGffv2YcCAAbhy5QqWLFliVFv5+fl49913odfr8dlnn+HXX3/F8uXL8b///Q/Tpk1DXl4e5s6dW81XREQV0euts5imtcZVWyic5PDp4AYn79r9xGZ2kZoFY22U1ae4u3btQlxcHHr06IHx48cbtjs4OOCjjz7CgAEDsHnzZrz88stwc3Mrt62///4b2dnZaNGiBUaMGGHYLpfL8eqrr2L9+vWIjIxEeno6PD09q+2aiKh8MpkMn3y+HXHxaVVuIz09t9TPL772rUkxBQbUwxuvjjapDSpf8JgG8O3gjqh9SbX6dt3OhNPI0qox0q8TAp28LB0OVYLVJ00HDhwAAAwePPiufR4eHujWrRsOHjyIw4cPl0qEyiKXywEA6enp0Gg0sLOzM+zLyMiATqeDUqmEs7OzGa+AiKoiLj4N128kVvn1RVpdqZ9NaYtqRvqVHKh8HaDJrb3zfAp0GmRp1SjQFcHTzsXS4VAlWf3tucjI4rWxgoODy9zfvHlzAMDly5crbKtdu3bw9PREamoqXnvtNdy4cQMFBQU4d+4cXnjhBQDAtGnTSiVTRERUMxLCM/DXkkjcPFb1EUZr5yC3w3NNBmNSYG+oFLX7NmRtZPUjTcnJyQAAX1/fMvf7+PiUOq48jo6OWL58OV599VX8/vvv+P333w37HBwcsGDBAkyYMMGkeAsLC6HTmfeRWUmS4OhovY+k5ufnQ4jatwYC+90y2O+WwX6vWW6wg1qtZr9bEWOexLf6pEmtLl6Px8Gh7MrFJdtLjqtIkyZNMHLkSKxduxYtW7aEn58fIiMjERcXh3Xr1qF169Zo165dleONiIio8mvvxdHREa1btzZ7u+YSFRWF/PzaV5yN/W4Z7HfLsLZ+VzorUHTHbbra0O8CAhJKP5lqbf3+X7Wh343VqVOnCo+x+qRJLpdDr9dDquARaGMy4YSEBEyZMgUZGRn49ttv0aNHD8Nrv//+e3z88cd44oknEBoaCn//qhX1a9u2bbWMNFmzJk2a1MpvIux3y2C/W4a19LuTjz1CnmoCSSbh6IeXDNtrQ7+fyLqOm4Xp6OTaBI0diieAW0u/30tt6HdzsvqkSaVSITMz856ZbkFBcRVjY4Y3P//8c8THx+PNN980JExA8Zt22rRpiIiIQGhoKNatW4c333yzSvHa29e9e9TWPLRcm7HfLYP9Xr0K0jVQqhSABDh4/Lukiq33uxACN5JTkFmUB7lSUemizJZiSr/HxcVh+fLlAICZM2ciMDDQXGFZjNVPBC+Zs5SSklLm/pK5TCXHlefYsWMAgD59+pS5v1+/fgCq5xYbERFVTK8VOLPqOg7P+wcFGUWWDsdsJEnCowE9McC7LZqqKv57VRusWLEC4eHhCA8Px8qVKy0djllYfdJU8tTctWvXytxfsv1eT9fdKSsrCwCgUJQ9wFZSkkCrrb2PuxIRWbucm/nQa2vfLSGVwgEd3ZvUmSrgsbGxhp9jYmIsGIn5WH3S1LdvXwDA3r1779qXkZGBkydPQqlUolevXhW21axZMwD/1n76ryNHjgAAWrVqVdVwiYiIqJay+qRp8ODBqF+/Po4cOYKNGzcathcUFGDu3LlQq9UYP348vLz+rapaVFSE69ev4/r16ygq+nd4d/LkyQCA5cuX49SpU6XO88svv2Dr1q1QKpWYMmVKNV8VERGVx6+jOzo82RgODZQVH2zl4vLTsDvxLGLVqZYOhUxk9WOEDg4OWLhwIaZPn44FCxZg69atCAgIwNmzZ5GcnIzWrVvj9ddfL/WapKQkDB8+HACwf/9+BAQEAAAefvhhXLhwAT///DOmTJmCdu3awc/PD9euXUNUVBSUSiU+/PBDQ8FMIiKyDLcmKtRr6YpsjXHlZKzZxex4XMyJh0KSoSGXTbFpVp80AUC3bt2wZcsWLF++HH/99ReuXbuGgIAAjB8/Hk899VSllj1ZsGAB7r//fvz000+IiIjApUuX4OHhgREjRuDpp5/mrTmiWkKSOUHoip+ulclVFo6GKivxdAbUKYVwyrUHxlk6GtO0d2sIhSRDS5cGlg6FTGQTSRMAtGjRAsuWLTPq2ICAAFy5cuWe+wcNGoRBgwaZKzQiskIKhyBoC64CAOT2HD22NVnRamRFq9GsqZ+lQzGZv4MH/B08LB0GmYHNJE1ERJUhk6tgpwqxdBhEVItY/URwIiKqmyS5BIcGSvyTHWfpUKpEo9fiRHokMjV5lg6FzIRJExERWSXn+g7wfcgD+1MioBN6S4dTadfzknAk7Qp+uXWCS5HUEkyaiIjIKuXE50OTpkVzlR80etsrOuwkt0MjJy+0cgmw+jXmyDic00RERNZJAAlb0vHWZ2MsHUmVNHLyRiMnb44y1SIcaSIiIqpGHGWqPZg0ERGR1cvTFtrULbpYdSqKbCheMg5vzxHVQXFxcVi+fDkAYObMmQgMDLRwRET39lvCaVzOvYUH/TqilQ0UiMzTFmDLzeNQyhR4pvFAOMrtLB0SmQlHmojqoBUrViA8PBzh4eFYuXKlpcMhKpeL0hEAkKbJsXAkxsksUsNF4Yh6di5MmGoZjjQR1UGxsbGGn2NiYiwYCVHFOrk3RWf3ZlAp7C0dilEaOHrimcYDka/TWDoUMjMmTUREZNWcFQ6WDqHSJEmCk40keWQ83p4jIiIyk3ydhiUGajGONBERkdVL1+TiRPpV6IUeI/w7WTqce/o5/hj0QuBBv47wdXCzdDhV5uGugk6vh1xmXWMrlo6JSRMREdmEiznxkEsyaPRa2Mms789XjjYfGUV5EELA7fbkdVulUjlALpNhUdghxGVmVamNdLW61M8zt+80KaZAdzfM7tfXpDZMZX3vOiIiov/wUKrQu14w6jt4QiFZ1+hHCReFI55vMgSJhZlwqCVPzcVlZuFaWlqVXluk15f6uartWBMmTUREZPUkSUJ3zxaWDqNC9nIlGjl5WzoMqibWma4TERERWRkmTUREZDNSCrPxV/o1ZBflWzqUUv5MvYTtt/7Czfx0S4dC1Yi354iIyGbsS76AmwXpsJMpEOLe2NLhAACEEPgnOw55ukK0dW1o6XCoGjFpIiIimxHk7Ac7mcKwtIo1kCQJ4xt0R2RuAhpzPlOtxqSJiIhsRmePZujs0czSYdzF294V3vaulg6DqhnnNBEREREZgUkTERHZnCK9FhmaPEuHgTh1Kg6lXkRKYbalQ6EawKSJiIhsSnReMpbf2IvfEk9bOhRcyI7FqYzruJAda+lQqAZwThMREdkUb3tX6IQe+ToNivQ6KGVyi8US7NIAWqFHK5cGFouBag6TJiIbw4U0qa5TKRzwdOMBcFM4QZIki8bSTOWLZipfi8ZANYdJE5GN4UKaRIC7UmXpEKgOYtJEZKO4kCaR5RTqihCZm4AgZ79aszgvVYxj6UREZJOOpF3GdzEHkVRQtRFXU1zNS8Te5HP4Of54jZ/bVkiubmX+bMtMSprOnz9vrjiIiIgqJbkwG2maXESrk2v83ApJDi87F7Rw9q/xc9sKeadOkPz8Ifn5Q96pk6XDMQuTbs898sgjCAoKwtixYzFy5EjUq1fPXHERERGVq7N7U7RxCUAjJ68aP3dLl/po6VIfeqGv+OA6SnJ1g2LAQEuHYVYmjTSpVCpcvXoVn3zyCfr27Yvnn38e+/btg06nM1d8REREZWro5IVgl/oWnVMkkzjLpS4xaaTp2LFj+P3337F9+3YcP34cBw4cwMGDB+Hh4YGRI0dizJgxCA4ONlesREREFpdcmAVvO1eLlzugmmdSimxvb4+HHnoI3377LQ4ePIhZs2ahadOmSE9Px7p16zB69GiMGzcOGzduRFZWzU/UIyKi2k2tLcSFrFhcyKqZitzZRWr8EPsnvo7ejyK9tkbOSdbDbOOKvr6+mD59Onbt2oVffvkFU6ZMgZ+fH/755x988MEH6NOnD1599VUcPnwYQghznZaIiOqwWwUZ2Jt8DiczrtXI+VI1ObCTKeCqdIJSxqo9dU213Ixt27Yt3n77bezfvx9z5syBo6MjioqKsGfPHkyfPh0DBw7EunXrUFRUVB2nJyKiOqKhkxfqO3jU2KTspipfPN9kCIb5hlT7ucj6VEua/NdffyE0NBT79+9HRkYGhBBQKBTo27cvEhIScPHiRSxatAjbtm3D2rVr+dQdERFViZ1MgUmBvWv0nAqZHG4ypxo9J1kHsyVNN27cwI4dOxAaGoqEhATDLbigoCCMGzcOI0eOhKenJwDgn3/+waxZsxAZGYn33nsPy5YtM1cYRERE1UKr10FhwcWByfJMSprS09Oxa9cu7NixA//88w8AQAgBZ2dnDB8+HOPHj0f79u3vel2bNm2wcOFCPProozh69KgpIRAREUEvBJIKs+Bj7wp5NZQBEELgh7g/oZLbY4hPB3jYce27usikpOn++++HTqeDEAKSJKFLly4YN24chg4dCnt7+3Jf6+VVXIzM2dnZlBCIiIjwfWwY0jS5mNCgBwKrodhlelEu0jW5yJbUUCnK//tGtZdJSZNWq4W/vz9Gjx6NsWPHIjAw0OjXSpKEWbNmsY4TERGZzMfeDTnaAuRoC6ql/Xp2Lnim8UAkF2bBjk/N1Vkm/ct/88036NWrV5UKfAUEBGD69OmmnJ6IiAgA0N+7DYb6hlTLrbkSbkonuCk5AbwuM+nd1bt3b0iShBs3buDzzz+/a/+nn36KN998E5GRkaachoiIqFxOcvtqTZiIADPUadqyZQtGjhyJr776CvHx8aX2nTt3Dtu2bcPYsWOxdetWU09FRERUIXMXUD6Sehl/JJ1HamGOWdsl22NS0nT8+HG888470Gq16NmzJ5RKZan9TzzxBAYNGgStVot58+YZnrAjIsuSXN3K/JmqV1xcHGbPno3Zs2cjLi7O0uHUOrHqVGyKP4q9yefM1qZO6HEuOwbnsmOQo803W7tkm0xKmr7//ntIkoRXX30V3377LXx9fUvtHzBgAJYvX45Zs2ZBq9Xi22+/NSlYIjIPeadOkPz8Ifn5Q96pk6XDqTNWrFiB8PBwhIeHY+XKlZYOp1aKz0/H9bwks402SZDwoG9HhLg1RqNqeCqPbItJE8HPnTuHevXqVTih++mnn8a3336Lv/76y5TTEZGZSK5uUAwYaOkw6pzY2H8XlY2JibFgJLVTA0dPDPJuh8Yq7yo9oFQWmSShscobjVXeZmmPbJtJI025ubnw9/ev8M0pk8kQGBiIzMxMU05HRER0T3JJhhD3xnBXsvAkVQ+TkiZvb2/ExcUZNQyamJgId3d3U05HRERUY2LVqQjPuI7caqr9RLbHpKTpvvvuQ1ZWFjZu3Fjucb/88gtSU1MREhJiyumIiIjKJYTA9dxE7Eu+gEJdkUltnc2KRljqRZzOvGGm6MjWmTSn6dFHH8Xu3buxcOFCJCcn45FHHkFAQIBh/61bt7B161Z89dVXkCQJkydPNjlgqn3i4uKwfPlyAMDMmTMrVVmeiOhOkiThYOo/yCxSo7GTN5o7+1W5rSZO3lBrC9DSuYEZIyRbZlLS1LlzZzzzzDP46quv8PXXX+Prr7+Gk5MTnJycoFaroVarARRn/k888QR69OhhlqCpdil5oggAVq5ciY8//tjCERGRLWvjGohcbQFclY4mtdPerRHauzUyU1RUG5i8gE7J+nHLly9HVFQU8vLykJeXZ9jfoEEDPP/88xg3bpypp6Jaik8UEZE59fBsYekQqJYyy6qDDz74IB588EHExsYiJiYGmZmZcHR0RJMmTdCsWTNznIKIiKhG5Os0uFWQgcZO3lyahUox61LNDRs2RMOGDc3ZpEFUVBRWrFiB06dPIy0tDX5+fhg2bBhmzJgBJ6fKLaCoVquxdu1a7N27F7GxsZDJZGjVqhUee+wxDB06tFriJyKimpVZpEaRXgtve9dKvS4y9xb+SL6Aho5eeCSA00roXzaRQp8/fx5jx45FaGgovLy80K9fP6jVaqxevRoTJ05Ebm6u0W2lpqbi4Ycfxpdffon09HT07t0bwcHBOHPmDF5++WWsW7eu+i6EiIhqxNnMaHwTvR9H0i5X+rV6IeAkt0cTlU81REa2zOSRpqSkJKxduxZnzpxBdnY2dDrdPes2SZKEffv2Vap9rVaLWbNmQa1W48MPP8T48eMBAAUFBXj11Vdx4MABLFmyBO+++65R7c2dOxfXrl3DAw88gMWLF8Pe3h4AcOzYMUyfPh2ffPIJhgwZgvr161cqTiIish71HTwggwS9EBBCVKpC+H3uTdDBrRF0Zl74l2yfSUnTzZs38cgjjyA9Pd2oApdVKWu/a9cuxMXFoUePHoaECQAcHBzw0UcfYcCAAdi8eTNefvlluLmVv/Do+fPnERYWhoYNG5ZKmACgZ8+eGDt2LI4cOYKIiAgmTURENszH3hUvNH0A9nJlxQeXQSbJIDPPSixUi5iUNK1atQppaWlQqVQYM2YMmjVrBgcHB3PFBgA4cOAAAGDw4MF37fPw8EC3bt1w8OBBHD58GCNGjCi3rT179gAApk6dWiphKrFgwQIzRExERJYmSVKVEqbsIjVclZWbJ0t1h0lJ0+HDhyFJEtauXYsOHTqYK6ZSIiMjAQDBwcFl7m/evDkOHjyIy5cvV5g0RUREAABCQkKgVquxd+9eXLhwATqdDu3atcOIESPMnvQREZFl6YWAzIg7HRmaXHwbcxD+Du54NKAXZHxyjv7DpKQpPT0dzZs3r7aECQCSk5MBAL6+vmXu9/HxKXVceaKjowEAGRkZGDFiBG7evGnYt2nTJqxatQqrV69GUFCQiVETEZGl6YUevyWeQYw6FU826geVovwvxQkFmZAgwV6mZMJEZTIpafL09ERRkWlr+1SkpKr4vUaASraXHFeekqfsXn/9ddSvXx8bNmxAq1atEB8fj08//RSHDx/GM888g99++w3Ozs5VirewsBA6na5Kr70XSZLg6GhaZdvqVFBQYNSctnu587VCCOTn55sckynxlLD2frdG+fn5Jve9tfe7Kdf43/e6MZ9bNaW29ntGYS4K9UWIzLyJYCf/co9trPDE4369UagvqrF/G2vvd2tkjs+ZshhTvsikpKlnz57YuXMnYmNjq60+k1wuh16vr3ASuTEdWFhYCACws7PDDz/8YJg43rJlS6xevRpjxoxBZGQkNm3ahKeffrpK8ZbcAjQnR0dHtG7d2uztmsrDXQWdXm/yLc07/23N8QGi1elw8Z9/TE7orbXfrVlUVJTJSa+197sp13jne7KoqAiXLl0yV1gmq6393kBuh4bCC7rsDFxCptGvS6z0marG2vvdGpnjc6YsnTp1qvAYk5KmmTNn4o8//sDs2bOxYsUKeHp6mtJcmVQqFTIzM+/ZQQUFBQBg1B9aBwcH5OXlYfTo0Xc9aadQKDBx4kQsWLAAx48fr3LS1LZt22oZabJGKpUD5DIZFoUdQlxmVpXbSb/jG126Wo2Z23dWua1AdzfM7tcXQUFBZhnxoMpp0qRJre93U65RqVSW+rlVq1bmCstktbnfjVHZsgTmYu39bo2q+71QHpOSpiNHjmD48OHYvHkz+vXrh7Zt28LX17fUB8OdJEnCokWLKnUOHx8fZGZmIiUlBYGBgXftL5nLVDK3qTxeXl7Iy8tDQEBAmftLtmdkZFQqxjuV9VRebReXmYVraWlVfn2RXl/qZ1PaKsHhbsuoC/1uyjX+d1S1sqsZ1GXV+d4SQmBd7CF427min3drOFcw94ksy5KfMyYlTfPmzTN8CGg0Gpw5c6bM4yRJMmTxlU2agoODERkZiWvXrqFjx4537b927ZrhOGPaiomJQVJSUpn7U1JSAKBaRsyIiMgy0jQ5uJxzC+5KJ7RxvfvLd2JhJtI0OcguUuMB3+p7sIlsn0lJ0+jRo6t9aLFv374IDQ3F3r178cgjj5Tal5GRgZMnT0KpVKJXr14VttWvXz/8/vvv2LNnD1544YW7RsT+/PNPAEDXrl3NdwFERGRRceo0HE+PRICDZ5lJk5+9OyYH9kaGJg9KmdwCEZKtMClpWrhwobniuKfBgwejfv36OHLkCDZu3IjJkycDKJ7LNHfuXKjVajz66KPw8vIyvKaoqAixsbEAihcRLkmOhg8fjpUrVyI6OhoLFizAu+++a9i3ZcsW7N27F25ubhg3bly1XxcREdWMJioftMj3R1NV2aVrJEmCv4MH/B08ajgysjUmrz1X3RwcHLBw4UJMnz4dCxYswNatWxEQEICzZ88iOTkZrVu3xuuvv17qNUlJSRg+fDgAYP/+/Ya5So6Ojvjiiy/w9NNPY/PmzQgLC0OHDh0QExODyMhI2NvbY+HChahXr16NXycREVUPN6UTRvp3tnQYVAuYrXrX9evX8e2332Lu3Ll46aWXABTXTtq2bZvJj35369YNW7ZswQMPPIBbt24hLCwMLi4ueP7557F+/fpK1VRq27YtQkNDMXXqVNjZ2SEsLAwZGRl48MEHsXnzZgwYMMCkWImIyHYcS7uCY2lXkFNk/kfYqfYxeaQpPz8f8+fPR2hoKMR/VpOOj4/Hm2++iRUrVmDt2rUm1XJq0aIFli1bZtSxAQEBuHLlyj33e3t74+2338bbb79d5XiIiMi2FOg0iM9PRzOVLyRJglavw+nMGyjUaxHoWA8uytr/9CeZxqSRJr1ej+effx47d+6EJElo3bo1PDz+vSdcWFgIhUKB+Ph4TJ48Genp6SYHTLWP5OpW5s9E1qSkmKu1scaYrJFWr8PqqH3YnnAKGUV5hu0DvNuilUsDBDhyWgZVzKSRpq1bt+L48eNo3LgxvvzySwQFBWHSpEk4e/YsAKBdu3bYs2cPpk+fjujoaKxbtw6zZs0yS+BUe8g7dYIuXBh+JrJG5ijmas5CrsC/xVypYgqZHPUdPJCrK4BaVwhPOEMhk6ONa2CZT9QRlcWkpGn79u2QJAlLly695yK3gYGB+OKLLzBq1CgcPHiQSRPdRXJ1g2LAQEuHQWQUU4q5VkchVzLemPpdWVKATGJS0hQZGYmGDRuiZcuW5R7XokULNGrUCHFxcaacjoiIqMqUMjn0QiA+Pw2x+amAADp7NIODvOxVLIj+y6SkqbCw0OhlAFQqlcXWiiEiIorMTcCBlAjkagsM285kRmGoXwhaOPtbMDKyFSZNBPfz80N0dDQ0Gk25x6nValy/fh2+vmUXFiMiIqpOkbkJ2JkQXiphAgCN0GJnQjgicxMsFBnZEpOSpp49e6KgoABr1qwp97hly5ahsLAQPXr0MOV0RERElaYXAgdSIso95mBKBPS8G0IVMOn23FNPPYXt27dj1apVyM3NxUMPPWQYdSooKEBkZCTWr1+P3377DQqFAo8//rhZgiYiIjJWfH7aXSNM/5WjLUB8fhoaOnmVexzVbSYlTYGBgVi8eDFee+01/PDDD/jhhx8M++677z4AgBACcrkcCxYsQNOmTU2LloiIqJLydIVmPY7qLpOXURk8eLBh+RGlUmmoCi6EgEwmQ48ePbB+/XqMGTPGHPESERFVikpub9bjqO4yy4K9LVu2xIoVK6DRaBATE4OcnBw4OTkhMDAQKpXKHKcgIiKqkgDHenBWOJR7i85F4cCq4FQhk0aatm/fjsOHDxv+287ODkFBQejYsSNatmxZKmH69ddfsWTJElNOR0REVGkyScIA77blHtPfuy1kt9dNJboXk5KmOXPmVPjkXImNGzdiw4YNppyOiIioSlo4+2Okf2c4KxxKbXdROGCkf2fWaSKjGH17LjU1FVevXr1re3Z2No4fP17ua2/evImrV69CoTDL3UAiIqJKa+Hsj+YqP8TnpyFPVwiV3B4BjvU4wkRGMzqLUSqVeOWVV5CdnW3YJkkSrl69iieffLLC1wsh0KVLl6pFSUREZAYySWJZAaoyo2/Pubm54bnnniv1dByAUv9d1v8AwMnJCV26dMH8+fOr5SKIiIiIqlul7pdNmzYN06ZNM/x3y5Yt0alTJ2zcuNHccRERERFZFZMmGY0ZMwZNmjQxVyxEREREVsukpOnjjz82VxxEREREVs0sj7MlJSUhOzsbQUFBhm3r1q3Dzp07odPp0K9fP8yYMQNOTk7mOB0RERFRjTN5GZVly5Zh4MCBWLt2rWHb6tWrsWjRIly8eBFXrlzBV199hSeffBI6nc7U0xERERFZhElJU1hYGFauXAmtVouCguLy9BqNBt988w0AoH///pg9ezb8/Pxw7tw5bN682fSIiYiIiCzApKTpl19+gSRJmDVrFj7//HMAwPHjx5Gbm4t69eph+fLleOKJJ/DVV18BAHbv3m16xERENkpydSvzZyKyDSbNaTp37hw8PT3xzDPPGLaVrEXXt29fyOVyAEBQUBAaNmyIyMhIU05HRGTT5J06QRcuDD8TkW0xKWnKzMxEy5YtId1Rgv7YsWOQJAndunUrdayzszNu3rxpyumIiGya5OoGxYCBlg6DiKrIpNtz7u7upZZVSUxMxI0bNwDgrqQpISEBLi4uppyOiIiIyGJMSpoaN26M2NhYXLt2DQCwc+dOAECLFi3g6+trOG7Hjh1IT09HcHCwKacjIiIishiTbs8NGzYMp06dwuOPP4777rsPYWFhkCQJY8aMAVA88vTNN99g06ZNkCQJo0ePNkfMRERERDXOpJGmiRMnYsiQIUhLS8O+ffug1WrRpUsXTJkyBUBx0csNGzZAq9Xi4YcfZtJERERENsukkSaZTIZly5bh8OHDuHz5Mho3bowBAwYYnppr0qQJBg0ahFGjRmHw4MFmCZiIiIjIEsyyjEqfPn3Qp0+fu7a7urpi+fLl5jgFERERkUWZvIwKERERUV1g9EjTG2+8AUmS8H//93/w8vIybKsMSZKwaNGiykVIREREZAWMTpp27twJSZLw7LPPGpKmkm1CiHJfW3IMkyYiIiKyVUYnTaNHj4YkSaUKVJZsIyIiIqrtjE6aFi5caNQ2IiIiotqIE8GJiIiIjGCWkgMajQaZmZnw8fExbNu3bx927NgBnU6Hfv36Yfz48ZDJmKMRERGRbTI5i/nll1/Qu3dvLFu2zLBty5YtePHFF7Fv3z4cOHAA8+bNw0svvWTqqYiIiIgsxqSk6cyZM3jnnXeQnZ2NxMREAIBOp8PSpUsBAG3atMGUKVPg7OyM/fv347fffjM5YCIiIiJLMClp2rhxI4QQmDJlClauXAkAOHXqFNLS0uDi4oL169fj7bffxqpVqyCEwI4dO8wSNBEREVFNMylpOnv2LNzc3PDGG2/Azs4OAHD48GEAwP333w9HR0cAQOfOnVG/fn1cvHjRxHCJiIiILMOkpCk1NRWBgYGGhAkAjh8/DkmS0K1bt1LHenp6Iisry5TTEREREVmMSUmTs7Mz8vPzDf+dnp6OS5cuAQC6d+9e6tjk5GSoVCpTTkdERERkMSYlTQ0bNkRMTAySkpIAAHv37oUQAg0bNkRgYKDhuMOHDyM5ORnNmjUzLVoiIiIiCzGpTtOAAQPw999/44knnkCfPn3wyy+/QJIkPPTQQwCAjIwMbN++HcuXL4ckSRg2bJhZgiYiIiKqaSYlTY8//jgOHDiAv//+G1FRURBCICgoCE8++SQA4Pr164YFevv27YtHH33U9IiJiIiILMCkpMne3h4//PADfv31V1y+fBmNGzfG+PHj4eTkBABo2rQpWrVqhVGjRuGxxx5jRXAiIiKyWSYvo2JnZ4eJEyeWuc/T0xPbtm0z9RREREREFsehHyIiIiIjmGXB3suXL+PMmTPIycmBVquFEOKex86cOdMcpyQiIiKqUSYlTVqtFrNnz8bu3buNfg2TJiIiIrJFJiVNGzZswK5duwAAbm5uaNSoEezt7c0SGBEREZE1MSlp2rFjByRJwvTp0/Hyyy/z6TgiIiKqtUxKmqKiolCvXj288sorkCTJXDERERERWR2ThoaUSiV8fHxqJGGKiorC66+/jv79+6N9+/YYMmQIPv/8c6jVapPbXrRoEYKDg/Hll1+aIVIiIiKqjUxKmlq2bIno6GhoNBpzxVOm8+fPY+zYsQgNDYWXlxf69esHtVqN1atXY+LEicjNza1y20ePHsV3331nxmiJiIioNjIpaXrsscegVquxcuVKc8VzF61Wi1mzZkGtVuPDDz/Eli1bsGzZMuzbtw8DBgzAlStXsGTJkiq1nZ6ejtmzZ5dbIoGIiIgIMHFOU9++ffHkk09izZo1uHr1Kvr16wdfX18olcp7vqZHjx6VOseuXbsQFxeHHj16YPz48YbtDg4O+OijjzBgwABs3rwZL7/8Mtzc3CrV9ltvvYWMjAx07NgRZ86cqdRriYiIqG4xKWnq0KGD4ecDBw7gwIED5R4vSRIuXrxYqXOUtDl48OC79nl4eKBbt244ePAgDh8+jBEjRhjd7saNG3Hw4EG8/PLLyMjIYNJERERE5TLp9pwQolL/0+v1lT5HZGQkACA4OLjM/c2bNwdQXJXcWFevXsWiRYvQsWNHzJgxo9IxERERUd1j0khTZRKVqkpOTgYA+Pr6lrnfx8en1HEVKSwsxKxZs6BUKrF48WLI5XLzBEpERES1mlnWnqtOJSUFHBwcytxfst3Y0gOffPIJIiMjsWjRIgQEBJgnyDsUFhZCp9OZtU1JkuDo6GjWNmu7/Px8kyf4s98rj/1uGXWh381xjdbI2vvdGlXXe8HJyanCY8yWNOXn5+Ovv/5CVFQUcnNz4ezsjEaNGqFLly5wdnaucrtyuRx6vb7CWlDGdGBYWBg2bNiA4cOHY/To0VWOqTwRERFmb9PR0RGtW7c2e7u1WVRUFPLz801qg/1eeex3y6gL/W6Oa7RG1t7v1qi63gudOnWq8BizJE3ff/89VqxYgZycnLv2OTg44IUXXsDTTz9dpbZVKhUyMzPv2UEFBQUAUGGmnpqaijfffBP+/v547733qhSLMdq2bVstI01UOU2aNDHLN2+qHPa7ZdSFfjfHNVoja+93a2TJ94LJSdPChQvx/fffQwgBpVKJJk2awNnZGdnZ2YiOjkZ+fj4+++wzJCUlYe7cuZVu38fHB5mZmUhJSUFgYOBd+0vmMpXMbbqXlStXIj09Ha1atcKCBQtK7fvnn38AAL///jtiYmLQrFkzPPfcc5WOFQAXLLYSHO62DPa7ZdSFfq8L10jGseR7waSk6cSJE1i3bh0UCgVeeuklTJ06tdTF5ObmYsOGDVi+fDk2bNiAIUOGoEuXLpU6R3BwMCIjI3Ht2jV07Njxrv3Xrl0zHFeekjlPly5dwqVLl8o8JjIyEpGRkejatWuVkyYiIiKqnUwqOfDjjz9CkiS89dZbmD59+l3Zn7OzM5599lm89dZbEELg559/rvQ5+vbtCwDYu3fvXfsyMjJw8uRJKJVK9OrVq9x2Fi5ciCtXrpT5v8ceewwAMHPmTFy5cgXr16+vdJxERERUu5mUNJ05cwYeHh549NFHyz3u0UcfhYeHR5UKSA4ePBj169fHkSNHsHHjRsP2goICzJ07F2q1GuPHj4eXl5dhX1FREa5fv47r16+jqKio0uckIiIi+i+Tbs9lZmaidevWFU5kkyQJgYGBVarr5ODggIULF2L69OlYsGABtm7dioCAAJw9exbJyclo3bo1Xn/99VKvSUpKwvDhwwEA+/fvr5bSAkRERFS3mDTS5OLigsTERKOOTUxMhEqlqtJ5unXrhi1btuCBBx7ArVu3EBYWBhcXFzz//PNYv369SSUNiIiIiIxh0khTmzZtcPToUfzxxx9lrg1XYu/evUhOTkbv3r2rfK4WLVpg2bJlRh0bEBCAK1euGN323Llzq/RkHxEREdUdJo00PfzwwxBCYM6cOdi9e3eZx+zevRtvvvkmJEnC+PHjTTkdERERkcWYNNL0wAMPYNCgQdi3bx9ee+01fPzxx2jdujWcnZ2Rm5uLixcvIjU1FUIIDBo0CEOHDjVX3EREREaJi4vD8uXLARQ/JV1WzT8iY5g00gQAS5cuxeTJkyGXy5GSkoJDhw5h165dOHToEFJSUiCXyzFp0iQsWbLEHPESERFVyooVKxAeHo7w8HCsXLnS0uGQDTO5IrhCocA777yD6dOn488//8SNGzeQm5sLlUqFpk2b4v7774efn585YiUiIqq02NhYw88xMTEWjIRsndkW7PX19cXDDz9calt8fDzc3d3NdQoiIiIiizH59pxWq8Xnn3+O/v37o7CwsNS+Tz75BD179sRnn30GjUZj6qmIiIiILMakpEmj0eCpp57CV199hcTERERFRZXan5KSArVajW+++QYzZ840KVAiIiIiSzIpaVq/fj1OnjwJDw8PLF68GM2bN79r/4oVK+Dl5YXDhw9jy5YtJgVLREREZCkmJU27du2CTCbD119/jREjRkChKD1FSqFQYODAgVixYgUAYOvWraacjoiIiMhiTEqaoqKi0LhxY7Rp06bc49q3b4+AgABERkaacjoiIiIiizEpaZIkCXZ2dkYd6+LiAr1eb8rpiIiIiCzGpKQpICAA165dQ3p6ernHZWVl4dq1a/D39zfldEREREQWY1LSNHDgQGi1Wrz11lv3LCmg1WrxzjvvoKioCH379jXldEREREQWY1Jxy8mTJ2Pz5s04dOgQhg8fjlGjRqFly5ZwcnJCXl4eIiMjERoaitjYWLi6uuLJJ580V9xERERENcqkpMnLywvLli3Dyy+/jPj4+DLX9BFCwN3dHcuXL4ePj48ppyMiIiKyGJOXUenUqRN2796NTZs2ISwsDLGxscjMzISjoyMaN26Mvn37YtKkSfD09DRHvEREREQWYZa151xdXTF9+nRMnz7dHM0RERERWR2T154jIiIiqguYNBEREREZgUkTERERkRGYNBEREREZgUkTERERkRGYNBERkdXycFdBZ4XrllpjTFT9zFJygIiIqDqoVA6Qy2RYFHYIcZlZVWojXa0u9fPM7TtNiinQ3Q2z+3FZsLqISRMREVm9uMwsXEtLq9Jri+4YFSrS66vcDhFvzxEREREZgUkTERERkRGYNBEREREZgUkTERERkRGYNBEREREZgUkTERERkRGYNBEREREZgUkTERERkRGYNBEREREZgUkTERERkRGYNBEREREZgUkTERERkRGYNBEREREZgUkTERERkRGYNBEREREZgUkTERHVapKrW5k/E1WWwtIBEBERVSd5p07QhQvDz0RVxaSJiIhqNcnVDYoBAy0dBtUCvD1HREREZAQmTURERERGYNJEREREZAQmTURERERGYNJEREREZAQmTURERERGYNJEREREZAQmTURERERGYNJEREREZAQmTURERERGYNJEREREZAQmTURERERGsJkFe6OiorBixQqcPn0aaWlp8PPzw7BhwzBjxgw4OTlVqq2wsDBs2LABERERyMnJgbu7Ozp27Iinn34aHTp0qKYrICIiIltmEyNN58+fx9ixYxEaGgovLy/069cParUaq1evxsSJE5Gbm2t0W0uWLMGMGTNw5MgRNGjQAP369YOrqyt+//13PProo9i2bVs1XgkRERHZKqsfadJqtZg1axbUajU+/PBDjB8/HgBQUFCAV199FQcOHMCSJUvw7rvvVthWeHg41qxZA0dHR6xZswbdunUz7Nu0aRPmzZuHd999F926dUP9+vWr7ZqIiIjI9lj9SNOuXbsQFxeHHj16GBImAHBwcMBHH30EJycnbN68GVlZWRW29csvvwAAnn766VIJEwBMnDgRffv2hUajwd69e817EURERGTzrD5pOnDgAABg8ODBd+3z8PBAt27dUFRUhMOHD1fYloODA1q0aIHu3buXub9p06YAgKSkJBMiJiIiotrI6pOmyMhIAEBwcHCZ+5s3bw4AuHz5coVtzZ8/H6GhoejcuXOZ+8+dOwcA8PPzq0qoREREVItZfdKUnJwMAPD19S1zv4+PT6njqurAgQM4c+YMlEplmaNaREREVLdZ/URwtVoNoPjWWllKtpccVxVXrlzBm2++CaB4vlODBg2q3FZhYSF0Ol2VX18WSZLg6Oho1jZru/z8fAghTGqD/V557HfLYL9bBvvdMszR72UxpnyR1SdNcrkcer0ekiSVe1xVO/D8+fOYPn06MjMz0b9/f7z00ktVaqdERESESa8vi6OjI1q3bm32dmuzqKgo5Ofnm9QG+73y2O+WwX63DPa7ZZij38vSqVOnCo+x+qRJpVIhMzPznh1UUFAAAFXK1P/3v/9hzpw5yM/Px6BBg/D5559DJjPtjmXbtm2rZaSJKqdJkyZm+QZIlcN+twz2u2Ww3y3DHP1eVVafNPn4+CAzMxMpKSkIDAy8a3/JXKaSuU3GWrFiBb788ksIITBp0iS8/fbbkMvlJsdrb29vchtkOg53Wwb73TLY75bBfrcMS/a71U8EL3lq7tq1a2XuL9l+r6fr/kuv12POnDlYtmwZZDIZ5s6di3nz5pklYSIiIqLay+qTpr59+wJAmQUnMzIycPLkSSiVSvTq1cuo9t5++21s27YNTk5OWLVqFR577DGzxktERES1k9UnTYMHD0b9+vVx5MgRbNy40bC9oKAAc+fOhVqtxvjx4+Hl5WXYV1RUhOvXr+P69esoKioybN++fTu2bt0KhUKBlStXGhIyIiIioopY/ZwmBwcHLFy4ENOnT8eCBQuwdetWBAQE4OzZs0hOTkbr1q3x+uuvl3pNUlIShg8fDgDYv38/AgICoNPpsHTpUgBAvXr1sHXrVmzdurXMc/bu3RujR4+uzssiIiIiG2P1SRMAdOvWDVu2bMHy5cvx119/4dq1awgICMD48ePx1FNPwdnZucI2rly5goSEBADFSVVoaOg9j3V1dWXSRERERKXYRNIEAC1atMCyZcuMOjYgIABXrlwpta1169Z3bSMiIiIyltXPaSIiIiKyBkyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICEyaiIiIiIzApImIiIjICApLB2CsqKgorFixAqdPn0ZaWhr8/PwwbNgwzJgxA05OTpVqKykpCStXrsSxY8eQmJgILy8vDBgwAC+88AI8PT2r6QqIiIjIltnESNP58+cxduxYhIaGwsvLC/369YNarcbq1asxceJE5ObmGt1WXFwcxo0bh02bNsHBwQH9+/eHXC7Hhg0bMGbMGCQmJlbjlRAREZGtsvqkSavVYtasWVCr1fjwww+xZcsWLFu2DPv27cOAAQNw5coVLFmyxOj25syZg5SUFLzwwgsIDQ3FsmXLsHfvXkycOBGJiYmYN29eNV4NERER2SqrT5p27dqFuLg49OjRA+PHjzdsd3BwwEcffQQnJyds3rwZWVlZFbYVHh6O8PBwNG7cGDNnzjRsl8vlePvtt1G/fn2EhYXh2rVr1XItREREZLusPmk6cOAAAGDw4MF37fPw8EC3bt1QVFSEw4cPG93WwIEDIZOVvnSlUokBAwYAAPbv329q2ERERFTLWH3SFBkZCQAIDg4uc3/z5s0BAJcvXza6rRYtWpjcFhEREdUtVv/0XHJyMgDA19e3zP0+Pj6ljquptspy5coVFBYWVum1FZEkCRPGtIVW16pa2q8KezslLly4gAkN/KHzL7tPa5pcJsOFCxcghDBLe+x347DfLYP9bhnsd8swd7//l729/T0HaEpYfdKkVqsBFM9hKkvJ9pLjjGnL0dHR5LbuRZKkKr+2Im5ulSutUFPc7/FvY0nm/HdgvxuP/W4Z7HfLYL9bRnX+na2I1SdNcrkcer2+wk4yJvOUy+UAKu7wqmaxFWWoREREZLusfk6TSqUCAOTn55e5v6CgAMC9R4+qqy0iIiKqW6w+aSqZZ5SSklLm/pL5RyXH1VRbREREVLdYfdJUcsvrXrWTSrYbc2vMnG0RERFR3WL1SVPfvn0BAHv37r1rX0ZGBk6ePAmlUolevXoZ3dYff/wBvV5fal9RUZGhPlP//v1NDZuIiIhqGatPmgYPHoz69evjyJEj2Lhxo2F7QUEB5s6dC7VajfHjx8PLy8uwr6ioCNevX8f169dRVFRk2N6xY0e0a9cO169fx5IlSwwTvnU6HT788EMkJCTg/vvvR6tW1vPYJxEREVkHSVRXwQMzOnnyJKZPn46CggK0adMGAQEBOHv2LJKTk9G6dWusX78ezs7OhuPj4+MxcOBAAMXVvQMCAgz7rl69iilTpiAzMxNNmzZFUFAQLl26hNjYWDRo0AA//fTTPes4ERERUd1l9SNNANCtWzds2bIFDzzwAG7duoWwsDC4uLjg+eefvythqkhQUBB+/fVXjB07Fjk5OTh48CAAYOrUqdi8eTMTJiIiIiqTTYw0EREREVmaTYw0EREREVkakyYiIiIiIzBpIiIiIjICkyYiIiIiIzBpIiIiIjICkyYiIiIiIzBpIrNh9Yqaxz63DPY7Ud3EpIlMVvIHRJIkAEB6erolw6kTStZOlCQJer0eUVFRFo6obmC/E9VtCksHQLXHnj17sHXrVqhUKjz77LOl1vDT6/WQJMmQWJFpSvpx48aN2LlzJ2QyGV577TV07tzZwpHVbux3qsv0ej1kMhmuX7+OsLAw+Pj4oFevXvD09LzrmNqq9l4Z1QidTgdJkvD3339jwYIFEEIgMDAQTZo0QUZGBg4cOIDIyEjIZDImTGZS0ufHjx/Hl19+iQYNGmDEiBHo3Lkz4uLi8Msvv+DcuXO8hWRm7Heqy3Q6HWQyGa5evYpXXnkFq1atwq5duwz7k5KSAKBWJ0wAl1EhMxk3bhyCgoLwzjvvQKVSISIiAq+99hri4uKg1+sxe/ZsTJs2DQCYPJnJyJEjcd9992HevHmQyWQ4fvw4Xn/9daSlpQEAXn75ZTz33HMWjrL2Yb9TXTZhwgR4enpi1qxZCAoKQkpKCj755BNcuHAB9evXx1tvvYXmzZtbOsxqU7tTQqoRly9fRm5uLsaPHw+VSoUjR47g2WefhUqlwptvvokhQ4Zg9+7dAJgwmcv169chSRLGjBkDmUyG//3vf3jllVfQvHlzrFy5ElOnTsW2bdtQWFho6VBrFfa75ZTMJyPLOX36NBITE/Haa68hKCgIsbGxeOaZZ3DkyBEEBgYiOjoaW7ZssXSY1YpzmqjKSu5d+/n5wdHREQcOHMClS5ewfv16ODs7Y/Xq1fDx8YGLiwu+/vprxMTEoHHjxpYOu1bw9vaGRqPB2rVr0apVK2zYsAENGjTAqlWr4OTkhLy8PPz555+4efMmmjZtaulwaw32e824c15MXl4eVCpVrb/tYwvkcjkcHR2RkpKCjIwMfPHFF0hMTMTq1asREhKCd955B5cuXUJubi5UKlWt/JLMpIkqRafT4ciRIwgJCYGbmxvS09Ph6emJjh07YufOncjMzETXrl3x2muvwcfHBxqNBteuXQMAJkxm5OrqisceewwbNmzAgQMHMGzYMDz77LNwcnKCWq3GxYsX4eDgwD/cZsZ+r35CCEOCtGnTJmzZsgUtWrTA//3f/5WacEw1z9/fH3q9Hh999BHi4+Ph6emJRYsWISQkBHq9Hq6urigqKoKzs7OlQ602TJqoUi5fvoyvvvoK3bt3R8uWLfHiiy8iLCwM7777Lrp37w57e3v06dMHWq0WZ8+exeXLl7FlyxbMmzfP0qHbtJJv3jqdDqmpqYiMjES3bt0QHBwMJycntGzZEunp6di7dy9u3LiBX375BYsXLwZQnOjK5XILX4FtYr/XPL1eD7lcjm+++Qbff/89WrRoge7du8PT0xM3btxASkoKXF1dSz2dS9VPCAFfX18sW7YMX331FXr37o1Ro0ahZcuWyMvLw8WLF7F161b83//9H4Da+/5n0kSVEhQUhIYNG2LFihVwd3dHo0aNkJGRAT8/PwwZMsRw3FtvvYW9e/fC09MTDz30EIYPH27BqG1fyTfvzz//HH/88QdiYmIwYcIEvPfee4ZjVqxYgR9//BGNGjXC448/jn79+gFArfzgqins95olhIBcLkdubi7WrFmDN954AyNGjICjoyO2b9+Ozz77DCkpKQgODsbSpUvRpEkTo9uu7Y/CV7fExERotVq0bNkSS5YsMWz/9NNPceLECWRmZuK+++7DuHHjANTe9z+fnqMqWbduHRYuXAgA6NmzJ8aNG4fu3bujXr16AIDz58/j0qVL6NSpExo0aABHR8dy2xNCQJIkw/+nf5V8Y9u3bx/efvttPPXUU+jbty98fHzg7u6O9PR0XLlyBRqNBiqVCv7+/vDx8YFSqeQfChOYu9+1Wi0UCn5PNcbOnTvx9ddf4/vvv4eTkxN27dqFefPmoVevXhg3bhzmzZuHgQMH4oMPPrB0qLVWyfs/NjYW69atw/79+5GRkYHAwEB0794djz76KAIDA7FkyRJcvnwZI0aMMPx+1NZRJoBJE1WSTqcDABw5cgSXLl2Cr68vFi9ejJycHDz00EMYOXIkOnbsCDs7O8TExKBRo0b3TITu3F5QUAAHB4cavRZbM3XqVDRp0gTz5s2DXC439N+NGzcwYcIETJ48Ga+88goAMPk0I3P0+/bt2/H333/jqaeeQkBAAP9tynBn350/fx7Tpk3D6tWr8ddff2HdunXo2bMnPvroIzg7O2PevHnIyMjA4sWLYW9vX267er0eJ06cwB9//IHExER07NgRTZo0QY8ePaBSqWri0mzaI488Arlcjg4dOmDAgAH45JNPEB8fj/Xr10Mmk6FZs2Z16ssZv/aQ0UqGzgGgb9++6Nu3LwBgzJgx+PTTT7F27VocO3YMw4YNQ1JSEg4fPoyjR4/Czs7urrZKvonk5ORg69atOH78OAYPHowRI0YYkqf/Ls9SlxUWFkKpVMLOzu6ub3BNmzbFhAkTcPDgQbz00kssJGpG5ur3mJgYbNq0CcePH8cTTzyBIUOGcFLzf5T03d69e9GrVy/06NEDjz32GABg1KhReOedd+Ds7Izc3FxER0fD39+/woQJAJYuXWooedKsWTN88cUX8PT0xMqVK9G4cWM4OzvzS8Z/lPTH7t27ER0djXXr1qF169YAim9ZDxo0CCkpKfj000+xYMECtG3b1sIR15y6kRqSyUqWQdHpdNi9ezdWrFiBn376CREREQCA119/HQcPHkTbtm3x3XffITw8HC+99FKZCRPw71yR+fPnY9OmTcjMzISPjw8cHBwQHx9vqIdTcsuursnNzS3130qlEkqlEgcPHjQUUSxZ/wwAmjdvDp1Oh5s3b9Z4rLVJdfX7yy+/bHgK7L333sNLL72EP//8E/n5+dVzITampD937tyJOXPm4NKlS3jrrbewfv16rFu3DosWLQIAxMXFYdWqVYiIiMCrr75aYbtRUVFYu3Ytnn/+efz0009Ys2YNhg4dioYNG0Kr1WLNmjXQarVMmG777xfVqKgoNGvWDPXr1wcAbNmyBVevXsUzzzyDli1bIioqCn///belwrUMQWQErVYrhBBi0aJFomvXrqJVq1aiZcuWYsKECeLbb78V8fHxhmMvXbokzp49e8+2dDqdEEKIY8eOifbt24ujR48a9n3xxReib9++Ijg4WLz66quisLCwei7IiiUnJ4vRo0eX6lMhhPjrr79E165dxcyZM8WxY8eEWq0WQgiRnp4u5s+fLx5++GFLhFtrVEe/37p1S3z33XciLi5OCCFEUVGR+P3338WYMWNE69atxRtvvCEuXrwoioqKqu/CrJxerzf8/PHHH4uPP/641DYhivtt5MiRIjg4WIwYMUJs2LDBqLa//PJLMX78eJGcnCyEEOLatWuiZcuW4o8//hBXr14Vbdq0EZs2bTLfxdQymzZtEm3atBHXrl0TQgjRtWtX8eWXXwohhEhLSxPjxo0Tn3/+uQUjrHm8PUdGkcvliI6Oxvr16zF79myMGjUKJ06cwKpVq7BmzRqcOnUKw4YNw8CBA9GyZcty2yoZZdqzZw8GDRqEnj17IiMjAxs3bsTKlSsxbNgwtGjRAmvWrMGff/6JQYMG1cQlWo20tDR07twZDRo0QEJCAm7evIn27dujS5cueO6557BkyRKcP38enTp1QmBgIC5cuICIiAgsX74cQO191Le6VUe/r1u3Dt9//z2OHDmCsWPHolevXhg8eDC6dOmCbdu2Ye3atQgLC8PUqVMxduxY+Pv719lRj927d+Pw4cNo0KCBoQ+KioqgUCigVqvxyiuvID8/H23btkXDhg2NatPZ2Rm3bt0ytDd//nz06dMHAwcORHp6Oho0aICsrKxquyZbsnLlSnTp0gVdunQxbGvVqhWcnJywdu1aAIC7uzsef/xxAMCtW7cQFxeH++67D0AdejrR0lkb2Y6TJ0+KGTNmiOzsbMM2nU4n1qxZI/r16ycGDBggXn75ZXH58uVy2yn5Fvnll1+KBx54QERERIh33nlHtGnTRixdulTodDqRlpYmHnroIfH1119X6zVZq5KRvdmzZ4vg4GDx0UcfGUYrbt68KWbNmiWGDBkievfuLZ5//nnx+++/CyHEXd/QqXKqo9937NghevXqJdq0aSPmzJkjTp48aRhBjYqKEvPnzxft2rUTDz30kPjpp59Efn5+NV+l9dFoNOKtt94S/fv3F8HBwWLJkiVmaffw4cMiJCREhIaGij179og2bdqIixcvCiGESEpKEg8++KD47rvvzHIuW5aeni7GjBkjDh06JIQQIjEx0bDvl19+EZ07dxbBwcFi3rx5IjY2Vmzfvl08/vjjYtKkSZYK2WKYNFG5Sv4Y/PXXX+L9998Xo0aNMvxCaTQaw3GJiYlizpw5olOnTqV+4cpz/PhxMXToUNGtWzfRsWNH8cEHHxj2xcXFiX79+olff/21VBx1TUxMjFi6dKno1KmT6NOnj/juu+8Mf3CTk5OFWq0udQuzrvaTuZmj30NDQ0v99yeffCJatWolevXqJb744gtx5coVw77w8HAxdepU0bFjxzp5S1oIIbKyssSOHTvEk08+KYKDg8W4cePEvn37DPt1Op3h1r6xcnNzxXPPPSfatGkjWrVqJebPny+EECIlJUUsWbJEdO3a1XBrtK7/7kRFRQkhhLhw4YJo166dWLRokcjJyRFCCPHnn3+KmTNniuDgYNGjRw8REhIiZs2aJWJjY4UQ/37ZqAtYcoCM8umnn+K7774DAMydOxePPvqoYWI48G8hs5s3b6JBgwb3bEf85ymVU6dOITIyEkFBQejatSuys7MRHR2Nn3/+GeHh4di7d281XpXtuHz5MlavXo3ff/8drVq1wnPPPYfevXuzTEM1q2q/b9q0CT///DM2bNhgePoOAJKSkvDee+/hwIEDaNmyJSZOnIj+/fvD19cXhYWFSEhIqPPLDaWlpWHXrl3YsWMHYmNj0aNHD7zwwgsIDg426vVqtRr/93//h5kzZxqqhi9atAi7d++GRqNBz549cfbsWTg5OeHpp5/G6NGjWUPrDqdOncLXX3+Ny5cvw8nJCTNmzMCYMWMAANHR0fj777/RokULNG3aFA4ODhU+eVjbbtsxaSKjJCUl4dy5c/j5559x9OhR9OvXD6+88oph/lJFHzr/XY7i+PHjkMlkaNKkCdq1awcAuHLlCl5++WXcunULTZs2xZtvvolu3brVqQ+0kn4qKChAZGQkCgoKoNfr0b17dwDAwYMHsXr1aly8eBG9evUq9W9AVWfufr9x4wbS09PRuXNnfP3114iMjMSzzz6LZs2aAQBOnDiBDz74ANeuXcOAAQPw0EMPoW/fvnBycqqR67Um0dHROHbsGKKjo9GhQwc0a9YMLVu2xOXLl7Fjxw6EhYUhKioKn3/+OYYNG1Zhe6dPn8aLL74IvV6P8ePHY8aMGXBxccHp06fx+++/4/r162jVqhUGDRqEDh061MAV2p709HT8/vvv2L17Ny5evIjWrVvjhRdeQLdu3SwdmsUxaaJKSU9Px+7du/HDDz/g1q1bmDRpEp599lmja84sWbIEv/32G27dugWFQgEXFxd06NABL730EgICArB27Vo0bNgQ7du3R/Pmzav5aqxLyUTijIwMvPvuuzh48CC0Wi1UKhUaNGiA2bNno1evXsjLy8OOHTuwePFiLF682OiJ8hV9I6yrqrPfCwsL8corr+DChQvw9fXFAw88gIkTJ8LV1RUA8OOPP2Lp0qXIzs7G7t2768xCvyVfhHbv3o1Vq1YhPj4e/v7+uHHjBry9vbF48WJ0794dQgjs37/fUJW9ooVgY2NjsWLFChw+fBg5OTmQy+Xw9vbGk08+iUcffRTA3b8Hdf334s7rLywsRF5enuHz/MaNG9i9ezf27t2LpKQk9O/fH88995xRo6G1tagokyYqU8kvUl5eHsLDw5GSkgJnZ2e0atUKjRo1wtWrV7Ft2zb8+uuvyM3NxWeffYYHHnigzLbuXI7irbfewpQpUzBhwgTExsbijz/+wKFDh6BSqbBkyZI6fWuipM9nzpyJuLg4TJ06FUFBQTh58iT27duHyMhIzJgxAzNmzIBer0daWhp8fX2NapPurTr6/U56vR6//vortm3bhri4ODRr1gxjxozByJEjARQnEL/++iseeeSR6rpEqyLuWDKpe/fuGDduHIYPH462bdti6dKl+PXXX7FmzRqcOHECTzzxBABAo9Hcs+bbncaNGwdfX1888sgj6Ny5M8LCwvDnn3/i0KFDCAoKwnPPPYdevXqViqOuK/l83rp1K3bv3o1Lly6hW7duePHFFw1JfHh4OHbv3o0dO3agXbt2WLduXYXtLlmypFRR0aNHj9aOoqI1OH+KbETJpL5r166JKVOmiHbt2omQkBDRo0cP0aVLF7F7924hRHHtlOPHj4sZM2aUW5epxCOPPCLef/99UVBQUGr7b7/9Jjp27CheeeWVUuevS0omuKanp4uhQ4eKsLCwUhNTL168KF566SXRq1cvw4TNitz5+j/++ENs3LhR7NmzRyQkJJg1dltWHf1+52TlOyd1JyUliS+++EKMGDFC9OvXT7z22mvi5MmT5rkQG7R582YxcOBAcePGDSFEcb9169ZNrFixQkRFRYnOnTuLNWvWGN3eqVOnRPfu3cW5c+dK/RveunVLfPTRRyI4OFiEhISIN998k78Dt5W8V69duybat28vpkyZImbNmiX69esn2rdvX2oyeGFhodixY4fh6ejyPqdv3Lgh2rRpI7Zu3WqokfXaa6+JyZMni7Nnz4pPP/3UZmuT1Z7ZWWQ2JZO633jjDTg5OWHJkiU4e/YspkyZgvz8fPj4+GD79u3IyclB9+7dsXjxYoSEhJTbZlZWFuRyOXx8fAxLH2i1WgDAgw8+iGnTpiE8PBwajaZO1hgqmSh56tQpeHl5IS0tDZIkQaPRACiul7J48WLk5+djz549RrVZUmX5q6++wptvvokFCxbgzTffxDvvvIPt27cjLy+vei7Ghpi73++c9Lp9+3YsWLAAs2fPxqVLl+Dj44OXXnoJH330EXr27IkLFy5g5syZOHToUPVdoBVTKpUoKCgw/L4vXLgQrq6umDJlCvz8/ODn54fMzEyj21MoFNBqtcjLyzP8Gwoh4O/vjzfffBNBQUFo27Ytzp07hzVr1hgeYqnLSt6ra9euRb9+/fDVV19h0aJFWLZsGSZOnIiff/4Zw4cPx5YtW2BnZ4eRI0caJuSX9zm9a9cutGrVCn369IG3tzeuX7+OXbt2Ydq0aXB2dsZ3332HrVu31sg1mhuTJirTyZMnER8fj+effx79+vUDABw9ehQDBgyAh4cHli1bhp9//hkA4OLiUmF7KpUKhYWFOH78uCFZUigU0Ov1EELAz88PKpUKSUlJ1XZN1i48PByzZ8/GqVOnDMvT2NnZGf6AA0BISAjS0tKMWlqmZJ7ON998g8cffxwHDhzAa6+9hqioKHz22Wf44IMPcOzYsWq7Hlth7n4HgG+++Qbvv/8+Dhw4gKNHj2LMmDF44403kJOTg3bt2uHDDz/Eq6++ij59+qB3797Vcl3WLjAwEHl5eThz5gxiYmKwfv16vPzyy3B1dYVOp0O9evVK/RtUxMvLC0VFRdi3bx+A4n9DIQT0ej3y8vLQqFEj9O3bF+PHj8dPP/2EyMjI6ro0m1CSNKanp0Mul6NJkyZwdHSEQqFAu3btMHPmTHzxxRcICQnBO++8g/Hjxxv971FeUVEPDw+bLipaNx5Joiqxs7ODvb09FAoFdu7ciYiICPz6669o2rQpGjRogGvXrhndlkwmw9ChQ/HZZ5/hhx9+wOjRo+Hq6gqFQoGcnBxcuHABLi4uCAwMrMYrsm7169fHyy+/jMOHD+PHH39ERkYG3nzzTfj4+AAorsB748YN9OnTxzAnpKL5AMnJyWjbti1GjRqF+vXrY8qUKRg2bBiWL1+OPXv2ICIiAl27dsW0adPqbN+bo9/Pnz8PR0dHBAUFQa/X4+uvv8ZLL72EYcOGISEhAXv37sXWrVvxv//9D88//zyeffZZDB06FEOGDKlVj2Pfy9mzZ9GiRQuoVCpER0dDkiS0atUK3bp1w3vvvQd3d3f06NEDw4cPh06nw/79+xEeHo65c+ca1b4QAgEBAXjllVewdOlSpKam4qmnnkL79u0BAJGRkTh+/DgmTpyItm3bYvPmzUhPT6/OS7Z6JSNFH330ESIiIu56GtTFxQW9e/dGixYt0KtXL2i1WkMiWtHnTvPmzaFWq3HixAkoFAqcPXsWW7ZsMZSpubMMh61h0kRl8vDwQFpaGg4cOIAWLVrgk08+wbRp09C0aVNoNBr4+PjctbhpeWQyGZ555hnExcVh8eLFOHLkCPr37w8HBwecOHECBw8exKpVqwDU3WVA6tevj2nTpqFHjx74448/sGvXLgwbNgwPPvggNBoNsrKy4Ovri2nTppXbTkn/aTQa2NvbQ61WIzs727C/Xr16mDdvHsaOHWsYMZw6dWo1X531MrXfc3Jy8MgjjyAkJARTpkyBo6Oj4daEj48PfHx80Lx5c3Tr1g3bt2/HihUr8M0332DlypXo2rVrzV6sBWRkZGDRokXQ6XR46623MH/+fEyaNAkTJkzA559/jo8//hi///47oqOj8cUXX+D06dNIS0vD5MmTERQUZNQ58vPzIUkSJk2ahMzMTOzcuRMvvPACmjVrBnt7e/zzzz8ICQlB7969cerUKaSlpRn9xG9tJYRAYWEhhBDIyMjA//73P8yfPx8TJkww1LcCAB8fHzz88MOVSu7vu+8+9OjRA3PmzIFerze0mZqaio0bNyIlJQVTpkwxxGFLk8H59ByVSaPRYN68eQgLC0Pz5s1x8+ZNbNu2DW5ubjh79iyeffZZvP3223jooYfKbSc5ORl//vknNBoNmjdvDi8vL1y/fh3ffvst/v77b9jb26NNmzYYO3Ysxo8fb3O/QKYoSW4KCwtx5coVJCcno3v37nB2doZGo8HZs2cRGhqKP/74A1lZWRgzZgw+/vhjAMX/PgqFotwPsnfffRdHjx7FzZs30a1bN7zwwgu47777oFQqSx134cIFQ62suqA6+n3nzp1YtmyZ4bHsCxcuYOXKlaX++ABASkoKjhw5gnXr1mHevHno2LFjjV23JR08eBBr1qzBuXPnIJfLsXbtWkPCmJSUhPPnz2PXrl24cuUKGjVqhIceeggPPPBAhfXZYmJisHbtWly/fh1NmjTByJEj0aVLF5w5cwaHDh1CREQEcnJyMHr0aIwYMQIajQazZs2Ck5MTVq9eXROXbhMuXryIL7/8EuHh4WjWrBmGDh2KYcOGVeop0bpSVJRJE91Tamoq5s6di0OHDqFRo0YYO3YsEhIScP78eTg7O+OHH34o83UlvwhhYWH45JNPEB0dDXt7exQWFqJ79+4YNWoUhg0bhhs3bkAIgcDAQEP9ldpWPbY8JQnim2++iX379iEnJwc+Pj6YOnUqxo8fDw8PD2RnZ+PIkSPYvXs39u3bhzZt2uDdd9+9Z1G+kv7bvHkzFixYgLFjxwIANm/ejICAADzyyCMYOHCgochiXVQd/V5i+fLlWLNmDYqKitC7d2+8+OKLaN++/V1fBJKTkw23/2qz1NRUeHp6QiaT4erVq3jooYfg5OQEIQSeeOIJPPnkk6VqLxUWFhoeFDHGo48+ilu3bsHf3x+XL1+Gr68vBgwYgEcffbTUor5qtRorVqzAnj174Orqiq+//hre3t5mvVZbUfL+1+l0uHXrFjw8PAz/Br/99hu++eYbJCQkoEuXLhg4cCAefPBBo26l1ZmiojX5qB5Zp5JHRxMSEsSWLVvEW2+9JY4cOWLYv3fvXjFy5EgxdOhQMXjwYPHtt98atb5cjx49xDvvvCOOHj0qsrOzxU8//SQGDRokunbtKnbt2lXq2Lq27lPJ9R47dkyEhISI1atXiwMHDoi5c+eKNm3aiFGjRondu3cb1veLj48X69evF2PGjBHBwcHi559/Lrf9GTNmiJUrV5ZaGPaJJ54QwcHBYsKECWLbtm118rHr6ur3O9dhzMjIEHPmzBHBwcFixIgRYv369YY1uuqSrKws0bNnTxERESGEKF7P74cffhAnT54UCxYsECEhIaJ///6G9SW1Wq04dOiQ0X31888/i+7du4tTp04JIYrX9uvdu7cICQkRI0aMEN99953h8fjk5GSxceNGsXr1ahEdHV0NV2sbSj7rY2NjxYsvvigGDRokRo0aJdavX1/qmBUrVoi+ffuKjh07GvXvERMTI9544w3Ro0cP0bZtW9GhQwcxaNAg8eOPPxqO+e9nvK1+5jNpIoNHH31UdOjQwbCi9WOPPSbOnz9v2J+QkCDUanW5bZT8Ipw6dUqMGTNGXLt2rdT+wsJC8eyzz4ru3buXWrC0rtq0aZOYO3euod9yc3PF4cOHxdSpU0XLli3FzJkzxV9//WU4/sKFC2Lp0qWG2il3KvlAvHr1qvjoo4/EqlWr7jomLCxMPPDAAyI4OFi8//771XRV1s+c/X6nO38/zpw5Y0i2nnzySbFnzx6RlpZWPRdkhRISEgx1luLj40VoaKjIyMgQQgiRk5MjwsLCxIwZM0RwcLCYNGmS+Pzzz0WnTp1EeHi4Ue1PmDBBfP7550Kr1Yrs7Gzx7rvviqVLl4p9+/YZFgFv06aNCAsLq65LtFmTJk0Sw4YNE7NmzRJjx44Vbdq0ERMmTBAHDhwwHBMXFyc2b94shKg4wRk7dqx47rnnxMGDB0VOTo4IDQ0V//d//ye6du0qJk+eXOpLuK0mSyWYNJEQQojdu3eL3r17i6NHj4orV66IzZs3iwcffFC0bt1avPPOOyIlJaXCNu78A7R27VrRv39/Q9Kk0WgMox6RkZGiQ4cOYv/+/dV3QVbszuKHoaGh4uGHHxbZ2dmljklPTxdbtmwRffr0Eb169SpVJLGsonAlfZ+WliaCg4NFcHCwGDNmjLh06VKZRei+/PJLcejQIXNdkk2ojn4v6duTJ0+Kzz77TMyfP1+sXbu2VHL0yy+/iF69eong4GCxaNEic1+WVSt5X65atUoEBweLGTNmiFOnThm2JyUliW3btolx48aJTp06iddff92oNtVqtXjsscfEl19+KYQoLqbYuXNnceLECSGEEAsWLBA9e/YUq1evviuWuqrkvfrbb7+Jzp07GwoSf/PNN+Khhx4Sw4YNEx06dBCzZ882jA6WuPN357/qWlFRJk0khCiuGD1nzhzDbYbCwkJx5coV8fnnn4vu3buLnj17ihUrVpT7y1Piyy+/NPzhfu+990RmZmap/ampqWLMmDHiu+++q45LsWol/VdYWCjmz58vRo0aJTp06CB++eWXMo+Njo4W586dE0KUvgVUnoiICPH6668bbsXt3bu3To1wlKU6+r2kzZSUFNG7d2/Ru3dv0bVrV9G9e3cxderUu9p+9913xZ49e8x5WVbtzpUFTp48KUJDQ8WDDz4o2rVrJ+bPn2+oBK7X60Vqaqq4efNmhSPZQvzb7y+++KIYNmyYEEKI9957T4wcOdJw3m+++Ua89tprhpEtYz636opnn31WvP/++0Kr1YrCwkLx4Ycfijlz5ogzZ86IkSNHGj67t2zZYlR7Z8+eFZ07dxbHjh0TQhT/jt2ZPI0YMUJMmTJFDB8+XMyfP9/mV3yoGzNuqUwlFaNzc3ORmZmJM2fOoLCwEEBxjaYWLVpg+vTp+PLLL9GpUyf89ttvRk3SnjlzJt5//300bdoU27Ztw/LlyxEeHm7YHx4ejkuXLhmenhF18FmE999/H7t374a7uztatWqFuXPn4vnnn8fVq1cNx8hkMjRq1MhQa+a/T70BxSvEf/vtt1i9ejW+++47nDt3Dm3atMHixYuxfPly5OTkYNasWfjkk08QHh4OtVpdY9dojczV7yXHAcWVrBs1aoQNGzYgLCwMjz/+OLKysvDll1/i5ZdfxvHjxwEA7733HoYOHVrNV2g9SsqGvPrqq7h48SL69euHb775BjNmzMDu3bsxdepUrF27FhkZGahXrx7q168PR0fHCtv99ttv8ddff+HFF180lMqIjo42PImo0Whw4cIFFBYWwt3dHQDqzMMl5Sn5nFWpVIaClpmZmdi2bRuGDBmC++67DxMnTkSjRo2wdOlSjB8/3qh261xRUQsnbWQhJdn+rVu3xKhRo0TXrl1FcHCweO2118qct5GSkmJYQ6g8d36jy8nJEQsXLhS9evUSw4cPF1OmTBFTpkwRkyZNMgyrGzt6UhuUfPvKyMgQAwcOFGFhYaKgoEDExsaKH374QQwePFi0bdtWLFy4UKSnp1fYXlhYmOjXr5/o2LGj6NOnj+jXr5/o0qWLmDFjhoiJiTEc9/XXX4uuXbuK3r17iw8//NCof8faxNz9LsS/vz8ajUa89dZbhrkfJa5duybmz58vhgwZIgYPHizeeeedUrf6aruSPv/jjz/EkCFDSs1tLCoqEpcuXRKzZ88W7du3F8OGDTN63lHJBP6Syd8lfbpo0SLRp08fsWHDBrFkyRLRoUMHERkZKYSom2tZlueDDz4Qffr0EUIIsXz5ctG/f38hRHFf/vrrr2LSpEni5s2bQoiKb2mW7P/uu+9Ehw4dxEsvvWQYoRWieF5fx44dxeHDh0VGRoYYOnRoqflNtojpdx1V8i3wrbfeglKpxMSJEzFx4kQcPnwYvXv3xnfffVfqeC8vL6Me0ZXJZNDr9dDpdHB2dsbs2bPxww8/oHnz5jh16hROnz6NNm3aGFZ4v9e3+Nqo5LHz69evo3Xr1pDJZLC3t0dgYCAmTJiAL774AlOmTMH27dvRq1cvHD16tNz2FixYgL59+2LLli34888/8cEHH2DChAmIjIzE8OHDsWTJEuh0Ojz99NMIDQ1F7969sX37dri5udXE5VoNc/c78O/vzyeffIIbN27cVR2/WbNmmDdvHubPnw9fX184OzvbbAXkytLr9ZAkCQUFBbhw4QJUKhWcnJwAFI92KBQKtGzZEvPmzcOKFSsMIx7GuHLlCvr06YPWrVsD+HcEqXfv3nBycsKnn36K0NBQPPvss4bq7HWxUG555s6di1mzZgEoLglRUkBUp9MhPDwcDg4OqF+/PgBUWDMvPz8f+fn5mDRpEqZNm4YLFy7ghRdewLRp0zBjxgy8+OKLhqKiV69erR1FRS2dtVHNKxkNSktLE9OmTTM8raJWq0VYWJh46aWXRHBwsBgyZEiVnjwpmTAbGhoqPvjgA8P2w4cPi/Hjx4tevXqJZ555RmzevFnk5eWZ4Ypsx7lz50SnTp1EcHCw+Pbbb+/an5OTI44ePSpefPFFcevWrXu2c/LkSdG/f/9S3+qEKP62ePLkSTF37lzRs2dP8eqrr4q4uDjD/ro6t8lc/X6n5ORkMW7cOMPvyuHDh8scTdJqtXVyRHXRokWiffv2hgngly5dKvP4/07Gv5fNmzeLgQMHij59+oh//vnnrv2JiYkiLCxMJCYmGmKo63OZ7rx+jUYjUlJSSm3bsGGDCA4OFh9//LF45513RLt27QyfKeWN0EVHR4t3331XTJ48Wbz99tuGJ01Pnz4tlixZIp588knx8MMPi40bN4qsrCyRkpIipk6dKmbMmFFNV1pzmDTVYd99952hLs2dkpKSxNatW8XkyZNFcHBwqVobFblzOLddu3Zi1apVd31wfffdd2LQoEGiTZs2pUoa1FZ39klKSopYunSpGDt2rAgODr7nxMjc3FwhxL0/9OPj40XXrl3Fjh07hBDFydKd7aSlpYnVq1eLbt26iRUrVpjzcmxGdfT7nW0eO3ZMZGRkiK1bt4ohQ4aIkJAQsXDhQhEVFWV4fcnxde3JrdjYWDFhwgSxdOlSMX36dMNDCT///LPRt0D/69ChQ2LAgAEiODhYjB49WuzZs8fw7/Vfda2/K/Lrr7+Kxx57TDz44INi+PDhYs2aNaKwsFDk5eWJV199VXTr1k089NBDhi8UFSWbEydOFPfff7+YMGGC6NChgxgyZIhYuHBhqWkBQgiRl5cnPvnkE9G/f38xatSoWjE1gElTHfXXX3+JwYMHi5CQEPF///d/Zf6SREZGirVr11ZqLkbJKNPChQvFgw8+WOoD8s5HtpOTk+vM03MlH+AbNmwwJDmnTp0S7733nggJCRGdO3cWP/30U6Xay8nJERMmTBBjxowpVWhUo9GU+oOxfPlyEfL/7d15QI1p+wfwb6tWU2lKaCp5nQgVFeotihZLUST7UiZkZqxlahBeEmZEWbNliJGIsRPJMtkrY2Rr0iZtWrWczun+/dHvPCOSMzOydK7PX5zlOc99d87zXPd23aamYiUjbWned72/ateuXczIyIjdv3+fMVafk2jt2rWsV69ezM7OjkVFRYmVpqOl+uOPP5izszPz9vZm+fn57OLFi2zQoEGsT58+bPbs2ezChQuNpnAQx86dO7l8ciEhISwlJUXie5QaI2oUREdHMwsLCzZp0iT2ww8/MB8fH9ajRw/m4ODA7t+/z2pqalhaWhorLS3l6rGp+pT0pKIUNEmoyspKFhMTw3x9fRmPx2ODBw9+o8eJMfFabK+/RpQrSHSjev21kjgxUygUMn9/fzZw4ED2+PFjxlh9PZ09e5ZNnz6d8Xg85urqyq5evSr2MR8+fMgcHR3ZgAED3ljeXl1dzRhj7NKlS8za2vqNYTxJ0Rz1zlh9A8DHx4c5ODhwQ9g1NTXs7t27bO7cuaxr165s8ODBEjscylh9Pbu5uTFnZ2f29OlTVlVVxX788UdmY2PDbGxs2IoVK/7WteDVdATl5eUsICCA8Xg85uzszHbt2sXS0tKaoxifvX79+rHw8HCu8VtQUMDi4uKYm5sbc3R0bNCgEud6L+lJRSlokjCv/ygKCgpYVFQUGzlyJDM3N2c+Pj5vJDZrjKiVmJ6ezsLDw9l3333HDh8+zOVeuX79epPzOCSx+zw1NZU5Ojoyd3f3BjfTnJwcduDAAWZvb8/c3Nz+1jETExPZuHHjmI2NDZs1axaXK4Wx+r/RiRMnmKWlJcvLy3tv5fjcNEe9M1bfm+Lt7c1Gjx7dYN5YeXk5O3XqFAsODn4v5/85eNvv+cmTJ2z8+PHs66+/ZqWlpYyx+h7sSZMmsZ07d4p17Dt37rCgoCA2bdo0Nn78eBYdHc1df1JTU9nYsWMZj8djYWFh76cwLYAoGK2pqWGLFy9+Y9sqoVDIbty4wYyNjbms7e9CSUXr0Ya9EkK0katQKER5eTlu3boFLS0tdOjQARoaGkhPT8ehQ4cQFxeH0tJS2NnZYdmyZe/cgdrV1RX5+flQVFREXl4eunfvjkmTJsHKyorLkUL+kpaWhlmzZuE///kPFi9eDHV1de65+/fvo02bNtDW1m509+/8/Hzcv38fV69ehb6+PvT19WFqaoqamhrs2LED586dw8uXL9GxY0c4Ozvjxo0bSElJgaOjIwIDAz90UT8p/6begb82oX7x4gVqamqgo6MDAMjOzsbs2bNRVFSEdevWNdiIVJI2nxaVNSIiAsnJyfDw8IC6ujqMjIyQn5+PqVOnQl1dHRs2bPhbG+VeuHABS5YsgYqKCgwMDMDn83H58mUEBQVhyJAhaN26NQDg1KlT6NmzJ7S1tbkNaSVdXl4eBg0ahMrKSkycOBGBgYEQCARgjHGrlr29vWFgYICFCxe+83iiv/F3332HJ0+e4OTJk1i2bBlu376No0ePQigUIjIyEqmpqVi4cCHU1NRa5G+AgiYJs2HDBvz6668oLCxEZWUldHR04OLiwi1B/e233xAVFQV1dXUsX768yWMdO3YMoaGhWL16Nbp164akpCT89NNP+OOPP+Ds7AxPT09069atwS7mkq62thaHDx9GcHAwpk6dim+//Vbs906YMAGpqanQ1NREbm4uFBQUwOPx4OHhgaFDh+Lx48c4deoUbt++jUePHkFbWxuOjo7w9fWV+JvIv6n3V3l6euLly5cwNTXFkCFDwOPxUFxcjJ07d6K8vBy+vr4wMjJ6z2f/eeDz+XB3d8eTJ0/QqlUrmJiYoLa2FnZ2dtDX18eyZctgamqKOXPmwNDQUKxj2tvbw97eHj4+PtDS0kJ0dDTWrFmDHTt2IDo6GtbW1hg0aBD3egqY/pKeno7Nmzfj9u3bqKiowIoVKzBw4EDu+YqKCnz33Xf48ssvsWrVqnceb9u2bTAxMYG6ujpu3bqFMWPGwMvLC3p6eggKCkJVVRUCAgIgFAoRHh7enEX7qJruRiAtglAohIyMDM6cOYPIyEhMmjQJ1tbWaNOmDZycnCAnJ4ekpCRoamrCysoKPXr0ECu3Sdu2bdG3b1+YmJhATk4Offv2xS+//IKYmBhs2LABiYmJcHR0xJw5cxq07CUBn8+HvLw8bt26heLiYqipqaFt27ZcbqAvvvgCq1evhrKyMiZOnAgATfbqhYaGorCwEOvXr4eFhQXq6uqwd+9enD59GqtWrUJcXBxmzJiBWbNmgc/ngzEGgUAAZWXlD1XkT8L7rvdXlZeXQ1FREQ8ePEBOTg5iYmJgaGiIjh07oqioCLm5uYiIiEBISIjE5GR6lby8PLZv344VK1YgLS0NHTp0gLq6OmJjY5GZmQlFRUXExcWhe/fuYgVN169fB2MMgwYNgpaWFgBg8+bNGDFiBPT09PD48WOUlpY2CJooYPqLgYEB/P39cffuXURFReGbb76Bvb09vLy8UFZWhuTkZNy5cwfR0dEA/rpPNCYxMRGbNm3Ctm3b8J///Ad6enoAACMjIxw/fhydOnVCfn4+Ll68iIMHD77zeJ8zCpokgOiLGxERAVdXV3h7e0NJSQlXrlyBrKwsBgwYgOjoaDx8+BBbtmxpMvmhqLv12bNnuHPnDhITE5Geno7OnTsDqL8BjR49Go6Ojli3bh3u3bsnUQHTvXv3oKenB1VVVQD1yQ/v3r0LDQ0N1NXVoV27duDxeFBWVoaenh7OnDkDKyurJnsnampqcP36ddjb28PKyoq7MUydOhWurq7Ys2cPTpw4gcWLFyMoKAjGxsYAJCtxaHPU++tUVVUxf/58bNiwAbKysvjmm2+QlJSEe/fuoaCgALm5ucjNzcWAAQMwZMiQ5irqJ+X1np22bdti2rRp+Pnnn6GkpIRx48bBz88Pd+/exaVLl/D48WMMGDBArGO3bdsWJSUlKCoqAlDfS15bWwsvLy988cUXMDIyQkZGBsrLy7m/O6lXXl4OeXl5aGpqwt7eHj169MD58+exe/duTJgwAYwxDBkyBBEREejcufNbh6VFmkoqeuHCBfz4449QV1eXiKSiNDzXgr0a6b948QK+vr6wtrbmhib69esHZ2dnBAQE4MCBA1izZg3OnDmDNm3aNHm8P//8E76+viguLkZpaSn69++PmTNngsfjvdHCrqqqgqKi4jt/lC1BXV0dXFxcUF5ejpkzZ8LT0xMFBQUQCoVISEhAdXU1/vjjD+Tk5KC2tha///47GGNo164d1q1bhx49erx1eOGbb74BYwwbN24EUD/cJC0tzf19b926BX9/fygrKyMyMvKtf8OWqDnrXfT4q88/evQIs2fPhoKCAsLCwtCuXTtUVVUhLS0NiYmJmDZt2oeugo/u/PnzMDc35xpcv//+O/z9/dG6dWuEh4dzPUUVFRViDdczxlBUVAQPDw906tQJixYtgqurKwIDAzFq1Cjw+XwEBASgrKwM27Zta9ayfS5E19j4+HhERUXh3r17sLe3x7Bhw2BmZgZ5eXmkpaXh1KlTOHnyJIqKiuDr64tx48Y1eW0+ePAgtm7dCj6fjy1btnCBk0heXh4ePHgAIyMjaGlpQUpKqkXOZeJ80Gnn5KMQ5dyYNGkSGzduHGOMsY0bNzIrKytuuWlcXByzs7Pj8s40ZezYsWz8+PFs7969LDIyktnZ2TFzc3MWGhrKsrKyJDZnSnV1NYuMjGReXl7MwsKCTZw4kSUmJnLPi1aRlJWVsbKyMpaTk8NOnDjBPD092aJFi5o89p49e1iXLl3Ynj17GqxKfPXfmZmZrFevXiw2Nvb9FuwT11z1LnpfVlYWi4yMZOfOnePyjpWXl7MFCxaw6dOns5s3b7a4FUJ/x7lz55ixsTHr2bMnmzFjBtu9ezdLSUlhxcXFzMfHh40bN44lJSWJfbzMzEwu4/fp06dZ7969mampKbO3t2c1NTWspqaGxcTEMBMTEy5XkCSmMXmV6PtXXV3NzM3NmYeHB1uwYAGztbVlvXr1YiEhIezhw4fca27evMn8/PyYiYkJs7Oza/B7eR0lFW2IgqYWqKioiI0fP77BxogCgYDt3r2b8Xg8tnTpUtajRw924MABxlh9FuSAgIAml12LLkoVFRVs6tSp7Pbt24yx+oAsPT2drVy5kpmamjJnZ2cWHR39j7P+tgRPnz5l4eHhbMiQIczCwoLNmTPnjcRur17kz5w5w3g8XpMXrvLycjZ37lzm4ODA1q9fz6V2YOyvoLi4uJi5ubmxVatWvecSfR6ao94ZYyw8PJzxeDxmZ2fHrK2t2ezZs9mhQ4fYpk2bmI+PD/v+++8lOpGlQCBgiYmJbPfu3WzcuHGsf//+XF2NHDmSWVpasilTprDCwkKxjufn58dMTU1Zamoq4/P5bO/evczNzY1ZW1szR0dH5uTkxOzt7dnKlSsZY5JxoxbX/v372ahRo7itgIqKilhwcDAzNjZmDg4ObO/evQ2eO336NHN1dWXx8fHvPDYlFa1Hw3MtUFxcHJYuXYq6ujpuOE5XVxdA/YTiX375BaWlpVi0aBHU1NRw5coVxMfHY8OGDTA3N2/y2Bs2bMDVq1fh5uaGUaNGcY/z+Xzcu3cPkZGROHv2LJYvX46RI0c2azk/NUKhENLS0g02iN27dy9u3LiBiooKuLm5YcaMGWjVqhWAvyYtx8fH49tvv8WiRYvg6enJHa+yshIKCgpcN3dOTg5WrVqF8+fPw8LCAsOHD0fv3r255e9ZWVmYOnUqxowZg8mTJ3/Ywn9E77veG1NUVITbt2/j8ePHuHDhAoqLi6Gjo4OUlBQIBALo6enh2LFjEjkB/HV//vknsrOzkZmZifPnzyMjIwOtW7fGkSNHxHr/3bt3sWjRIujp6SEsLAwA8OTJE1y9ehUFBQWorKyEh4cHDA0NIS8v37KHgsTAXhk6/v333xEZGYng4GDu+w7Uz0lat24d4uPj0aVLF+zcuRPq6upgjKGwsLDJNBCiKRZA/fBqcHAwDh8+DAMDA3h6esLW1hYdO3Zs3kJ+QihoaqHu3LmD48eP4+LFi6itrYW7uzt8fX3RqlUrxMXFISYmBgkJCWjTpg06duyIkSNHwtXVtclj3rt3D97e3lwep0WLFnG7YYuUlZXh1q1bsLOzk9iVLHl5ebh06RKEQiEUFRWRmZmJp0+fIikpCUpKSvDx8eHqWiAQ4MSJE1ixYgVOnDjBXbzq6uowadIkuLm5wdbWFhoaGtyN4eTJk1i9ejXKyspgYmICQ0NDqKmp4cKFCygtLcX58+c/Wtk/pvdR76+rq6vDgwcPIBQK8eWXX6Jt27YAgAcPHuDp06coLy/HoUOHGswVJA2lp6dDRUXlb+VnOn36NObPnw8nJycsX76cu2m/ilF6gQauXLmCX3/9FfHx8QgLC4O5ufkbi0FOnz6NpKQkBAQEvDPYTEpKwtGjR/H8+XO8fPkSrq6ucHNzg6ysLB48eID//e9/uH37NmbOnClR330KmlqYVy8kfD4fCQkJ+PXXX3Hr1i20adMG06dPx9ChQwHUt56fP3+OLl26iNVSq66uxr1793D+/HkcOHAAampqmDVrFlxdXZucRCsJRJMwz507h82bN+Phw4dQVlZGRUUFLC0tYWZmBjk5OSQmJiIpKQn//e9/sWnTJq7eX7x4AQ0NDQB/1duwYcPw6NEjODg4YMyYMejevXuDSbS7d+/GlStXkJubi+fPn2Po0KFwc3NrkGCxpXuf9S4iWvBw/vx5REdHIyEhAbKyslBQUMCAAQMQGBjY5ApTUu/f9gAdP34ca9euxahRozB9+nQA9QsgJGlV6LuIvqtxcXGYO3cuWrdujZcvX0JdXR2+vr4YMGDAW1cvN5USgJKKvh0FTS2U6ILFGENJSQmOHj2Ky5cv4/79++jatStmzJjRYCiuqS/8688JhUJcv34dO3fuxJUrV2BhYYF58+bB1NS0uYv1SRLVj0AggI2NDZydneHk5ISePXsiISEBYWFhyM/Ph7+/P6ytrbFjxw5069YNw4YNe2No6VVbtmzBunXr0KFDBzx//hzu7u4YN24c9PT0oKCgAKC+61wgEEBaWlricjI1R72LfjclJSVwcnKCmZkZPDw8UFFRgeTkZJw+fRqtW7fGkiVL0LdvXy5ok5QbRnOprq5GZWUl6urqoKmpCaC+0bd582Zs2bIFa9as4Rp75E2Ojo6ws7PDoEGDIBAIsGvXLly4cAG9e/eGj48PzMzMGu2textKKvp2FDS1MKLWw+3bt3Hy5Elcu3YNqqqqsLKygrS0NKqqqnDp0iUUFxejT58++P7777mL1NuOxefzkZWVhQsXLkBWVha6uroYOHAgamtrcezYMezevRsPHz7EkCFDEBISIrEtwZ07d+Lw4cOIiIh4Y9hy3rx5uHDhAmJjY6Gvr8893tTF5syZM1i6dCkiIyNx8eJFREZGAgAmT54MFxcXfPnlly0+jYM43me9ix7/3//+x80DEf0+Xr58iQsXLmDVqlXo1asX1q9f36zlkiT+/v749ddfYWJiAg0NDdja2qJdu3YwMzPDli1bcPr0aSxevBj9+/f/2Kf6yXj1+rx8+XIMHz4cPXv2BFCfp+nKlSuIiIhAeno6Bg4ciBkzZoidVPT777/Hjz/+iF69egEA7Ozs4OTkhBkzZnCBVEvO+t0UuuK2IKKEYnl5eZg/fz7U1NTQu3dvAMCmTZvg5eWFKVOmwMzMDJcuXUJsbCxcXFzQr1+/Ro8n6roNDg7m9qSTlZWFvLw8du/ejblz58Ld3R1mZmaIiYlBTU2NxAZMQH0SShUVFW7PPfb/Wbnl5OTg5eWFM2fO4NGjR9DX1+d6NJpqnWloaKCsrAxCoRA+Pj7o378/9u7di7CwMJw4cQLTpk2DjY2NxCf2e5/1LiUlhZqaGuTl5cHAwACampqoq6sDYwzKyspwcXFBRUUFli5diuzsbHTo0OEDlrRlEgqFcHZ2Rp8+fXDjxg3k5ubip59+gpKSEkpKSqChoYHnz59j3bp1sLS0hJKS0sc+5U+CjIwMysvLMXHiRGRnZ0NPT48LmlRVVeHs7IyePXvixIkTWLNmDQYMGCBW0ERJRZtGQVMLIpo/sGLFCujq6mLlypVo3749+Hw+9u3bBxMTExw8eBCqqqpYuHAhBg0ahL59+zZ6LFEr5ujRozh9+jRmzpyJESNGIDs7G2fPnsXp06cRGBiIkJAQmJiY4JtvvuECppaaPv9dVFRU8PDhQ2RlZYHH44ExxvUEqaurw9DQEIWFhQAg1lwPQ0NDeHp6citTOnfujMDAQAwaNAjbtm2Dv78/evfujZkzZ3IXS0n0b+s9PT0dtbW1XFb7Vq1aQUZGBikpKQ3eIxqK++qrr6Cjo4OioiIKmt6DpKQk3Lx5E926dYO/vz80NDRQWVmJpKQkMMZw7do1ZGRkgMfjQUlJSaKGgt4lNzcX6urqePbsGWJiYmBkZARra2sA9Q0AbW1tTJw4EQ4ODtwK6qaIGgdqamo4ePAgjIyMsH37dgQGBkJLSwt8Ph8VFRWQk5OTyIAJACW3bGmKioqYu7s727FjB/fYtGnT2OjRo5lQKGRbtmxh3bt3f2tyste5u7uzVatWsZqamgaPJyQksD59+jAfHx9WW1v7Xsvwubp//z6ztLRko0aNYpmZmQ2eO3HiBOvRoweXN0jc3DI1NTWNvra4uJjFxsYyCwsLtnfv3n9/8p+xf1vvM2fOZDwej/3000+suLiYMcbYsWPHmLGxMdu4cSOXAJax+mSisbGxrE+fPtxryT93+PBhZm5uzkxNTVnXrl2Zo6Mji4mJeeN1FRUV3N9OEnMDNSUzM5MdPnyYDRs2jBkZGbGAgAAuF9PfPQ4lFX036mlqYVRUVMAYQ3JyMgDg9u3bSEhIwJ49eyAtLY327dtDQ0MDmZmZ6NKlS5PHKisrg7y8PFRVVbn8M6LWtq2tLSZNmoT9+/ejrq6uuYv1WejSpQtCQ0OxZMkSODk5YeTIkejRoweuX7+OlJQUuLu7Q09P72+tKnpb3h81NTUMGzYMffv2hba29vssxmfn39R7XV0dvv76a+jq6iIqKgqHDx+Gn58frKysYGdnh7CwMKSlpaFfv34wMjLCpUuXsGfPHowcOZIbDiT/jEAgwMqVKzFy5EgMHz4ciYmJ2LBhA3744QdER0dj3rx56NmzJ+rq6hoscpDknEyN0dXVha6uLqytrREdHY19+/bh7NmzmD59OsaOHSv2cGZ4eDjOnTuH/fv3w97eHoWFhTh06BDy8/Ph4uICKSkp1NbWYvTo0TA3NwdjTCJHFGgieAsUEhKCY8eOISwsDEFBQejWrRtCQkLAGMO+ffuwfft2nDhxotEfE3ul65sxhjFjxkBGRgZbt27llruz+kzyiI2NRUREBLZv3y5W168kYIwhLS0N+/btQ3x8PIqKiqCjowNXV1dMmzYNsrKyEp+Mrzn823oXCoW4evUqQkNDkZqaCgcHB3z99de4d+8eNm3axA3vqampwdnZGUuWLPmApWuZNm3ahLi4OEREREBTUxMxMTE4c+YMrK2tsWvXLhQUFMDS0hKurq5wd3f/2Kf7SRBNfcjNzUViYiKOHDkCZWVlGBgYYPTo0fjqq69w//597Nu3DzExMTA0NERsbKxYSVcpqah4KGhqgfLy8uDj44O0tDQIBALs378fBgYGuHnzJpYvXw5PT0/4+vo2+l5R0HT58mUYGxtzCQB9fX3h6ekJdXV1yMvL48WLFwgODkZubi6ioqI+cAk/fUKhEHw+H8+ePYOmpiaX10dSLzQfyt+td1Hen7KyMnz//ffIzc2FQCBARUUFcnNz4eHhgSlTpiA9PR1ycnLQ0dGBgYEBrVr8l4RCIZYuXYry8nKsXr0acnJymDZtGrS1tbFs2TJERkYiJCQE+vr6CAwMhK2t7cc+5Y/u1QbtmDFjkJWVha+++goVFRUoKSmBtLQ0tyNAeXk5fvvtN9TV1WHo0KFiX3coqei70S+/BdLW1samTZuwd+9exMbGYtq0adDQ0EBFRQX69Onz1oBJNPR27do1zJkzB9u2bcOECROQk5ODrVu34vLly7CxsYGMjAySk5ORkpKCXbt2AZDcyd9vIyMjA0VFRRgaGuLVdgkFTM3r79a7KPjx9/dHZWUlli1bhu7duyMjIwN37tzhhizmzp2LQYMGSe7k1/dMRkYGWlpauHLlCmpqapCeno6EhAScPXsWAGBgYAAzMzNs2LABbdq0+chn+2kQBSuhoaEoLS1FaGgoLCwsIBQKuYTD4eHhaNu2LQYOHIjBgwdz7xU3yHF2doZAIMDatWuxe/fuRpOKSnLABFDQ1GK1b98es2fPhpubG65cuQI+nw9LS0tuhdCrcnNzoaOjw91AkpOTMXDgQG7V1vz589G3b1/s3r0bERERUFFRgampKVasWAFjY2Mu1QFpnKRfZD4WcepdSkoKeXl5SE9Px9dff43u3bsDAPT09NCuXTu0a9cOc+fOxeLFi7F+/XqcOnWKy4ZM/h0bGxskJyejtrYW8fHxMDQ0xFdffQWhUIjKykoUFhZCKBR+7NP8ZEhLS6OyshIXL16Ei4sLzMzMuHlFjo6OMDMzg6+vL0JDQ9G3b98G88De9ltoLKmoo6Mj0tLSsH79enTo0AFDhw6V6FQyr6OgqQVr1aoVOnfu3CBQamw01sPDA15eXvDy8kJcXBySk5NRUlKC2tpaAPWt8X79+qFfv37c8EXbtm2p5UFaBCUlJUhLS+Pu3bsYOXIk16KXk5ND79694eDggKKiIgwaNIgCpvfIxMQEmzdvhpycHLS0tFBYWIiHDx+ipKQEW7duhZWVFbS0tGhI+xVSUlJo1aoVysvLuUZuXV0d6urq8OWXX2Lw4MGIiopq9DrfmMWLFzeaVHTKlCmoqanB2rVroaKiQklFX0FBk4R5PcA5cuQIqqurMXz4cAD1kzMfPHiAuro67Nu3Dx4eHtDR0QFQH3CJ/i36v5SUFAVN5LOmqqoKExMTXLx4ETdv3oSFhQX33RYKhVBTU0Nubi4cHR0/9qm2OKKGV8+ePaGqqorhw4dDRUUFbdu2xQ8//PCRz+7To6CgAF1dXVy+fBkjRoyAgYEBpKWlIS0tDaFQiNatW3M9Uq/uU9kYSir6z1DQRFBVVYUjR44gMzMTWlpa8PPzw969e7F582bcuXMH06dPh6mpKbffmQgFS6SlmDVrFlJTU+Hr6wtvb284ODhAUVERd+/eRUxMDCZMmECTv5uRnp4etm7disTERLRq1Qp9+vSBvLw8zZV8RWlpKR4/fgxlZWX8+eefWL16NaZPnw49PT1oaGjg0aNH2L9/P3r06AEtLa13TtimpKL/DK2ek3DZ2dlYs2YNrl27htLSUnh7e8PPzw9A/S7jW7duRXp6Otzd3TF+/Hh07NiRbh6kRXr27Bm2bNmC48ePo7a2FqqqqmCMwdLSkvaZIx+FaHFOfHw8du7cibt373I5xLZv347i4mL06tULJSUlKCgogIKCAn755Reoqqo2OawZGxuL4OBgCAQC8Pl8dOjQAT4+PhgxYkSD1718+RJKSkqQkpKiYdL/R0ETQW1tLZycnPDy5UsIhUL07t0b3377LYyMjFBRUYGff/4Z+/btQ2FhIdatWwdnZ+ePfcqENAs+n4/s7Gxcu3YNfD4fJiYm6NSpE62aIx+cqGdHIBDA1taW20uuU6dOMDIyQlpaGtasWYOkpCRYWlrC2NgYDg4OMDQ0bLKHTiAQwMrKCiNGjGiQVLSiogImJiYNkoqKk99J0lCXAQEA8Hg8DB48GKmpqTh//jymTJkCFxcX+Pr6wtfXF87OztiyZQu6du36sU+VkGYjLy+Pjh07citHCflYRENhP//8M9TU1DB16lS0a9eOe15bWxv6+vrIy8vDwoULG+wM0NSQZkREBDp06ABvb29oamri999/h5mZGZdUdPLkyZRUtAkUNBHIyclh8+bNAAAXFxf07dsXx48fx8mTJxEfH48pU6Zg7NixWL16NQBK0EgIIR9KVVUVFBQUuG17RINDKioqMDY2xr59+1BQUABtbe13zjsSCoV4/vw59PT0uMSv586dg46ODiZPngygfkeJ58+fcykISEN05yNvsLGxwaJFizBv3jzo6+sjODgY3t7e3I+VAiZCCPkwdHR0kJ6ejrt37wJAgxXLvXr1gqGhIZ49e8Y91xRRUtGUlBTU1NTg0aNHSEhIwNSpUwH8lVQ0KiqKsrC/Bd39SKNUVFTg5uaGoKAgeHt7Y8yYMdxkQEIIIR+GmZkZ1NTUsGTJEly/fr1BDqbk5GQ8ffqU23xdnCnKNjY26NixIyUV/YdoIjghhBDyCXv69CkCAgKQkpICe3t7mJiY4P79+7h//z6srKwQFBT0t6ZNiLZFiY2NRUhICH7++WeUlJRg5cqVMDExwdKlS2kaxltQ0EQIIYR84vLy8nDs2DEcPXoU+fn50NHRwZAhQzBlyhTIysr+oyAnIyMD3t7eyMnJ4ZKKHjp0CPLy8hQ0vQUFTYQQQshnQDTROysrC5qamlBUVATw7xbnpKWlNUgqqqurS0lFm0BBEyGEEEKIGKjvjRBCCCFEDBQ0EUIIIYSIgYImQgghhBAxUNBECCGEECIGCpoIIYQQQsRAQRMhhBBCiBgoaCKEEEIIEQMFTYQQQgghYqCgiRBCCCFEDBQ0EUIIIYSIgYImQgghhBAxUNBECCGEECIGCpoIIYQQQsRAQRMhhBBCiBgoaCKEEEIIEQMFTYSQFquqqgrZ2dkf5bN5PB54PB5+++23j/L5hJD3j4ImQkiLdOzYMTg5OSExMfFjnwohpIWgoIkQ0iKFhoYiLy/vY58GIaQFoaCJEEIIIUQMFDQRQgghhIiBgiZCSIsSHh4OHo+HnJwcAMDChQvB4/EQHh7OvebGjRvw8/PDwIEDYWpqim7dusHGxgbffvvtW+dApaWlISAgAEOGDIGpqSl69eqF4cOHIzQ0FEVFRWKfX0ZGBmxsbMDj8TBp0iRUVVX9uwITQj4Y2Y99AoQQ8j7p6OigZ8+euHfvHvh8PvT09NCmTRvo6OgAAH766SdEREQAADQ0NNCxY0dUVFQgJycHZ8+exdmzZ7Fs2TJ4enpyx0xKSoKXlxcqKyvRunVrGBgYoKamBo8ePUJqaipiY2Nx4MAB7jPeJisrCxMnTkR+fj7++9//YuPGjVBQUGi+yiCEvFdSjDH2sU+CEELeN3t7e+Tk5GD58uXw8PAAAFy/fh0TJ06EtLQ0li9fDjc3N0hL13e4P3/+HH5+frhx4wY0NTVx+fJl7rlRo0YhJSUFEyZMgL+/P+Tl5QHUB0FTp07F06dP4enpiWXLlnGfz+PxAAC7du2ClZUVcnJyMGHCBOTk5KB///4IDw/njkMI+TzQ8BwhRGJcvnwZ8vLycHBwwIgRI7igCADatm2LWbNmAQAKCwsbDLk9ePAAADBixIgGgY6uri4WLFgAOzs7tG/f/q2fm5ubi4kTJyInJwcODg4UMBHymaLhOUKIxJg/fz7mzZsHPp/f6POvDpVVV1dz/9bT08OjR48QFBSEOXPmwNzcHHJycgDqe7Ts7e3f+pl5eXkICgpCdnY2BgwYgHXr1kFWli69hHyO6JdLCJEoUlJSkJaWxq1bt/DkyRNkZWUhMzMTDx8+REZGBve6uro67t9+fn6YMWMGUlJSMHnyZCgpKcHCwgJWVlbo378/9PX13/p5S5Ys4QKwoqIiSElJNVvZCCHNi4bnCCESgzGGyMhI2NvbY9y4cQgKCsL27dtx7tw5yMjIYNiwYY2+z9bWFjExMRgyZAiUlZVRWVmJhIQErFy5Ek5OThg7diyePHnS6Hurq6thb28PBQUFJCcnY9euXc1ZREJIM6KeJkKIxNi4cSOXemDw4MGwtbVFp06d0LFjRygrK+Pp06c4evRoo+/t0qUL1q5di9raWqSkpOD69ev47bffcOfOHdy+fRuTJ0/G2bNnoaSk1OB9bm5uCA4Oxq5du7B69WqEhYXBzs4OhoaGzV5eQsj7RT1NhBCJUFtbix07dgAAZs6cidDQULi5uaF79+5QVlYGUL+C7nVCoRAZGRm4efMmAEBOTg7m5uaYOXMmoqKiEBUVBSkpKRQUFDS6Oa+rqyukpaUxefJkdOvWDTU1NQgMDIRQKGzG0hJCmgMFTYSQFkk0d0iUVaW4uBiVlZUAAGNj40bfc/DgQe7fAoEAAPD48WM4Ojpi0qRJKCgoeOM9ZmZmXND16jyo18nIyGDFihWQlZWlYTpCPlMUNBFCWiTRMJkoM7iGhgbU1NQAAJGRkSgtLeVe++LFCyxZsgTHjx/nHhNN3jYyMkLnzp0hFAoxd+7cBr1RfD4foaGhqKiogJKSEszNzZs8JyMjI0ydOhUAEBYWhrS0tH9fUELIB0PJLQkhLdKCBQtw5MgRyMrKonPnznBwcICamhqWLl0KAFBUVIS+vj74fD4yMjIgEAjQtWtX5Obmori4GJs2bcKAAQMAAE+ePMHo0aNRXl4OOTk5dOjQAYqKisjOzkZZWRlkZGSwatUquLi4cJ//enJLET6fD1dXV6Snp8PExAT79++HjIzMB6wZQsg/RT1NhJAWacGCBXBycoKioiL+/PNPpKWlYezYsYiMjIS1tTVUVVXx+PFjFBUVwcTEBIsXL0Z0dDT69esHAIiPj+eO1alTJ8TGxmLMmDFo3749nj17hidPnqB169YYMWIEjh492iBgaoq8vDyWL18OKSkppKSkYOfOnc1SfkLI+0c9TYQQQgghYqCeJkIIIYQQMVDQRAghhBAiBgqaCCGEEELEQEETIYQQQogYKGgihBBCCBEDBU2EEEIIIWKgoIkQQgghRAwUNBFCCCGEiIGCJkIIIYQQMVDQRAghhBAiBgqaCCGEEELEQEETIYQQQogYKGgihBBCCBEDBU2EEEIIIWL4PwEHNt5AlsHAAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "palette_name = \"mako\"\n",
    "num_colors = 6\n",
    "\n",
    "colors = sns.color_palette('rocket_r', n_colors=2)\n",
    "line_color = colors[0]\n",
    "\n",
    "# Get the color palette\n",
    "colors = sns.color_palette(palette_name, n_colors=num_colors)\n",
    "\n",
    "color1=colors[0]\n",
    "color2=colors[5]\n",
    "\n",
    "plot_tasks = acc_df['task'].unique()[2:]\n",
    "filtered_df = acc_df[acc_df['task'].isin(plot_tasks)]\n",
    "\n",
    "CC_new_f = CC_new[CC_new['task'].isin(plot_tasks)]\n",
    "\n",
    "sns.set(font_scale=1.5)\n",
    "sns.set_style(\"whitegrid\")\n",
    "\n",
    "#plt.figure(figsize=(6,4))\n",
    "xvals = [-0.21, 0.25, 0.80, 1.24, 1.77, 2.26, 2.77, 3.23, 3.81, 4.24]\n",
    "ax1 = sns.barplot(data=CC_new_f[CC_new_f[\"language\"]!=\"en_2nd_run\"], x=\"task\", y=\"consistency\", hue=\"correct\", palette=\"mako\")\n",
    "ax3 = plt.plot(xvals, CC_new_f[CC_new_f[\"language\"]==\"en_2nd_run\"][\"consistency\"].values, linestyle=\"dotted\", marker='o', color=color2)\n",
    "\n",
    "# Add max consistencies\n",
    "for i, task in enumerate(CC_new_f['task'].unique()):\n",
    "    for j, condition in enumerate(['correct', 'incorrect']):\n",
    "        xmin = 0.05 + i/5.3 + j*0.09\n",
    "        xmax = xmin+1/14\n",
    "        # print(i, j, task, condition, upper_bounds[task][condition])\n",
    "        plt.axhline(y=upper_bounds[task][condition], xmin=xmin, xmax=xmax, linewidth=2, color=line_color)\n",
    "\n",
    "sns.despine(left=True, right=True, top=True, bottom=False)\n",
    "plt.legend([],[], frameon=False)\n",
    "\n",
    "plt.xticks(ticks= range(len(plot_tasks)), labels=[\"elements\\nfrom-pos\", \"olympics\\n100m\", \"olympics\\ndownhill\", \"writers\\naverage\", \"companies\\naverage\"], fontsize=12, rotation=60)\n",
    "plt.yticks([0, 0.2, 0.4, 0.6, 0.8, 1])\n",
    "plt.text(3, 1.03, \"id baseline\", fontsize=15, color=color2)\n",
    "plt.ylim(0, 1.03)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "ef0dd906",
   "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>task</th>\n",
       "      <th>language</th>\n",
       "      <th>consistency</th>\n",
       "      <th>correct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>en_2nd_run</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>en_2nd_run</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>0.991952</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>en-paraphrase</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>arithmetics_addition</td>\n",
       "      <td>de_from_en</td>\n",
       "      <td>0.450704</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>companies_sv</td>\n",
       "      <td>it_from_en</td>\n",
       "      <td>0.850000</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>176</th>\n",
       "      <td>companies_sv</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>0.937500</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>177</th>\n",
       "      <td>companies_sv</td>\n",
       "      <td>nl_from_en</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>178</th>\n",
       "      <td>companies_sv</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>0.925000</td>\n",
       "      <td>correct</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>179</th>\n",
       "      <td>companies_sv</td>\n",
       "      <td>sv_from_en</td>\n",
       "      <td>0.850000</td>\n",
       "      <td>incorrect</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>180 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     task       language  consistency    correct\n",
       "0    arithmetics_addition     en_2nd_run     1.000000    correct\n",
       "1    arithmetics_addition     en_2nd_run     1.000000  incorrect\n",
       "2    arithmetics_addition  en-paraphrase     0.991952    correct\n",
       "3    arithmetics_addition  en-paraphrase     0.666667  incorrect\n",
       "4    arithmetics_addition     de_from_en     0.450704    correct\n",
       "..                    ...            ...          ...        ...\n",
       "175          companies_sv     it_from_en     0.850000  incorrect\n",
       "176          companies_sv     nl_from_en     0.937500    correct\n",
       "177          companies_sv     nl_from_en     0.900000  incorrect\n",
       "178          companies_sv     sv_from_en     0.925000    correct\n",
       "179          companies_sv     sv_from_en     0.850000  incorrect\n",
       "\n",
       "[180 rows x 4 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(conditional_consistency)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "2b642677",
   "metadata": {},
   "outputs": [],
   "source": [
    "CC = conditional_consistency"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "524e8452",
   "metadata": {},
   "outputs": [],
   "source": [
    "new_dict = {\"en_2nd_run\": [0, 0], \"en-paraphrase\": [0, 0], \"de_from_en\": [0, 0], \"it_from_en\": [0, 0], \"nl_from_en\": [0, 0], \"sv_from_en\": [0, 0]}\n",
    "for subtask in [\"english\", \"german\", \"italian\", \"dutch\", \"swedish\"]:\n",
    "    for language in [\"en_2nd_run\", \"en-paraphrase\", \"de_from_en\", \"it_from_en\", \"nl_from_en\", \"sv_from_en\"]:\n",
    "        new_dict[language][0] += CC[((CC.task == \"writers_\" + subtask) \n",
    "                                    & (CC.language == language)) \n",
    "                                    & (CC.correct == \"correct\")][\"consistency\"].values[0] /5\n",
    "        new_dict[language][1] += CC[((CC.task == \"writers_\" + subtask) \n",
    "                                    & (CC.language == language)) \n",
    "                                    & (CC.correct == \"incorrect\")][\"consistency\"].values[0] / 5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "bdb311ef",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'en_2nd_run': [0.9854960488522133, 0.5525452196382429],\n",
       " 'en-paraphrase': [0.9765383398945042, 0.3759560723514212],\n",
       " 'de_from_en': [0.9732138865700509, 0.32894056847545217],\n",
       " 'it_from_en': [0.962470064182393, 0.37484496124031],\n",
       " 'nl_from_en': [0.9704834739081315, 0.2424935400516796],\n",
       " 'sv_from_en': [0.9521424922109853, 0.24589147286821705]}"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "9bdbf840",
   "metadata": {},
   "outputs": [],
   "source": [
    "new_dict = {\"en_2nd_run\": [0, 0], \"en-paraphrase\": [0, 0], \"de_from_en\": [0, 0], \"it_from_en\": [0, 0], \"nl_from_en\": [0, 0], \"sv_from_en\": [0, 0]}\n",
    "for subtask in [\"us\", \"de\", \"it\", \"nl\", \"sv\"]:\n",
    "    for language in [\"en_2nd_run\", \"en-paraphrase\", \"de_from_en\", \"it_from_en\", \"nl_from_en\", \"sv_from_en\"]:\n",
    "        new_dict[language][0] += CC[((CC.task == \"companies_\" + subtask) \n",
    "                                    & (CC.language == language)) \n",
    "                                    & (CC.correct == \"correct\")][\"consistency\"].values[0] / 5\n",
    "        new_dict[language][1] += CC[((CC.task == \"companies_\" + subtask) \n",
    "                                    & (CC.language == language)) \n",
    "                                    & (CC.correct == \"incorrect\")][\"consistency\"].values[0] / 5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "8d1e6e83",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'en_2nd_run': [0.9897528089887639, 0.9118621378621379],\n",
       " 'en-paraphrase': [0.9416928082177112, 0.819964701964702],\n",
       " 'de_from_en': [0.9336151696102326, 0.7565654345654345],\n",
       " 'it_from_en': [0.9146601455576642, 0.691073593073593],\n",
       " 'nl_from_en': [0.9213835909935303, 0.7710782550782551],\n",
       " 'sv_from_en': [0.8887373393818154, 0.664966366966367]}"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d6d397aa",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}