evaluate.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 190,
   "id": "a50d72a3-6ee6-42e1-8406-7d024e0d0fb0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c0724570-12d6-49df-9c07-34d9b8008e74",
   "metadata": {},
   "source": [
    "## Load sample train data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "id": "4ce4d764-fd31-4d92-a925-dd1e681eeee5",
   "metadata": {},
   "outputs": [],
   "source": [
    "from create_Tatoeba_train_test import get_language_mappings, get_sentence_word_char_len"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "id": "7d928008-850b-4bbc-862c-77bc59b05279",
   "metadata": {},
   "outputs": [],
   "source": [
    "language_mappings = get_language_mappings(\"language_mappings.csv\")\n",
    "iso_639_2_English = language_mappings[[\"ISO 639-2\", \"English Name\"]].set_index(keys=\"ISO 639-2\")[\"English Name\"].to_dict()\n",
    "# iso_639_2_English"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "id": "fe032851-6103-4379-81d7-b62fa9e1aec9",
   "metadata": {},
   "outputs": [],
   "source": [
    "THIS_FILE = \"Tatoeba_stratify_train_2021-8-4_1049.csv\"\n",
    "sample = pd.read_csv(f\"output/{THIS_FILE}\")\n",
    "sample[\"Language_English_Name\"] = sample[\"Language\"].map(iso_639_2_English)\n",
    "sample[\"Sentence_len\"] = sample.apply(lambda row: get_sentence_word_char_len(row), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "id": "c5c19830-29f8-4ec2-9731-f2e945c64a39",
   "metadata": {},
   "outputs": [],
   "source": [
    "bin_labels = [\"1\",\"2\",\"3\",\"4\",\"5\",\"6\",\"7\",\"8\",\"9\",\"10\",\"11 to 16\", \"17 to 27\", \"28 to 48\", \"49 to 99\"]\n",
    "\n",
    "sample[\"Sentence_len_bin\"] = pd.cut(\n",
    "    x=sample[\"Sentence_len\"],\n",
    "    bins=[0,1,2,3,4,5,6,7,8,9,10,16,27,48,99],\n",
    "    labels=bin_labels\n",
    ")\n",
    "# sample.sample(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "id": "9cdc763f-a29e-4541-9d86-c02bfcbfc10f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Language\n",
       "cmn    4317\n",
       "dan    4317\n",
       "deu    4317\n",
       "eng    4317\n",
       "fra    4317\n",
       "ita    4317\n",
       "jpn    4316\n",
       "kor    4316\n",
       "nld    4317\n",
       "nob    4316\n",
       "por    4317\n",
       "spa    4317\n",
       "Name: Sentence, dtype: int64"
      ]
     },
     "execution_count": 195,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample.groupby([\"Language\"]).count()[\"Sentence\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "41e738cc-8006-41eb-8b96-d1bdae6ff56f",
   "metadata": {},
   "source": [
    "> There are 12 languages in the sample & ~4,317 sentences per language for a total of 51,801 sentences."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "id": "811ce511-105b-4871-9252-446ebeb7b076",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Sentence_len_bin\n",
       "1              58\n",
       "2             570\n",
       "3            1236\n",
       "4            2161\n",
       "5            3130\n",
       "6            3851\n",
       "7            4804\n",
       "8            5798\n",
       "9            6486\n",
       "10           5847\n",
       "11 to 16    13303\n",
       "17 to 27     3749\n",
       "28 to 48      645\n",
       "49 to 99      163\n",
       "Name: Sentence, dtype: int64"
      ]
     },
     "execution_count": 196,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sample.groupby([\"Sentence_len_bin\"]).count()[\"Sentence\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8dd76176-3555-4a55-8ddf-0d72b2d1136c",
   "metadata": {},
   "source": [
    "> Due to how we performed the stratified sampling each language has the same number of sentences per sentence length bin. For instance, each language should have 58 sentences of one word/character."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b4bcc949-1ae2-4093-82a9-72e58f854297",
   "metadata": {},
   "source": [
    "# langid"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c043a4f0-c452-4655-bfe4-0edd91846423",
   "metadata": {},
   "source": [
    "#### Setup for langid"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "id": "66d7b4c9-a3b1-40fe-8817-e764065284b4",
   "metadata": {},
   "outputs": [],
   "source": [
    "from langid.langid import LanguageIdentifier, model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "id": "64fa32d9-bf0d-47a4-a315-c7133d769704",
   "metadata": {},
   "outputs": [],
   "source": [
    "iso_639_2_iso_639_1 = language_mappings[[\"ISO 639-2\", \"ISO 639-1\"]].set_index(keys=\"ISO 639-2\")[\"ISO 639-1\"].to_dict()\n",
    "# iso_639_2_iso_639_1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "id": "e18cd086-c6ce-4e0c-9a48-8945a8e13952",
   "metadata": {},
   "outputs": [],
   "source": [
    "# find all languages to include in langid predictions\n",
    "set_languages = [iso_639_2_iso_639_1[i] for i in sample[\"Language\"].unique().tolist()]\n",
    "\n",
    "# unlimited langid preditions\n",
    "lang_identifier = LanguageIdentifier.from_modelstring(model, norm_probs=True)\n",
    "\n",
    "# limiting langid predictions to expected languages (sorry Galician & Afrikaans)\n",
    "lang_identifier_limited = LanguageIdentifier.from_modelstring(model, norm_probs=True)\n",
    "lang_identifier_limited.set_languages(set_languages)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7494894-8426-4544-b2e3-e6289365cd58",
   "metadata": {},
   "source": [
    "#### Generate langid predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "id": "c203b4a9-cf17-4dc3-9dbe-a26e7cd24c12",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 2min 54s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# 2min, 45s, with no limits on languages\n",
    "\n",
    "sample[\"langid\"] = sample.apply(lambda row: list(lang_identifier.classify(row[\"Sentence\"])), axis=1)\n",
    "\n",
    "sample[\"langid_language\"], sample[\"langid_score\"] = sample[\"langid\"].str\n",
    "sample.drop(columns=[\"langid\"], inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "id": "fa81be89-577d-414c-a185-5fb079902908",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 12.3 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# 12s, with limited languages\n",
    "\n",
    "sample[\"langid_limited\"] = sample.apply(lambda row: list(lang_identifier_limited.classify(row[\"Sentence\"])), axis=1)\n",
    "\n",
    "sample[\"langid_limited_language\"], sample[\"langid_limited_score\"] = sample[\"langid_limited\"].str\n",
    "sample.drop(columns=[\"langid_limited\"], inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "id": "e21af4d8-d09c-44c0-b8a5-00b4d0408036",
   "metadata": {},
   "outputs": [],
   "source": [
    "# map Language to English Name & langid_language to English Name\n",
    "ISO_639_2 = language_mappings.set_index(\"ISO 639-2\")\n",
    "ISO_639_2 = ISO_639_2[\"English Name\"].to_dict()\n",
    "sample[\"Language_English_Name\"] = sample[\"Language\"].map(ISO_639_2)\n",
    "\n",
    "ISO_639_1 = language_mappings.set_index(\"ISO 639-1\")\n",
    "ISO_639_1 = ISO_639_1[\"English Name\"].to_dict()\n",
    "sample[\"langid_language_English_Name\"] = sample[\"langid_language\"].map(ISO_639_1)\n",
    "sample[\"langid_limited_language_English_Name\"] = sample[\"langid_limited_language\"].map(ISO_639_1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "id": "6dd454f7-97e4-490f-a2e8-d42b1b9d567d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Norwegian    10.17\n",
       "Japanese      8.79\n",
       "English       8.53\n",
       "German        8.35\n",
       "Korean        8.33\n",
       "French        8.28\n",
       "Italian       8.15\n",
       "Dutch         8.14\n",
       "Spanish       7.90\n",
       "Chinese       7.88\n",
       "Name: langid_language_English_Name, dtype: float64"
      ]
     },
     "execution_count": 203,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "round(sample[\"langid_language_English_Name\"].value_counts(normalize=True, dropna=False) * 100,2)[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "id": "125e7204-9b40-4908-ab1a-bafc7bbe04c7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Japanese      8.79\n",
       "English       8.64\n",
       "Norwegian     8.50\n",
       "French        8.46\n",
       "German        8.44\n",
       "Spanish       8.40\n",
       "Korean        8.33\n",
       "Italian       8.28\n",
       "Dutch         8.26\n",
       "Portuguese    8.12\n",
       "Name: langid_limited_language_English_Name, dtype: float64"
      ]
     },
     "execution_count": 204,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "round(sample[\"langid_limited_language_English_Name\"].value_counts(normalize=True, dropna=False) * 100,2)[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fcae5725-d9d8-46df-853b-1d8d5890a9c4",
   "metadata": {},
   "source": [
    "# langdetect"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "id": "1c07cf0b-5edc-4a70-8ec5-cb186751b73e",
   "metadata": {},
   "outputs": [],
   "source": [
    "from langdetect import DetectorFactory\n",
    "DetectorFactory.seed = 7"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "id": "6f260389-5b6b-4283-bf0b-50a5e464557b",
   "metadata": {},
   "outputs": [],
   "source": [
    "from langdetect import detect_langs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "id": "41c64f4f-7e0e-4c23-b9f3-cebdd8ebe5dd",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_langdetect_prediction(this_string: str):\n",
    "    \"\"\"\n",
    "    Run langdetect on a sentence and get the results.\n",
    "    \n",
    "    Args:\n",
    "    this_string - str; the sentence to run through langdetect.\n",
    "    \n",
    "    Returns:\n",
    "    result - langdetect.language.Language; the langdetect output.\n",
    "    \"\"\"\n",
    "    assert isinstance(this_string, str), \"this_string is not of type str.\"\n",
    "    \n",
    "    result = None\n",
    "    if len(this_string)>0:\n",
    "        try:\n",
    "            result = detect_langs(this_string)[0]\n",
    "        except:\n",
    "            result = \"Error\"\n",
    "            print(f\"Error processing sentence: {this_string}\")\n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "id": "ea2be769-e42a-4c80-9de4-3b17a9e8e602",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Error processing sentence: 咦！？\n",
      "Wall time: 3min 33s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# 3min 34s\n",
    "\n",
    "sample[\"langdetect\"] = sample.apply(lambda row: get_langdetect_prediction(row[\"Sentence\"]), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "id": "207658b2-60b1-4e19-9580-ae7e9275d452",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_langdetect_lang(this_result):\n",
    "    \"\"\"\n",
    "    Get the language-component from the langdetect results.\n",
    "    \n",
    "    Args:\n",
    "    this_result - langdetect.language.Language; the langdetect output.\n",
    "    \n",
    "    Returns:\n",
    "    result - str; 2-character identifier for the predicted language.\n",
    "    \"\"\"\n",
    "    result = \"!!\"\n",
    "    if this_result:\n",
    "        try:\n",
    "            result = this_result.lang\n",
    "            result = result.split(\"-\")[0] # mainly for zh-cn\n",
    "        except:\n",
    "            pass\n",
    "    return result\n",
    "\n",
    "def get_langdetect_prob(this_result):\n",
    "    \"\"\"\n",
    "    Get the probability-component from the langdetect results.\n",
    "    \n",
    "    Args:\n",
    "    this_result - langdetect.language.Language; the langdetect output.\n",
    "    \n",
    "    Returns:\n",
    "    result - float; the probability for the predicted language.\n",
    "    \"\"\"\n",
    "    result = 0.0\n",
    "    if this_result:\n",
    "        try:\n",
    "            result = this_result.prob\n",
    "        except:\n",
    "            pass\n",
    "    return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "id": "e31301f0-bcd4-45b6-81b9-e6bf8b5e9c32",
   "metadata": {},
   "outputs": [],
   "source": [
    "sample[\"langdetect_language\"] = sample.apply(lambda row: get_langdetect_lang(row[\"langdetect\"]), axis=1)\n",
    "sample[\"langdetect_score\"] = sample.apply(lambda row: get_langdetect_prob(row[\"langdetect\"]), axis=1)\n",
    "# sample.drop(columns=[\"langdetect\"], inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "id": "d82ab74a-1c30-4791-ad42-1f6d4057bbcc",
   "metadata": {},
   "outputs": [],
   "source": [
    "sample[\"langdetect_language_English_Name\"] = sample[\"langdetect_language\"].map(ISO_639_1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "id": "3950862f-5e36-4620-9b8b-8392294027cf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ko    10.66\n",
       "no     8.93\n",
       "pt     8.36\n",
       "fr     8.35\n",
       "ja     8.26\n",
       "de     8.22\n",
       "en     8.16\n",
       "it     8.15\n",
       "es     7.87\n",
       "nl     7.51\n",
       "Name: langdetect_language, dtype: float64"
      ]
     },
     "execution_count": 212,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "round(sample[\"langdetect_language\"].value_counts(normalize=True, dropna=False) * 100,2)[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3ef3f61f-2cb4-4c94-9679-a6991c1f631a",
   "metadata": {},
   "source": [
    "# textblob\n",
    "> Currently, requests from IP are blocked by Google"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "id": "d34c3027-46d8-4841-ba0a-18e1b86728ad",
   "metadata": {},
   "outputs": [],
   "source": [
    "# from textblob import TextBlob"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "id": "b04855ca-339c-423f-832e-be9f2bdc823d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# sample[\"textblob\"] = sample.apply(lambda x: TextBlob(str(sample[\"Sentence\"])).detect_language(), axis=1) # HTTP Error 429: Too Many Requests"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "02df5ec8-e7af-4424-b755-a962007de017",
   "metadata": {},
   "source": [
    "# googletrans\n",
    "#### Unofficial & unsupported access to Google Translate API - support could be terminated at any time\n",
    "#### Only used to evaluate results & approximate differences in results\n",
    "> Currently, requests from IP are blocked by Google ?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "id": "e70b2fc1-9bd7-46f9-a8a2-11fb163f08ac",
   "metadata": {},
   "outputs": [],
   "source": [
    "# from googletrans import Translator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "id": "2c896805-c31d-4404-99ec-c32481744626",
   "metadata": {},
   "outputs": [],
   "source": [
    "# translator = Translator()\n",
    "\n",
    "# translator.detect(\"this is an English sentence.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "id": "59814272-3371-4c7b-a11b-d6515a224f02",
   "metadata": {},
   "outputs": [],
   "source": [
    "# def get_googletrans_prediction(this_string: str):\n",
    "#     \"\"\"\n",
    "#     Run googletrans on a sentence and get the results.\n",
    "    \n",
    "#     Args:\n",
    "#     this_string - str; the sentence to run through langdetect.\n",
    "    \n",
    "#     Returns:\n",
    "#     result - ???; the googletrans output.\n",
    "#     \"\"\"\n",
    "#     assert isinstance(this_string, str), \"this_string is not of type str.\"\n",
    "    \n",
    "#     result = None\n",
    "#     if len(this_string)>0:\n",
    "#         try:\n",
    "#             result = translator.detect(this_string)\n",
    "#         except Exception as e:\n",
    "#             result = \"Error\"\n",
    "#             print(repr(e))\n",
    "# #             print(f\"Error processing sentence: {this_string}\")\n",
    "#     return result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "id": "5bf3d65a-ad07-478e-8e25-86ac8d5ee3d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# short_sample = sample.sample(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "id": "276eed6c-78cd-4231-920c-74e3844abe0b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# short_sample[\"googletrans\"] = short_sample.apply(lambda x: get_googletrans_prediction(x[\"Sentence\"]), axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c4c9c128-6d68-4ea9-9d88-e2d9b8fba88f",
   "metadata": {},
   "source": [
    "# Evaluate overall results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "id": "b256ee49-7f7e-431a-8844-8ab2cbb79962",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import f1_score"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f086c3b8-53ae-4825-b3af-3b9c0a4607be",
   "metadata": {},
   "source": [
    "## F1 scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "id": "fc7dd39d-549f-47e1-b6c4-73ba86f7b870",
   "metadata": {},
   "outputs": [],
   "source": [
    "labels = sample[\"Language_English_Name\"].unique().tolist()\n",
    "# labels"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cf37edb5-b5fb-4fbe-8cf7-2cbed7e91e95",
   "metadata": {},
   "source": [
    "#### langid"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "id": "96b50468-fe12-4e63-bf9e-c6f9ad4fb870",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "langid F1: 0.94\n"
     ]
    }
   ],
   "source": [
    "langid_f1 = round(f1_score(\n",
    "    sample[\"Language_English_Name\"],\n",
    "    sample[\"langid_language_English_Name\"],\n",
    "    labels=labels,\n",
    "    average=\"weighted\"\n",
    "),2)\n",
    "\n",
    "print(f\"langid F1: {langid_f1}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "id": "b9ba210c-5d1c-4409-b49a-50869dd4e0e7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "langid limited F1: 0.95\n"
     ]
    }
   ],
   "source": [
    "langid_limited_f1 = round(f1_score(\n",
    "    sample[\"Language_English_Name\"],\n",
    "    sample[\"langid_limited_language_English_Name\"],\n",
    "    labels=labels,\n",
    "    average=\"weighted\"\n",
    "),2)\n",
    "\n",
    "print(f\"langid limited F1: {langid_limited_f1}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0719a3d2-1d6b-4c3f-a4f6-09d78e1a80e0",
   "metadata": {},
   "source": [
    "#### langdetect"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 224,
   "id": "f852d79c-6d7f-4a9d-9916-cf24cabcdf51",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "langdetect F1: 0.93\n"
     ]
    }
   ],
   "source": [
    "langdetect_f1 = round(f1_score(\n",
    "    sample[\"Language_English_Name\"],\n",
    "    sample[\"langdetect_language_English_Name\"],\n",
    "    labels=labels,\n",
    "    average=\"weighted\"\n",
    "),2)\n",
    "\n",
    "print(f\"langdetect F1: {langdetect_f1}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "555c933c-1830-421d-aac5-8213dbc56c82",
   "metadata": {},
   "source": [
    "> Overall, langid performs slightly better than langdetect, .94 & .93 respectively. Even more so when langid predictions are limited to expected languages (.95). Also, langid runs much faster when limited."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6cb53e87-a0cc-44d4-bd60-0b119f8473e4",
   "metadata": {},
   "source": [
    "## Full, simple confusion matrices"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "94758fcc-6090-4d86-9b14-18c1fc2f1de4",
   "metadata": {},
   "source": [
    "#### langid confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "id": "95538b7b-33d8-435d-961b-e791223e205e",
   "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>langid_language_English_Name</th>\n",
       "      <th>Afrikaans</th>\n",
       "      <th>Albanian</th>\n",
       "      <th>Aragonese</th>\n",
       "      <th>Basque</th>\n",
       "      <th>Bosnian</th>\n",
       "      <th>Breton</th>\n",
       "      <th>Catalan</th>\n",
       "      <th>Chinese</th>\n",
       "      <th>Croatian</th>\n",
       "      <th>Czech</th>\n",
       "      <th>...</th>\n",
       "      <th>Slovenian</th>\n",
       "      <th>Spanish</th>\n",
       "      <th>Swahili</th>\n",
       "      <th>Swedish</th>\n",
       "      <th>Tagalog</th>\n",
       "      <th>Turkish</th>\n",
       "      <th>Vietnamese</th>\n",
       "      <th>Volapük</th>\n",
       "      <th>Walloon</th>\n",
       "      <th>Welsh</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Language_English_Name</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Chinese</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4063</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Danish</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>18</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dutch</th>\n",
       "      <td>28</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>9</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>English</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>French</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>13</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>German</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Italian</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>33</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Japanese</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>15</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Korean</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Norwegian</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Portuguese</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>26</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>70</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Spanish</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>59</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>3969</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12 rows × 54 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "langid_language_English_Name  Afrikaans  Albanian  Aragonese  Basque  Bosnian  \\\n",
       "Language_English_Name                                                           \n",
       "Chinese                               0         0          0       0        0   \n",
       "Danish                                0         0          0       0        1   \n",
       "Dutch                                28         3          0       1        3   \n",
       "English                               0         0          0       2        0   \n",
       "French                                1         0          0       2        0   \n",
       "German                                4         0          1       2        0   \n",
       "Italian                               0         0          0       4        0   \n",
       "Japanese                              0         0          0       0        0   \n",
       "Korean                                0         0          0       0        0   \n",
       "Norwegian                             0         0          0       3        0   \n",
       "Portuguese                            0         0          2       0        0   \n",
       "Spanish                               1         0          3       6        0   \n",
       "\n",
       "langid_language_English_Name  Breton  Catalan  Chinese  Croatian  Czech  ...  \\\n",
       "Language_English_Name                                                    ...   \n",
       "Chinese                            0        0     4063         0      0  ...   \n",
       "Danish                             0        0        0         3      0  ...   \n",
       "Dutch                              1        0        1         2      0  ...   \n",
       "English                            0        0        0         0      0  ...   \n",
       "French                            12       13        0         1      0  ...   \n",
       "German                             1        0        0         0      0  ...   \n",
       "Italian                            2        8        0         1      1  ...   \n",
       "Japanese                           0        0       15         0      0  ...   \n",
       "Korean                             0        0        0         0      1  ...   \n",
       "Norwegian                          0        1        0         2      0  ...   \n",
       "Portuguese                         2       26        0         0      2  ...   \n",
       "Spanish                            0       59        3         1      4  ...   \n",
       "\n",
       "langid_language_English_Name  Slovenian  Spanish  Swahili  Swedish  Tagalog  \\\n",
       "Language_English_Name                                                         \n",
       "Chinese                               0        0        0        0        0   \n",
       "Danish                                0        3        0       18        1   \n",
       "Dutch                                 9        2        0        1        0   \n",
       "English                               0        5        0        2        0   \n",
       "French                                0        6        0        0        0   \n",
       "German                                0        2        0       12        0   \n",
       "Italian                               2       33        0        2        0   \n",
       "Japanese                              0        0        0        0        0   \n",
       "Korean                                0        0        0        0        0   \n",
       "Norwegian                             2        2        0       17        0   \n",
       "Portuguese                            1       70        1        0        1   \n",
       "Spanish                               2     3969        0        0        1   \n",
       "\n",
       "langid_language_English_Name  Turkish  Vietnamese  Volapük  Walloon  Welsh  \n",
       "Language_English_Name                                                       \n",
       "Chinese                             0           0        0        0      0  \n",
       "Danish                              0           0        0        0      3  \n",
       "Dutch                               0           0        0        0      0  \n",
       "English                             0           0        0        0      0  \n",
       "French                              1           0        0        2      1  \n",
       "German                              0           0        0        0      1  \n",
       "Italian                             0           0        0        1      0  \n",
       "Japanese                            0           0        0        0      0  \n",
       "Korean                              0           0        0        0      0  \n",
       "Norwegian                           2           0        1        0      2  \n",
       "Portuguese                          0           1        0        2      2  \n",
       "Spanish                             0           0        0        1      1  \n",
       "\n",
       "[12 rows x 54 columns]"
      ]
     },
     "execution_count": 225,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(sample[\"Language_English_Name\"], sample[\"langid_language_English_Name\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "851c4fde-1d10-4182-b6f3-4b8e4551aaba",
   "metadata": {},
   "source": [
    "#### langdetect confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "id": "c48161eb-6a47-4713-a20c-744f58422c38",
   "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>langdetect_language_English_Name</th>\n",
       "      <th>Afrikaans</th>\n",
       "      <th>Albanian</th>\n",
       "      <th>Catalan</th>\n",
       "      <th>Chinese</th>\n",
       "      <th>Croatian</th>\n",
       "      <th>Czech</th>\n",
       "      <th>Danish</th>\n",
       "      <th>Dutch</th>\n",
       "      <th>English</th>\n",
       "      <th>Error</th>\n",
       "      <th>...</th>\n",
       "      <th>Slovak</th>\n",
       "      <th>Slovenian</th>\n",
       "      <th>Somali</th>\n",
       "      <th>Spanish</th>\n",
       "      <th>Swahili</th>\n",
       "      <th>Swedish</th>\n",
       "      <th>Tagalog</th>\n",
       "      <th>Turkish</th>\n",
       "      <th>Vietnamese</th>\n",
       "      <th>Welsh</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Language_English_Name</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Chinese</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3106</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Danish</th>\n",
       "      <td>17</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>3394</td>\n",
       "      <td>10</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>47</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dutch</th>\n",
       "      <td>391</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>3832</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>English</th>\n",
       "      <td>28</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>14</td>\n",
       "      <td>20</td>\n",
       "      <td>4142</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>French</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>19</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>German</th>\n",
       "      <td>21</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>15</td>\n",
       "      <td>23</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Italian</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>18</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>16</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>22</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Japanese</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>11</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Korean</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Norwegian</th>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>434</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>44</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Portuguese</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>56</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>76</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Spanish</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>74</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>3967</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12 rows × 35 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "langdetect_language_English_Name  Afrikaans  Albanian  Catalan  Chinese  \\\n",
       "Language_English_Name                                                     \n",
       "Chinese                                   3         0        0     3106   \n",
       "Danish                                   17         0        0        0   \n",
       "Dutch                                   391         1        0        0   \n",
       "English                                  28         1        2        0   \n",
       "French                                    4         0       19        0   \n",
       "German                                   21         0        1        0   \n",
       "Italian                                   0         0       18        0   \n",
       "Japanese                                  0         0        0       11   \n",
       "Korean                                    0         0        0        0   \n",
       "Norwegian                                11         1        0        0   \n",
       "Portuguese                                0         0       56        0   \n",
       "Spanish                                   1         0       74        0   \n",
       "\n",
       "langdetect_language_English_Name  Croatian  Czech  Danish  Dutch  English  \\\n",
       "Language_English_Name                                                       \n",
       "Chinese                                  0      0       0      0        4   \n",
       "Danish                                   6      0    3394     10        7   \n",
       "Dutch                                   10      0       6   3832       10   \n",
       "English                                  1      0      14     20     4142   \n",
       "French                                   3      1       1      4        7   \n",
       "German                                   0      0       5     15       23   \n",
       "Italian                                  1      0       4      1       16   \n",
       "Japanese                                 0      0       0      0        0   \n",
       "Korean                                   0      0       0      0        0   \n",
       "Norwegian                                3      0     434      5        4   \n",
       "Portuguese                               4      0       0      1        3   \n",
       "Spanish                                  1      1       1      0        9   \n",
       "\n",
       "langdetect_language_English_Name  Error  ...  Slovak  Slovenian  Somali  \\\n",
       "Language_English_Name                    ...                              \n",
       "Chinese                               1  ...       4          0       0   \n",
       "Danish                                0  ...       1          3       1   \n",
       "Dutch                                 0  ...       0          6       4   \n",
       "English                               0  ...       3          4       4   \n",
       "French                                0  ...       2          2       0   \n",
       "German                                0  ...       1          0       0   \n",
       "Italian                               0  ...       2          4       0   \n",
       "Japanese                              0  ...       0          0       0   \n",
       "Korean                                0  ...       0          0       0   \n",
       "Norwegian                             0  ...       0          5       1   \n",
       "Portuguese                            0  ...       2          0       0   \n",
       "Spanish                               0  ...       5          3       5   \n",
       "\n",
       "langdetect_language_English_Name  Spanish  Swahili  Swedish  Tagalog  Turkish  \\\n",
       "Language_English_Name                                                           \n",
       "Chinese                                 1        2        1        1        1   \n",
       "Danish                                  1        0       47        2        1   \n",
       "Dutch                                   1        1        4        0        1   \n",
       "English                                 3        0        2       10        1   \n",
       "French                                  6        0        0        1        1   \n",
       "German                                  2        0        9        3        2   \n",
       "Italian                                22        1        0        6        0   \n",
       "Japanese                                0        0        0        0        0   \n",
       "Korean                                  0        0        0        0        0   \n",
       "Norwegian                               0        1       44        1        1   \n",
       "Portuguese                             76        1        0        0        0   \n",
       "Spanish                              3967        1        0        1        1   \n",
       "\n",
       "langdetect_language_English_Name  Vietnamese  Welsh  \n",
       "Language_English_Name                                \n",
       "Chinese                                    6      0  \n",
       "Danish                                     0      5  \n",
       "Dutch                                      0      0  \n",
       "English                                    1     11  \n",
       "French                                     1      0  \n",
       "German                                     0      0  \n",
       "Italian                                    1      1  \n",
       "Japanese                                   0      0  \n",
       "Korean                                     0      0  \n",
       "Norwegian                                  0      2  \n",
       "Portuguese                                 1      0  \n",
       "Spanish                                    0      1  \n",
       "\n",
       "[12 rows x 35 columns]"
      ]
     },
     "execution_count": 226,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(sample[\"Language_English_Name\"], sample[\"langdetect_language_English_Name\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9ada44c4-73ba-43bd-aeec-7f412aacd010",
   "metadata": {},
   "source": [
    "## Limited, colorful confusion matrices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 227,
   "id": "8f6df35f-3b37-4d96-b8db-ea2a1a9faa64",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import itertools"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "id": "15e2fe22-c051-46a3-a710-83c6a0ad0b53",
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"\n",
    "Below is a modified function from user Calvin Duy Canh Tran on Stack Overflow to plot a confusion matrix with any number of labels.\n",
    "You can check out the original post on SO here: https://stackoverflow.com/a/50386871/1576558.\n",
    "\"\"\"\n",
    "\n",
    "def plot_confusion_matrix(cm,\n",
    "                          target_names,\n",
    "                          title='Confusion matrix',\n",
    "                          cmap=None,\n",
    "                          normalize=True):\n",
    "    \"\"\"\n",
    "    given a sklearn confusion matrix (cm), make a nice plot\n",
    "\n",
    "    Arguments\n",
    "    ---------\n",
    "    cm:           confusion matrix from sklearn.metrics.confusion_matrix\n",
    "\n",
    "    target_names: given classification classes such as [0, 1, 2]\n",
    "                  the class names, for example: ['high', 'medium', 'low']\n",
    "\n",
    "    title:        the text to display at the top of the matrix\n",
    "\n",
    "    cmap:         the gradient of the values displayed from matplotlib.pyplot.cm\n",
    "                  see http://matplotlib.org/examples/color/colormaps_reference.html\n",
    "                  plt.get_cmap('jet') or plt.cm.Blues\n",
    "\n",
    "    normalize:    If False, plot the raw numbers\n",
    "                  If True, plot the proportions\n",
    "\n",
    "    Usage\n",
    "    -----\n",
    "    plot_confusion_matrix(cm           = cm,                  # confusion matrix created by\n",
    "                                                              # sklearn.metrics.confusion_matrix\n",
    "                          normalize    = True,                # show proportions\n",
    "                          target_names = y_labels_vals,       # list of names of the classes\n",
    "                          title        = best_estimator_name) # title of graph\n",
    "\n",
    "    Citiation\n",
    "    ---------\n",
    "    http://scikit-learn.org/stable/auto_examples/model_selection/plot_confusion_matrix.html\n",
    "\n",
    "    \"\"\"\n",
    "    \n",
    "    np.fill_diagonal(cm, 0) # fill the diagonal with zeroes to emphasize errors not correct predictions\n",
    "    \n",
    "    accuracy = np.trace(cm) / np.sum(cm).astype('float')\n",
    "    misclass = 1 - accuracy\n",
    "\n",
    "    if cmap is None:\n",
    "        cmap = plt.get_cmap('Purples')\n",
    "\n",
    "    plt.figure(figsize=(20, 10))\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "\n",
    "    if target_names is not None:\n",
    "        tick_marks = np.arange(len(target_names))\n",
    "        plt.xticks(tick_marks, target_names, rotation=90)\n",
    "        plt.yticks(tick_marks, target_names)\n",
    "\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "\n",
    "\n",
    "    thresh = cm.max() / 1.5 if normalize else cm.max() / 2\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        if normalize:\n",
    "            plt.text(j, i, \"{:0.4f}\".format(cm[i, j]),\n",
    "                     horizontalalignment=\"center\",\n",
    "                     color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "        else:\n",
    "            plt.text(j, i, \"{:,}\".format(cm[i, j]),\n",
    "                     horizontalalignment=\"center\",\n",
    "                     color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "id": "8a098a10-8c17-409c-86a0-9849be4b9019",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "13d44862-52fb-4174-926d-8975c14909e2",
   "metadata": {},
   "source": [
    "#### langid confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "id": "403bf4b8-ee5a-48d2-ba05-8bb7f2589531",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cm = confusion_matrix(sample[\"Language_English_Name\"], sample[\"langid_language_English_Name\"], labels)\n",
    "plot_confusion_matrix(cm=cm, target_names=labels, normalize=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ad5262c4-5975-4030-b60b-222e0a62cbf3",
   "metadata": {},
   "source": [
    "> langid struggled the most with Danish (1,345 errors), Norwegian (308), & Chinese (254)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "baebec83-3d8e-4737-b538-2a37b31240d9",
   "metadata": {},
   "source": [
    "#### langid limited confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "id": "24dab101-e79a-4997-8e9e-2b8b1bf066d1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cm = confusion_matrix(sample[\"Language_English_Name\"], sample[\"langid_limited_language_English_Name\"], labels)\n",
    "plot_confusion_matrix(cm=cm, target_names=labels, normalize=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d8617e6f-db74-4c9e-8c23-527b7f49a999",
   "metadata": {},
   "source": [
    "> There are some unexpected differences between the limited langid predictions & the unlimited ones. Namely, the performance on Norwegian is significantly worse (2x), Portuguese suddenly becomes more problematic, & it performs exactly the same (errors on the same sentences?) on Chinese mistaking 254 sentences for Japanese. Regardless, the limited version of langid also struggled with Danish (820 errors), Norwegian (653), & Chinese (254)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cded3a6c-bf88-48de-ab34-935a2c12f902",
   "metadata": {},
   "source": [
    "#### langdetect confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "id": "4d7f755f-806c-4a08-bea1-8bdf788764b6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0wAAALWCAYAAAB1FxMmAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8rg+JYAAAACXBIWXMAAAsTAAALEwEAmpwYAACYmUlEQVR4nOzdeZxVdf348dcbRjRFQ3NhExFUQJB9sQVyTUFNLQVxQUXEft/K0jJb1XI3W8C1stxzycoAFXJfUNkElXBDodhSySVxA4fP74+5jDPkVZjt3Ht4PXncB/cs95z3Z865Z+Z935/zuZFSQpIkSZL0v5plHYAkSZIklSoTJkmSJEkqwoRJkiRJkoowYZIkSZKkIkyYJEmSJKkIEyZJkiRJKsKESZLKTER8KiImRsSbEfGnemznqIj4e0PGlpWIGBwRz2UdhyQpf8LvYZKkxhERRwKnAl2Bt4A5wLkppUfqud1jgG8Cn0spfVDfOEtdRCRg55TS/KxjkSRteKwwSVIjiIhTgV8D5wHbAR2Ay4GDG2DzOwDPbwjJ0rqIiIqsY5Ak5ZcJkyQ1sIj4NPAz4Osppb+klN5OKa1KKU1MKZ1WWGfjiPh1RCwtPH4dERsXlu0REYsj4jsR8UpELIuI4wvLfgqcAYyIiBURcUJEnBURN9TYf8eISGsSiYg4LiJeioi3ImJBRBxVY/4jNV73uYiYUejqNyMiPldj2QMRcXZETC1s5+8RsXWR9q+J/3s14j8kIoZFxPMR8VpE/LDG+gMj4rGIeKOw7qUR0aKw7KHCak8W2juixvZPj4h/A1evmVd4TefCPvoWpttGxKsRsUd9jqskacNkwiRJDe+zwCbAXz9mnR8BuwO9gV7AQODHNZa3Bj4NtANOAC6LiC1TSmdSVbW6JaXUMqX0+48LJCI2A8YDQ1NKmwOfo6pr4NrrbQXcUVj3M8AvgTsi4jM1VjsSOB7YFmgBfPdjdt2aqp9BO6oSvN8BRwP9gMHATyJix8K6lcApwNZU/ez2Bv4PIKU0pLBOr0J7b6mx/a2oqraNrbnjlNKLwOnADRGxKXA1cG1K6YGPiVeSpI9kwiRJDe8zwPJP6DJ3FPCzlNIrKaVXgZ8Cx9RYvqqwfFVK6U5gBdCljvGsBnpExKdSSstSSv/4iHUOAF5IKV2fUvogpXQT8CxwUI11rk4pPZ9Sehe4lapkr5hVVN2vtQq4mapkaFxK6a3C/udRlSiSUpqVUnq8sN+FwG+AL65Dm85MKb1fiKeWlNLvgPnANKANVQmqJEnrzYRJkhref4CtP+HemrbAP2tM/7Mwr3obayVc7wAt1zeQlNLbwAjga8CyiLgjIrquQzxrYmpXY/rf6xHPf1JKlYXnaxKal2ssf3fN6yNil4iYFBH/joj/UlVB+8jufjW8mlJ67xPW+R3QA7gkpfT+J6wrSdJHMmGSpIb3GPA+cMjHrLOUqu5ka3QozKuLt4FNa0y3rrkwpTQlpbQvVZWWZ6lKJD4pnjUxLaljTOvjCqri2jmltAXwQyA+4TUfO8RrRLSkatCN3wNnFbocSpK03kyYJKmBpZTepOq+ncsKgx1sGhEbRcTQiLiosNpNwI8jYpvC4AlnADcU2+YnmAMMiYgOhQEnfrBmQURsFxEHF+5lep+qrn2rP2IbdwK7RMSREVERESOAXYFJdYxpfWwO/BdYUah+/b+1lr8MdFrPbY4DZqaUxlB1b9aV9Y5SkrRBMmGSpEaQUvoFVd/B9GPgVWAR8A3g9sIq5wAzgaeAp4EnCvPqsq+7gVsK25pF7SSnWSGOpcBrVN0btHZCQkrpP8CBwHeo6lL4PeDAlNLyusS0nr5L1YASb1FV/bplreVnAdcWRtEb/kkbi4iDgf35sJ2nAn3XjA4oSdL68ItrJUmSJKkIK0ySJEmSVIQJkyRJkiQVYcIkSZIkSUWYMEmSJElSER/3pYqqg6233jp13KFj1mHoI+RteJNP+pKacpOn45O3Y7N40RtZh9Cg2m/fKusQVESergOQr2tB3o7NE0/MWp5S2ibrOErRVrFTWsU7Tb7fFSybklLav8l3vA5MmBpYxx06Mm3a9KzD0EfI24iQEXn6VZyv45O3Y/PDb03MOoQGdd64g7IOQUXk6ToA+boW5O3YbNSi4p9Zx1CqVvEO/RjT5Pt9kLO3bvKdriO75EmSJElSEVaYJEmSJFXLpDpawkVMK0ySJEmSVIQVJkmSJEkfyuL2OytMkiRJklR+TJgkSZIkqQi75EmSJEkCqnrjRbMM+uStbvpdrisrTJIkSZJUhBUmSZIkSVUCcvSdyw3CCpMkSZIkFWGFSZIkSdKHLDHVYoVJkiRJkoowYZIkSZKkIuySJ0mSJKmaPfJqs8IkSZIkSUVYYZIkSZJUENl8cW0Js8IkSZIkSUWYMJWhyVMms2v3bnTpugsXXnRh1uHUS57a8txzz9Gvf7/qx1af2ZJx48dlHVadLVq0iL332ZvdevagZ6/dGD9+fNYh1Vnejs2YMSfQpm1revXumXUo6+ULe3bi1B/vySk/2pORx/ejoqIZW35mU75+2hBOO2tvjhzdn+bNqz7V3HGnz3Dy6V/kvPEHsVufNhlHvn7ydF0r9+vAmBPH0LZdG3r37lU977XXXmP/ofvRbdeu7D90P15//fUMI6yfPJ1rb7zxBiNGDKdHj+7stlsPHnv8saxDkqqZMJWZyspKTj75m0yaeAdPPzWXW26+mXnz5mUdVp3kqS0AXbp0YdbMWcyaOYvp06az6aabcsjBh2QdVp1VVFTw84t+ztNPzWXqI49yxZWXl+3xyduxGXXssdwx6c6sw1gvW3x6Ez6/RyfGX/ggvzr3fpo1C3r1b8ewQ3blkfte5Odn3cu776xkwOd2AOCN197h1utnM2fmkowjXz95u66V+3Xg2FGjmDTpjlrzLrroQvbacy+emfcse+25FxeVaaKRt3PtlFNP4Uv77cfcuf9g1qwn6Na1W9YhbbiCqlEfmvpRwkyYysz06dPp3LkznTp1okWLFgwfMYIJEydkHVad5Kkta7vvvnvp1KkTO+ywQ9ah1FmbNm3o27cvAJtvvjldu3ZlydLy+uP1o+Th2AwZPIStttoq6zDWW7Pmzdhoo+Y0axZstFFz3nrzPTrvsjVPz14KwKxpi+jes6qa9Ppr7/Lvpf8lpZRlyOstb9e1cr8ODB48hK22rP1emThxIsccMwqAY44ZxYQJ5Xl88nSuvfnmmzzyyMOMPn40AC1atKBVq1bZBiXVYMJUZpYuXcL27bevnm7frh1Ll5TPL6+a8tSWtd1y662MGHFE1mE0mIULFzJnzhwGDRyUdSj1lrdjUy7+++Z7PHTPfH5wzpf40Xn78d57q1j8rzd5991VrF5dlRS9+fq7bNFqk4wjrZ88X9fych14+ZWXadOmKjFv3bo1L7/ycsYR1U2ezrUFCxaw9dZbc8KYE+g/oD9jTxrL22+/nXVYGzQLTLWVXcIUEZURMSci/hERT0bEdyKizu2IiKsiYtePWf5ARPSv6/a14Vm5ciWTJk3ksK8elnUoDWLFihUMH344v/zFL9liiy2yDqde8nZsysmnPrURu/ZszYVn3M25P5xCixYVdNl126zD0jrK03WgpoggSv0vtQ3AB5UfMHv2bE466SRmzpjJZpttVrZdJZVPZZcwAe+mlHqnlLoD+wJDgTPrurGU0piUUtl0+m3bth2LFi+qnl68ZAlt27XLMKK6y1Nbapo8eTJ9+vRhu+22yzqUelu1ahWHDz+MkSOP5NBDv5J1OPWWp2NTbnbqug2v/+cd3l6xktWrE3PnLGOHzlvxqU9tRLPC8LWf3vJT/PeN9zKOtH7yeF3L23Vgu223Y9myZQAsW7aMbbcpz8Q9T+da+3btad++fXX18qtf+Qqz58zOOCrpQ+WYMFVLKb0CjAW+EVU6RsTDEfFE4fE5gIjYo1Apui0ino2IG6PwkdKaClJENI+IayJibkQ8HRGn1NjV4RExPSKej4jBGTS12oABA5g/fz4LFixg5cqV3HrLLRx04EFZhlRneWpLTbfccnMuunyllDjxxDF069qNU0455ZNfUAbycmzK0Ruvv0uHHbdko42aA7BTl615ZdlbvPj8cnbr0xaAfoO25x9PLcsyzHrL23Utj9eBAw86kOuvvw6A66+/joMOKs/jk6dzrXXr1rRv357nnnsOgPvuu49u3Rz0IStVYz5Ekz9KWdl/cW1K6aWIaA5sC7wC7JtSei8idgZuAtZ0p+sDdAeWAlOBzwOP1NhUb6BdSqkHQES0qrGsIqU0MCKGUVXN2qdmDBExlqrEjQ4dOjRo+9ZWUVHBuHHjGXbAUCorKznuuOPp3r17o+6zseSpLWu8/fbb3HPvPVx++RVZh1JvU6dO5YYbb2C3HrvRr1/VTd9nn3MOw4YOyziyusnTsTnq6CN58MEHWb58OTt07MCZZ5zJ6NEnZB3Wx1q08HWenr2Uk7//RVavTixd/CbTpv6TZ+e+zJGj+/Olg7qydNGbzHjsXwC079CKUWMH8qlNN6Jbj9bse0BXfnnO/Rm34pPl7bpW7teBo48+igcfqnqvdNxxB84440y+d9rpjDzyCK6+5mo6dOjATX+8Oesw6yRv59qvfzWOUceOYuXKlXTacUeuuur3WYckVYtyG4EoIlaklFquNe8NoAvwHnApVclPJbBLSmnTiNgD+FFKad/C+lcAU1NKN0TEA8B3gReBmcCdwB3A31NKqwvLf5RSmhoR2xVet1Ox+Pr365+mTZvecA1Wgym3c/2TlPqnMesrT8cnb8fmh9+amHUIDeq8ceX5KfyGIE/XAcjXtSBvx2ajFhWzUkreo/4RPt28Xdp9s//X5Pv9+1s/KdljUtZd8gAiohNVydErwCnAy0AvqipLLWqs+n6N55WsVV1LKb1eeN0DwNeAqz7itf/zOkmSJEn5VdZ//EfENsCVwKUppRQRnwYWFypDxwLN12NbWwMrU0p/jojngBsaJ2pJkiSpRAVEs/xURxtCOSZMn4qIOcBGwAfA9cAvC8suB/4cEaOAycD6DOLfDri6xhDlP2iYcCVJkiSVq7JLmFJKRatGKaUXgJ41Zp1emP8AVV3t1qz3jRrP96ixft+P2OYeNZ4vBzqub8ySJEmSylPZJUySJEmSGk+OxitpEGU/6IMkSZIkNRYrTJIkSZI+ZImpFitMkiRJklSECZMkSZIkFWGXPEmSJEnV7JFXmxUmSZIkSSrCCpMkSZKkKhFEM0tMNVlhkiRJkqQiTJgkSZIkqQi75EmSJEn6kKM+1GKFSZIkSZKKsMIkSZIkCYDAAtParDBJkiRJUhFWmCRJkiRVC0tMtVhhkiRJkqQiTJgkSZIkqQi75EmSJEn6kD3yarHCJEmSJElFWGGSJEmSVCUgmlliqsmESRsMR3wpbR6f0nXeuIOyDqFBVVauzjqEBtO8eb46iuTtOpBSyjqEBpO3YyOtj3xdaSVJkiSpAVlhkiRJkvQhC4q1WGGSJEmSpCKsMEmSJEmq5j1rtVlhkiRJkqQirDBJkiRJKggrTGuxwiRJkiRJRZgwSZIkSVIRdsmTJEmSVCWwpLIWfxySJEmSVIQVJkmSJElAVYHJQR9qs8IkSZIkSUWYMEmSJElSEXbJkyRJklTNHnm1WWGSJEmSpCKsMEmSJEn6kCWmWqwwSZIkSVIRJkySJEmSVIQJUxmaPGUyu3bvRpeuu3DhRRdmHU695KktYHtKWZ7aAvlqT7m3ZdGiRXzpS/vQq3dPevfpxSWXjgfgySfnMHjI5xkwsB+f/dwgZsyYnnGkdVPux6emcm/LmBPH0LZdG3r37lU977bbbqNXr5602HgjZs6amWF09VPuxyZXoqpHXlM/SpkJU5mprKzk5JO/yaSJd/D0U3O55eabmTdvXtZh1Ume2gK2p5TlqS2Qr/bkoS0VFRVceOFFPDnnKR5+6BGuvPJKnnlmHj/44Q/40Y9+wozpszjjjLP44Q9/kHWo6y0Px2eNPLTl2FGjmDTpjlrzunfvzq23/onBgwdnFFX95eHYKN9MmMrM9OnT6dy5M506daJFixYMHzGCCRMnZB1WneSpLWB7Slme2gL5ak8e2tKmTRv69OkLwOabb07Xrl1ZsmQpEcFb//0vAP99803atGmbZZh1kofjs0Ye2jJ48BC22nKrWvO6detGly5dMoqoYeTh2ORNNIsmf5QyE6Yys3TpErZvv331dPt27Vi6ZEmGEdVdntoCtqeU5aktkK/25KktAAsXLuTJOXMYOHAgF1/8C37wg+/TufOOfP8Hp3P22edkHd56y9PxyVNb8sZjo1JXMglTRKzIOgZJkupqxYoVHDFyOBdf/Au22GILfvvb3/Dzn1/Miy8u4OcXXcxJXxubdYiStG68iamWkkmYtG7atm3HosWLqqcXL1lC23btMoyo7vLUFrA9pSxPbYF8tScvbVm1ahUjjhjOEUeM5JBDDgXghhuur37+1a8exsyZM7IMsU7ycnwgX23JG4+NSl1JJUwR0TIi7o2IJyLi6Yg4uDC/Y0Q8GxE3RsQzEXFbRGxaWHZGRMyIiLkR8duIqhQ1Ih6IiAsjYnpEPB8Rgwvzm0fEzwuveSoiTirMbxMRD0XEnMK21qz/pYh4rBDTnyKiZTY/nSoDBgxg/vz5LFiwgJUrV3LrLbdw0IEHZRlSneWpLWB7Slme2gL5ak8e2pJS4qSTTqRr1658+1unVM9v06YtDz30EAD3338/O+20U1Yh1lkejs8aeWpL3nhsVOoqsg5gLe8Bh6aU/hsRWwOPR8Sau/66ACeklKZGxB+A/wMuBi5NKf0MICKuBw4EJhZeU5FSGhgRw4AzgX2AE4A3U0oDImJjYGpE/B34CjAlpXRuRDQHNi3E8GNgn5TS2xFxOnAq8LOaQUfEWGAsQIcOHRrlB7NGRUUF48aNZ9gBQ6msrOS4446ne/fujbrPxpKntoDtKWV5agvkqz15aMujj07lxj/eSI8ePRgwsB8AP/vZOVxx+RV857un8sEHH7DJJptw+WVXZBzp+svD8VkjD205+uijePChB1m+fDkdd9yBM844k6223Ipvn/ItXn31VQ4++Mv06tWLO++4K+tQ10sejk2eBCXfQ67JRUop6xiA6nuYtgR+BQwBVlOVJO0IbAI8lFLqUFh3L+DklNIhEfFV4HvApsBWwCUppQsi4gHgR4UEaztgakppp4i4DegJvFPY9aeBk6hK1v4A3ADcnlKaExEHAtcAiwvrtgAeSymdUKwd/fv1T9Omled3bUjShqCycnXWITSY5s1LqqOI1lIqf2M1hMjZX9AVGzWflVLqn3UcpWirTbZPe3f8TpPv97bnTinZY1JqFaajgG2AfimlVRGxkKpkCWDtq06KiE2Ay4H+KaVFEXFWjfUB3i/8X8mHbQ3gmymlKWvvPCKGAAcA10TEL4HXgbtTSiPr3TJJkiSp1EXkLkGur1L7aOrTwCuFZGlPYIcayzpExGcLz48EHuHD5Gh54d6iw9ZhH1OA/xcRGwFExC4RsVlE7AC8nFL6HXAV0Bd4HPh8ROxUWHeziNilnm2UJEmSVCZKosIUERVUVYNuBCZGxNPATODZGqs9B3y9cP/SPOCKlNI7EfE7YC7wb2BdhiC6CugIPFEYIOJV4BBgD+C0iFgFrABGpZRejYjjgJsK9ztB1T1Nz9e9tZIkSZLKRUkkTEB34MWU0nLgs2svjIiOwAcppaPXXpZS+jFVScza8/eo8Xw5VUkSKaXVwA8Lj5quLTzW3s59wIB1bokkSZJUzkqtD1rGMv9xRMTXgJv4iKRHkiRJkrKUeYUppXQlcOUnrLMQ6NEkAUmSJEkbMAd9qC3zCpMkSZIklarMK0ySJEmSSocVptqsMEmSJElSESZMkiRJklSEXfIkSZIkVQkISyq1+OOQJEmSVPIi4g8R8UpEzK0xb6uIuDsiXij8v2VhfkTE+IiYHxFPRUTfGq85trD+CxFx7Cft14RJkiRJ0ocimv6xbq4B9l9r3veBe1NKOwP3FqYBhgI7Fx5jgSuqmhZbAWcCg4CBwJlrkqxiTJgkSZIklbyU0kPAa2vNPhi4tvD8WuCQGvOvS1UeB1pFRBtgP+DulNJrKaXXgbv53ySsFu9hkiRJkpS1rSNiZo3p36aUfrsOr9supbSs8PzfwHaF5+2ARTXWW1yYV2x+USZMkiRJkgAI1qeHXINanlLqX58NpJRSRKSGCmgNu+RJkiRJKlcvF7raUfj/lcL8JcD2NdZrX5hXbH5RJkySJEmSqgREs2jyRz1MANaMdHcs8Lca80cVRsvbHXiz0HVvCvCliNiyMNjDlwrzirJLniRJkqSSFxE3AXtQdb/TYqpGu7sAuDUiTgD+CQwvrH4nMAyYD7wDHA+QUnotIs4GZhTW+1lKae2BJGoxYZIkSZJU8lJKI4ss2vsj1k3A14ts5w/AH9Z1vyZMkiRJkgrW63uRNgjewyRJkiRJRVhhkiRJklTNAlNtJkxSmarqmis1vsjZb87mzfPTuSJv14G8nWt5a4+0oTJhkiRJklStnsN8505+PmaTJEmSpAZmwiRJkiRJRdglT5IkSVKVwFEf1mKFSZIkSZKKsMIkSZIkqZoFptqsMEmSJElSESZMkiRJklSEXfIkSZIkAYUxH/weplqsMEmSJElSEVaYJEmSJH3IAlMtVpgkSZIkqQgrTJIkSZKqBITjitdihUmSJEmSijBhkiRJkqQi7JInSZIkqSAcVnwtVpgkSZIkqQgrTJIkSZKqOeZDbVaYysyYMSfQpm1revXumXUoDWLylMns2r0bXbruwoUXXZh1OPVW7u0Zc+IY2rZrQ+/evarnnf7979GjR3f69O3DYYd9lTfeeCO7AOvoueeeo1//ftWPrT6zJePGj8s6rHp54403GDFiOD16dGe33Xrw2OOPZR1SnSxatIi999mb3Xr2oGev3Rg/fnzWIdVbuV8Hahp/yXh69+5Fr149y/49s0ZlZSX9+/fjywcflHUo9ZaXcy2P1wHliwlTmRl17LHcMenOrMNoEJWVlZx88jeZNPEOnn5qLrfcfDPz5s3LOqw6y0N7jh01ikmT7qg1b5+992HOnCeZ/cRsdt55Zy688IKMoqu7Ll26MGvmLGbNnMX0adPZdNNNOeTgQ7IOq15OOfUUvrTffsyd+w9mzXqCbl27ZR1SnVRUVPDzi37O00/NZeojj3LFlZeX3fumpjxcB9aYO3cuf/j973n00ceYNesJ7rzzDubPn591WPU2fvx4unbrmnUY9Zancy1v1wHljwlTmRkyeAhbbbVV1mE0iOnTp9O5c2c6depEixYtGD5iBBMmTsg6rDrLQ3sGDx7CVlvWPr/23fdLVFRU9d4dNGh3Fi9ZkkVoDea+++6lU6dO7LDDDlmHUmdvvvkmjzzyMKOPHw1AixYtaNWqVbZB1VGbNm3o27cvAJtvvjldu3ZlydLyPcfycB1Y49lnn2XAwIFsuummVFRUMGTwEG6//a9Zh1Uvixcv5s677mT06BOyDqXe8nSu5e06kAsRTf8oYSZMyszSpUvYvv321dPt27VjaRn/MZ639nyUa665mv332z/rMOrllltvZcSII7IOo14WLFjA1ltvzQljTqD/gP6MPWksb7/9dtZh1dvChQuZM2cOgwYOyjqUOsvTdaB79+5MfeQR/vOf//DOO+9w1+S7WLR4cdZh1cup3zmFC86/gGbNyv/PnzydazXl4Tqg/CnLK0ZEVEbEnBqPjo20nz0iYlJjbFsqN+effx4VFRUceeSRWYdSZytXrmTSpIkc9tXDsg6lXj6o/IDZs2dz0kknMXPGTDbbbDMuKuP7FwBWrFjB8OGH88tf/JItttgi63AEdOvWje+edhpDhw3lgAOH0atXb5o3b551WHU26Y5JbLvNtvTr1y/rUFSE14ESERDNoskfpawsEybg3ZRS7xqPhWsWRJVybdcGpW3bdixavKh6evGSJbRt1y7DiOonb+2p6drrruWOO+/guuuuJ0q8bP5xJk+eTJ8+fdhuu+2yDqVe2rdrT/v27as/gf3qV77C7DmzM46q7latWsXhww9j5MgjOfTQr2QdTr3k7Tow+vjRTJ82nfvve4AtW7Vi5513zjqkOnv00UeZOGkinXfqxFFHHcn999/PqFHHZB1WneXtXMvTdUD5k4vEIiI6RsRzEXEdMBfYPiJOi4gZEfFURPy0xnrPRMTvIuIfEfH3iPhUYdlOEXFPRDwZEU9EROfC5ltGxG0R8WxE3Bjl/NdiiRkwYADz589nwYIFrFy5kltvuYWDDizfUYvy1p41pkyZzC8uvpi//uV2Nt1006zDqZdbbrm57LvjAbRu3Zr27dvz3HPPAXDffffRrVt5DvqQUuLEE8fQrWs3TjnllKzDqbe8XQdeeeUVAP71r39x++23M/KIkRlHVHfnnXse/1z4L16c/xI33vhH9txzT6677vqsw6qzPJ1rebsOKH/KNWH6VI3ueGvuQN0ZuDyl1B3oUpgeCPQG+kXEkBrrXVZY7w3gq4X5Nxbm9wI+BywrzO8DfBvYFegEfH7tYCJibETMjIiZry5/tUEburajjj6SLwz+PM899xw7dOzAH/7w+0bdX2OqqKhg3LjxDDtgKD12685hhx9O9+7dsw6rzvLQnqOPPorBQ77Ac88/R8cdd+APV/+Bb337W7y14i32H7o//fr34/++/n9Zh1knb7/9Nvfcew+HHnJo1qE0iF//ahyjjh1Fn759ePLJJ/n+6T/IOqQ6mTp1KjfceAP3338//fr1pV+/vtx5V/mOBJqH60BNw0ccTs+eu3HIoYcwfvz4sh1cJI/ydK7l7TpQ7gLHfFhbpJSyjmG9RcSKlFLLGtMdgftTSjsWpi8GDqMqIQJoCZwP3AvcnVLaubDe6cBGwDjgmZRS+7X2swfwo5TSvoXpK4CpKaUbisXWv1//NG3a9Po3UvoE5fjeVXmysF668nYd8FxTU6nYqPmslFL/rOMoRdts0TEdPPAnTb7f3987pmSPSUXWATSgmkNEBXB+Suk3NVcoJFbv15hVCXzqE7a79vp5+plJkiRJtfnhRS3l2iXvk0wBRkdES4CIaBcR2xZbOaX0FrA4Ig4prL9xRJT3zRqSJEmS6i2X1ZKU0t8johvwWKG8vwI4mqoKUTHHAL+JiJ8Bq4DDGz1QSZIkqcTYPba2skyYat6/VJheCPRYa944qu5NWluPGutcXOP5C8Bea637EvBAjXW+UdeYJUmSJJWfvHbJkyRJkqR6K8sKkyRJkqRGEBCWVGrxxyFJkiRJRVhhkiRJklRQBt8k28SsMEmSJElSESZMkiRJklSEXfIkSZIkVbNHXm1WmCRJkiSpCCtMkiRJkqoERDNLTDVZYZIkSZKkIkyYJEmSJKkIu+RJkiRJ+pCjPtRihUmSJEmSirDCJEmSJAmAwALT2qwwSZIkSVIRVpgkSZIkVXNY8dqsMEmSJElSESZMkiRJklSEXfIkSZIkVYlw1Ie1WGGSJEmSpCKsMGmDsXr16qxDaFDNmvl5R6l6/71VWYfQoDbeZKOsQ1ARKWUdQcPyQ22pNPherM2/uCRJkiSpCBMmSZIkSSrCLnmSJEmSqvk9TLVZYZIkSZKkIqwwSZIkSaoSEI76UIsVJkmSJEkqwgqTJEmSpA9ZYKrFCpMkSZIkFWHCJEmSJElF2CVPkiRJElDVG89hxWuzwiRJkiRJRVhhkiRJklTNYcVrs8IkSZIkSUWYMEmSJElSEXbJkyRJklQlAhz0oRYrTJIkSZJUhBUmSZIkSdUc86E2K0ySJEmSVIQJUxmaPGUyu3bvRpeuu3DhRRdmHU69lHtbFi1axL5f2oeevXrSq3cvLrlkPABnnnUmffv1of+AfgwbNpSlS5dmHGndlPvxWeO9995j98/uTt++fejZazfO+ulZWYe03p5//nl2/9zA6kfrtttw6WWXAHDFlZfTp29P+g/ow49+/MOMI11/Y8acQJu2renVu2fWoTSIcn/fnDh2DO3at6F3n17V83529k/puGMH+g/oR/8B/bjrrjszjLDuPNdKV57aovyJlFLWMeRK/37907Rp0xtt+5WVlXTbtSuT75pC+/bt2X33Qdxww43suuuujbbPxtLUbVm9enWDb3PZsmX8+9/L6NOnL2+99RaDdh/EbbfdRvt27dliiy0AuPTSS3jmmWe47LLLG3TfzZo17ucdeTrXUkq8/fbbtGzZklWrVjHki0P41S9/xe67794o+3v/vVWNst01Kisr2WmXTjx4/0MsWLiAi35+IX+57XY23nhjXnn1FbbdZtsG3d/Gm2zUoNtb20MPP0TLzVpy/OjjeHLOU426r8bW9Ne1hv8d/vDDD9GyZUuOH308c2Y/CVQlTC03a8mpp36nwfdXU7NGvtHcc600ZdGWio2az0op9W+0HZSx7T7TOR057Pwm3++vbxhRssfEClOZmT59Op07d6ZTp060aNGC4SNGMGHihKzDqpM8tKVNmzb06dMXgM0335yuXbuydMnS6mQJ4O133inLL4DLw/FZIyJo2bIlAKtWreKDVavK8piscf8D99Fpxx3p0GEHrrrqd3zn1O+y8cYbAzR4stQUhgwewlZbbZV1GA0iD++bwYOHsOWW+Tgea/NcK015aovyyYSpzCxduoTt229fPd2+XTuWLlmSYUR1l6e2ACxcuJAnn5zDwIEDAfjJGT+hU+cduemmmzjzzLOyDa4O8nZ8Kisr6devL23atmbvffZh0KBBWYdUZ7fd9icOP3wEAC/Mf4FHH53KF/cczH7778OsWTMzjm7Dlrf3TU1XXHk5ffv14cSxY3j99dezDmeDl6dzLU9tyYWgaljxpn6UsEZLmCKiMiLmRMTciPhTRGy6Hq/tHRHDGis2qaGtWLGCEUcM5+KLf1FdXTr7Z2fz0osLGDlyJJdf0bDd8bT+mjdvzqxZT/DPhf9ixowZzJ07N+uQ6mTlypXceecdHHroVwD44IMPeP3113ngvoc495zzOebYo7CrtRraSWO/xrPPPM/MGbNo3bo13zv9tKxDkqQm05gVpndTSr1TSj2AlcDX1uVFEVEB9AZMmD5C27btWLR4UfX04iVLaNuuXYYR1V1e2rJq1SpGjBjOyCNGcughh/7P8pFHjOSvf/1rBpHVT16Oz9patWrFHnvswZS/T8k6lDr5+9+n0Kt3b7bbdjsA2rVrx5e/fDARQf/+A2jWrBnLly/POMoNV17fN9tttx3NmzenWbNmnDB6DDNmzMg6pA1ens61PLUlLyKa/lHKmqpL3sPAThGxVUTcHhFPRcTjEdETICLOiojrI2IqcD3wM2BEoUI1orD8u2s2VqhadSw8/0lEPBcRj0TETWvWi4gHIqJ/4fnWEbGw8Lx5RPw8ImYU4jipML9NRDxUoyo2uDD/SxHxWEQ8UaiUtWyin9lHGjBgAPPnz2fBggWsXLmSW2+5hYMOPCjLkOosD21JKTH2pBPp2rUr3/72KdXzX3jhhernEydOoEuXLlmEVy95OD5rvPrqq7zxxhsAvPvuu9xzzz1leUwA/nTbrRx+2PDq6YMO/DIPPfQgUHXerVy5kq233jqr8DZ4eXrf1LRs2bLq53/72+107949w2gE+TrX8tQW5VOjf3FtoWI0FJgM/BSYnVI6JCL2Aq6jqpoEsCvwhZTSuxFxHNA/pfSNwjbOKrLtAcBXgV7ARsATwKxPCOkE4M2U0oCI2BiYGhF/B74CTEkpnRsRzYFNI2Jr4MfAPimltyPidOBUqhK6mnGMBcYCdOjQYd1+MHVUUVHBuHHjGXbAUCorKznuuOPL9hdXHtry6KNTufHGG+nRowf9B/QD4OyfncPV11zN888/T7NmQYcOO3DZpZdlHOn6y8PxWWPZsmWMHn08lZWVrE6rOeywwznwgAOzDmu9vf3229x3372MH3dp9bxRxxzL1/5vLP0H9qVFixb89jdXld2AFkcdfSQPPvggy5cvZ4eOHTjzjDMZPfqErMOqkzy8b44+5igeeqjqeOzYaQfO+MmZPPjQgzz55JNEBDvssAOXX3ZF1mHWiedaacpTW5RPjTaseERUAk8XJh8GvgNMA76aUnqpsM4ioDtVSUhKKf20MP84/jdhWpFSurgwPRc4EDgE2DKldGZh/i+BpSmliyPiAeC7KaWZhcRnZkqpY0TcBvQE3inE9mngJOA94A/ADcDtKaU5EXEgcA2wuLBuC+CxlFLRq2tjDyuuumuMYcWz1NjDiqvuGntY8abW2MOKq+4aY1jxLDX2sOLSGg4rXlzrrTuno77c9N+F9curDy/ZY9KYFaZ3U0q9a874hE893/6YZR9Qu/vgJuuw/5qvqbl+AN9MKf3PDQwRMQQ4ALimkHy9DtydUhq5DvuTJEmSlDNN/RH1w8BRABGxB7A8pfTfj1jvLWDzGtMLgb6F1/UFdizMnwocFBGbFO4tOnCt1/QrPD+sxvwpwP+LiI0K29slIjaLiB2Al1NKvwOuKuzvceDzEbFTYd3NImKX9W+2JEmSVCYc9aGWpk6YzgL6RcRTwAXAsUXWux/Ydc2gD8Cfga0i4h/AN4DnAVJKM4AJwFPAXVR1AXyzsI2LqUqMZgM174C+CpgHPFHo2vcbqiptewBPFtYfAYxLKb0KHAfcVIj5MaBrPX8GkiRJkspEo93D1FQiomVKaUXhe54eAsamlJ7IKh7vYSpd3sOkpuI9TGoq3sMk1Y33MBXXeuvO6aiDL2ry/f7yD4eV7DFp9FHymsBvI2JXqu5TujbLZEmSJEkqaxFlN9pqYyv7hCmldGTWMUiSJEnKp7JPmCRJkiQ1nLDXfy3+OCRJkiSpCCtMkiRJkqp5D1NtVpgkSZIkqQgTJkmSJEkqwi55kiRJkj5kl7xarDBJkiRJUhEmTJIkSZKqRNWw4k39WKfQIk6JiH9ExNyIuCkiNomIHSNiWkTMj4hbIqJFYd2NC9PzC8s71vVHYsIkSZIkqaRFRDvgZKB/SqkH0Bw4ArgQ+FVKaSfgdeCEwktOAF4vzP9VYb06MWGSJEmSVA4qgE9FRAWwKbAM2Au4rbD8WuCQwvODC9MUlu8ddRwv3UEfJEmSJAEQZPY9TFtHxMwa079NKf12zURKaUlEXAz8C3gX+DswC3gjpfRBYbXFQLvC83bAosJrP4iIN4HPAMvXNzATJkmSJElZW55S6l9sYURsSVXVaEfgDeBPwP5NEZgJkyRJkqQPNSvJYcX3ARaklF4FiIi/AJ8HWkVERaHK1B5YUlh/CbA9sLjQhe/TwH/qsmPvYZIkSZJU6v4F7B4RmxbuRdobmAfcDxxWWOdY4G+F5xMK0xSW35dSSnXZsQmTJEmSpJKWUppG1eANTwBPU5XH/BY4HTg1IuZTdY/S7wsv+T3wmcL8U4Hv13XfdsmTJEmSVCUiq0EfPlFK6UzgzLVmvwQM/Ih13wMOb4j9WmGSJEmSpCKsMEmSJEmqVqIFpsyYMDWwBKxeXaf7yUpSs9IcJUUqaRtvslHWITSo1atXZx1Cg2rWLD+dK/L2R43nmqRSZMIkSZIk6UN+YF6LH31IkiRJUhEmTJIkSZJUhF3yJEmSJFUr1WHFs2KFSZIkSZKKsMIkSZIkqUrkbwTO+rLCJEmSJElFmDBJkiRJUhF2yZMkSZL0Ib+HqRYrTJIkSZJUhBUmSZIkSQAEDiu+NitMkiRJklSEFSZJkiRJVSII72GqxQqTJEmSJBVhwiRJkiRJRdglT5IkSdKH7JFXixUmSZIkSSrChKkMnDh2DO3at6F3n17/s+xXv/olLTauYPny5RlEVn+Tp0xm1+7d6NJ1Fy686MKsw1lvixYtYt8v7UPPXj3p1bsXl1wyHoAzzzqTvv360H9AP4YNG8rSpUszjnT9vffee+z+2d3p27cPPXvtxlk/PSvrkOql3M+1mvJwbKqua23p3ad39bwnn3qSwUO+QJ++vTnk0EP473//m12A9ZCncw1gp50707tPb/r178eg3QdlHc56KXaNfu211xg6dH923bUbQ4fuz+uvv55xpHWTl3Nt0aJF7L3P3uzWswc9e+3G+PHjsw5pgxcRTf4oZSZMZWDUMaOYNPGO/5m/aNEi7rnnbjp06JBBVPVXWVnJySd/k0kT7+Dpp+Zyy803M2/evKzDWi8VFRVcdOFFPPXkUzzy8CNcceWVzHtmHt859Ts8MWs2M2fMYtiwYZx77jlZh7reNt54Y+65+x6eeGI2s2Y+wZQpU3j88cezDqtO8nCu1ZSHYzPqmGOZNHFSrXlf+9pJnHvOecx+Yg6HHHwwv/jlLzKKru7ydq6tcc/d9zBr5iymPT4t61DWS7Fr9EU/v4g999qLefOeYc+99uKin1+UdajrLU/nWkVFBT+/6Oc8/dRcpj7yKFdceXnZtkX5ZMJUBgYPHsKWW271P/O/e9p3OO/8C0o+Ky9m+vTpdO7cmU6dOtGiRQuGjxjBhIkTsg5rvbRp04Y+ffoCsPnmm9O1a1eWLlnKFltsUb3O2++8U5bHKCJo2bIlAKtWreKDVavKsh2Qj3Otpjwcm8GDB//Pde2FF15g8ODBAOy99z789a9/zSK0esnbuVbuil2jJ06cyDFHHwPAMUcfw4QJ5XeM8nSutWnThr59ax+nJUuXZByV9CETpjI1YcIE2rVtR6+e/9tNr1wsXbqE7dtvXz3dvl07li4p3wvkwoULefLJOQwcOBCAn5zxEzp13pGbbrqJM888K9vg6qiyspJ+/frSpm1r9t5nHwYNKq/uOGvk7VyD/BybmnbdddfqP1z//OfbWLx4UcYRrb88nmsRwdBhQxk4aCC/u+p3WYdTZzWv0a+88jJt2rQBoHXr1rzyyssZR7f+8niuQdVxmjNnDoMGlv81rZxFs2jyRykrm4QpIiojYk6Nx/frsa0Vhf/bRsRtH7Nex4iYW9f9NJZ33nmHCy86v2z/CM+jFStWMOKI4Vx88S+qq0tn/+xsXnpxASNHjuTyKy7POMK6ad68ObNmPcE/F/6LGTNmMHduyb0dNlh5PDa//c3v+M1vrmTQ7gN5a8UKWrRokXVIAh64/0FmTJ/BpImTuOKKK3j44YeyDmm9fdQ1eo1yuH9iQ7FixQqGDz+cX/7il/9znKQslU3CBLybUupd43FBfTeYUlqaUjqsIYJrSi++9CILFy6k/4C+7LxLZxYvXsyg3Qfw73//O+vQ1kvbtu1YVOMT5MVLltC2XbsMI6qbVatWMWLEcEYeMZJDDzn0f5aPPGJkWXYtqqlVq1bsscceTPn7lKxDqZO8nGsfpdyPTU1du3blzjvvYtrj0xkxfASdOnXKOqT1lsdzrV0h/m233ZZDDj6YGTNmZBzR+vmoa/S2227HsmXLAFi2bBnbbLNtliHWSd7OtVWrVnH48MMYOfJIDj30K1mHs2ELB31YWzklTB8pIhZGxE8j4omIeDoiuhbmbxMRd0fEPyLiqoj4Z0RsvdZrqytIEdE9IqYXqldPRcTOhdWaR8TvCtv5e0R8qomb+D9267EbSxYv44XnX+SF51+kffv2THt8Bq1bt846tPUyYMAA5s+fz4IFC1i5ciW33nILBx14UNZhrZeUEmNPOpGuXbvy7W+fUj3/hRdeqH4+ceIEunTpkkV49fLqq6/yxhtvAPDuu+9yzz33lGU7IB/nWk15OjY1vfLKKwCsXr2a8y84j7Enjs04ovWXt3Pt7bff5q233qp+fvc9d9O9e/eMo1p3xa7RBx14INffcD0A199wPQcdVH7HKE/nWkqJE08cQ7eu3TjllFM++QVSEyunL679VETMqTF9fkrplsLz5SmlvhHxf8B3gTHAmcB9KaXzI2J/4IRP2P7XgHEppRsjogXQHNgO2BkYmVI6MSJuBb4K3NBwzfpkRx9zFA899CDLly9nx047cMZPzuT440c3ZQiNoqKignHjxjPsgKFUVlZy3HHHl9UvYoBHH53KjTfeSI8ePeg/oB8AZ//sHK6+5mqef/55mjULOnTYgcsuvSzjSNffsmXLGD36eCorK1mdVnPYYYdz4AEHZh1WneThXKspD8fm6GOOrnFd68gZPzmDFStWcMWVVwJwyCGHcOyxx2UbZB3k7Vx7+eWXOezwqo4YlR98wBFHHMF+++2fcVTrrtg1+rTTvseRR47kmquvpkOHDvzxjzdlHOn6y9O5NnXqVG648QZ267Eb/fpVDf5w9jnnMGzosIwjk6pESinrGNZJRKxIKbX8iPkLgc+nlJZExCDg3JTSPoXk6tCU0oLCeq8Bu6SUlq/ZVkR0BCallHpExJHAj4DrgL+klF4oLL87pbRzYRunAxullM5ZK4axwFiADh069Jv/wkuN8jPIQrMSvwlvfaxevTrrEBpUs2ZlXyBWmfC9U7rK5Xf4uspbe/J0ruVNxUbNZ6WU+mcdRylq23aXdNKYpv+g96yzv1SyxyQv7+T3C/9XUseqWUrpj8CXgXeBOyNir7W2XXT7KaXfppT6p5T6b731NnXZvSRJkqQSlJeE6aNMBYYDRMSXgC0/buWI6AS8lFIaD/wN6NnoEUqSJEklJHDQh7WVU8L0qbWGFf+kUfJ+CnypMKjD4cC/gbc+Zv3hwNxCV74eVHXNkyRJkrQBK5tBH1JKzYvM71jj+Uxgj8Lkm8B+KaUPIuKzwICU0vuF9VoW/l9IVXJEYZjytZOw19YsL6xzcQM0RZIkSSpZJV7waXJlkzDVQQfg1ohoBqwETsw4HkmSJEllJrcJU0rpBaBP1nFIkiRJKl+5TZgkSZIkrT+75NVWToM+SJIkSVKTssIkSZIkqUoZDPPd1KwwSZIkSVIRJkySJEmSVIRd8iRJkiRVs0debVaYJEmSJKkIK0ySJEmSAAhw0Ie1WGGSJEmSpCJMmCRJkiSpCLvkSZIkSapmj7zarDBJkiRJUhFWmCRJkiRVc9CH2qwwSZIkSVIRVpgkSZIkVQnvYVqbFSZJkiRJKsKESZIkSZKKsEueJEmSpGqBffJqMmFqYAE0a+ZJVopSyjqChvXfN9/NOoQGtfkWm2QdQoPJ2+hCzZrZGaFUVVauzjqEBlVR0TzrECTpf5gwSZIkSQKqPvzP2ed+9ebHhpIkSZJUhAmTJEmSJBVhlzxJkiRJ1eySV5sVJkmSJEkqwgqTJEmSpGp5G+21vqwwSZIkSVIRVpgkSZIkVbPAVJsVJkmSJEkqwoRJkiRJkoqwS54kSZKkKhH2yVuLFSZJkiRJKsIKkyRJkqRqFphqs8IkSZIkSUWYMEmSJElSEXbJkyRJkgRAAGGfvFqsMEmSJElSEVaYJEmSJFWzwFSbFaYyM2bMCbRp25pevXtmHUqDmDxlMrt270aXrrtw4UUXZh3Oenvvvff4/Bc+S/8Bfendpxc/+9lPAViwYAFfGPw5uu3alaOOPpKVK1dmHGlx3zj5/7FL14587gsDquddcOG5dO+xM0P2+CxD9vgsd989pdZrFi9exPY7bMcll45r6nDrbPwl4+nduxe9evVk3PjyibuYcn/v1OR1rfRUVlYyaPcBHPqVQwC44orL2bV7Nzb5VAuWL1+ebXD1kIdjU1Ne2rNo0SL23mdvduvZg569dmP8+PFZhyTVYsJUZkYdeyx3TLoz6zAaRGVlJSef/E0mTbyDp5+ayy0338y8efOyDmu9bLzxxkyZfDczZzzBjOkz+fvdU5g27XF+9OMfcvI3v8Uz856lVatWXH3NH7IOtagjjziKP91y+//M/9rXvsFDDzzGQw88xr777ldr2Y9+8n323nvfJoqw/ubOncsffv97Hn30MWbNeoI777yD+fPnZx1WneXhvVOT17XSc+mll9ClS9fq6c9+9rPceedddOiwQ4ZR1U9ejs0aeWpPRUUFP7/o5zz91FymPvIoV1x5edm2RflkwlRmhgwewlZbbZV1GA1i+vTpdO7cmU6dOtGiRQuGjxjBhIkTsg5rvUQELVu2BGDVqlWsWrWKiOCBB+7nK1/5KgDHHH0MEyaUbrs+97kvsOWWW67z+nfcOZEdOnSka5dujRhVw3r22WcZMHAgm266KRUVFQwZPITbb/9r1mHVWR7eOzV5XSstixcv5q7Jd3H88aOr5/Xu3YeOO3TMLqgGkIdjU1Oe2tOmTRv69u0LwOabb07Xrl1ZsnRJxlFt2CKiyR+lzIRJmVm6dAnbt9++erp9u3YsXVJ+F8jKykoGDOxH++3bsvfe+9CpU2c+/elWVFRU3SLYrl17li5dmnGU6++q3/+GLwwZxDdO/n+88cbrAKxYsYJx43/F9077QcbRrZ/u3bsz9ZFH+M9//sM777zDXZPvYtHixVmHVWd5ee/kUR6OzWmnfYfzzj2fZs3y9SdCHo5NTXlrzxoLFy5kzpw5DBo4KOtQpGolfzWMiB9FxD8i4qmImBMRDfoOiohHP2H5iobcn/KnefPmzJg+i5deXMjMGTN47rlnsw6p3kYfP4YnZj7NQw88RuvttuPHZ/wQgAsvOo//97WvV1fVykW3bt347mmnMXTYUA44cBi9evWmefPmWYcllZw777yDbbbdtvrTfqkprVixguHDD+eXv/glW2yxRdbhbLiiatCHpn6UspIeJS8iPgscCPRNKb0fEVsDLRpyHymlzzXk9rTu2rZtx6LFi6qnFy9ZQtt27TKMqH5atWrFF7+4B49Pm8abb77BBx98QEVFBUuWLKZt27ZZh7dett12u+rno445niOOPAyAWU/MYMLE2znrpz/hzTffpFmzZmyyycacOOZrWYW6zkYfP5rRhS5GP/7xj2jXvn3GEdVd3t47eVLux+bRxx7ljkmTmDx5Mu+//x7//e9/Oe74Y7nm6muzDq3eyv3YrC1v7Vm1ahWHDz+MkSOP5NBDv5J1OFItpV5hagMsTym9D5BSWp5SWhoRCyPiooh4OiKmR8ROABFxUERMi4jZEXFPRGxXmH9WRPwhIh6IiJci4uQ1O1hTQYqINhHxUKGKNTciBtdY59yIeDIiHl+zTdXfgAEDmD9/PgsWLGDlypXcesstHHTgQVmHtV5effVV3njjDQDeffdd7r33Hrp27coXv7gHf/nLnwG4/obrOeig8mrXv//97+rnk+6YSLeuuwJw56S7eXL2PJ6cPY+vnfR/nPLt75ZFsgTwyiuvAPCvf/2L22+/nZFHjMw4orrLw3snr8r92Jxz9rm8+OICnn/uBa677gb22GPPXCRLUP7HZm15ak9KiRNPHEO3rt045ZRTsg5HFL68tokfpazUE6a/A9tHxPMRcXlEfLHGsjdTSrsBlwK/Lsx7BNg9pdQHuBn4Xo31uwL7AQOBMyNio7X2dSQwJaXUG+gFzCnM3wx4PKXUC3gIOHHtICNibETMjIiZry5/tc6NXRdHHX0kXxj8eZ577jl26NiBP/zh9426v8ZUUVHBuHHjGXbAUHrs1p3DDj+c7t27Zx3Wevn3v5fxpf32oV//Pnzu859l77334YBhB3DuOecxbvyv6bZrV1577TWOP270J28sI2NOPI799t+L+fNfoPtuu3D9Dddy1k9/zOcHD+QLQwbxyCMPce455Ttc7RrDRxxOz567ccihhzB+/HhatWqVdUh1lof3Tk1e10rfZZddSufOO7JkyWIGDOjH1/7fSVmHtN7ydmzy1J6pU6dyw403cP/999OvX1/69evLnXflY+RM5UOklLKO4WNFRHNgMLAncBLwfeAsYK+U0kuFxOffKaXPRMRuwC+oqky1ABaklPaPiLOAVSmlcwvbfAbYN6W0OCJWpJRaRsQQ4A/ADcDtKaU5hXXfBzZJKaWIGFF43Zhi8fbv1z9Nmza9EX4Sqq/KytVZh9Cg3l7xftYhNKjNt9gk6xAaTKmP9qP8+OCDyqxDaFAVFd5bqKZRsVHzWSml/lnHUYo6dOiaTv/OVU2+3298e3DJHpNSrzCRUqpMKT2QUjoT+Abw1TWLaq5W+P8S4NJC5ekkoOZfYDX/uqxkrfu3UkoPAUOAJcA1ETGqsGhV+jCr/J/XSZIkSXkROKz42ko6YYqILhGxc41ZvYF/Fp6PqPH/Y4Xnn6Yq4QE4dj33tQPwckrpd8BVgEMESZIkSRu4otWSiLiE2lWcWlJKJxdb1oBaApdERCvgA2A+MJaqkfO2jIinqKocrbl7+yzgTxHxOnAfsON67GsP4LSIWAWsAEZ9/OqSJElS/pR4wafJfVz3splNFkURKaVZwP8M+10o2/08pXT6Wuv/DfjbR2znrLWme9R43rLw/7XA/wwFtGZ54fltwG3r2QxJkiRJZapowlRIIKpFxKYppXcaPyRJkiRJKg2feA9TRHw2IuYBzxame0XE5Y0e2cdIKXVMKS3PMgZJkiQpjxz0obZ1GfTh11R9f9F/AFJKT1I1mpwkSZIk5do6DZGdUlq0VuaXry9+kCRJkgQ46MPa1iVhWhQRnwNS4UtivwU807hhSZIkSVL21iVh+howDmgHLAWmAF9vzKAkSZIkZaAM7ilqap+YMBUGVziqCWKRJEmSpJKyLqPkdYqIiRHxakS8EhF/i4hOTRGcJEmSJGVpXUbJ+yNwK9AGaAv8CbipMYOSJEmS1PSCqkEfmvpRytYlYdo0pXR9SumDwuMGYJPGDkySJEmSslb0HqaI2Krw9K6I+D5wM5CAEcCdTRCbJEmSpCZW6hWfpvZxgz7MoipBWvMjO6nGsgT8oLGCkiRJkqSaIqIVcBXQg6p8ZDTwHHAL0BFYCAxPKb0eVUP9jQOGAe8Ax6WUnqjLfosmTCmlHeuyQUmSJElqBOOAySmlwyKiBbAp8EPg3pTSBYVecd8HTgeGAjsXHoOAKwr/r7d1+R4mIqIHsCs17l1KKV1Xlx1KkiRJKl2l+D1MEfFpYAhwHEBKaSWwMiIOBvYorHYt8ABVCdPBwHUppQQ8HhGtIqJNSmnZ+u77ExOmiDizEMSuVN27NBR4BDBhkiRJktQQto6ImTWmf5tS+m2N6R2BV4GrI6IXVbcPfQvYrkYS9G9gu8LzdsCiGq9fXJjX8AkTcBjQC5idUjo+IrYDbljfHUmSJEkqfRkVmJanlPp/zPIKoC/wzZTStIgYR1X3u2oppRQRqaEDW5dhxd9NKa0GPoiILYBXgO0bOhBJkiRJKmIxsDilNK0wfRtVCdTLEdEGoPD/K4XlS6ids7QvzFtv65IwzSyMSPE7qkpfTwCP1WVnkiRJkrS+Ukr/BhZFRJfCrL2BecAE4NjCvGOBvxWeTwBGRZXdgTfrcv8SrEOXvJTS/xWeXhkRk4EtUkpP1WVnkiRJkkpYlOagDwXfBG4sjJD3EnA8VQWgWyPiBOCfwPDCundSNaT4fKqGFT++rjv9uC+u7ftxy+o6jrkkSZIkra+U0hzgo+5z2vsj1k3A1xtivx9XYfrFxyxLwF4NEUDevLXifR58eEHWYTSYLw7Oz9dxNWtWsp+W1MkWn/5U1iFIylhFRfOsQ2hQVX/f5EcJf0q/3vJ2bPQJ8nPqNoiP++LaPZsyEEmSJEkqNev0xbWSJEmS8i/IV3W0IazLKHmSJEmStEEyYZIkSZKkIj6xS15U1eSOAjqllH4WER2A1iml6Y0enSRJkqQmZZe82talwnQ58FlgZGH6LeCyRotIkiRJkkrEugz6MCil1DciZgOklF4vfFmUJEmSpJyxwFTbulSYVkVEc6q+e4mI2AZY3ahRSZIkSVIJWJeEaTzwV2DbiDgXeAQ4r1GjkiRJkqQS8Ild8lJKN0bELGBvqoZmPySl9EyjRyZJkiSpaUU46MNa1mWUvA7AO8DEmvNSSv9qzMAkSZIkKWvrMujDHVTdvxTAJsCOwHNA90aMS5IkSVITCxz0YW3r0iVvt5rTEdEX+L9Gi0iSJEmSSsS6VJhqSSk9ERGDGiMYSZIkSdnyHqba1uUeplNrTDYD+gJLGy0iSZIkSSoR61Jh2rzG8w+ouqfpz40TjiRJkiSVjo9NmApfWLt5Sum7TRSPJEmSpAzZJa+2ol9cGxEVKaVK4PNNGI8kSZIklYyiCRMwvfD/nIiYEBHHRMRX1jyaIrgNXWVlJSeNPYgf/vBEAH7+8+9z4pgDGTPmAM466+u8++7btdZ/6KHJ7L3XTjz33NNZhFsnk6dMZtfu3ejSdRcuvOjCrMNZb2NOHEPbdm3o3btX9bzXXnuN/YfuR7ddu7L/0P14/fXXM4yw7saMOYE2bVvTq3fPrENpMJWVlfTv348vH3xQ1qHUW7m/d2rKU1sgX+1577332P2zu9O3bx969tqNs356VtYh1ctOO3emd5/e9Ovfj0G7l//4VXk618ZfMp7evXvRq1dPxo0fl3U4G7aoGla8qR+l7OMSpjU2Af4D7AUcCBxU+F+N7C9/uYYOHXaqnv6///sRv7tqEldddQfbbtuW2/96ffWyd95ZwV/+fC3duvX6qE2VpMrKSk4++ZtMmngHTz81l1tuvpl58+ZlHdZ6OXbUKCZNuqPWvIsuupC99tyLZ+Y9y1577sVFZfpLbNSxx3LHpDuzDqNBjR8/nq7dumYdRr3l4b2zRp7aAvlrz8Ybb8w9d9/DE0/MZtbMJ5gyZQqPP/541mHVyz1338OsmbOY9vi0rEOplzyda3PnzuUPv/89jz76GLNmPcGdd97B/Pnzsw5LqvZxCdO2hRHy5gJPF/7/R+H/uU0Q2wbt1VeXMe3xBxg2bHj1vM02qxp/I6XEyvffr5WOX/2HX3PEyLG0aLFxk8daV9OnT6dz58506tSJFi1aMHzECCZMnJB1WOtl8OAhbLXlVrXmTZw4kWOOGQXAMceMYsKE8mrTGkMGD2Grrbb65BXLxOLFi7nzrjsZPfqErEOptzy8d9bIU1sgf+2JCFq2bAnAqlWr+GDVKu9tKBF5OteeffZZBgwcyKabbkpFRQVDBg/h9tv/mnVYUrWPS5iaAy0Lj81rPF/zUCO67LJzGHvS6USz2r+YLrrwdA47bHf+tehFDj206o/y55+fy6uvLmP33ffMItQ6W7p0Cdu33756un27dixdsiTDiBrGy6+8TJs2bQBo3bo1L7/ycsYRCeDU75zCBedfQLNm61JYL215eu/kqS2Qv/ZAVSWjX7++tGnbmr332YdBg8q3K1tEMHTYUAYOGsjvrvpd1uHUS57Ote7duzP1kUf4z3/+wzvvvMNdk+9i0eLFWYe1QYuIJn+Uso8bJW9ZSulnDbGTiEjAL1NK3ylMfxdomVI6qyG2X4+4rirEVVI17Mceu48tW32GXXbpwZw5tbs+fO/0C6msrOTSS37KA/ffwZf2+wpXXnEe3zv9ooyi1ccph4vAhmDSHZPYdptt6devHw88+EDW4UhlpXnz5sya9QRvvPEGXz3sq8ydO5cePXpkHVadPHD/g7Rr145XXnmF/YfuT9cuXRg8eEjWYW3wunXrxndPO42hw4ay2Wab0qtXb5o3b551WFK1j/uotSH/ynsf+EpEbF2XF0fEunxf1HpLKY0ptWQJ4B9zZ/Hoo/dy5Mgvcs7Z32bO7Mc477wPvz+4efPm7LnngTz08BTeeedtFix4gVNPOYojR36RefPm8JMfn1QWAz+0bduORYsXVU8vXrKEtu3aZRhRw9hu2+1YtmwZAMuWLWPbbbbNOCI9+uijTJw0kc47deKoo47k/vvvZ9SoY7IOq87y9N7JU1sgf+2pqVWrVuyxxx5M+fuUrEOps3aFY7HttttyyMEHM2PGjIwjqru8nWujjx/N9GnTuf++B9iyVSt23nnnrEPaYAUQzaLJH6Xs4xKmvRtwPx8AvwVOWXtBRHSMiPsi4qmIuDciOhTmXxMRV0bENOCiiHg6IlpFlf9ExKjCetdFxL4R0Twifh4RMwrbOqmwvFlEXB4Rz0bE3RFxZ0QcVlj2QET0Lzy/IiJmRsQ/IuKnNeJbGBE/jYgnCjE0+h3jY048jVtuncofb3qQH//k1/Tu81l+8INfsGTJQqDqHqZHH72XDtt3omXLzfnr7TP4400P8sebHmTXXXtz9jm/oUuX3Ro7zHobMGAA8+fPZ8GCBaxcuZJbb7mFgw4s/9HLDjzoQK6//joArr/+Og46qPzbVO7OO/c8/rnwX7w4/yVuvPGP7Lnnnlx33fWf/MISlaf3Tp7aAvlrz6uvvsobb7wBwLvvvss999xDly5dsg2qjt5++23eeuut6ud333M33bt3zziqusvbufbKK68A8K9//Yvbb7+dkUeMzDgi6UNFKzcppdcaeF+XAU9FxNp9xy4Brk0pXRsRo4HxwCGFZe2Bz6WUKiPiSqq+E+qfwEvAYOA64LPA/wNOAN5MKQ2IiI2BqRHxd6Af0BHYFdgWeAb4w0fE96OU0muFL+u9NyJ6ppSeKixbnlLqGxH/B3wXGFPzhRExFhgLsO12bevwo/lkKSUuvOB7vPPOClJKdO7cjW99+6ef/MISVlFRwbhx4xl2wFAqKys57rjjy+6X19FHH8WDDz3I8uXL6bjjDpxxxpl877TTGXnkEVx9zdV06NCBm/54c9Zh1slRRx/Jgw9WtW2Hjh0484wzczFgQh7k4b2zRp7aAvlrz7Jlyxg9+ngqKytZnVZz2GGHc+AB5TlQ7ssvv8xhhx8GQOUHH3DEEUew3377ZxxV3eXtXBs+4nBe+89rVGy0EePHj6dVq1ZZhyRVi5RS4+8kYkVKqWVE/AxYBbxL4R6miFgOtEkprYqIjai6d2rriLgGuD+ldG1hG0cBPalKmN6jKkH5KvDXlNLAiLitsPydwm4/DZwEDAOeTCldXdjOX4A/ppRui4gHgO+mlGZGxNcK26wA2gDfTCndHBELgc+nlJZExCDg3JTSPsXa2qXLbumKK29vmB9cCfji4B2zDqHBNMW53pS8N0pS3nidLl15OzYbtaiYlVLqn3Ucpahzp13T+ef9scn3O2Jkn5I9Jk09XNSvqaoEbbaO69f8ZtaHqKoqDQYeAF4FDgMeLiwPqpKc3oXHjimlv6/LTiJiR6oqR3unlHoCd1D1/VNrvF/4v5KPHyhDkiRJUo40acJU6OZ3K1VJ0xqPAkcUnh/FhwnQ2q9dBGwN7JxSegl4hKok56HCKlOA/1eoUhERu0TEZsBU4KuFe5m2A/b4iM1vQVVy9mZhnaF1bqQkSZJUrjIYUrzUq7FZVEt+AXyjxvQ3gasj4jSqqkbHf8xrp1H1/VBQlVidT1XiBHAVVfcqPRFVP/VXqboX6s9UDWAxD1gEPAG8WXOjKaUnI2I28Gxhnal1a5okSZKkPGmShCml1LLG85eBTWtM/xPY6yNec9xHzDumxvNHqVEhSymtBn5YeNQSEd9NKa2IiM8A04GnC6/Z4+P2V5jfscbzmXx0hUqSJEnKhRIv+DS5DeV+nEkR0QpoAZydUvp3xvFIkiRJKgMbRMJUs5IkSZIkSetqg0iYJEmSJK2bUh+Eoak19bDikiRJklQ2rDBJkiRJAqq+2NQKU21WmCRJkiSpCBMmSZIkSSrCLnmSJEmSqtkjrzYrTJIkSZJUhBUmSZIkSVWqRn3IOoqSYoVJkiRJkoowYZIkSZKkIuySJ0mSJKma38NUmxUmSZIkSSrCCpMkSZKkahaYarPCJEmSJElFWGGSJEmSVBBEM0tMNVlhkiRJkqQiTJgkSZIkqQi75EmSJEkCqgZ8cNCH2kyYGtjmLTfmi4N3zDoMfQS/U0CSSpvX6dLlsdGGzIRJkiRJUjUT5Nq8h0mSJEmSijBhkiRJkqQi7JInSZIkqZpd8mqzwiRJkiRJRVhhkiRJklTNAlNtVpgkSZIkqQgrTJIkSZKqeQ9TbVaYJEmSJKkIEyZJkiRJKsIueZIkSZKAqgEf7JJXmxUmSZIkSSrCCpMkSZKkahaYarPCJEmSJElFmDBJkiRJUhF2yZMkSZJUEA76sBYrTJIkSZJUhAlTGZo8ZTK7du9Gl667cOFFF2YdTr3kqS1ge0pZntoC+WpPntoCtqeU5aktkK/25KkteRARTf4oZSZMZaayspKTT/4mkybewdNPzeWWm29m3rx5WYdVJ3lqC9ieUpantkC+2pOntoDtKWV5agvkqz15aovyyYSpzEyfPp3OnTvTqVMnWrRowfARI5gwcULWYdVJntoCtqeU5aktkK/25KktYHtKWZ7aAvlqT57aonwyYSozS5cuYfv221dPt2/XjqVLlmQYUd3lqS1ge0pZntoC+WpPntoCtqeU5aktkK/25KkteRHR9I9SVnYJU0SsqPF8WEQ8HxE7ZBmTJEmSpHwq22HFI2JvYDywX0rpn+uwfgCRUlrd6ME1orZt27Fo8aLq6cVLltC2XbsMI6q7PLUFbE8py1NbIF/tyVNbwPaUsjy1BfLVnjy1JQ8iIJqVeMmniZVdhQkgIoYAvwMOTCm9WJh3akTMLTy+XZjXMSKei4jrgLnA9hFxWkTMiIinIuKnNbZ5e0TMioh/RMTYGvNXRMS5EfFkRDweEds1aWPXMmDAAObPn8+CBQtYuXIlt95yCwcdeFCWIdVZntoCtqeU5aktkK/25KktYHtKWZ7aAvlqT57aonwqxwrTxsDtwB4ppWcBIqIfcDwwCAhgWkQ8CLwO7Awcm1J6PCK+VJgeWFhvQkQMSSk9BIxOKb0WEZ8CZkTEn1NK/wE2Ax5PKf0oIi4CTgTOqRlQIcEaC9ChQ4dGbXxFRQXjxo1n2AFDqays5Ljjjqd79+6Nus/Gkqe2gO0pZXlqC+SrPXlqC9ieUpantkC+2pOntuRFqd9T1NQipZR1DOslIt4B7gNeTCl9qzDvW8BnUkpnFKbPBl4FJgD3p5R2LMy/GDgMeKOwuZbA+Sml30fEWcChhfkdqerq93hEvA9sklJKETEC2DelNKZYfP379U/Tpk1vyCZLkiSpAVVs1HxWSql/1nGUoi5ddktXXHl7k+937712KtljUo4VptXAcODeiPhhSum8T1j/7RrPg6oE6Tc1V4iIPYB9gM+mlN6JiAeATQqLV6UPs8pKyvNnJkmSJKkOyvIeppTSO8ABwFERcQLwMHBIRGwaEZtRVSl6+CNeOgUYHREtASKiXURsC3waeL2QLHUFdm+ShkiSJEklJjL4V8rKtlpSuN9of+Ah4FvANcCavnBXpZRmR0THtV7z94joBjxWNWgeK4CjgcnA1yLiGeA54PEmaYQkSZKkklZ2CVNKqWWN54uAHWss/uVa6y4Eeqw1bxww7iM2PXQd9ncbcNt6By1JkiSVi9Iu+DS5suySJ0mSJElNwYRJkiRJkooouy55kiRJkhpP+EVMtVhhkiRJkqQirDBJkiRJqhJggak2K0ySJEmSVIQVJkmSJElA4UtrLTHVYoVJkiRJkoowYZIkSZKkIuySJ0mSJKmaPfJqs8IkSZIkSUVYYZIkSZJUzUEfarPCJEmSJElFmDBJkiRJKgsR0TwiZkfEpML0jhExLSLmR8QtEdGiMH/jwvT8wvKOdd2nCZMkSZKkahFN/1gP3wKeqTF9IfCrlNJOwOvACYX5JwCvF+b/qrBenZgwSZIkSSp5EdEeOAC4qjAdwF7AbYVVrgUOKTw/uDBNYfneUcebsxz0QZIkSVKVyGzQh60jYmaN6d+mlH671jq/Br4HbF6Y/gzwRkrpg8L0YqBd4Xk7YBFASumDiHizsP7y9Q3MhEmSJElS1panlPoXWxgRBwKvpJRmRcQeTRYVJkySJEmSSt/ngS9HxDBgE2ALYBzQKiIqClWm9sCSwvpLgO2BxRFRAXwa+E9dduw9TJIkSZIACEpz0IeU0g9SSu1TSh2BI4D7UkpHAfcDhxVWOxb4W+H5hMI0heX3pZRSXX4mVpi0wajje6RkVVauzjqEBtWsWZ4+v8nXuZavYyNJypnTgZsj4hxgNvD7wvzfA9dHxHzgNaqSrDoxYZIkSZJULZsxH9ZdSukB4IHC85eAgR+xznvA4Q2xPz82lCRJkqQirDBJkiRJqhaUeImpiVlhkiRJkqQiTJgkSZIkqQi75EmSJEmqVuqDPjQ1K0ySJEmSVIQVJkmSJEnVwhJTLVaYJEmSJKkIEyZJkiRJKsIueZIkSZKAqgEf7JFXmxUmSZIkSSrCCpMkSZKkgnDQh7VYYZIkSZKkIqwwSZIkSapmgak2K0ySJEmSVIQJkyRJkiQVYZc8SZIkSdUc9KE2K0xlqrKykv79+/Hlgw/KOpR6mTxlMrt270aXrrtw4UUXZh1OvTz33HP069+v+rHVZ7Zk3PhxWYe13iorKxm0+wAO/cohAIw58QS6dN2FgYP6M3BQf558ck6m8a2rE8eOoV37NvTu06vW/Msuu5Qeu3WnV++efP8Hp2cU3fpbtGgR+35pH3r26kmv3r245JLxAHz/+6fTY7ce9O3Xh8MOP4w33ngj20DrKC/XNMjXdQ3y1Z4xY06gTdvW9OrdM+tQGkRejs17773H7p/dnb59+9Cz126c9dOzsg5JqsWEqUyNHz+ert26Zh1GvVRWVnLyyd9k0sQ7ePqpudxy883Mmzcv67DqrEuXLsyaOYtZM2cxfdp0Nt10Uw45+JCsw1pvl156CV261D63zj/vfKZPm8n0aTPp1at3NoGtp1HHjGLSxDtqzXvggfuZOHECs2Y+wZNznuLUU76TUXTrr6KigosuvIinnnyKRx5+hCuuvJJ5z8xj7733Yc7sOTwxazY777xz2f7RlIdrGuTvupa39ow69ljumHRn1mE0iDwdm4033ph77r6HJ56YzayZTzBlyhQef/zxrMPasEUGjxJmwlSGFi9ezJ133cno0SdkHUq9TJ8+nc6dO9OpUydatGjB8BEjmDBxQtZhNYj77ruXTp06scMOO2QdynpZvHgxd02+i+OPH511KPU2ePAQttxyq1rzfvPb33Daad9j4403BmDbbbfNIrQ6adOmDX369AVg8803p2vXrixdspR9992Xioqq3tWDBg1iyZLFWYZZJ3m5pkH+rmt5a8+QwUPYaqutPnnFMpCnYxMRtGzZEoBVq1bxwapVdglTSTFhKkOnfucULjj/Apo1K+/Dt3TpErZvv331dPt27Vi6ZEmGETWcW269lREjjsg6jPV22mnf4bxzz/+fc+vMs86g/4C+nHbad3n//fcziq7+XnjhBR6Z+gif/8Jn2XufPZk5c0bWIdXJwoULefLJOQwcOLDW/GuuuYb99ts/o6jqLi/XNMjfdS1v7cmTvB2byspK+vXrS5u2rdl7n30YNGhQ1iFJ1cr6t1NErCj83zEijlyH9TtGxNzC8/4RMb6xY2xok+6YxLbbbEu/fv2yDkVFrFy5kkmTJnLYVw/LOpT1cuedd7DNttvSt2/fWvPP/tk5PPXkXKY+8hivvf4aF//i5xlFWH8ffPABr7/2Oo88/CgXnH8hRx45kpRS1mGtlxUrVjDiiOFcfPEv2GKLLarnn3/B+VRUVHDkyE+8FJYUr2mSAJo3b86sWU/wz4X/YsaMGcydOzfrkDZcUVX1a+pHKSvrhKmGjsB6/ZWQUpqZUjq5ccJpPI8++igTJ02k806dOOqoI7n//vsZNeqYrMOqk7Zt27Fo8aLq6cVLltC2XbsMI2oYkydPpk+fPmy33XZZh7JeHn3sUe6YNIlduuzMqFFH88AD93Pc8cfSpk0bIoKNN96YUaOOZebMmVmHWmft27XjkEMOISIYMGAgzZo1Y/ny5VmHtc5WrVrFiBHDGXnESA495NDq+ddddy133nkH1117Xcn/0llbnq5pkL/rWt7akyd5PTatWrVijz32YMrfp2QdilQtLwnTBcDgiJgTEacUKkkPR8QThcfn1n5BROwREZMKzwdGxGMRMTsiHo2ILoX5x0XEXyJickS8EBEXNXG7/sd5557HPxf+ixfnv8SNN/6RPffck+uuuz7rsOpkwIABzJ8/nwULFrBy5UpuveUWDjqw/EfIuuWWm8uyO945Z5/Liy8u4PnnXuC6625gjz325Jqrr2XZsmUApJSYOGEC3XfdNeNI6+7LXz6YBx58AIDnn3+elatWsvXWW2cb1DpKKTH2pBPp2rUr3/72KdXzp0yZwsW/+AV/+fNf2XTTTTOMsG7ydE2D/F3X8taePMnTsXn11VerR/h89913ueeee+jSpUu2QW3AAoho+kcpy8v3MH0f+G5K6UCAiNgU2Del9F5E7AzcBPT/mNc/CwxOKX0QEfsA5wFfLSzrDfQB3geei4hLUkqLar44IsYCYwE6dOjQcK3KuYqKCsaNG8+wA4ZSWVnJcccdT/fu3bMOq17efvtt7rn3Hi6//IqsQ2kwxx1/LMuXv0pKiZ49e3HpJZdlHdI6OfqYo3jooQdZvnw5O3bagTN+cibHHXc8J44dQ+8+vWjRogW/v+oPZVORefTRqdx444306NGD/gOquq+d/bNzOPXUU3h/5fsMHVZ179KggYO47LLLswx1g5a361re2nPU0Ufy4INV14UdOnbgzDPOLNvBRvJ0bJYtW8bo0cdTWVnJ6rSaww47nAMPODDrsKRqUW7992uKiBUppZYRsQe1E6ZPA5dSlexUAruklDaNiI7ApJRSj5qviYjtgfHAzkACNkopdY2I44DPp5ROLGz3LuDclNIjxWLq369/mjZtemM0V/VUzuf6R6msXJ11CA0qDzf8fyhf51q+jo0kQcVGzWellD7uw/QNVvfuvdIfb5rc5Pvt3attyR6TvFSY1nYK8DLQi6puh+99wvpnA/enlA4tJFUP1FhWc0iwSvL7M5MkSZLKpvdFU8nLx4ZvAZvXmP40sCyltBo4Bmj+Ca//NLBmLM7jGjw6SZIkSWUpLwnTU0BlRDwZEacAlwPHRsSTQFfg7U94/UXA+RExGytIkiRJ2oBFBo9SVtbJQUqpZeH/VcBeay3uWeP56YX1FgI9Cs8foND1LqX0GLBLjfV/XJh/DXBNjf15B6IkSZK0ASnrhEmSJElSw/Ieptry0iVPkiRJkhqcCZMkSZIkFWGXPEmSJEnV7JFXmxUmSZIkSSrCCpMkSZIkoGrABwd9qM0KkyRJkiQVYcIkSZIkSUXYJU+SJElSNXvk1WaFSZIkSZKKsMIkSZIkqZoVptqsMEmSJElSESZMkiRJklSEXfIkSZIkVfN7mGqzwiRJkiRJRVhhkiRJklTNAlNtVpgkSZIkqQgrTJIkSZKAquqS9zDVZoVJkiRJkoowYZIkSZKkIuyS18ASkFLKOowGk6eSbJ7aAlBR0TzrEBpUvt43+fosavXq1VmH0KCaNcvX8ZGaQp6u0dL68reGJEmSJBVhhUmSJElStbz1yqkvK0ySJEmSVIQJkyRJkiQVYZc8SZIkSdXskVebFSZJkiRJKsKESZIkSZKKMGGSJEmSpCK8h0mSJElSNe9hqs0KkyRJkiQVYcIkSZIkSUXYJU+SJElStcA+eTVZYZIkSZKkIqwwSZIkSfqQBaZarDBJkiRJUhEmTJIkSZJUhF3yJEmSJAFV38Hk9zDVZoWpDL3xxhuMGDGcHj26s9tuPXjs8ceyDqnOJk+ZzK7du9Gl6y5ceNGFWYdTL2PGnECbtq3p1btn1qE0mHI+PmNOHEPbdm3o3btX9bzXXnuN/YfuR7ddu7L/0P14/fXXM4yw7sr9XFu0aBH7fmkfevbqSa/evbjkkvEA3Pbn2+jVuxcbb9KCWbNmZhxl3ZXz+2ZtixYtYu999ma3nj3o2Ws3xo8fn3VI9ZKnYwPl356Puk6feeYZ9Onbh379+zF02P4sXbo0wwilKiZMZeiUU0/hS/vtx9y5/2DWrCfo1rVb1iHVSWVlJSef/E0mTbyDp5+ayy0338y8efOyDqvORh17LHdMujPrMBpMuR+fY0eNYtKkO2rNu+iiC9lrz714Zt6z7LXnXlxUhn9gQPmfaxUVFVx04UU89eRTPPLwI1xx5ZXMe2Ye3Xftzq233MrgwYOzDrHOyv19s7aKigp+ftHPefqpuUx95FGuuPLysm1P3o5NHtrzUdfp73znu8x+YjazZs5i2LADOOfcczKKbkMWmfwrZSZMZebNN9/kkUceZvTxowFo0aIFrVq1yjaoOpo+fTqdO3emU6dOtGjRguEjRjBh4oSsw6qzIYOHsNVWW2UdRoMp9+MzePAQttqy9vGYOHEixxwzCoBjjhnFhAnl056ayv1ca9OmDX369AVg8803p2vXrixdspRu3brRpUuXjKOrn3J/36ytTZs29O1b+1gtWbok46jqJm/HJg/t+ajr9BZbbFH9/J233ybsG6YSYMJUZhYsWMDWW2/NCWNOoP+A/ow9aSxvv/121mHVydKlS9i+/fbV0+3btWPpkvL8RZxHeTw+L7/yMm3atAGgdevWvPzKyxlHpIULF/Lkk3MYOHBg1qE0iDy+b9ZYuHAhc+bMYdDAQVmHUid5OzZ5a09NP/nJj9mxU0duuukmzjrzrKzDkfKRMEVEZUTMiYh/RMSTEfGdiPjEtkXED9dhnWsi4rCGibT+Pqj8gNmzZ3PSSScxc8ZMNttss7LtViRlKSL85DJjK1asYMQRw7n44l/U+lRZpWfFihUMH344v/zFLz1WanRnn30OC15ayMiRI7n88suyDmfDFBk8SlguEibg3ZRS75RSd2BfYChw5jq87hMTplLTvl172rdvX/0J31e/8hVmz5mdcVR107ZtOxYtXlQ9vXjJEtq2a5dhRKopj8dnu223Y9myZQAsW7aMbbfZNuOINlyrVq1ixIjhjDxiJIcecmjW4TSYPL5vVq1axeHDD2PkyCM59NCvZB1OneXt2OStPR9l5Mgj+etf/5p1GFJuEqZqKaVXgLHAN6LKcRFx6ZrlETEpIvaIiAuATxUqUzcWlo2KiKcKVarra2x2SEQ8GhEvZV1tat26Ne3bt+e5554D4L777qNbt/Ic9GHAgAHMnz+fBQsWsHLlSm695RYOOvCgrMNSQR6Pz4EHHcj1118HwPXXX8dBB5V3e8pVSomxJ51I165d+fa3T8k6nAaVt/dNSokTTxxDt67dOOWU8j5WeTs2eWvPGi+88EL18wkTJ5T9fY3lygJTbbn8HqaU0ksR0Rwo+vFxSun7EfGNlFJvgIjoDvwY+FxKaXlE1LwLsQ3wBaArMAG4rdGCXwe//tU4Rh07ipUrV9Jpxx256qrfZxlOnVVUVDBu3HiGHTCUyspKjjvueLp37551WHV21NFH8uCDD7J8+XJ26NiBM884k9GjT8g6rDor9+Nz9NFH8eBDVcej4447cMYZZ/K9005n5JFHcPU1V9OhQwdu+uPNWYdZJ+V+rj366FRuvPFGevToQf8B/QA4+2fn8P7K9znllG/z6quvcvAhB9OrZy/uuKO8RgMs9/fN2qZOncoNN97Abj12o1+/qsEfzj7nHIYNHZZxZOsvb8cmD+35qOv05Lvu4vnnnyeaNWOHDh247LLLsw5TIlJKWcdQbxGxIqXUcq15bwBdqOqe1z+l9I3C/EnAxSmlB2q+LiK+CbROKf1ore1cA9ydUlpThXorpbT5WuuMpaqqRYcOHfq9OP+lRmhlNrzHQ00lD9eiNfL2vlm9enXWITSoZs1y17lCanR5ukYDbNSiYlZKqX/WcZSinj37pDsm3d/k++2ww5Yle0xy+VsjIjoBlcArwAfUbucmddjk+zU3v/bClNJvU0r9U0r9t956mzpsXpIkSVIpyl3CFBHbAFcCl6aqj0MWAr0jollEbA/UHLt2VURsVHh+H3B4RHymsJ3y/ZITSZIkSQ0iL/cwfSoi5gAbUVVRuh74ZWHZVGABMA94Bniixut+CzwVEU+klI6KiHOBByOiEpgNHNc04UuSJEklIl89y+stFwlTSqn5xyxLwFFFlp0OnF5j+lrg2rXWOW6t6Vr3SkmSJEnKr1wkTJIkSZIahgWm2nJ3D5MkSZIkNRQTJkmSJEkqwi55kiRJkoCq7nh5+z7B+rLCJEmSJElFmDBJkiRJUhEmTJIkSZJUhPcwSZIkSaoS4C1MtVlhkiRJkqQiTJgkSZIkqQi75EmSJEmq5rDitVlhkiRJkqQiTJgkSZIkqQgTJkmSJEkqwoRJkiRJkoowYZIkSZJULaLpH58cU2wfEfdHxLyI+EdEfKswf6uIuDsiXij8v2VhfkTE+IiYHxFPRUTfuv48TJgkSZIklboPgO+klHYFdge+HhG7At8H7k0p7QzcW5gGGArsXHiMBa6o645NmCRJkiQVRCb/PklKaVlK6YnC87eAZ4B2wMHAtYXVrgUOKTw/GLguVXkcaBURberyEzFhkiRJklQ2IqIj0AeYBmyXUlpWWPRvYLvC83bAohovW1yYt9784lpJkiRJWds6ImbWmP5tSum3a68UES2BPwPfTin9t+aX7KaUUkSkhg7MhEmSJEnSh9ZhEIZGsDyl1P/jVoiIjahKlm5MKf2lMPvliGiTUlpW6HL3SmH+EmD7Gi9vX5i33kyYGlgAsS5DfUiqxfdN6WrWLF+9t/fe6KysQ2gw9646K+sQtIHwGq2sRdVJ+HvgmZTSL2ssmgAcC1xQ+P9vNeZ/IyJuBgYBb9bourdeTJgkSZIkAWs+/M86io/0eeAY4OmImFOY90OqEqVbI+IE4J/A8MKyO4FhwHzgHeD4uu7YhEmSJElSSUspPULxzoJ7f8T6Cfh6Q+zbhEmSJElStdIsMGUnXx3TJUmSJKkBmTBJkiRJUhF2yZMkSZJUpYRHfciKFSZJkiRJKsIKkyRJkqRq1pdqs8IkSZIkSUWYMEmSJElSEXbJkyRJklTNMR9qs8IkSZIkSUVYYZIkSZL0IUtMtVhhkiRJkqQirDBJkiRJqmZ9qTYrTJIkSZJUhAmTJEmSJBVhwlSGJk+ZzK7du9Gl6y5ceNGFWYdTL2PGnECbtq3p1btn1qE0iDwdG8hfeyorK+nfvx9fPvigrEOpl7y9b6C0j833fn8If335e1z99Nc/cnnPwTvw21lf495VZ/LFr+5aPb/3Hjty1ez/V/34+7s/4QsHdwWgz5478ttZX+Pqp7/O9685lObNS/fXcZ6uA3lqC+SrPXlqS7kLqsZ8aOpHKSvdK7Q+UmVlJSef/E0mTbyDp5+ayy0338y8efOyDqvORh17LHdMujPrMBpE3o5N3toDMH78eLp265p1GPWWp/fNGqV8bCZfM5vv7X990eWv/OtNLjjur9zzx6drzZ/zwALG9LmCMX2u4JS9ruG9d1Yx4+8vEhH84Nqv8LMj/sTxu13Gy/98k/2O7d3IraibPF0H8tQWyFd78tQW5ZMJU5mZPn06nTt3plOnTrRo0YLhI0YwYeKErMOqsyGDh7DVVltlHUaDyNuxyVt7Fi9ezJ133cno0SdkHUq95el9A6V/bJ56+J+89dq7RZf/+59v8NLTL5NWp6LrfPGwXZl21wu8/+4qtvjMp1i1spLFL/wHgJl3v8iQGpWpUpKn60Ce2gL5ak+e2pIfkcGjdJkwlZmlS5ewffvtq6fbt2vH0iVLMoxIa+Tt2OStPad+5xQuOP8CmjXzsldqNoRjs9cRu3HfTVUVqDeXv0PzimZ06dcWqEqmtt3+01mGV1SergN5agvkqz15aovyqaR/O0VE64i4OSJejIhZEXFnRIyNiElF1r8qIkrzYzpJmZl0xyS23WZb+vXrl3UoWsuGcGy2at2STrttx/Qp86vn/eyIP/H1Xw3limljefetlayuXJ1hhJKkj1Oy38MUEQH8Fbg2pXREYV4v4MvFXpNSGtNE4WWmbdt2LFq8qHp68ZIltG3XLsOItEbejk2e2vPoo48ycdJE7pp8F++99x7//e9/GTXqGK67rvh9KWoaG8Kx2XN4Dx7+6zNUfvBhUjTv8UWcPOT3APTftzPtd/lMVuF9rDxdB/LUFshXe/LUlrwo9UEYmlopV5j2BFallK5cMyOl9CTwMNAyIm6LiGcj4sZCckVEPBAR/QvPV0TEuRHxZEQ8HhHbFeZvExF/jogZhcfnC/O/GBFzCo/ZEbF5Yf5phfWeioifNvHP4H8MGDCA+fPns2DBAlauXMmtt9zCQQeW3qhSG6K8HZs8tee8c8/jnwv/xYvzX+LGG//Innvumas/yMtZuR6bQ78+kEO/PnCd1t175G7ce1PtASFabbMZABu1aM7I0wcz4coZDR5jQ8jTdSBPbYF8tSdPbVE+lXLC1AOYVWRZH+DbwK5AJ+DzH7HOZsDjKaVewEPAiYX544BfpZQGAF8FrirM/y7w9ZRSb2Aw8G5EfAnYGRgI9Ab6RcSQtXdU6CY4MyJmvrr81fVs5vqpqKhg3LjxDDtgKD12685hhx9O9+7dG3Wfjemoo4/kC4M/z3PPPccOHTvwhz/8PuuQ6ixvxyZv7cmTPL1vysFP/ngYlz12Itt32Zo/LfoOw0b3pUPXbXjzP1UDQXTp35Y/LfoOXzy8O6f+5stcPfcb1a9tvUMrttn+0zz54MJa2zzitM9z7bxv8vunvs5jE59j9v0LmrJJ6yxP14E8tQXy1Z48tUX5FCkVH9UnSxFxMrBjSumUtebvAfwopbRvYfoKYGpK6YaIeAD4bkppZkS8D2ySUkoRMQLYN6U0JiJeAZbW2OQ2QBfgG8ChwI3AX1JKiyPiYuAw4I3Cui2B81NKRf866d+vf5o2bXo9Wy9Jaix7b3RWvbdx/sSj+MlXbuaDVZX1D6ge7l11Vqb7l8pVxUbNZ6WU+mcdRynq07tvuv++R5p8v1t+ZrOSPSYlew8T8A+qkpWP8n6N55V8dDtWpQ+zwZrrNAN2Tym9t9b6F0TEHcAwYGpE7EfVGIfnp5R+U5cGSJLy6QcH3Zh1CJKkJlLKXfLuAzaOiLFrZkRET6q6y9XH34Fv1thm78L/nVNKT6eULgRmAF2BKcDoiGhZWKddRGxbz/1LkiRJpSmqBn1o6kcpK9mEqVAdOhTYpzCs+D+A84F/13PTJwP9C4M4zAO+Vpj/7YiYGxFPAauAu1JKfwf+CDwWEU8DtwGb13P/kiRJkspEKXfJI6W0FBj+EYt+V2Odb9R4vkeN5y1rPL+NqmSHlNJyYMRH7Ouba88rzB9H1UARkiRJ0gagxEs+TaxkK0ySJEmSlLWSrjBJkiRJalqlfk9RU7PCJEmSJElFmDBJkiRJUhEmTJIkSZJUhAmTJEmSJBXhoA+SJEmSPuSgD7VYYZIkSZKkIkyYJEmSJKkIu+RJkiRJAiAK//QhK0ySJEmSVIQJkyRJkiQVYcIkSZIkSUWYMEmSJElSEQ76IEmSJKlaOOZDLVaYJEmSJKkIEyZJkiRJKsKESZIkSZKK8B4mSZIkSVUCb2JaixUmSZIkSSrCClMDS0BKKeswGkz4CYOayOrVq7MOocE0a5avz6LydE0DuHfVWVmHoCJWr87Xudasmb9DpTwwYZIkSZJUzVS/tnx9DCpJkiRJDcgKkyRJkqQPWWKqxQqTJEmSJBVhwiRJkiRJRdglT5IkSVI1e+TVZoVJkiRJkoqwwiRJkiTpQ34PZy1WmCRJkiSpCBMmSZIkSSrChEmSJEmSijBhkiRJkqQiHPRBkiRJUjWHfKjNCpMkSZIkFWGFSZIkSdKHLDHVYoVJkiRJkoowYZIkSZKkIuySJ0mSJAmo6o0X9smrxQpTGdpp58707tObfv37MWj3QVmHUy+Tp0xm1+7d6NJ1Fy686MKsw2kQlZWV9O/fjy8ffFDWodRbOR+fRYsWse+X9qFnr5706t2LSy4ZD8CZZ51J33596D+gH8OGDWXp0qUZR7r+Fi1axN777M1uPXvQs9dujB8/PuuQ6uWNN95gxIjh9OjRnd1268Fjjz+WdUj1Us7vm49Szu05cewY2rVvQ+8+varn/ezsn9Jxxw70H9CP/gP6cdddd2YYYf2U87FZW57aovyJlFLWMeRKv37907THpzXqPnbauTOPPzaNrbfeulH3AxDReJ8wVFZW0m3Xrky+awrt27dn990HccMNN7Lrrrs22j6bwq9+9StmPTGT//73v0z428Ssw6mzpj4+q1evbtDtLVu2jH//exl9+vTlrbfeYtDug7jtttto3649W2yxBQCXXnoJzzzzDJdddnmD7rtZs8b9LGrZsmUsW7aMvn2r2jZw0AD+fNtfGu3YNPbvieNHH88XvvAFThh9AitXruSdd96hVatWjbY/r2vrrumvAw17rj388EO0bNmS40cfz5zZTwJVCVPLzVpy6qnfadB9fZRmzTzX1kUWbanYqPmslFL/RttBGevXt1+aOvXxJt/vpzZtUbLHxAqTMjN9+nQ6d+5Mp06daNGiBcNHjGDCxAlZh1Uvixcv5s677mT06BOyDqXeyv34tGnThj59+gKw+eab07VrV5YuWVqdLAG8/c47jfrHc2Np06YNffvWbtuSpUsyjqpu3nzzTR555GFGHz8agBYtWjRqstTYyv19s7Zyb8/gwUPYcsutsg6jUZT7sakpT21RPpkwlaGIYOiwoQwcNJDfXfW7rMOps6VLl7B9++2rp9u3a8fSJeX5R98ap37nFC44/4JGrzA0hTwdn4ULF/Lkk3MYOHAgAD854yd06rwjN910E2eeeVa2wdXTwoULmTNnDoMGlmf33AULFrD11ltzwpgT6D+gP2NPGsvbb7+ddVh1lqf3DeSvPWtcceXl9O3XhxPHjuH111/POpw6ydOxyVNblE8l/VddRGwXEX+MiJciYlZEPBYRh2YdV9YeuP9BZkyfwaSJk7jiiit4+OGHsg5JwKQ7JrHtNtvSr1+/rENRDStWrGDEEcO5+OJfVFeXzv7Z2bz04gJGjhzJ5Vc0bHe8prRixQqGDz+cX/7il7UqZ+Xkg8oPmD17NieddBIzZ8xks8024yLvX1AjOmns13j2meeZOWMWrVu35nunn5Z1SFLJiQwepaxkE6ao6idzO/BQSqlTSqkfcATQfh1fn9sRANu1awfAtttuyyEHH8yMGTMyjqhu2rZtx6LFi6qnFy9ZQttC28rRo48+ysRJE+m8UyeOOupI7r//fkaNOibrsOosD8dn1apVjBgxnJFHjOTQQ/73s5aRR4zkr3/9awaR1d+qVas4fPhhjBx5JIce+pWsw6mz9u3a0759++oK2Ve/8hVmz5mdcVR1l4f3TU15aw/AdtttR/PmzWnWrBknjB7j79ASkKe2KJ9KNmEC9gJWppSuXDMjpfTPlNIlEdE8In4eETMi4qmIOAkgIvaIiIcjYgIwrzD9YET8rVCluiAijoqI6RHxdER0LrzuoIiYFhGzI+KeiNiuMP+siPhDRDxQeP3JWfwganr77bd56623qp/ffc/ddO/ePeOo6mbAgAHMnz+fBQsWsHLlSm695RYOOrB8R5Y779zz+OfCf/Hi/Je48cY/sueee3LddddnHVadlfvxSSkx9qQT6dq1K9/+9inV81944YXq5xMnTqBLly5ZhFcvKSVOPHEM3bp245RTTvnkF5Sw1q1b0759e5577jkA7rvvPrp165ZxVHVX7u+bteWtPVA1aMoaf/vb7f4OLQF5aksuZFFeKvESUylXYboDTxRZdgLwZkppQERsDEyNiL8XlvUFeqSUFkTEHkAvoBvwGvAScFVKaWBEfAv4JvBt4BFg95RSiogxwPeANcPndAX2BDYHnouIK1JKq2oGExFjgbEAHTp0qHfDP87LL7/MYYcfBkDlBx9wxBFHsN9++zfqPhtLRUUF48aNZ9gBQ6msrOS4444v219ceVTux+fRR6dy44030qNHD/oPqOomefbPzuHqa67m+eefp1mzoEOHHbjs0ssyjnT9TZ06lRtuvIHdeuxGv35Vgz+cfc45DBs6LOPI6ubXvxrHqGNHsXLlSjrtuCNXXfX7rEOqs3J/36yt3Ntz9DFH8dBDD7J8+XJ27LQDZ/zkTB586EGefPJJIoIddtiByy+7Iusw66Tcj01NeWqL8qlkhxUvVHN2TCmdUpi+DPgCsBL4J9ATeKew+qeBkwrLzkwp7Vl4zR7Aj1JK+xamHwJ+kFKaGhF7ASenlA6JiN2AXwBt4P+3d+dhklbl3ce/P4Zl2BFB4gKKCBoCioALILygwYDGBeMSATWCoDEiLolb3kQ0iUSN24sGRQ2v4hKdAIoruCEgqMBkAAG9VJZgICKrCLLM8Msf59RMdU/3wFQ1/dR5+H2ua66uOtXddT9T1VV1P+c+92Ft4HLb+0k6CrjL9j/Vn78U2Nf2r2aLez7ais+nFjuIRZvmuq14l/rQ9GPYpL5PjCqva5NrrtuKd+2+bCse40lb8dntsssuPvsH8/9ZduG6a03sYzLJ7+oXU2aLALD9V8DTgM0pE3dH2N6p/tva9mCGaXp7pTuGLt89dP1uVsywHQN82PaOlMRr4Sw/v4zJnpWLiIiIiIg5NMkJ03eBhZL+cmhsvfr1VOAvJa0FIGk7SeuPcV8bA4P+lS8b4/dERERERESPTOxsSV1P9FzgA5LeBPyGMnv0ZmAR8Ahgce2m9xvguWPc3VHAIkk3UhK1rcf4XRERERERzUox6VQTu4apVVnDFDGarGGaXH17n8jr2uTKGqaYL1nDNLtddtnF53SwhmmdCV7DNLEzTBERERER0YHk+lP06zRoRERERETEHErCFBERERERMYuU5EVERERExHKpyJsqM0wRERERERGzyAxTRERERERUgnQTnSIzTBEREREREbNIwhQRERERETGLJEwRERERERGzSMIUERERERExizR9iIiIiIiI5dLzYarMMEVERERERMwiCVNERERERMQskjBFRERERETMIglTRERERETELNL0ISIiIiIiABCgdH2YIjNMEREREREx8STtJ+lnkn4h6S3zdb9JmCIiIiIiYqJJWgB8BNgf2B54saTt5+O+kzBFRERERMSkeyLwC9uX2b4T+HfgOfNxx0mYIiIiIiJi0j0UuGro+q/q2H0uTR/m2OLF51+31tprXjkPd7UZcN083M986dPx9OlYIMczyfp0LNCv4+nTsUCOZ5L16Vhg/o7n4fNwH006f/H5p6651oLNOrjrhZLOG7p+nO3jOohjJUmY5pjtzefjfiSdZ3vX+biv+dCn4+nTsUCOZ5L16VigX8fTp2OBHM8k69OxQP+Op0W29+s6hln8N7Dl0PWH1bH7XEryIiIiIiJi0p0LbCtpa0lrA38OnDIfd5wZpoiIiIiImGi2l0p6DXAqsAD4N9sXz8d9J2Fq10TUdM6hPh1Pn44FcjyTrE/HAv06nj4dC+R4JlmfjgX6dzwxh2x/Hfj6fN+vbM/3fUZERERERDQha5giIiIiIiJmkYQpIiIiIiJiFkmYIiIiIiIiZpGmDzHvJK0HvBHYyvZhkrYFHm37qx2HFhERQyQ9lLLB5/LPC7bP6C6iGJD0FGBb28dL2hzYwPblXcc1CkkLgC2Y+jz7r+4iipgqTR8aImkP4ChWvHkJsO1HdhnX6pL0BeB84KW2d6gJ1Nm2d+o2svFI2h14BFNf8D/dWUBj6FtSK2kd4M9Y+fF5Z1cxjap+MDqMlY/lkK5iGoek7YC/YeUP5U/tLKgR9fCxeTfwIuASYFkdtu1ndxfV6CQ9D3g38CDK++fgPXSjTgMbgaS3A7tSXpe3k/QQYJHtPToObbVJOgJ4O/Br4O46bNuP7S6qiKkyw9SWTwKvpyQby+7heyfZNrZfJOnFALZvk6SugxqHpBOAbYAlDH2wAJpMmIDjKc+z3er1/wYWAU0mTMCXgZspx3RHx7GM68vAmcC3aft1YGAR8FHg47R/PH17bJ5L+UDe+t/MwHuAZ9m+tOtA5sABwOOBxQC2r5a0YbchjexIyvPs+q4DiZhNEqa23Gz7G10HMQfulLQuJaFA0ja0/yF2V2B792fKtm9J7cNs79d1EHNkPdtv7jqIObTU9rFdBzFH+vbYXAasRfuvzwO/7kmyBHCnbUsavI+u33VAY7iKckIrYmIlYWqApJ3rxe9Jei9wEkNvYLYXdxLY6N4OfBPYUtJngT2Av+g0ovH9BPgD4JquA5kjfUtqz5a0o+2Lug5kDnxV0jPq5n198BVJrwZOZurr2g3dhTSyvj02twFLJH2HqY/Na7sLafXVUjyA82pJ+JeYejwndRHXmL4o6WPAJpIOAw6hzNK26DLgdElfY+rj8v7uQoqYKmuYGiDpe6u42Y3W+j8QeDKlhvyHtq/rOKSRSPoKJanYENgJ+DFTX/BbrfXfF/i/wPbAadSk1vbpXca1uiRdRHl81gS2pbwx38GKtQvN1MhLuoVyLALWpxzHXTS8DgNA0kyL1JtbmwnLH6M+PTYvm2nc9qfmO5ZxSDp+FTe74TVm+wJPpzzPTrX9rY5DGkldj7US2++Y71giZpOEKeZdbV6xxPatkg4GdgY+ZPvKjkNbbZL+z6put/39+YplrvUhqZX08FXd3uJzLiKiluDdbnuZpEcDjwa+YfuujkOL6KUkTA2RdCRlMf4tlKn3nYG32D6t08BWk6QLgccBj6UczyeBF9peZfIxySRtDVxj+/Z6fV1gC9tXdBrYiPqU1MLyksJf2b5D0t6U596nbd/UZVyjmOWx+WDLLXgl7UCZzVw4GGu4w+QDKLOZw8fSZBvu2h3zaFZ+bJqb/QOQ9CngyMHffX2s3tfiDJOk84E9gQcAZwHnUdY1HdRpYCOo3SXfBPwRU59nzVXPRH9l49q2HGL7t5Qp+AcCLwH+uduQRrK0Nkd4DvAR2x+hlLS1bBEr2qFC6ZC1qKNY5sKxwG2SHge8Afgl7Xb8AzgRWCbpUcBxwJbA57oNaWTDj80bKY/NCd2GNLpajnNM/bcPpZNZq6WsrwDOAE4F3lG/HtVlTGM6nvJ8W0p5bD4NfKbTiMbz2OGTJLZvpHSaa5Fs3wY8DzjW9gsoCUeLPgv8FNia8ndzBXBulwFFTJeEqS2DLmXPoJwdv3horCW3SHorJeH7mqQ1KJ2YWram7TsHV+rltTuMZ1x9S2rvtr2U8uHiGNt/Azy445hGNfzYfLgHj83zgacB/2P75ZTZ5427DWlkRwJPAK60vQ/lw/hNnUY0nnVtf4fy4fxK20cBz+w4pnGsUWeVAJC0Ke02v5Kk3YCDgK/VsQUdxjOOB9r+JHCX7e/XGb/MLsVEafWF4v7qfEmnUc7CvLXuuXD3PfzMJHoRcCBlxux/JG0FvLfjmMb1G0nPtn0KgKTnAM2t+RkynNTu2YOk9q7aIv2lwLPqWKvHM3hsDgb26sFj83vbd0taKmkj4FrKDGCLbrd9uyQkrWP7p3V9SavuqM+vn0t6DWU/tg06jmkc7wPOkTSY/X8B8K4O4xnH64C3AifbvljSI4FVNYiaZIN1V9dIeiZwNbBph/FErCRrmBpS37h2Ai6zfVNdlP9Q2xd2G9nqq4vxt7X9bUnrAQts39J1XKOqa2Q+CzyEMut3FfBS27/oNLARSfoDSlJ7ru0za1K7d8PrSrYHXgWcY/vzdc3ZC22/u+PQVlsPH5t/Bd4G/DmlxPB3lDVaL+80sBFIOhl4OeXD7FOBG4G1bD+jy7hGJekJwKXAJsA/UGb+3mP7h13GNY76WjCYvfiu7Uu6jGdcktarpXnNkvSnlA2ft6SU5m4EvGNwAjJiEiRhaoCkx9QzlTvPdHtr+zDVPSMOBza1vU1dWPxR20/rOLSxSdoAwPbvuo5lXH1LamPySXoEsFGLJ4Gmqx00Nwa+OVyuG92RdILtl9zTWAtqOd4ngQ1sb1XXNL7S9qs7Di2il1KS14Y3UBKM981wm2mv1vevgCcCPwKw/XNJD+o2pPFIWgf4M+ARwJpSWVpm+50dhjWy4aQW2AZ4KPBRylqTZkj6ou0XDu3HNEVj+zCdZfspQ/sxLb+JBvf6WdWJIEk7t3YiaEDSUygnGo6v3b8eCsy019TEkvRB268b2mduilb3l2NaUwRJC4BdOoplXB8E/gQ4BcD2BZL26jSiEdUZ/yOo75+D8YafZ9FDSZgaYPvw+nWfrmOZI3fYvnOQVEhakxnelBvzZeBm4HyGNq5tWF+S2iPr1z/tNIo5YPsp9WvLDR6G9e1E0KDj366UPXGOp6wt+wxl4+eWDLou/kunUcyRuubvbcC6kn7LimZJd1K6ZjbJ9lWD99FqWVexjOlLlNmyr9Dmuuy4H0jC1BhJu7PyWZjW1i58X9LgzWtf4NWUF8qWPcz2fl0HMYd6kdTavqZ+bXL/qNnUM+NbMPV1oKl9mHp4IgjgAEpnvMUAtq+uzXmaYvv8+rXZjbeH2T4aOFrS0bbf2nU8c+Sq+nnAktainBy6tOOYRnW77f/XdRARq5KEqSGSTqCURy1hxZkk097+OG8BDgUuAl4JfB34RKcRje9sSTvavqjrQOZIL5LaGcrXlt9Eg2VsAJKOAN4O/JoVZ2NN2Yy3OZKeN8PwzcBFtq+d73jGdKdtSzKApPW7DmgUs5WwDrRUyjrNN2YqW2t0Y+FXAR+ilHz+N3AapTKgRR+qs7OnMVSh0WpZbvRTmj40RNKlwPbOgzZxJF0CPIqyVuEOVnwgb/KDRe3IeChlk2RRNuD8RJ573ZP0C+BJtq/vOpa5IOlrwG6saIm8N6W0dWvgnbab2ZRX0l8D2wL7AkcDhwCfs31Mp4GtptrwZVatztjWNVkDCyllx+fbbq78s08kHU3ZwuKXDJ0EyuMSkyQJU0Pq3hGvHZQZtUrSHsBRwMMps5yD5OKRXcY1jtk+YLT6waJv6gaV091i+64ZxieapO8B+9aNeJsn6VRKC/5f1+tbUGbNXwycYXuHLuO7t1TqVx8GPIahEw22v9VpYDErSVsCH7T9Z13HsrpqQ5HDWLlE/5CuYhpVPQm0fbpJxiRLSV5bNgMukfRjpk5bt9ZJ5pPA6ylnkVtdpDqF7Stn6I7V7AaPki5n5u5YrSa1iyl7fNxI+SC7CfA/kn4NHDZYs9GIy4DT68zM8OvA+7sLaSxbDpKl6to6doOkZhLaWor3dds7Ar1IkiQ9mbIvzh8CawMLgFtbLGWdxa8ox9aiL1P2Lvo27b+P/oTymtxaCW7cjyRhastRXQcwR262/Y2ug5hLPeqONbDr0OWFwAtoe+f1bwH/YftUAElPp7SBPx74V+BJHca2uv6r/lu7/mvd6ZK+Ciyq159fx9YHbuosqtEslvQE2+d2Hcgc+TBlQ+FFlNeElwLbdRrRGCQdw4oTQYON4FtdJ7Oe7Td3HcQc2QT4qaRzaftkcPRYSvJi3kn6Z8qZypPoyQJPSUuo3bFsP76OXdjqGqaZSDrfdpN7lki6qJ75Hx670PZjJS2xvVNHod3v1VK25wFPqUM/AE5scb2cpJ9S1jJeCdxK+2sZz7O96/BrmaT/HLzGtUbSy4auLgWusP2DruIZh6R/BM62/fWuYxlX3eR5JX3p0hj9kBmmhvSoPGJwNn94FqPJfVeG9KI71sC0zUTXoDxWLb9eXCPpzcC/1+svAn5d23M3te9HLfd8E2UTzoWD8VYXSNe/m/MoM8/flrQepZz1lo5DG8WfdB3AHLtN0trAEknvAa6hvB40yfan6t8Ptn/TdTxjOhJ4m6Q7gLtotPNnfQ3+mO3HdB1LxKq0/AHo/qgX5RE923dl4IuSPgZsIukwSnesj3cc0ziGNxNdClwBvLCbUObEgZRW3F+q139QxxbQ3nF9FvgCZTPeVwEvA5r98Ff/Xg6nlHxuQ2mT/FHgaV3GNYq6lvFxwJ516EzbF3QZ05heQkmQXkNZd7olZTawKXUW8+2U41ijDi0FjrH9zk6DG1FfNrC2vUzSzyRt1dpecnH/kpK8hvSpPELSM1n5DHmTb1zpjhXzaVAaOe114FzbT+g6tlHUctYnAj8aKmddqYSyBZKOpHQuO6kOHQAc11pb8QFJR9r+0D2NTTpJbwD2Bw63fXkdeyRwLPBN2x/oMr7VIekxtn86rQpguRZL2yWdQSlp/zGllBXIGqaYLJlhaksvyiMkfRRYD9iHsmHt8ykvlE3qaXesN8wwfDNlz5Il8xzO2CRtB/w1K7fgbbGMbdA57pp64uFq2m7IcYftO8t5B5C0JqvYNHXCHUrZI+tWAEnvBs6hlFK36GWUzVGH/cUMY5PuJZRW/NcNBmxfJulgymapzSRMwBsoM7Lvm+G2Vkvb/67rACLuSRKmtsxUHtHc/hHA7nWx/YW23yHpfUDrXfP61h1r1/pvsNHjnwIXAq+StMj2ezqLbDSLKGVen6D9Frz/KGlj4I2UD+IbUV4PWvV9SW8D1pW0L/BqVjzvWiOmPr+W1bGmSHoxpWR1a0mnDN20IXBDN1GNZa3hZGnA9m8krdVFQKOyfXj92pvSdtvfr3sZbju0jnFB13FFDEvC1JBaHz9YsPqOruMZw+/r19skPQS4Hnhwh/HMhScBB0u6gh50x6KUGO5s+3ewvG3614C9KPtntZYwLbV9bNdBzAXbX60Xb6bM0rbuLZSZmYuAVwJfpyS2LToe+JGkk+v151L2nWvN2ZQKhs2YOpNxC+XESWtWtSFqs5ulStqdlWfNP91ZQCPq0zrG6K8kTA3o4YLVr0raBHgvZQ8M02iDhKGFqn3rjvUghlq+U8rAtrD9+9qVqTVfkfRq4GSmtrJv7mx5XXvxIWA3Soe/c4DX276s08BGtw/wGdtNvgYASNrS9lW23y/pdFa0SH855cNfU2xfSWmNvlvXscyRx0n67QzjYmgdbUsknUBJLpawYlbTQHMJE/BX1HWMALZ/LulB3YYUMVWaPjSgTwtWp5O0DrDQ9s1dxzIKSYtt71wvn2i7xRLJlUj6O8qC9S/XoWcBp1DONh9n+6CuYhuFpMtnGLbtR857MGOS9EPgI8Dn69CfA0fYbmnz3eUkfYrywfwG4EzgDOAs2zd2GthqqPsv7Wf7imnjhwB/a3ubTgIbkaRbmHkdWZOtq/tI0qXA9i3uVzadpB/ZftKgiVVdx7i44QqN6KEkTA2Q9J9MW7BaxzcHTmutS56khZR1Ck+hvCmfBRxr+/ZOAxvBcJfCVjsWzkbSrsAe9eoPbJ/XZTxRaIYNkSVdYPtxXcU0F2p57vMpzTkeYruZCghJzwA+CDzT9s/r2FuAg4D9bf+qw/CihyQtAl5r+5quYxlXbWJ1E2WrlCMonw8usf23XcYVMSwJUwMk/cT2Dqt726SS9EVKLfxn6tCBwCa2X9BdVKOZNsO0/HLrJG0103hr+2RIetOgQYWkF9heNHTbu2y/rbvoRlM7r91I2YTXlE14H0ApcW2uzLB2KtsT2BG4jnIC5Uzb53Qa2GqS9DTgY5R1S6+glBg9s6WZsph8kr5C+bvfENiJ0mF2uMy4uVbcktagrGN8eh061Xar6xijp5IwNWBVH8Rb/JAu6RLb29/TWAskLWNFk4d1gdsGN9Fw6Yqki1hRkrMusDXwM9t/1F1Uq29VCW2Lfzswpbxw8PgMd2FrrsxQ0nXALymLvL83vaytJZL2pKyTOxt4YYuz5jHZaoOELSjlq8P2BK6x3UyTEUnPAR5m+yP1+o+BzSmvbW+y/R9dxhcxrJmSh/u5vi1YXSzpybZ/CCDpSUCT5V62e9n6dPqmoXWTxFd3FM44NMvlma5PNElPAK6yvXW9/jLKtgJXAEe1NrM0YHszSX9E6cD4T5K2pSTnL+k4tHttaM2PgHUo3b2urQ17mj1xEhPpOcBbbV80PCjpBuBdtNWV8U2UNZgDawO7ABtQOk4mYYqJkYSpAT38UL4LcLakQXnXVsDPBrMaWeg5eWwvroltazzL5ZmuT7qPAX8MIGkv4GhKvf9OwHGU9T/NkbQR5TXg4ZQWyRtTuv81w/aGXccQ9xtbTE+WAGxfJOkRHcQzjrVtXzV0/ax64ucGSet3FVTETJIwRRf26zqAWLXamXFgDWBn4OqOwhnHYHZWlI1RBzO1Lc7OLhiaRXoRpVvhicCJkpZ0F9bYzhr69+E0SIhYpU1Wcdu68xXEHHnA8BXbrxm6uvk8xxKxSkmYYt7VPT6o+ywsHBpvqqFAzw2fMV9K2bT2xI5iGVnPZmcXSFrT9lJKydfhQ7c1+VouaQHwLdtv7DqWiEacJ+mw6fuWSXoFZVPxlvxolmN5JaWZRcTEaPJNNtom6dmU/XweAlxLKcW5FGiqoUDPXTLcUQ5Klzlg0SzfH/e9zwPfr00Sfk9d9C3pUUCT+5jZXiZp967jiGjI64CTJR3EigRpV8r6nwO6CmpErwe+JOlAyib2UEr216F0m4yYGOmSF/NO0gXAU4Fv103q9gEOtn1ox6FFNVMHuVa7yvWJpCcDD6bsv3ZrHdsO2MD24lX+8ISSdCzwUEoyfutg3PZJnQUVMeHq++ZgS5GLbX+3y3jGIemprDhh2vSxRH8lYYp5J+k827vWxOnxtu/uw8abfSBpf+AZwAuBLwzdtBFlV/kndhJY9Jak42cYtu1D5j2YiIiIGaQkL7pwk6QNKCVFn5V0LUNnlqNTV1NavD+bqfXwt1DKJyLmlO2Xdx1DRETEqmSGKeadpPWA2ymdyg6mzF58ttV9ZPqmLsQ/wfaBXccS/SfpYcAxwB516EzgyHTLi4iISZGEKebN0OaOU4br19uBXwJ/a/s78xpYrETSmcDTbN/ZdSzRb5K+BXwOOKEOHQwcZHvf7qKKiIhYIQlTTIQ6q7EDZaZph3v6/rhvSfo08IfAKUxdiP/+zoKKXpK0xPZO9zQWERHRlTW6DiACSnth2xdQSnOie78Evkp5jdhw6F/EXLte0sGSFtR/BwPXdx1URETEQGaYImJWtTkHtn/XdSzRT5IeTjlRshulZPds4LXZyDoiIiZFEqaIWImkHShrSjatQ9cBL7V9cXdRRURERMy/JEwRsRJJZ1MacHyvXt8beJft3buMK/pD0t+v4mbb/od5CyYiImIVsg9TRMxk/UGyBGD7dEnrdxlQ9M5Me6+tDxwKPBBIwhQRERMhM0wRsRJJJwOLmdrqeRfbB3QXVfSVpA2BIynJ0heB99m+ttuoIiIiinTJi4iZHAJsDpwEnAhsVsci5oykTSX9I3AhpeJhZ9tvTrIUERGTJCV5EbGcpIXAq4BHARcBb7R9V7dRRR9Jei/wPOA4YMd0YoyIiEmVkryIWE7SF4C7gDOB/YErbL+u06CilyTdDdwBLKW0E19+E6Xpw0adBBYRETFNEqaIWE7SRbZ3rJfXBH5se+eOw4qIiIjoTNYwRcSw5eV3tpd2GUhERETEJMgMU0QsJ2kZK9o9C1gXuI2USUVERMT9VBKmiIiIiIiIWaQkLyIiIiIiYhZJmCIiIiIiImaRhCkiYsJIWiZpiaSfSFokab0xftf/l/T8evkTkrZfxffuLWn3Ee7jCkmb3dvxad+zWvsvSTpK0l+vbowRERGjSsIUETF5fm97J9s7AHdSNhNerrZ8X222X2H7klV8y97AaidMERERfZaEKSJisp0JPKrO/pwp6RTgEkkLJL1X0rmSLpT0SgAVH5b0M0nfBh40+EWSTpe0a728n6TFki6Q9B1Jj6AkZq+vs1t7Stpc0on1Ps6VtEf92QdKOk3SxZI+QemiuEqSviTp/Pozh0+77QN1/DuSNq9j20j6Zv2ZMyU9Zk7+NyMiIlbTSGcpIyLivldnkvYHvlmHdgZ2sH15TTputv0ESesAP5B0GvB44NHA9sAWwCXAv037vZsDHwf2qr9rU9s3SPoo8Dvb/1K/73PAB2yfJWkr4FTgD4G3A2fZfqekZwKH3ovDOaTex7rAuZJOtH09sD5wnu3XS/r7+rtfAxwHvMr2zyU9CfhX4Kkj/DdGRESMJQlTRMTkWVfSknr5TOCTlFK5H9u+vI4/HXjsYH0SsDGwLbAX8Hnby4CrJX13ht//ZOCMwe+yfcMscfwxsL20fAJpI0kb1Pt4Xv3Zr0m68V4c02slHVAvb1ljvR64G/hCHf8McFK9j92BRUP3vc69uI+IiIg5l4QpImLy/N72TsMDNXG4dXgIOML2qdO+7xlzGMcawJNt3z5DLPeapL0pyddutm+TdDqwcJZvd73fm6b/H0RERHQha5giItp0KvCXktYCkLSdpPWBM4AX1TVODwb2meFnfwjsJWnr+rOb1vFbgA2Hvu804IjBFUk71YtnAAfWsf2BB9xDrBsDN9Zk6TGUGa6BNYDBLNmBlFK/3wKXS3pBvQ9Jetw93EdERMR9IglTRESbPkFZn7RY0k+Aj1GqBk4Gfl5v+zRwzvQftP0b4HBK+dsFrCiJ+wpwwKDpA/BaYNfaVOISVnTrewcl4bqYUpr3X/cQ6zeBNSVdCvwzJWEbuBV4Yj2GpwLvrOMHAYfW+C4GnnMv/k8iIiLmnGx3HUNERERERMREygxTRERERETELJIwRUREREREzCIJU0RERERExCySMEVERERERMwiCVNERERERMQskjBFRERERETMIglTRERERETELP4X/n+qAAoA9xMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1440x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cm = confusion_matrix(sample[\"Language_English_Name\"], sample[\"langdetect_language_English_Name\"], labels)\n",
    "plot_confusion_matrix(cm=cm, target_names=labels, normalize=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "54c98b27-b082-4c81-accd-de9d3b83a37b",
   "metadata": {},
   "source": [
    "> langdetect struggled with Chinese (1,179 errors), Danish (801), & Norwegian (434)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d5deb118-1a04-4bb3-9803-cb71799e53e1",
   "metadata": {},
   "source": [
    "# Takeaways\n",
    "On a high-level langid & langdetect performed similarly in terms of overall F1, .94 & .93 respectively for the unlimited versions. They even struggled with the same languages (Danish, Chinese, & Norwegian) albeit to different extents.\n",
    "\n",
    "We can perform these evaluations on any number of samples using get_predictions.ipynb to see how they generalize but the results are not drastically different than for 1 sample.\n",
    "\n",
    "Based on these results I would opt for using a limited version of langid because 1) it's an order of magnitude faster than langdetect, 2) it inherently provides deterministic results, 3) its F1 score is at least as good as, if not better, than langdetect's (being charitable here by generalizing beyond this sample), & 4) it appears to be less problematic with Chinese than langdetect which is much more common than Danish or Norwegian.\n",
    "\n",
    "We'll extend this evaluation by looking at aggregated results from 100 samples by languages & as a function of sentence length in plot_results.ipynb."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e72259ae-ff7c-47d2-9cf5-a18b251dba4e",
   "metadata": {},
   "source": [
    "## Capture results as a function of sentence length [optional]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "id": "e6a59939-a855-4b37-a067-e7e22e75b491",
   "metadata": {},
   "outputs": [],
   "source": [
    "results = []\n",
    "for this_bin in bin_labels:\n",
    "    bin_results = sample[sample[\"Sentence_len_bin\"]==this_bin]\n",
    "    for lang in bin_results[\"Language_English_Name\"].unique().tolist():\n",
    "        these_results = bin_results[(bin_results[\"Language_English_Name\"]==lang)]\n",
    "        for prediction in [\"langid\", \"langdetect\"]:\n",
    "            these_labels = these_results[\"Language_English_Name\"].unique().tolist()\n",
    "            this_f1 = f1_score(these_results[\"Language_English_Name\"], these_results[f\"{prediction}_language_English_Name\"], labels=these_labels, average=\"weighted\")\n",
    "            result = (this_bin, lang, prediction, round(this_f1,2))\n",
    "            results.append(result)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb58c85b-c8b5-47f4-bba6-2730cba8ab72",
   "metadata": {},
   "source": [
    "#### Export results above for plotting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "id": "4c555d60-bc5c-4377-84d0-0cfddbbc6c55",
   "metadata": {},
   "outputs": [],
   "source": [
    "from datetime import datetime"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "id": "5781bd65-b477-4ea5-978f-449b05bd64e0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Results file saved: output/results_df_2021-8-19_1928.csv\n"
     ]
    }
   ],
   "source": [
    "now = datetime.now()\n",
    "datetime_stamp = f\"{now.year}-{now.month}-{now.day}_{now.hour}{now.minute}\"\n",
    "\n",
    "results = pd.DataFrame(data=results, columns=[\"Bin\", \"Language\", \"Source\", \"F1\"])\n",
    "RESULTS_FILE = f\"output/results_df_{datetime_stamp}.csv\"\n",
    "results.to_csv(RESULTS_FILE, index=False)\n",
    "\n",
    "print(f\"Results file saved: {RESULTS_FILE}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "39a0816d-24c5-4b1c-a2a4-cf0a35520493",
   "metadata": {},
   "source": [
    "# End"
   ]
  }
 ],
 "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": "3.9.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}