movies_english_level.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "5504c7fc",
   "metadata": {},
   "source": [
    "# INTRO\n",
    "\n",
    "Для заказчика необходимо найти решение определения уровня сложности английского языка в фильмах по субтитрам.\n",
    "\n",
    "Заказчик предоставил данные:\n",
    "- английские словари Oxford, в которых слова распределены по уровню сложности\n",
    "- набор файлов-субтитров, рассортированных по каталогам в соответствии с уровнем сложности\n",
    "- excel файл со список несортированных фильмов и указанием их уровня\n",
    "\n",
    "Цель:\n",
    "- создание модели определения уровня сложности\n",
    "\n",
    "Задачи:\n",
    "- изучение и обработка предоставленного материала\n",
    "- обработка текста субтитров и подготовка для машинного обучения\n",
    "- расчет и добавление дополнительных признаков\n",
    "- тестирование модели и подбор гиперпараметров\n",
    "- выделение наиболее удачных признаков\n",
    "- обучение оптимальной модели на датасете с наиболее полезными признаками\n",
    "- создание приложения на платформе streamlit для демонстрации решения\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "a16cd5ab",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import re\n",
    "import warnings\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from joblib import dump, load\n",
    "\n",
    "from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer\n",
    "from sklearn.model_selection import train_test_split, cross_val_score\n",
    "from sklearn.preprocessing import OneHotEncoder, StandardScaler\n",
    "from sklearn.pipeline import Pipeline, FeatureUnion\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.cluster import KMeans\n",
    "\n",
    "import xgboost\n",
    "\n",
    "try:\n",
    "    import spacy\n",
    "except:\n",
    "    !pip install spacy\n",
    "    !python -m spacy download en\n",
    "    import spacy\n",
    "\n",
    "try:\n",
    "    import chardet\n",
    "except:\n",
    "    !pip install chardet\n",
    "    import chardet\n",
    "\n",
    "try:\n",
    "    from pypdf import PdfReader\n",
    "except:\n",
    "    !pip install pypdf\n",
    "    from pypdf import PdfReader\n",
    "\n",
    "try:\n",
    "    import pysrt\n",
    "except:\n",
    "    !pip install pysrt\n",
    "    import pysrt\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "8da1824c",
   "metadata": {},
   "outputs": [],
   "source": [
    "warnings.filterwarnings(\"ignore\")\n",
    "plt.style.use('ggplot')\n",
    "nlp = spacy.load(\"en_core_web_sm\")\n",
    "\n",
    "SCORES_PATH    = 'English_level/English_scores'\n",
    "SUBTITLES_PATH = 'English_level/English_scores/Subtitles_all'\n",
    "OXFORD_PATH    = 'English_level/Oxford_CEFR_level'\n",
    "\n",
    "ENGLISH_LEVELS = ['A1', 'A2', 'B1', 'B2', 'C1', 'C2']\n",
    "\n",
    "RANDOM_STATE = 7"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "26f98292",
   "metadata": {},
   "source": [
    "# LOAD DATA"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "de84bdd3",
   "metadata": {},
   "source": [
    "## The Oxford core by CEFR level\n",
    "\n",
    "В словарях Oxford 3000 и 5000 содержатся наиболее важные слова, которые должен знать каждый, кто учит английский.\n",
    "\n",
    "Словарь Oxford соответствует стандартам CEFR – The Common European Framework of Reference for Languages. Это общеевропейские компетенции владения иностранным языком: изучение, преподавание, оценка — система уровней владения иностранным языком, используемая в Европейском Союзе.\n",
    "\n",
    "У нас словари представлены в виде pdf-файлов.\n",
    "- прочитаем их,\n",
    "- разобьем на строки,\n",
    "- выделим слова,\n",
    "- определим уровень\n",
    "- сохраним, из какого словаря взяли слово\n",
    "- удалим дубликаты, оставим те случаи, где слову присвоена более высокая сложность\n",
    "- выгрузим в joblib для дальнейшего использования в приложении Streamlit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f00b3d3d",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "American_Oxford_3000_by_CEFR_level.pdf\n",
      "The_Oxford_3000_by_CEFR_level.pdf\n",
      "The_Oxford_5000_by_CEFR_level.pdf\n",
      "American_Oxford_5000_by_CEFR_level.pdf\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_46cd5\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_46cd5_level0_col0\" class=\"col_heading level0 col0\" >word</th>\n",
       "      <th id=\"T_46cd5_level0_col1\" class=\"col_heading level0 col1\" >source</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >level</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_46cd5_level0_row0\" class=\"row_heading level0 row0\" >A1</th>\n",
       "      <td id=\"T_46cd5_row0_col0\" class=\"data row0 col0\" >735 слов</td>\n",
       "      <td id=\"T_46cd5_row0_col1\" class=\"data row0 col1\" >['American_Oxford_5000_by_CEFR_level.pdf']</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_46cd5_level0_row1\" class=\"row_heading level0 row1\" >A2</th>\n",
       "      <td id=\"T_46cd5_row1_col0\" class=\"data row1 col0\" >747 слов</td>\n",
       "      <td id=\"T_46cd5_row1_col1\" class=\"data row1 col1\" >['American_Oxford_5000_by_CEFR_level.pdf']</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_46cd5_level0_row2\" class=\"row_heading level0 row2\" >B1</th>\n",
       "      <td id=\"T_46cd5_row2_col0\" class=\"data row2 col0\" >758 слов</td>\n",
       "      <td id=\"T_46cd5_row2_col1\" class=\"data row2 col1\" >['American_Oxford_5000_by_CEFR_level.pdf']</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_46cd5_level0_row3\" class=\"row_heading level0 row3\" >B2</th>\n",
       "      <td id=\"T_46cd5_row3_col0\" class=\"data row3 col0\" >1,446 слов</td>\n",
       "      <td id=\"T_46cd5_row3_col1\" class=\"data row3 col1\" >['American_Oxford_5000_by_CEFR_level.pdf']</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_46cd5_level0_row4\" class=\"row_heading level0 row4\" >C1</th>\n",
       "      <td id=\"T_46cd5_row4_col0\" class=\"data row4 col0\" >1,198 слов</td>\n",
       "      <td id=\"T_46cd5_row4_col1\" class=\"data row4 col1\" >['American_Oxford_5000_by_CEFR_level.pdf']</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x28b623940>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def process_pdf(content):\n",
    "    \n",
    "    current_level = ''\n",
    "    words_by_level = {key:[] for key in ENGLISH_LEVELS}\n",
    "    \n",
    "    for line in content:\n",
    "        line = line.strip()\n",
    "\n",
    "        if 'Oxford' in line or 'English' in line:\n",
    "            pass\n",
    "        \n",
    "        elif line in ENGLISH_LEVELS:\n",
    "            level = line\n",
    "            \n",
    "        elif ' ' in line:\n",
    "            line = re.sub(r'\\d|,', '', line.lower())   # remove digits and commas and transform to lowercase\n",
    "            words_by_level[level] += [line.split()[0]] #get the first occurency in line as word\n",
    "            \n",
    "        else:\n",
    "            pass\n",
    "    \n",
    "    return words_by_level\n",
    "\n",
    "\n",
    "content = []\n",
    "oxford_words = pd.DataFrame(columns=['word', 'level', 'source', 'type'])\n",
    "\n",
    "# load all pdfs\n",
    "for dirname, _, filenames in os.walk(OXFORD_PATH):\n",
    "    for filename in filenames:\n",
    "        print(filename)\n",
    "        reader = PdfReader(f'{OXFORD_PATH}/{filename}')\n",
    "        for page in reader.pages:\n",
    "            content += page.extract_text().splitlines()\n",
    "        processed = process_pdf(content)\n",
    "        for level in processed:\n",
    "            dict_type = 'us' if 'American' in filename else 'uk'\n",
    "            oxford_words = pd.concat([oxford_words,\n",
    "                                      pd.DataFrame({'word'   : processed[level],\n",
    "                                                    'level'  : level,\n",
    "                                                    'source' : filename,\n",
    "                                                    'type'   : dict_type\n",
    "                                                   })\n",
    "                                     ])\n",
    "\n",
    "# sort words and keep unique with higher levels\n",
    "oxford_words = oxford_words.sort_values(by=['word', 'level'], ascending=True)\n",
    "oxford_words = oxford_words.drop_duplicates(subset=['word'], keep='last')\n",
    "\n",
    "# после сортировки и удаления дубликатов остался только американский инглиш\n",
    "# похоже там все уровни слов считаются выше\n",
    "\n",
    "oxford_words.groupby('level') \\\n",
    "            .agg({'word':'count', 'source': 'unique'}) \\\n",
    "            .style.format({'word':'{:,.0f} слов'})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "22134c28",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['english_level_oxford.joblib']"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dump(oxford_words,'english_level_oxford.joblib')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6af22f02",
   "metadata": {},
   "source": [
    "## Subtitles .srt\n",
    "\n",
    "Субтитры у нас представлены в виде файлов .srt – это текстовые файлы, в которых информация представлены в таком виде:\n",
    "\n",
    "        18\n",
    "        00:04:59,796 --> 00:04:59,921\n",
    "        - ( police radio chatter )\n",
    "        - Man: What's the difference\n",
    "    \n",
    "То есть номер реплики, время и текст.\n",
    "\n",
    "Файлы рассортированы по папкам, соответствующим уровням сложности языка.\n",
    "\n",
    "Просканируем все файлы построчно:\n",
    "- удалим из текста по возможности тэги, скобки, указания говорящего лица\n",
    "- нетекстовые знаки, \n",
    "- лишние пробелы и переводы строки\n",
    "\n",
    "Сохраним временные характеристики реплик:\n",
    "- длительность реплики\n",
    "- скорость речи в знаках в секунду\n",
    "- скорость речи в словах в секунду\n",
    "\n",
    "Соберем все в виде датасета для дальнейшей обработки.\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "9555659b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "посторонний файл .DS_Store\n",
      "посторонний файл .DS_Store\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>filename</th>\n",
       "      <th>content</th>\n",
       "      <th>duration</th>\n",
       "      <th>charsrate</th>\n",
       "      <th>wordsrate</th>\n",
       "      <th>level</th>\n",
       "      <th>lemmas</th>\n",
       "      <th>A1</th>\n",
       "      <th>A2</th>\n",
       "      <th>B1</th>\n",
       "      <th>B2</th>\n",
       "      <th>C1</th>\n",
       "      <th>C2</th>\n",
       "      <th>A1ratio</th>\n",
       "      <th>A2ratio</th>\n",
       "      <th>B1ratio</th>\n",
       "      <th>B2ratio</th>\n",
       "      <th>C1ratio</th>\n",
       "      <th>C2ratio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Crown, The S01E01 - Wolferton Splash.en</td>\n",
       "      <td>in seeking his british nationalization, his ro...</td>\n",
       "      <td>[4.56, 3.4, 4.28, 4.0, 6.56, 1.0, 2.4, 1.0, 2....</td>\n",
       "      <td>[12.500000000000002, 11.176470588235295, 10.74...</td>\n",
       "      <td>[1.7543859649122808, 2.058823529411765, 1.4018...</td>\n",
       "      <td>B2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Suits.Episode 1- Denial</td>\n",
       "      <td>you're the most amazing woman i have ever met....</td>\n",
       "      <td>[2.034, 4.069, 1.901, 1.334, 2.201, 1.5, 1.667...</td>\n",
       "      <td>[22.123893805309738, 9.584664536741213, 11.572...</td>\n",
       "      <td>[4.424778761061947, 1.9660850331776847, 1.5781...</td>\n",
       "      <td>B2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Crazy4TV.com - Suits.S06E06.720p.BluRay.x265.H...</td>\n",
       "      <td>i've been after sutter for three years now. t...</td>\n",
       "      <td>[1.46, 1.21, 1.46, 1.71, 1.71, 1.752, 1.543, 2...</td>\n",
       "      <td>[10.95890410958904, 34.710743801652896, 26.027...</td>\n",
       "      <td>[0.0, 6.6115702479338845, 5.47945205479452, 2....</td>\n",
       "      <td>B2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                            filename  \\\n",
       "0            Crown, The S01E01 - Wolferton Splash.en   \n",
       "1                            Suits.Episode 1- Denial   \n",
       "2  Crazy4TV.com - Suits.S06E06.720p.BluRay.x265.H...   \n",
       "\n",
       "                                             content  \\\n",
       "0  in seeking his british nationalization, his ro...   \n",
       "1  you're the most amazing woman i have ever met....   \n",
       "2   i've been after sutter for three years now. t...   \n",
       "\n",
       "                                            duration  \\\n",
       "0  [4.56, 3.4, 4.28, 4.0, 6.56, 1.0, 2.4, 1.0, 2....   \n",
       "1  [2.034, 4.069, 1.901, 1.334, 2.201, 1.5, 1.667...   \n",
       "2  [1.46, 1.21, 1.46, 1.71, 1.71, 1.752, 1.543, 2...   \n",
       "\n",
       "                                           charsrate  \\\n",
       "0  [12.500000000000002, 11.176470588235295, 10.74...   \n",
       "1  [22.123893805309738, 9.584664536741213, 11.572...   \n",
       "2  [10.95890410958904, 34.710743801652896, 26.027...   \n",
       "\n",
       "                                           wordsrate level  lemmas  A1  A2  \\\n",
       "0  [1.7543859649122808, 2.058823529411765, 1.4018...    B2     NaN NaN NaN   \n",
       "1  [4.424778761061947, 1.9660850331776847, 1.5781...    B2     NaN NaN NaN   \n",
       "2  [0.0, 6.6115702479338845, 5.47945205479452, 2....    B2     NaN NaN NaN   \n",
       "\n",
       "   B1  B2  C1  C2  A1ratio  A2ratio  B1ratio  B2ratio  C1ratio  C2ratio  \n",
       "0 NaN NaN NaN NaN      NaN      NaN      NaN      NaN      NaN      NaN  \n",
       "1 NaN NaN NaN NaN      NaN      NaN      NaN      NaN      NaN      NaN  \n",
       "2 NaN NaN NaN NaN      NaN      NaN      NaN      NaN      NaN      NaN  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# process line\n",
    "def process_line(line):\n",
    "    if re.search(r'[A-Za-z]',line): \n",
    "        line = line.lower()\n",
    "        line = re.sub(r'\\n', ' ', line)                            # remove new lines\n",
    "        line = re.sub(r'- ', ' ', line)                            # remove dash\n",
    "        line = re.sub(r'\\<[^\\<]+?\\>', '', line)                    # remove html tags\n",
    "        line = re.sub(r'\\([^\\(]+?\\)', '', line)                    # remove () parenthesis\n",
    "        line = re.sub(r'\\[[^\\[]+?\\]', '', line)                    # remove [] parenthesis\n",
    "        line = re.sub(r'^([\\w#\\s]+\\:)', ' ', line)                 # remove speaker tag\n",
    "        line = re.sub(r'[^[:alnum:][:punct:][:blank:]]',' ', line) # remove all other non-speach shars\n",
    "        line = re.sub(r'\\s\\s+', ' ', line).strip()                 # remove extra spaces\n",
    "    return line\n",
    "\n",
    "\n",
    "# process text line by line   \n",
    "def process_text(content):\n",
    "    text = []\n",
    "    duration  = []\n",
    "    charsrate = []\n",
    "    wordsrate  = []\n",
    "    for item in content:\n",
    "        if not hasattr(item, 'duration'):\n",
    "            print('no')\n",
    "        if item.duration.ordinal>0:\n",
    "            line = process_line(item.text_without_tags)\n",
    "            text.append(line)\n",
    "            duration.append(item.duration.ordinal/1000)\n",
    "            charsrate.append(item.characters_per_second)\n",
    "            wordsrate.append(len(line.split())/item.duration.ordinal*1000.0)\n",
    "    return ' '.join(text), duration, charsrate, wordsrate\n",
    "\n",
    "\n",
    "# process file\n",
    "def process_srt(dirname, filename):\n",
    "    global count\n",
    "    if not filename.endswith('.srt'):                        # skip non srt files\n",
    "        print('посторонний файл', filename)\n",
    "        return False\n",
    "    fullpath = os.path.join(dirname,filename)\n",
    "    try:\n",
    "        enc = chardet.detect(open(fullpath, \"rb\").read())['encoding']\n",
    "        content = pysrt.open(fullpath, encoding=enc)\n",
    "    except:\n",
    "        print('не прочиталось', filename)\n",
    "        return False\n",
    "    return process_text(content)                            # clean text and return\n",
    "\n",
    "\n",
    "# movies dataset template\n",
    "movies = pd.DataFrame(columns=['filename', \n",
    "                               'content', \n",
    "                               'duration', \n",
    "                               'charsrate', \n",
    "                               'wordsrate', \n",
    "                               'level', \n",
    "                               'lemmas'] + ENGLISH_LEVELS + [l+'ratio' for l in ENGLISH_LEVELS]\n",
    "                     )\n",
    "\n",
    "# recursive walk through dirs\n",
    "for dirname, _, filenames in os.walk(SUBTITLES_PATH):\n",
    "    for filename in filenames:\n",
    "        level  = dirname.split('/')[-1]                   # get level name from dir\n",
    "        result = process_srt(dirname, filename)           # process file\n",
    "        if result:                                        # add movie to dataframe\n",
    "            subs, duration, charsrate, wordsrate = result\n",
    "            movies.loc[len(movies)] = \\\n",
    "                {'filename' : filename.replace('.srt', ''),\n",
    "                 'content'  : subs,\n",
    "                 'duration' : duration, \n",
    "                 'charsrate': charsrate, \n",
    "                 'wordsrate': wordsrate, \n",
    "                 'level'    : level\n",
    "                }\n",
    "\n",
    "movies.head(3)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cca37763",
   "metadata": {},
   "source": [
    "## Excel labels processing\n",
    "\n",
    "Часть файлов лежит в общем каталоге без указания уровня сложности. Для них заказчик предоставил excel-файл со списком фильмов и уровнями.\n",
    " \n",
    "Загрузим файл и изучим его:\n",
    "- удалим дубликаты\n",
    "- поправим несовпадения названий\n",
    "- из указанных уровней оставим более сложный вариант\n",
    "- удалим фильмы, для которых нет информации по уровню сложности"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "a5e329ef",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 241 entries, 0 to 240\n",
      "Data columns (total 2 columns):\n",
      " #   Column  Non-Null Count  Dtype \n",
      "---  ------  --------------  ----- \n",
      " 0   Movie   241 non-null    object\n",
      " 1   Level   241 non-null    object\n",
      "dtypes: object(2)\n",
      "memory usage: 5.6+ KB\n",
      "\n",
      "Дубликаты полные 2\n",
      "Дубликаты в названиях фильмов 4\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>movie</th>\n",
       "      <th>level</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>10_Cloverfield_lane(2016)</td>\n",
       "      <td>B1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10_things_I_hate_about_you(1999)</td>\n",
       "      <td>B1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A_knights_tale(2001)</td>\n",
       "      <td>B2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A_star_is_born(2018)</td>\n",
       "      <td>B2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Aladdin(1992)</td>\n",
       "      <td>A2/A2+</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                               movie   level\n",
       "id                                          \n",
       "0          10_Cloverfield_lane(2016)      B1\n",
       "1   10_things_I_hate_about_you(1999)      B1\n",
       "2               A_knights_tale(2001)      B2\n",
       "3               A_star_is_born(2018)      B2\n",
       "4                      Aladdin(1992)  A2/A2+"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "movie_labels = pd.read_excel(f'{SCORES_PATH}/movies_labels.xlsx', index_col='id')\n",
    "movie_labels.info()\n",
    "movie_labels.columns = ['movie', 'level']\n",
    "print('\\nДубликаты полные', movie_labels.duplicated().sum())\n",
    "print('Дубликаты в названиях фильмов', movie_labels.movie.duplicated().sum())\n",
    "movie_labels = movie_labels.drop_duplicates()\n",
    "movie_labels.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "7b0885dd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_47666\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"blank level0\" >&nbsp;</th>\n",
       "      <th id=\"T_47666_level0_col0\" class=\"col_heading level0 col0\" >movie</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >level</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_47666_level0_row0\" class=\"row_heading level0 row0\" >A2</th>\n",
       "      <td id=\"T_47666_row0_col0\" class=\"data row0 col0\" >32 фильмов</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_47666_level0_row1\" class=\"row_heading level0 row1\" >B1</th>\n",
       "      <td id=\"T_47666_row1_col0\" class=\"data row1 col0\" >58 фильмов</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_47666_level0_row2\" class=\"row_heading level0 row2\" >B2</th>\n",
       "      <td id=\"T_47666_row2_col0\" class=\"data row2 col0\" >109 фильмов</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_47666_level0_row3\" class=\"row_heading level0 row3\" >C1</th>\n",
       "      <td id=\"T_47666_row3_col0\" class=\"data row3 col0\" >40 фильмов</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x104776d90>"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# correct some mistakes in movies names\n",
    "movie_labels.movie = movie_labels.movie.str.replace('.srt', '', regex=False)\n",
    "movie_labels.loc[movie_labels.movie == 'Up (2009)', 'movie'] = 'Up(2009)'\n",
    "movie_labels.loc[movie_labels.movie == 'The Grinch', 'movie'] = 'The.Grinch'\n",
    "\n",
    "# удалим из level лишние символы, разделители оставим пробел\n",
    "# из мультиуровней выберем наибольший\n",
    "movie_labels.level = movie_labels.level \\\n",
    "                                 .str.replace(',', '', regex=False) \\\n",
    "                                 .str.replace('+', '', regex=False) \\\n",
    "                                 .str.replace('/', ' ', regex=False) \\\n",
    "                                 .str.split().transform(lambda x: max(x))\n",
    "\n",
    "movie_labels.groupby('level').count().sort_index().style.format({'movie':'{:.0f} фильмов'})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "919b2f0a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "не найден текст The Secret Life of Pets.en\n",
      "не найден текст Glass Onion\n",
      "не найден текст Matilda(2022)\n",
      "не найден текст Bullet train\n",
      "не найден текст Thor: love and thunder\n",
      "не найден текст Lightyear\n"
     ]
    }
   ],
   "source": [
    "# excel processing\n",
    "for row in movie_labels.itertuples():\n",
    "\n",
    "    n = movies.loc[movies.filename.str.contains(row.movie, regex=False)].shape[0]\n",
    "\n",
    "    if n == 0:\n",
    "        print('не найден текст', row.movie)\n",
    "\n",
    "    elif n == 1:\n",
    "        selected_movie_level = movies.loc[\n",
    "            movies.filename.str.contains(row.movie, regex=False), 'level'].values[0]\n",
    "\n",
    "        if selected_movie_level == 'Subtitles':         # replace Subtitles with excel level\n",
    "             movies.loc[\n",
    "                 movies.filename.str.contains(row.movie, regex=False), 'level'] = row.level\n",
    "\n",
    "        elif selected_movie_level != row.level:          # replace with max current level or excel\n",
    "             movies.loc[\n",
    "                 movies.filename.str.contains(row.movie, regex=False), 'level'\n",
    "             ] = max(selected_movie_level, row.level)\n",
    "\n",
    "    else:\n",
    "        print('не единственный текст, уже есть', row.movie)\n",
    "\n",
    "\n",
    "movies = movies[movies.level!='Subtitles']\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "75709cf2",
   "metadata": {},
   "source": [
    "## Lemmatization and feature engineering\n",
    "\n",
    "Подготовим текст для машинного обучения:\n",
    "- фильтруем имена собственные, стоп-слова, пунктуацию, цифры, урлы\n",
    "- выполним лемматизацию – приведение слов к нормальной форме\n",
    "- выделим морфологическую разметку\n",
    "\n",
    "Добавим числовые признаки.\n",
    "\n",
    "Слова (леммы):\n",
    "- общее количество\n",
    "- количество уникальных слов\n",
    "- доля уников от общего количества\n",
    "\n",
    "Предложения:\n",
    "- количество предложений\n",
    "- количество слов в предложениях\n",
    "- медианная длина предложения\n",
    "\n",
    "Сложность:\n",
    "- подсчитаем количество уникальных слов каждого уровня\n",
    "- долю слов каждой сложности относительно общего количества уников\n",
    "\n",
    "Целевой признак перекодируем в порядковые числа от 0 до 5."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "3a9f509c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 6min 56s, sys: 14.5 s, total: 7min 10s\n",
      "Wall time: 7min 16s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# если выделять и леммы, и части речи, то так быстрее – \n",
    "# в один проход их отфильтровать и потом разделить\n",
    "movies['lemmas_pos'] = movies.content.transform(lambda x: [[token.lemma_ , token.pos_]\n",
    "                                                           for token in nlp(re.sub(r'-', '', x))\n",
    "                                                           if  not token.ent_type \n",
    "                                                           and not token.is_punct\n",
    "                                                           and not token.is_currency\n",
    "                                                           and not token.is_digit\n",
    "                                                           and not token.is_space\n",
    "                                                           and not token.is_stop\n",
    "                                                           and not token.like_num\n",
    "                                                           and not token.like_url\n",
    "                                                           and not token.like_email\n",
    "                                                           and token.pos_!='INTJ'\n",
    "                                                          ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "25914282",
   "metadata": {},
   "outputs": [],
   "source": [
    "movies['lemmas'] = movies.lemmas_pos.transform(lambda x: ' '.join(item[0] for item in x))\n",
    "movies['pos']    = movies.lemmas_pos.transform(lambda x: ' '.join(item[1] for item in x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "4a01f723",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 6min 48s, sys: 8.38 s, total: 6min 57s\n",
      "Wall time: 6min 58s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "# sentences\n",
    "movies['sents'] = movies.content.transform(lambda x: list(nlp(x).sents))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "5b12f7a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# stop words count and ratio\n",
    "movies['stops'] = movies.content.transform(lambda x: sum(token.is_stop for token in nlp(x)))\n",
    "movies['total'] = movies.content.transform(lambda x: len(x.split()))\n",
    "movies['stops_ratio'] = movies.stops/movies.total"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "6f13aff0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# total lemmas count\n",
    "# and unique lemmas count\n",
    "# and unique ratio\n",
    "movies['lemmas_count']  = movies.lemmas.transform(lambda x: len(x.split()))\n",
    "movies['lemmas_unique'] = movies.lemmas.transform(lambda x: len(set(x.split())))\n",
    "movies['lemmas_unique_ratio'] = movies.lemmas_unique / movies.lemmas_count\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "959dd990",
   "metadata": {},
   "outputs": [],
   "source": [
    "# count of sentences\n",
    "# lengths of sentences\n",
    "# median length\n",
    "movies['sents_count']  = movies.sents.transform(lambda x: len(x))\n",
    "movies['sents_length'] = movies.sents.transform(lambda x: [len(sent) for sent in x])\n",
    "movies['sents_median'] = movies.sents_length.transform(lambda x: np.median(x))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "8a484bce",
   "metadata": {},
   "outputs": [],
   "source": [
    "# oxford words in movies lemmas\n",
    "def count_words_by_level(row):\n",
    "    row_lemmas = row.lemmas.split()\n",
    "    total_score = oxford_words.loc[oxford_words.word.isin(row_lemmas)] \\\n",
    "                              .groupby('level').word.count().to_dict()\n",
    "    row[total_score.keys()] = total_score\n",
    "    return row\n",
    "\n",
    "movies = movies.apply(count_words_by_level, axis=1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "b714a27d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# ratio oxford words in content\n",
    "for l in ENGLISH_LEVELS:\n",
    "    movies[l+'ratio'] = movies[l]/movies.lemmas_unique\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "58899604",
   "metadata": {},
   "outputs": [],
   "source": [
    "# count words by length\n",
    "# and ratio to total count\n",
    "for i in range(1,10):\n",
    "    equal_col = 'len_equal_'+str(i)\n",
    "    more_col  = 'more_than_'+str(i)\n",
    "    less_col  = 'less_than_'+str(i)\n",
    "    \n",
    "    movies[equal_col] = movies.lemmas.transform(lambda x: sum(len(word)==i for word in x.split()))\n",
    "    movies[more_col]  = movies.lemmas.transform(lambda x: sum(len(word)>i for word in x.split()))\n",
    "    movies[less_col]  = movies.lemmas.transform(lambda x: sum(len(word)<i for word in x.split()))\n",
    "    \n",
    "    movies[equal_col + 'ratio'] = movies[equal_col] / movies.lemmas_count\n",
    "    movies[more_col + 'ratio']  = movies[more_col] / movies.lemmas_count\n",
    "    movies[less_col + 'ratio']  = movies[less_col] / movies.lemmas_count\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "3eafc858",
   "metadata": {},
   "outputs": [],
   "source": [
    "# speed rates in seconds\n",
    "for c in ['duration', 'charsrate', 'wordsrate']:\n",
    "    movies[c+'_median'] = movies[c].transform(lambda x: np.median(x))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "5c297e46",
   "metadata": {},
   "outputs": [],
   "source": [
    "# encode level to ordinal values\n",
    "# fill missing values\n",
    "# export data movies to csv\n",
    "movies['target'] = movies.level.map({'A1':0, 'A2':1, 'B1':2, 'B2':3, 'C1':4, 'C2':5})\n",
    "movies = movies.fillna(0)\n",
    "movies.to_csv('movies_df.csv', index=False)\n",
    "\n",
    "try:\n",
    "    os.system('say \"data processing completed\"')\n",
    "except:\n",
    "    pass\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aafb55cf",
   "metadata": {},
   "source": [
    "# MODELS TEST"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "bcf1f55d",
   "metadata": {},
   "outputs": [],
   "source": [
    "movies = pd.read_csv('movies_df.csv')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "84702de7",
   "metadata": {},
   "outputs": [],
   "source": [
    "lemmas_num = ['lemmas_count', 'lemmas_unique', 'lemmas_unique_ratio']\n",
    "\n",
    "levels_ratio = [l+'ratio' for l in ENGLISH_LEVELS]\n",
    "\n",
    "length_equal = ['len_equal_'+str(i) for i in range(1,10)]\n",
    "length_more  = ['more_than_'+str(i) for i in range(1,10)]\n",
    "length_less  = ['less_than_'+str(i) for i in range(1,10)]\n",
    "\n",
    "sents_num = ['sents_count', 'sents_median']\n",
    "\n",
    "equal_ratio = [c+'ratio' for c in length_equal]\n",
    "more_ratio  = [c+'ratio' for c in length_more]\n",
    "less_ratio  = [c+'ratio' for c in length_less]\n",
    "\n",
    "time_rate   = ['duration', 'charsrate', 'wordsrate']\n",
    "time_median = [c+'_median' for c in time_rate]\n",
    "\n",
    "num = [\n",
    "#     *lemmas_num,\n",
    "#     *sents_num,\n",
    "#     *levels_ratio,\n",
    "#     *equal_ratio,\n",
    "#     *more_ratio,\n",
    "#     *less_ratio,\n",
    "#     *time_median,\n",
    "#     'stops_ratio',\n",
    "#     'more_than_4ratio',\n",
    "#     'wordsrate_median',\n",
    "#     'duration_median',\n",
    "\n",
    "    \n",
    "# при тестировании использовались разные наборы признаков,\n",
    "# в итоге были оставлены только эти наборы\n",
    "\n",
    "    'B1ratio',\n",
    "    'len_equal_6ratio',\n",
    "    'charsrate_median',\n",
    "    'sents_median',\n",
    "]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "74532b29",
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train, X_test, y_train, y_test = train_test_split(movies, \n",
    "                                                    movies.target,\n",
    "                                                    stratify=movies.target,\n",
    "                                                    random_state=RANDOM_STATE,\n",
    "                                                    test_size=0.3\n",
    "                                                   )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "0ecd2e50",
   "metadata": {},
   "outputs": [],
   "source": [
    "kmeans = KMeans(n_clusters=3, n_init=10, random_state=RANDOM_STATE)\n",
    "scaler = StandardScaler()\n",
    "bow    = CountVectorizer()\n",
    "tfidf  = TfidfVectorizer()\n",
    "regressor = xgboost.XGBRegressor(random_state=RANDOM_STATE, eval_metric='rmse')\n",
    "\n",
    "# предобработка данных\n",
    "preprocessor = ColumnTransformer([\n",
    "    ('num', scaler, num),\n",
    "    ('txt', bow, 'lemmas'),\n",
    "    ('pos', tfidf, 'pos'),\n",
    "], remainder='drop')\n",
    "\n",
    "# кластеризация мне нужна после предобработки,\n",
    "# поэтому под kmeans в пайплайне отдельная ветка\n",
    "clusters = Pipeline(steps=[\n",
    "    ('pre', preprocessor),\n",
    "    ('kmeans', kmeans),\n",
    "])\n",
    "\n",
    "# объединяем фичи – предобработка плюс кластеры\n",
    "union = FeatureUnion([\n",
    "    ('pre', preprocessor),\n",
    "    ('clu', clusters)\n",
    "])\n",
    "\n",
    "# основной пайплайн модели\n",
    "pipeline = Pipeline(steps=[\n",
    "    ('uni', union),\n",
    "    ('reg', regressor)\n",
    "])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d9d3402d",
   "metadata": {},
   "source": [
    "## Regressor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "a6838d19",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.36171413 0.30447564 0.35471672 0.2572934  0.41407655]\n",
      "mean score 0.3384552901187087\n",
      "CPU times: user 26.2 s, sys: 5.05 s, total: 31.2 s\n",
      "Wall time: 7.68 s\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-3 {color: black;background-color: white;}#sk-container-id-3 pre{padding: 0;}#sk-container-id-3 div.sk-toggleable {background-color: white;}#sk-container-id-3 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-3 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-3 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-3 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-3 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-3 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-3 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-3 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-3 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-3 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-3 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-3 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-3 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-3 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-3 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-3 div.sk-item {position: relative;z-index: 1;}#sk-container-id-3 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-3 div.sk-item::before, #sk-container-id-3 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-3 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-3 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-3 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-3 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-3 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-3 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-3 div.sk-label-container {text-align: center;}#sk-container-id-3 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-3 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-3\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;uni&#x27;,\n",
       "                 FeatureUnion(transformer_list=[(&#x27;pre&#x27;,\n",
       "                                                 ColumnTransformer(transformers=[(&#x27;num&#x27;,\n",
       "                                                                                  StandardScaler(),\n",
       "                                                                                  [&#x27;B1ratio&#x27;,\n",
       "                                                                                   &#x27;len_equal_6ratio&#x27;,\n",
       "                                                                                   &#x27;charsrate_median&#x27;,\n",
       "                                                                                   &#x27;sents_median&#x27;]),\n",
       "                                                                                 (&#x27;txt&#x27;,\n",
       "                                                                                  CountVectorizer(),\n",
       "                                                                                  &#x27;lemmas&#x27;),\n",
       "                                                                                 (&#x27;pos&#x27;,\n",
       "                                                                                  TfidfVectorizer(),\n",
       "                                                                                  &#x27;pos&#x27;)])),\n",
       "                                                (&#x27;clu&#x27;,\n",
       "                                                 Pipeline(steps=[(&#x27;pre&#x27;,\n",
       "                                                                  ColumnTransformer(transformers=[(&#x27;num&#x27;,\n",
       "                                                                                                   StandardScaler(),\n",
       "                                                                                                   [&#x27;B1ratio&#x27;,...\n",
       "                              feature_types=None, gamma=None, gpu_id=None,\n",
       "                              grow_policy=None, importance_type=None,\n",
       "                              interaction_constraints=None, learning_rate=None,\n",
       "                              max_bin=None, max_cat_threshold=None,\n",
       "                              max_cat_to_onehot=None, max_delta_step=None,\n",
       "                              max_depth=None, max_leaves=None,\n",
       "                              min_child_weight=None, missing=nan,\n",
       "                              monotone_constraints=None, n_estimators=100,\n",
       "                              n_jobs=None, num_parallel_tree=None,\n",
       "                              predictor=None, random_state=7, ...))])</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-35\" type=\"checkbox\" ><label for=\"sk-estimator-id-35\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;uni&#x27;,\n",
       "                 FeatureUnion(transformer_list=[(&#x27;pre&#x27;,\n",
       "                                                 ColumnTransformer(transformers=[(&#x27;num&#x27;,\n",
       "                                                                                  StandardScaler(),\n",
       "                                                                                  [&#x27;B1ratio&#x27;,\n",
       "                                                                                   &#x27;len_equal_6ratio&#x27;,\n",
       "                                                                                   &#x27;charsrate_median&#x27;,\n",
       "                                                                                   &#x27;sents_median&#x27;]),\n",
       "                                                                                 (&#x27;txt&#x27;,\n",
       "                                                                                  CountVectorizer(),\n",
       "                                                                                  &#x27;lemmas&#x27;),\n",
       "                                                                                 (&#x27;pos&#x27;,\n",
       "                                                                                  TfidfVectorizer(),\n",
       "                                                                                  &#x27;pos&#x27;)])),\n",
       "                                                (&#x27;clu&#x27;,\n",
       "                                                 Pipeline(steps=[(&#x27;pre&#x27;,\n",
       "                                                                  ColumnTransformer(transformers=[(&#x27;num&#x27;,\n",
       "                                                                                                   StandardScaler(),\n",
       "                                                                                                   [&#x27;B1ratio&#x27;,...\n",
       "                              feature_types=None, gamma=None, gpu_id=None,\n",
       "                              grow_policy=None, importance_type=None,\n",
       "                              interaction_constraints=None, learning_rate=None,\n",
       "                              max_bin=None, max_cat_threshold=None,\n",
       "                              max_cat_to_onehot=None, max_delta_step=None,\n",
       "                              max_depth=None, max_leaves=None,\n",
       "                              min_child_weight=None, missing=nan,\n",
       "                              monotone_constraints=None, n_estimators=100,\n",
       "                              n_jobs=None, num_parallel_tree=None,\n",
       "                              predictor=None, random_state=7, ...))])</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-36\" type=\"checkbox\" ><label for=\"sk-estimator-id-36\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">uni: FeatureUnion</label><div class=\"sk-toggleable__content\"><pre>FeatureUnion(transformer_list=[(&#x27;pre&#x27;,\n",
       "                                ColumnTransformer(transformers=[(&#x27;num&#x27;,\n",
       "                                                                 StandardScaler(),\n",
       "                                                                 [&#x27;B1ratio&#x27;,\n",
       "                                                                  &#x27;len_equal_6ratio&#x27;,\n",
       "                                                                  &#x27;charsrate_median&#x27;,\n",
       "                                                                  &#x27;sents_median&#x27;]),\n",
       "                                                                (&#x27;txt&#x27;,\n",
       "                                                                 CountVectorizer(),\n",
       "                                                                 &#x27;lemmas&#x27;),\n",
       "                                                                (&#x27;pos&#x27;,\n",
       "                                                                 TfidfVectorizer(),\n",
       "                                                                 &#x27;pos&#x27;)])),\n",
       "                               (&#x27;clu&#x27;,\n",
       "                                Pipeline(steps=[(&#x27;pre&#x27;,\n",
       "                                                 ColumnTransformer(transformers=[(&#x27;num&#x27;,\n",
       "                                                                                  StandardScaler(),\n",
       "                                                                                  [&#x27;B1ratio&#x27;,\n",
       "                                                                                   &#x27;len_equal_6ratio&#x27;,\n",
       "                                                                                   &#x27;charsrate_median&#x27;,\n",
       "                                                                                   &#x27;sents_median&#x27;]),\n",
       "                                                                                 (&#x27;txt&#x27;,\n",
       "                                                                                  CountVectorizer(),\n",
       "                                                                                  &#x27;lemmas&#x27;),\n",
       "                                                                                 (&#x27;pos&#x27;,\n",
       "                                                                                  TfidfVectorizer(),\n",
       "                                                                                  &#x27;pos&#x27;)])),\n",
       "                                                (&#x27;kmeans&#x27;,\n",
       "                                                 KMeans(n_clusters=3,\n",
       "                                                        random_state=7))]))])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>pre</label></div></div><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-37\" type=\"checkbox\" ><label for=\"sk-estimator-id-37\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">num</label><div class=\"sk-toggleable__content\"><pre>[&#x27;B1ratio&#x27;, &#x27;len_equal_6ratio&#x27;, &#x27;charsrate_median&#x27;, &#x27;sents_median&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-38\" type=\"checkbox\" ><label for=\"sk-estimator-id-38\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">StandardScaler</label><div class=\"sk-toggleable__content\"><pre>StandardScaler()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-39\" type=\"checkbox\" ><label for=\"sk-estimator-id-39\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">txt</label><div class=\"sk-toggleable__content\"><pre>lemmas</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-40\" type=\"checkbox\" ><label for=\"sk-estimator-id-40\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CountVectorizer</label><div class=\"sk-toggleable__content\"><pre>CountVectorizer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-41\" type=\"checkbox\" ><label for=\"sk-estimator-id-41\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">pos</label><div class=\"sk-toggleable__content\"><pre>pos</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-42\" type=\"checkbox\" ><label for=\"sk-estimator-id-42\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">TfidfVectorizer</label><div class=\"sk-toggleable__content\"><pre>TfidfVectorizer()</pre></div></div></div></div></div></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><label>clu</label></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-43\" type=\"checkbox\" ><label for=\"sk-estimator-id-43\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">pre: ColumnTransformer</label><div class=\"sk-toggleable__content\"><pre>ColumnTransformer(transformers=[(&#x27;num&#x27;, StandardScaler(),\n",
       "                                 [&#x27;B1ratio&#x27;, &#x27;len_equal_6ratio&#x27;,\n",
       "                                  &#x27;charsrate_median&#x27;, &#x27;sents_median&#x27;]),\n",
       "                                (&#x27;txt&#x27;, CountVectorizer(), &#x27;lemmas&#x27;),\n",
       "                                (&#x27;pos&#x27;, TfidfVectorizer(), &#x27;pos&#x27;)])</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-44\" type=\"checkbox\" ><label for=\"sk-estimator-id-44\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">num</label><div class=\"sk-toggleable__content\"><pre>[&#x27;B1ratio&#x27;, &#x27;len_equal_6ratio&#x27;, &#x27;charsrate_median&#x27;, &#x27;sents_median&#x27;]</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-45\" type=\"checkbox\" ><label for=\"sk-estimator-id-45\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">StandardScaler</label><div class=\"sk-toggleable__content\"><pre>StandardScaler()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-46\" type=\"checkbox\" ><label for=\"sk-estimator-id-46\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">txt</label><div class=\"sk-toggleable__content\"><pre>lemmas</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-47\" type=\"checkbox\" ><label for=\"sk-estimator-id-47\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">CountVectorizer</label><div class=\"sk-toggleable__content\"><pre>CountVectorizer()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-label-container\"><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-48\" type=\"checkbox\" ><label for=\"sk-estimator-id-48\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">pos</label><div class=\"sk-toggleable__content\"><pre>pos</pre></div></div></div><div class=\"sk-serial\"><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-49\" type=\"checkbox\" ><label for=\"sk-estimator-id-49\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">TfidfVectorizer</label><div class=\"sk-toggleable__content\"><pre>TfidfVectorizer()</pre></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-50\" type=\"checkbox\" ><label for=\"sk-estimator-id-50\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">KMeans</label><div class=\"sk-toggleable__content\"><pre>KMeans(n_clusters=3, random_state=7)</pre></div></div></div></div></div></div></div></div></div></div><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-51\" type=\"checkbox\" ><label for=\"sk-estimator-id-51\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">XGBRegressor</label><div class=\"sk-toggleable__content\"><pre>XGBRegressor(base_score=None, booster=None, callbacks=None,\n",
       "             colsample_bylevel=None, colsample_bynode=None,\n",
       "             colsample_bytree=None, early_stopping_rounds=None,\n",
       "             enable_categorical=False, eval_metric=&#x27;rmse&#x27;, feature_types=None,\n",
       "             gamma=None, gpu_id=None, grow_policy=None, importance_type=None,\n",
       "             interaction_constraints=None, learning_rate=None, max_bin=None,\n",
       "             max_cat_threshold=None, max_cat_to_onehot=None,\n",
       "             max_delta_step=None, max_depth=None, max_leaves=None,\n",
       "             min_child_weight=None, missing=nan, monotone_constraints=None,\n",
       "             n_estimators=100, n_jobs=None, num_parallel_tree=None,\n",
       "             predictor=None, random_state=7, ...)</pre></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('uni',\n",
       "                 FeatureUnion(transformer_list=[('pre',\n",
       "                                                 ColumnTransformer(transformers=[('num',\n",
       "                                                                                  StandardScaler(),\n",
       "                                                                                  ['B1ratio',\n",
       "                                                                                   'len_equal_6ratio',\n",
       "                                                                                   'charsrate_median',\n",
       "                                                                                   'sents_median']),\n",
       "                                                                                 ('txt',\n",
       "                                                                                  CountVectorizer(),\n",
       "                                                                                  'lemmas'),\n",
       "                                                                                 ('pos',\n",
       "                                                                                  TfidfVectorizer(),\n",
       "                                                                                  'pos')])),\n",
       "                                                ('clu',\n",
       "                                                 Pipeline(steps=[('pre',\n",
       "                                                                  ColumnTransformer(transformers=[('num',\n",
       "                                                                                                   StandardScaler(),\n",
       "                                                                                                   ['B1ratio',...\n",
       "                              feature_types=None, gamma=None, gpu_id=None,\n",
       "                              grow_policy=None, importance_type=None,\n",
       "                              interaction_constraints=None, learning_rate=None,\n",
       "                              max_bin=None, max_cat_threshold=None,\n",
       "                              max_cat_to_onehot=None, max_delta_step=None,\n",
       "                              max_depth=None, max_leaves=None,\n",
       "                              min_child_weight=None, missing=nan,\n",
       "                              monotone_constraints=None, n_estimators=100,\n",
       "                              n_jobs=None, num_parallel_tree=None,\n",
       "                              predictor=None, random_state=7, ...))])"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "cv_score = cross_val_score(pipeline, X_train, y_train)\n",
    "print(cv_score)\n",
    "print('mean score', np.mean(cv_score))\n",
    "\n",
    "try:\n",
    "    os.system('say \"Model training completed\"')\n",
    "except:\n",
    "    pass\n",
    "\n",
    "pipeline.fit(X_train, y_train)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "68dc234a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rmse на тесте 0.6583479405610527\n"
     ]
    }
   ],
   "source": [
    "print('rmse на тесте', np.sqrt(mean_squared_error(y_test, pipeline.predict(X_test))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "ecdc639d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['english_level_model.joblib']"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dump(pipeline, 'english_level_model.joblib')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "119aaf59",
   "metadata": {},
   "source": [
    "# Feature importances"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "90b3afb9",
   "metadata": {},
   "source": [
    "Посмотрим, какие фичи вносят наибольший вклад в определение уровня ангийского.\n",
    "- медианная длина предложения, количество слов\n",
    "- медианная скорость речи, букв в секунду. Есть еще похожий признак – количество слов с секунду, но он не отражает длину слов.\n",
    "- леммы, отфильтрованные от стоп-слов, именованых сущностей\n",
    "- из частей речи заметно появление местоимений в числе 20 самых важных признаков.\n",
    "- также проиллюстрировано заметно разное распределение слов уровня В1\n",
    "- выявлена положительная корреляция между уровнем сложности и словами длиной 6 знаков"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "931194de",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAs0AAAGdCAYAAADpKU7rAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAACSu0lEQVR4nOzde1hVZfr/8fc+IIc4SQIbPGJKqKlUHpLRxtFssKTJDjiToWIM5YHGwUPDNI04TjEEv2nKyFNOaGlmGmZqjuZoU1lhXyNFPKVjhm4ENEUz47R/f/h1f91tFBAQhM/rurgu13rWeta9b0Du/exnPctgs9lsiIiIiIjIZRkbOwARERERkaZORbOIiIiISDVUNIuIiIiIVENFs4iIiIhINVQ0i4iIiIhUQ0WziIiIiEg1VDSLiIiIiFRDRbOIiIiISDVUNIuIiIiIVENFs4iIiIhINcyNHYBIc/Ldd99RXl7e2GE0Cf7+/hQVFTV2GE2KcuJMOXGmnDhSPpwpJ86uNidms5nWrVvX7Nha9y4il1VeXk5ZWVljh9HoDAYDcCEfNputkaNpGpQTZ8qJM+XEkfLhTDlxdq1youkZIiIiIiLV0EizSD2qmD2FioP7GjuMJuHbxg6gCVJOnCknzpQTR8qHs5aWE9PCNY0dAqCRZhEH0dHRZGdnN3YYIiIi0sSoaBYB3bwnIiIiV6TpGXJZycnJtG/fHoCPPvoIo9HI3XffzahRozAYDJw9e5bMzEz+53/+h7KyMrp3705sbCxBQUEAFBUVsWjRIvbt20d5eTn+/v48+uij3HbbbVcd0+7du5k1axZ/+MMfePPNNzl27BgdO3bkiSeeoEOHDgCcOXOGRYsWsXfvXs6ePUtgYCAjR45k4MCBTq/NbDbzn//8h3bt2lFcXAxAeno6cOFO3IyMjKuOVURERJoPFc1yRR9++CFDhgzhueee4+DBgyxYsIA2bdpw11138corr2C1WpkxYwbu7u4sXbqUlJQU/v73v2M2m1m0aBHl5eXMmjULV1dX8vPzcXNzq5e4Xn/9dWJjY/H19WXZsmWkpqby4osvYjabKSsro3Pnztx///24u7uzY8cOXn75ZQIDA+natavDa7v77ruZPXs2NpsNLy8v4uLimDhxIuHh4RiNl/8gpqyszGGVDIPBgLu7e728NhEREfk/F1fHqK69uuPqSkWzXNGNN97I2LFjMRgMBAcHc+TIEdatW0ePHj344osvmD17NjfffDMATz75JBMmTGD79u0MGDCA4uJi+vfvbx8BDgwMrLe4Hn74YXr16gXA5MmTeeKJJ8jOziYiIgI/Pz/uu+8++7HDhw8nJyeHTz/91KFotlgsPProo059e3h44Ovre8XrZ2VlsXLlSvt2SEgIqampdXxVIiIi8lMXP8GujsViadA4VDTLFXXt2tXhnVtoaChr164lPz8fk8nkUIR6eXkRHBzM0aNHgQvF6quvvsrOnTvp2bMn/fv3p2PHjvUSV2hoqP3fnp6eDtetrKxk9erVbNu2jZMnT1JWVkZ5eTmurq4OfXTu3Pmqrz9y5EhGjBhh327od7ciIiItldVqvWK7wWDAYrFQUFBQ63WazWYz/v7+NTu2Vj2LVOPSH9ahQ4fSu3dvduzYwc6dO8nKymLMmDEMHz68Qa59sXB97733WLduHWPHjqVDhw64ubmRmZnpdLNfXaaKuLi44OLiUqd4RUREpHo1LYRtNpsebiKN58CBA07bFouFdu3aUVFR4dB+5swZrFYr7dq1s+9r06YNd999N9OmTSMqKorNmzfXS1z79++3//vs2bNYrVaCg4MB2LNnD3369OHOO++kU6dOBAQEVPsu9SKTyURlZWW9xCgiIiLNh0aa5YpOnDjB4sWLGTZsGIcOHeL9999nzJgxBAUF0adPH+bPn098fDxubm4sW7YMPz8/+vTpA0BmZibh4eEEBwdz9uxZcnNzadu2bb3EtWrVKry8vPDx8WH58uV4eXnRr18/4MKcps8//5x9+/Zxww03sHbtWk6dOlWjawcEBJCbm0tYWBhmsxlPT896iVdERESubyqa5YruvPNOSktLSUpKwmg0Mnz4cO666y4AJk6cSGZmJn/7298oLy+nW7duJCUlYTZf+LGqrKxk0aJFnDx5End3d8LDwxk7dmy9xPXII4+QmZmJ1WqlY8eOzJgxw37dhx56iMLCQp599llcXV0ZOnQoffv25dy5c9X2GxMTw5IlS9i8eTN+fn61XnLO9Mw/qLxkVY2WymAwEBQUhNVqbdCPyq4nyokz5cSZcuJI+XCmnDQeFc1yRWazmXHjxvHb3/7Wqc3T05PJkydf9tzx48c3WFxhYWH8v//3/6ps8/T0ZMaMGVc8Pzk5ucr9ffr0sY+Ui4iIiFykOc0iIiIiItXQSLNccwsWLOCjjz6qsm3QoEFXbPvZz37WkKGJiIiIVMlg04QYucZOnz7NDz/8UGWbu7v7Fdt8fHwaMrQ6KyoqcnhSYEulOXfOlBNnyokz5cSR8uFMOXFWl5y4uLhonWZpunx8fK5Y/Db1wlhERERaHs1pFhERERGphopmEREREZFqqGgWEREREamG5jSL1KOK2VOoOLivscNoEr5t7ACaIOXEmXLiTDlxpHw4a+o5MS1c09ghNAiNNEuzVFlZSWVlZWOHISIiIs2ERprFQXJyMu3btwfgo48+wmg0cvfddzNq1CgMBgNnz54lMzOT//mf/6GsrIzu3bsTGxtLUFAQcGHJtUWLFrFv3z7Ky8vx9/fn0Ucf5bbbbrvqmHbv3s2sWbP4wx/+wJtvvsmxY8fo2LEjTzzxBB06dABg69atZGZmkpCQwBtvvIHVauWll17Cw8PjivFePG/KlCksXryY4uJiwsLCmDhxIq1bt65jNkVERKS50EizOPnwww8xmUw899xzxMbGsm7dOjZv3gzAK6+8wsGDB5kxYwZ//etfsdlspKSkUF5eDsCiRYsoLy9n1qxZpKenM3r0aNzc3Oolrtdff52YmBhSUlLw9vYmNTXVfl2AH3/8kdWrV/PEE0/w97//HW9v72rjvXjee++9x+TJk5k1axbFxcW8/vrrV4ylrKyMc+fO2b8ut7a0iIhIS2MwGK7519VetzY00ixObrzxRsaOHYvBYCA4OJgjR46wbt06evTowRdffMHs2bO5+eabAXjyySeZMGEC27dvZ8CAARQXF9O/f3/7CHBgYGC9xfXwww/Tq1cvACZPnswTTzxBdnY2ERERAFRUVPDYY4/RqVMnAKxWa7XxXjzvt7/9LRaLBYDIyEhWrlx5xViysrIcjgkJCSE1NbXeXquIiMj16uKnudfaxb/jDUVFszjp2rWrw7uv0NBQ1q5dS35+PiaTia5du9rbvLy8CA4O5ujRowAMHz6cV199lZ07d9KzZ0/69+9Px44d6yWu0NBQ+789PT0drgtgNpsdrnX06NFq4wVwdXV1+EVr3bo1JSUlV4xl5MiRjBgxwr5d23erIiIizZXVar2m1zMYDFgsFgoKCmr9RECz2awnAsq1c+kP6NChQ+nduzc7duxg586dZGVlMWbMGIYPH94g1760WG3VqpXD9uV+cX6632QyVXvMT7m4uODi4lKbUEVERFqExnq8t81ma9Bra06zODlw4IDTtsVioV27dlRUVDi0nzlzBqvVSrt27ez72rRpw9133820adOIioqyz4euq/3799v/ffbsWaxWK8HBwZc9vqbxioiIiFRHI83i5MSJEyxevJhhw4Zx6NAh3n//fcaMGUNQUBB9+vRh/vz5xMfH4+bmxrJly/Dz86NPnz4AZGZmEh4eTnBwMGfPniU3N5e2bdvWS1yrVq3Cy8sLHx8fli9fjpeXF/369bvs8TWJV0RERKQmVDSLkzvvvJPS0lKSkpIwGo0MHz6cu+66C4CJEyeSmZnJ3/72N8rLy+nWrRtJSUmYzRd+lCorK1m0aBEnT57E3d2d8PBwxo4dWy9xPfLII2RmZmK1WunYsSMzZsywX/dyqou3vpme+QeVZWUN0vf1xGAwEBQUhNVqbbSP6Zoa5cSZcuJMOXGkfDhTThqPwaaMyyWSk5Pp1KkT48aNa+xQ7C6u0/zaa69xww03NHY4V1RUVESZimb9p14F5cSZcuJMOXGkfDhTTpzVJScuLi41vhFQc5pFRERERKqh6RlyTSxYsICPPvqoyrZBgwZdse1nP/tZQ4YmIiIiUi1Nz5Br4vTp05d9ap67u/sV23x8fBoytHql6RkX6ONDZ8qJM+XEmXLiSPlwppw4u1bTMzTSLNeEj4/PFYvf66kwFhERkZZHc5pFRERERKqhollEREREpBoqmkVEREREqqE5zdKkrFixgu3bt5OWltbYoVyVitlTqDi4r7HDaBK+bewAmiDlxJly4kw5caR8OGuInJgWrmmAXpsXjTSLg+TkZDIzM2t1TnR0NNnZ2Q0TkIiIiEgToKJZWoTy8vIa7RMRERGpiqZniF1GRgZ5eXnk5eWxfv164MIo8qZNm0hPT8fLywuA1NRUzp07x8yZM0lISAAgPT0dAH9/fzIyMuocy3/+8x/eeustzp49y6233srjjz+Ou7s7AJMmTeKee+7h3nvvtR8/ffp0+vbtS3R0tD3uuLg4cnJy2LVrF1FRUQBs376d4cOH884771BUVMSECRNYsmQJ8+fPx8XFxd5feno6bm5uTJ48uc6vRURERK5/KprFLjY2FqvVSvv27Rk1ahQA3t7e5OTkMG/ePKZPn87GjRvZs2cPaWlpGI1GUlJSiIuLY+LEiYSHh2M01v3Di+PHj5Odnc1TTz3F999/zwsvvMDq1av5zW9+U6t+3n77bX7zm98wduxYjEYjW7ZsoaCggE8//ZSpU6diNBoJCgoiMzOTL774ggEDBgBQUlLCjh07+OMf/3jZvsvKyhweYmIwGOxFvYiIyPXGYDA0dghX7WLsDf0aVDSLnYeHB2azGVdXV3x9fe37ExISmDFjBkuXLmXDhg3Ex8fbn57j7e1tP/fSc+rCZrMxadIkexF65513kpubW+t+fvaznzFkyBCHfeXl5SQkJNjjBhg4cCBbt261F80ff/wxN954Iz169Lhs31lZWaxcudK+HRISQmpqaq1jFBERaQqCgoIaO4Q6s1gsDdq/imapVmBgIDExMSxYsICIiAgGDRrUoNfz9/d3GLX19fXl9OnTte7npptuqrLvSwtmgKFDh5KUlMTJkyfx8/Njy5Yt/PznP7/iO9aRI0cyYsQI+/b1/A5dRETEarU2dghXzWAwYLFYKCgoqPVjtM1msx6jLfUrLy8Po9FIUVERFRUVmEymBrvWT/s2GAwOvwQ/3QaoqKhw6sfV1bVG+0JCQujYsSMffvghvXv35siRIzz11FNXjNHFxcVhDrSIiMj1rLbFZlNks9ka9HVo9QxxYDabqaysdNi3bds2srOzSU5Opri4mFWrVjm0m0wmp3Makre3N6dOnbJvnzt3jsLCwjr1OXToULZu3cqWLVvo1asXbdq0qWOUIiIi0pxopFkc+Pv7c+DAAQoLC3Fzc6O0tJSFCxcyevRowsLCmDRpEikpKYSHhxMaGgpAQEAAubm5hIWFYTab8fT0bNAYb7nlFrZu3crtt9/ODTfcwFtvvVXnGxAHDhzI66+/zubNm+u0YobpmX9QeckNgi2VwWAgKCgIq9XaLEYv6oNy4kw5caacOFI+nCknjUcjzeIgKioKo9FIYmIicXFxzJ07ly5duhAZGQlAr169iIyMZM6cOZw/fx6AmJgYdu7cyYQJE6qd1lAf7r//frp168bf/vY3UlJS6Nu3L4GBgXXq08PDg/79++Pm5kbfvn3rKVIRERFpLgw2vU0RAWD27Nm0bduW8ePHX3UfRUVFDkvRtVQaCXGmnDhTTpwpJ46UD2fKibO65MTFxaXGNwJqpFlavLNnz/LJJ5+Qm5trH1EXERERuZTmNEu9S0xMpKioqMo2b29vSkpKqmyLj49v8OXsqvLUU09x9uxZRo8eTXBw8DW/voiIiDR9Kpql3iUlJVW5BBxcWGnjcm0+Pj4NGdZl1cdjv0VERKR5U9Es9a6mc4NERERErhea0ywiIiIiUg0VzSIiIiIi1VDRLI2usLCQ6OhoDh8+DMDu3buJjo7m+++/b5R4kpOTyczMbJRri4iISNOkOc3NSHJyMp06dWLcuHE1Pic6Oppp06bRr1+/hgusGm3atGHBggV4eXk1Wgz1pWL2FCoO7mvsMJqEbxs7gCZIOXGmnDhTThy11HyYFq5p7BDkJzTSLA2mvLy8RscZjUZ8fX0xmUw1Or6yspLKysq6hCYiIiJSKxppbiYyMjLIy8sjLy+P9evXAxdGkTdt2kR6erp9FDc1NZVz584xc+ZMEhISAEhPTwcurHpRl+XXVqxYwfbt2xk+fDjvvPMORUVFLF++nK+++opVq1bx7bffYjQaCQ0NZdy4cVgsFuDC9IzJkyfz/PPP06lTJ6d+t27dSmZmJgkJCbzxxhtYrVZeeuklPDw8yMzM5H/+538oKyuje/fuxMbGEhQUBMCZM2dYtGgRe/fu5ezZswQGBjJy5EgGDhxo7/v8+fO8+uqrfP7557i7uxMVFXXVr19ERESaLxXNzURsbCxWq5X27dszatQo4MKDRHJycpg3bx7Tp09n48aN7Nmzh7S0NIxGIykpKcTFxTFx4kTCw8MxGuv+wUNBQQGffvopU6dOtfd3/vx5RowYQYcOHfjxxx956623SE9P5/nnn6/xNX/88UdWr17NE088gZeXF97e3rz00ktYrVZmzJiBu7s7S5cuJSUlhb///e+YzWbKysro3Lkz999/P+7u7uzYsYOXX36ZwMBAunbtCsAbb7zB7t27mT59Or6+vixbtoxDhw5VWbxfqqyszOFx2QaDAXd396tLmoiIyE8YDIYr7r9ce0t0rXKiormZ8PDwwGw24+rqiq+vr31/QkICM2bMYOnSpWzYsIH4+Hj7Osre3t72cy89py7Ky8tJSEiw9w1wxx13OBwzYcIE4uLiyM/Pp0OHDjXqt6Kigscee8xezFqtVr744gtmz57NzTffDMCTTz7JhAkT2L59OwMGDMDPz4/77rvP3sfw4cPJycnh008/pWvXrpw/f55///vfTJ48mV69egEwefJknnjiiWrjycrKYuXKlfbtkJAQUlNTa/RaREREqnPxU9PLufhprfyfhs6JiuZmLjAwkJiYGBYsWEBERESDP6ba39/foWCGC6PPb731FgcOHODMmTP2+cjFxcU1LprNZjMdO3a0bx89ehSTyWQfMQbw8vIiODiYo0ePAhfmPq9evZpt27Zx8uRJysrKKC8vx9XV1R5XeXk5oaGh9j48PT1r9CjtkSNHMmLECPu23vGLiEh9slqtVe43GAxYLBYKCgqw2WzXOKqmqS45MZvNNX4om4rmFiAvLw+j0UhRUREVFRU1vuHualwsSC+VmppKmzZtePzxx2ndujU2m42pU6fW+EZBgFatWjkUppf7pbh0/3vvvce6desYO3YsHTp0wM3NjczMzFpd93JcXFxwcXGpcz8iIiJVqa74s9lsKpp/oqFzotUzmhGz2ey0qsS2bdvIzs4mOTmZ4uJiVq1a5dBuMpkadCWKM2fOcPToUR544AF69uxJu3bt6mX95Xbt2lFRUcGBAwccrmW1WmnXrh0Ae/bsoU+fPtx555106tSJgIAAh3fuFosFk8nE/v377fvOnj172Xf3IiIi0nKpaG5G/P39OXDgAIWFhZSUlFBcXMzChQsZPXo0YWFhTJo0iaysLIciMSAggNzcXE6dOsXZs2frPaYbbrgBLy8vPvjgAwoKCsjNzWXx4sV17jcoKIg+ffowf/589u7dy+HDh5kzZw5+fn706dMHuFAU79y5k3379pGfn8+CBQs4deqUvQ83NzeGDBnCG2+8wa5duzhy5AivvPKKplqIiIiIE03PaEaioqLIyMggMTGR0tJSevbsSZcuXYiMjASgV69eREZGMmfOHNLS0nBzcyMmJoYlS5awefNm/Pz86rTkXFWMRiO/+93veO2115g6dSrBwcHExsaSnJxc574nTpxIZmYmf/vb3ygvL6dbt24kJSVhNl/4sX7ooYcoLCzk2WefxdXVlaFDh9K3b1/OnTtn7yMmJobz58/z/PPP4+bmRlRUlEN7bZme+QeVl6yq0VIZDAaCgoKwWq36+PB/KSfOlBNnyokj5UOaEoNNP4Ui9aaoqMhhKbqWSn/onCknzpQTZ8qJI+XDmXLirC45cXFxqfGNgJqeISIiIiJSDU3PEAeJiYkUFRVV2ebt7U1JSUmVbfHx8Q2+nJ2IiIhIY1HRLA6SkpKoqKioss1kMl22zcfHpyHDEhEREWlUKprFQU3n9YiIiIi0JJrTLCIiIiJSDRXNIiIiIiLVUNEsIiIiIlINzWmW696KFSvYvn07aWlpAGRkZPD9998zY8YMAJKTk+nUqRPjxo1r8FgqZk+h4uC+Br/O9eDbxg6gCVJOnCknzpQTRw2RD9PCNQ3QqzR3GmmWOktOTiYzM7NW50RHR5Odnd0g8cTGxjJp0qQG6VtERERaJo00S7Pj4eHR2CGIiIhIM6OiWeokIyODvLw88vLyWL9+PXBhFHnTpk2kp6fj5eUFQGpqKufOnWPmzJkkJCQAkJ6eDlxY5i4jI6NeY7p0esZP5eTk8MILLzB+/Hh+/vOfc/LkSRYvXszOnTsxGAyEhYUxbtw4AgIC6i0mERERub6paJY6iY2NxWq10r59e0aNGgVceHJgTk4O8+bNY/r06WzcuJE9e/aQlpaG0WgkJSWFuLg4Jk6cSHh4OEbjtZsl9Mknn7BgwQImT55M3759+fHHH5k1axZhYWHMmjULo9HIO++8w3PPPUd6ejpmc9W/ImVlZZSVldm3DQYD7u7u1+pliIhIHRgMhsYO4apdjP16fg317VrlREWz1ImHhwdmsxlXV1d8fX3t+xMSEpgxYwZLly5lw4YNxMfH2x+c4u3tbT/30nMa2r/+9S/efPNNpk+fzi233AJcKKINBgNPPPGE/Zdt4sSJjBs3jt27d9O7d+8q+8rKymLlypX27ZCQEFJTUxv+RYiISJ0FBQU1dgh1ZrFYGjuEJqehc6KiWRpEYGAgMTExLFiwgIiICAYNGtSo8Xz++eecOnWK2bNn06VLF/v+Q4cOUVBQwJgxYxyOLysr4/jx45ftb+TIkYwYMcK+rXf8IiLXD6vV2tghXDWDwYDFYqGgoACbzdbY4TQJdcmJ2Wyu8dOQVTRLg8nLy8NoNFJUVERFRQUmk6nRYunYsSOVlZVs2bKFm266yV7k2mw2OnfuzJNPPul0zsUR8aq4uLjg4uLSYPGKiEjDaQ7Fps1maxavoz41dE605JzUmdlsprKy0mHftm3byM7OJjk5meLiYlatWuXQbjKZnM5pSBaLhZkzZ/LFF1/wz3/+074/JCQEq9WKt7c3FovF4UurcIiIiMhFGmmWOvP39+fAgQMUFhbi5uZGaWkpCxcuZPTo0YSFhTFp0iRSUlIIDw8nNDQUgICAAHJzcwkLC8NsNuPp6dngcQYHBzNz5kySk5MxmUyMGzeOQYMG8d5775GWlkZ0dDQ33ngjxcXFfP7559x3333ceOONtbqG6Zl/UHnJDYItlcFgICgoCKvVqpGQ/6WcOFNOnCknjpQPaUo00ix1FhUVhdFoJDExkbi4OObOnUuXLl2IjIwEoFevXkRGRjJnzhzOnz8PQExMDDt37mTChAk89dRT1yzW4OBg/vznP/PJJ5+wZMkSXF1dmTVrFjfeeCPp6en8/ve/Z+7cuZSWlmo1DBEREbEz2PTWTaTeFBUVOSxF11JpdMiZcuJMOXGmnDhSPpwpJ87qkhMXF5ca3wiokWYRERERkWpoTrM0CYmJiRQVFVXZ5u3tTUlJSZVt8fHxjb6cnYiIiDR/KpqlSUhKSqKioqLKNpPJdNk2Hx+fhgxLREREBFDRLE1ETecTiYiIiDQGzWkWEREREamGimYRERERkWqoaBYRERERqYbmNDczycnJdOrUiXHjxtX4nOjoaKZNm0a/fv0aLrDrxIoVK9i+fTtpaWlXdX7F7ClUHNxXz1Fdn75t7ACaIOXEmXLiTDlxdGk+TAvXNFocIhppFhERERGphkaam5GMjAzy8vLIy8tj/fr1wIVR5E2bNpGeno6XlxcAqampnDt3jpkzZ5KQkABAeno6cGEVi4yMjKuO4fDhwyxevJiDBw9iMBiwWCzEx8dz0003VTmKu27dOtavX2+/5u7du3njjTfIz8/HZDLRvn17nnzySfz9/e3n33333bzzzjucOXOG2267jccff5wbbrjB3ueWLVtYs2YNhYWF+Pv7M3z4cH75y1/a20+cOMHrr7/OV199RXl5OW3btuWxxx7j6NGjrFy50p43gIkTJzJ48OCrzoeIiIg0Dyqam5HY2FisVivt27dn1KhRwIUHg+Tk5DBv3jymT5/Oxo0b2bNnD2lpaRiNRlJSUoiLi2PixImEh4djNNbtw4c5c+bQqVMn4uLiMBqNHD58GJPJVKNzKyoqSEtLY+jQofzud7+jvLycr7/+GoPBYD+moKCATz/9lKeeeopz584xb948Fi1axJNPPgnABx98wNtvv8348eMJCQnhv//9L/Pnz8fV1ZXBgwdz/vx5kpOT8fPz46mnnsLX15dDhw5hs9mIiIjgyJEjfPXVVzzzzDMAeHh4VBlrWVmZw+OyDQYD7u7uV5s2ERGpgUv/HrRUF3OgXPyfa5UTFc3NiIeHB2azGVdXV3x9fe37ExISmDFjBkuXLmXDhg3Ex8fb10X29va2n3vpOVeruLiYqKgo2rZtC0BQUFCNz/3hhx84d+4ct99+OxaLBYB27do5HFNWVsakSZO48cYbARg/fjwpKSmMGTMGX19fVq1aRUxMDP379wcgICCA/Px8PvjgAwYPHszHH39MSUkJKSkpeHp6AtivBeDm5obRaKw2F1lZWfZRaYCQkBBSU1Nr/FpFRKT2avM3pbm79G+XXNDQOVHR3AIEBgYSExPDggULiIiIaNDHTt97773Mnz+fjz76iJ49e3LHHXfU+IfY09OTwYMH8+yzz9KzZ0969erFgAEDaN26tf2YNm3a2AtmgNDQUGw2G8eOHcNoNHLixAnmzZvH/Pnz7cdUVlbaR4wPHz5Mp06d7AXz1Ro5ciQjRoywb+sdv4hIw7NarY0dQqO7OPWxoKAAm83W2OE0CXXJidlsrvED1lQ0txB5eXkYjUaKioqoqKio8ZSJ2oqOjmbgwIHs2LGDnJwcVqxYwZQpU+jXr1+VUz9++njsiRMnMnz4cHJycti2bRvLly/nT3/6E6GhodVeu7KyEoDHH3+crl27OrRdvHarVq2u9qU5cHFxwcXFpV76EhGRmlGR+H9sNpvy8RMNnROtntHMmM1me/F40bZt28jOziY5OZni4mJWrVrl0G4ymZzOqYvg4GBGjBjBn/70J/r168eWLVuAC1NBTp065fADffjwYafzQ0JCGDlyJH/9619p3749H3/8sb2tuLiYkydP2rf379+PwWAgODgYX19f/Pz8OH78OBaLxeErICAAgA4dOnD48GHOnj1bZexV5U9ERERERXMz4+/vz4EDBygsLKSkpITi4mIWLlzI6NGjCQsLY9KkSWRlZbF//377OQEBAeTm5nLq1KnLFpM1UVpayqJFi9i9ezdFRUXs3buXgwcP2uc3d+/enZKSEt59910KCgrYsGEDX375pf38wsJCli1bxv79+ykqKuKrr77CarU6zGt2cXEhIyODw4cPs2fPHl577TUGDBhgn4P88MMPs3r1atavX8+xY8c4cuQIW7ZsYe3atQAMHDgQX19f0tLS2Lt3L8ePH+ezzz6z5yMgIIDCwkIOHz5MSUmJw81+IiIi0nIZbBrbb1aOHTtGRkYG33zzDaWlpfTs2ROj0cgf//hH+7zbxYsX88UXX5CWloabmxtffPEFS5YsoaioCD8/v6tecq68vJyXX36Zffv2cfr0aby8vOjfvz+PPvqofVrExo0bycrK4uzZs/Tv35/g4GA2b95MRkYGp06dYuHChXz99decOXOG1q1b8/Of/5yHHnoIo9FoX3Ju2LBhrFq1irNnz3LrrbfyxBNPOMxR/vjjj1mzZg35+fm4urrSoUMH7r33XvvDW4qKiliyZAk7d+6ksrKSdu3a8dhjj9GlSxfKysp46aWXyM3N5fvvv6/1knNFRUUqtLkwvywoKAir1aqPD/+XcuJMOXGmnDhSPpwpJ87qkhMXF5caz2lW0SzXjbo+re9aUNF8gf5Td6acOFNOnCknjpQPZ8qJs2tVNGt6hoiIiIhINbR6hjhJTEykqKioyjZvb29KSkqqbIuPj2/Q5exEREREGouKZnGSlJTktBTcRSaT6bJtPj4+DRkW0dHR9sdbi4iIiFxLKprFSU3n9oiIiIi0FJrTLCIiIiJSDRXNIiIiIiLVUNEsIiIiIlINzWmWJu2nazNnZGTw/fffM2PGjEaOrGoVs6dQcXBfY4fRJHzb2AE0QcqJM+XE2fWcE9PCNY0dgkiD0UhzC5GcnExmZmatzomOjiY7O7thArpKsbGxTJo0qbHDEBERkRZGI83SKMrLyzGba//j5+Hh0QDRiIiIiFyZiuYWICMjg7y8PPLy8li/fj1wYRR506ZNpKen4+XlBUBqairnzp1j5syZJCQkAJCeng5cWIYuIyPjqmO4OM1i+PDhvPPOOxQVFbF8+XJ++OEHXn/9dbZv305ZWRmdO3dm7NixdOrU6bKv5dLpGTabjTVr1rBp0ya+++47goODefDBB7njjjuorKxk4sSJPPDAA9x99932Pg4dOsQf/vAH5syZQ2BgIOfOnatVDCIiItLyqGhuAWJjY7FarbRv355Ro0YBF57sl5OTw7x585g+fTobN25kz549pKWlYTQaSUlJIS4ujokTJxIeHo7RWPeZPAUFBXz66adMnTrV3l9KSgqenp4kJSXh4eHBpk2bmD17Ni+++CKenp7V9rl8+XKys7OJi4sjKCiIPXv2MGfOHLy9venevTsRERF8/PHHDkXzxx9/TGhoKIGBgdhstquKoaysjLKyMvu2wWDA3d29jhkSEbm+GQyGBumvvvu9niknzq5VTlQ0twAeHh6YzWZcXV3x9fW1709ISGDGjBksXbqUDRs2EB8fb3+wibe3t/3cS8+pi/LychISEux95+bmcuTIEV599VVcXFwAGDNmDNu3b+ezzz7jrrvuumJ/58+fZ+3atcycOZPQ0FAAAgMD2bt3L5s2baJ79+4MGjSIdevWUVRUhL+/P5WVlWzbto2RI0cCsHv37quKISsri5UrV9q3Q0JCSE1NrVuCRESuc0FBQQ3Sr8ViaZB+r2fKibOGzomK5hYsMDCQmJgYFixYQEREBIMGDWrQ6/n7+9sLZrgwTeL8+fOMHz/e4bjS0lIKCgqq7S8/P5+ysjJmz57tsL+8vJyQkBDgQjEbHBzMJ598wv33309eXh6nT59mwIABdYph5MiRjBgxwr6td/wiImC1Wuu1P4PBgMVioaCgAJvNVq99X6+UE2d1yYnZbK7xk5BVNLdweXl5GI1GioqKqKiowGQyNdi1XF1dHbYrKytp3bo1ycnJTsfW5Ia/i78YSUlJ+Pn5ObRdepPhoEGD+Pjjj7n//vv5+OOP6d27t714v9oYXFxc7CPTIiJyQUMVcTabTQXiTygnzho6J1pyroUwm81UVlY67Nu2bRvZ2dkkJydTXFzMqlWrHNpNJpPTOfWpc+fOnDp1CqPRiMVicfi6dET6ctq1a4eLiwvFxcVO57dp08Z+3MCBAzly5AiHDh3i888/dxhRr2sMIiIi0jJopLmF8Pf358CBAxQWFuLm5kZpaSkLFy5k9OjRhIWFMWnSJFJSUggPD7fPDw4ICCA3N5ewsDDMZnONbsyrjZ49exIaGkpaWhqjR48mODiY7777ji+//JK+ffty0003XfF8d3d3oqKiWLx4MZWVlYSFhfHDDz+wb98+3NzcGDx4sP113HzzzcydO5eKigr69u1bbzH8lOmZf1B5yQ2CLZXBYCAoKAir1aqRkP+lnDhTTpwpJyJNl4rmFiIqKoqMjAwSExMpLS2lZ8+edOnShcjISAB69epFZGQkc+bMIS0tDTc3N2JiYliyZAmbN2/Gz8+vTkvOVcVgMJCUlMSbb77J3LlzKSkpwdfXl27duuHj41OjPkaNGoW3tzerV6/m+PHj3HDDDYSEhNhv9Lto4MCBLFq0iDvvvJNWrVrVawwiIiLS/BlseisrUm+KiooclqJrqTRa5kw5caacOFNOHCkfzpQTZ3XJiYuLS41vBNScZhERERGRamh6htRYYmIiRUVFVbZ5e3tTUlJSZVt8fHyDL2cnIiIi0pBUNEuNJSUlUVFRUWWbyWS6bJvmBouIiMj1TkWz1FhN5/yIiIiINDea0ywiIiIiUg0VzSIiIiIi1VDRLNdcYWEh0dHRHD58+LLHbN26lXHjxtWov9ocKyIiInI1NKe5mUhOTqZTp061Kh6jo6OZNm0a/fr1a7jArlJERAS33nprY4dRaxWzp1BxcF9jh9EkfNvYATRBLS0npoVrGjsEEZF6o6JZmqRWrVo5PLlPREREpDFpekYzkJGRQV5eHuvXryc6Opro6GhWrlzJ448/zpkzZ+zHpaamMnPmTCorK5k0aRIA6enpREdH27evxrlz5xg1ahSHDh0CwGazERsbS1JSkv2Yjz/+mPj4eIfzjh8/zqxZs3j00UeZPn06+/fvt7f9dMrF4cOHmTVrFmPGjGHs2LE89dRTHDx4sMp4zpw5Q1JSEqmpqZSWlmKz2Xj33XeZPHkyo0ePZvr06Xz22Wf2WBMSElizxnFE7MiRI4waNYqCgoKrzouIiIg0HxppbgZiY2OxWq20b9+eUaNGARceNpKTk8O8efOYPn06GzduZM+ePaSlpWE0GklJSSEuLo6JEycSHh6O0Xj17588PDzo1KkTu3fvpnPnznzzzTcAfPPNN5w7dw4PDw/y8vLo3r27w3nLly8nJiYGi8XC8uXLefHFF3nppZcwmUxO15gzZw6dOnUiLi4Oo9HI4cOHqzzuxIkT/PWvf+Wmm25iwoQJmEwm3nzzTbKzs4mLiyMoKIg9e/YwZ84cvL296d69O7/4xS/YunUr9913n72fLVu2EBYWhsViqfI1l5WVOTwu22Aw4O7uflX5E2muDAZDjdqrO64lUU4cKR/OlBNn1yonKpqbAQ8PD8xmM66urvj6+tr3JyQkMGPGDJYuXcqGDRuIj4+3r7Xs7e1tP/fSc65Wjx492L17N1FRUeTl5dGzZ0+OHz/O3r17ue2229i9ezf33nuvwzlRUVHcdtttwIX51YmJiRQUFNC2bVun/ouLi4mKirK3BQUFOR1z7Ngx/vrXv9KnTx9iY2MxGAycP3+etWvXMnPmTEJDQwEIDAxk7969bNq0yV40r1ixgq+//pouXbpQXl7ORx99xKOPPnrZ15uVlcXKlSvt2yEhIaSmptY+cSLNWFW/p1W53JvTlkw5caR8OFNOnDV0TlQ0N2OBgYHExMSwYMECIiIiGvRR1t27d+ff//43lZWV5OXlccstt9CmTRvy8vLo3LkzVqvVaaS5Q4cO9n9fLNxPnz5dZdF87733Mn/+fD766CN69uzJHXfc4fDLUVpayp///Gd+9rOfERsba9+fn59PWVkZs2fPduivvLyckJAQAFq3bs1tt93Gv//9b7p06cKOHTsoKytjwIABl329I0eOZMSIEfZtveMXcWa1Wq/YbjAYsFgsFBQUYLPZrlFUTZty4kj5cKacOKtLTsxmc40f3qaiuZnLy8vDaDRSVFRERUVFlVMa6kP37t354Ycf+O9//8uePXsYNWoUbdq0ISsri06dOuHj40O7du0czjGb/+/H72LRebkf9ujoaAYOHMiOHTvIyclhxYoVTJkyxb7yh4uLCz179mTHjh3cd9993HjjjQ79JSUl4efnd9nrDxkyhJdffplx48axZcsWBgwYgKur62Vfr4uLCy4uLjVNj0iLVNM/XjabTX/8f0I5caR8OFNOnDV0TnQjYDNhNpuprKx02Ldt2zays7NJTk6muLiYVatWObSbTCanc67WxXnNGzZswGAw0K5dO8LCwvjvf//Ljh076NatW52vERwczIgRI/jTn/5Ev3792LJli73NYDCQkJBA586d+ctf/sLJkycBaNeuHS4uLhQXF2OxWBy+2rRpYz//tttuw9XVlY0bN5KTk8OQIUPqHK+IiIg0Hyqamwl/f38OHDhAYWEhJSUlFBcXs3DhQkaPHk1YWBiTJk0iKyvLYYWKgIAAcnNzOXXqFGfPnq1zDD169OCjjz6ie/fuGAwGPD09adeuHdu2baNHjx5X3W9paSmLFi1i9+7dFBUVsXfvXg4ePOg0jcNoNPLkk0/SsWNH/vKXv3Dq1Cnc3d2Jiopi8eLFbN26lYKCAv773/+yYcMGtm7d6nDu4MGDWbZsGRaLxT7/WURERAQ0PaPZiIqKIiMjg8TEREpLS+nZsyddunQhMjISgF69ehEZGcmcOXNIS0vDzc2NmJgYlixZwubNm/Hz8yMjI6NOMfTo0YO1a9c6zF3u3r07hw8fdprPXBtGo5EzZ87w8ssvc/r0aby8vOjfvz/R0dFOx5pMJn73u9/xwgsvMGvWLJKTkxk1ahTe3t6sXr2a48ePc8MNNxASEsLIkSMdzh0yZAhZWVn84he/uOpYTc/8g8pLVtVoqQwGA0FBQVitVn18+L+UExGR65vBpv+9RQDYu3cvs2bNYu7cuVe9okhRUZHDUnQtlQpEZ8qJM+XEmXLiSPlwppw4q0tOXFxcdCOgSE2VlZVx4sQJ3nrrLQYMGFAvS/CJiIhI86KiWewSExMpKiqqss3b25uSkpIq2+Lj4xt0ObuG9sknnzB37lw6depEQkJCY4cjIiIiTZCKZrFLSkqioqKiyjaTyXTZNh8fn4YMq8ENHjyYwYMHN3YYIiIi0oSpaBa7ms7pEREREWlptOSciIiIiEg1VDSLiIiIiFRDRbOIiIiISDU0p1majOjoaKZNm0a/fv0a7BqFhYVMnjyZ559/nk6dOtV7/xWzp1BxcF+993s9+raxA2iCWlJOTAvXNHYIIiL1SiPNzVBycjKZmZm1Oic6Oprs7OyGCaiGFixYwK233tqoMYiIiIhURSPN0mRU91CR8vJyzGb9yIqIiMi1pwqkmcnIyCAvL4+8vDzWr18PXBhF3rRpE+np6Xh5eQGQmprKuXPnmDlzpv2BHunp6cCFpecyMjKuOoYVK1awfft2hg8fzttvv83Zs2e58847eeyxx3jvvfdYu3YtNpuNe+65hwceeMB+3qXTMy5Oo5gyZQobN27kwIEDxMXF0adPHxYtWsTevXs5e/YsgYGBjBw5koEDB9r7qaysZM2aNWzevJkTJ07g4+PDsGHDHK51qfz8fF5//XXy8vJwc3OjV69ejB07Fm9v76vOgYiIiDQvKpqbmdjYWKxWK+3bt2fUqFHAhaf55eTkMG/ePKZPn87GjRvZs2cPaWlpGI1GUlJSiIuLY+LEiYSHh2M01n3WzvHjx8nJyeHpp5+moKCAv//97xQWFhIUFMSsWbPYt28fc+fO5ZZbbiE0NPSy/SxdupQxY8YwceJEzGYzZWVldO7cmfvvvx93d3d27NjByy+/TGBgIF27dgVg2bJlbN68mbFjxxIWFsapU6c4evRolf1/9913zJw5k6FDhzJmzBhKS0tZunQpL7zwAjNnzrxsXGVlZZSVldm3DQYD7u7uV5ktkebHYDDU+JiaHNtSKCeOlA9nyomza5WTqy6az507x/79+zlz5gy33nornp6e9RmXXCUPDw/MZjOurq4O0x0SEhKYMWMGS5cuZcOGDcTHx9sfZnJxRNXDw6PaKRI1ZbPZmDBhAu7u7rRr144ePXpw7NgxkpKSMBqNBAcH8+6775KXl3fFovnee++lf//+Dvvuu+8++7+HDx9OTk4On376KV27duWHH37g/fffZ/z48fan/FksFsLCwqrsf+PGjXTu3JlHHnnEvm/ChAlMmDCBY8eOERwcXOV5WVlZrFy50r4dEhJCampqtXkRaSmCgoJqfKzFYmnASK5Pyokj5cOZcuKsoXNyVUXzypUreffddyktLQUgJSUFT09P/vKXv9CrVy/uv//++oxR6kFgYCAxMTEsWLCAiIgIBg0a1KDX8/f3dxh59fHxwWg0Ooxi+/j4cPr06Sv207lzZ4ftyspKVq9ezbZt2zh58iRlZWWUl5fj6uoKwNGjRykrK6Nnz541ivPQoUPk5uYSExPj1Hb8+PHLFs0jR45kxIgR9m294xdxZLVaqz3GYDBgsVgoKCjAZrNdg6iaPuXEkfLhTDlxVpecmM3mGj8RudZF87/+9S9WrlzJ3Xffza233srf/vY3e9ttt91Gdna2iuYmKi8vD6PRSFFRERUVFZhMpga71k/7NhgMVe6r7ofbzc3NYfu9995j3bp1jB07lg4dOuDm5kZmZibl5eUAtGrVqlZx2mw2br/9dh599FGntiuNuru4uODi4lKra4m0JLX5w2Wz2fTH/yeUE0fKhzPlxFlD56TWk1c3bNjAiBEjGD9+PL1793ZoCwoKqtHogjQss9lMZWWlw75t27aRnZ1NcnIyxcXFrFq1yqHdZDI5ndMU7dmzhz59+nDnnXfSqVMnAgICHH7mLBYLrVq1YteuXTXqLyQkhPz8fPz9/bFYLA5fPy3YRUREpOWq9UhzYWGhU7F8kbu7O+fOnatzUFI3/v7+HDhwgMLCQtzc3CgtLWXhwoWMHj2asLAwJk2aREpKCuHh4fb5xAEBAeTm5hIWFobZbG6yc9QtFguff/45+/bt44YbbmDt2rWcOnWKtm3bAhdGmn/1q1/xxhtvYDabufnmmykpKSE/P58hQ4Y49ffLX/6SzZs38+KLL3Lffffh5eVFQUEBn3zyCU888UStb4o0PfMPKi+5QbClMhgM9jfRGgm5QDkREbm+1bpo9vDwuOw81MLCQi3T1QRERUWRkZFBYmIipaWl9OzZky5duhAZGQlAr169iIyMZM6cOaSlpeHm5kZMTAxLlixh8+bN+Pn51WnJuYb00EMPUVhYyLPPPourqytDhw6lb9++Dm/WHnzwQUwmEytWrODkyZO0bt2aYcOGVdmfn58fs2fPZunSpTz77LOUlZXh7+9P7969NU9ZRERE7Ay2Wg55vPjii+Tn5zN79mxatWrFb37zG/72t7/RoUMH/vznP9O+fXueeOKJhopXpEkrKipyWIqupdKoqjPlxJly4kw5caR8OFNOnNUlJy4uLg13I2B0dDR//OMfSUxMpF+/fsCFec6HDx+muLiY3//+97XtUkRERESkSat10RwUFMTs2bNZvHgx//rXvwD4z3/+Q48ePUhISKBNmzb1HqRce4mJiRQVFVXZ5u3tTUlJSZVt8fHxDb6cnYiIiMi1VquiubS0lP/85z+EhYXx9NNPU1ZWxpkzZ/D09Kz1Ul/StCUlJVFRUVFlm8lkumybj49PQ4YlIiIi0ihqVTS3atWK1157jaeffhq4MA/Ez8+vQQKTxlXT+T0iIiIiLUGt12kOCAjg1KlTDRCKiIiIiEjTVOui+Z577mH16tVaj1lEREREWoxa3wj47bffcubMGSZNmsQtt9xC69atHdoNBgOxsbH1FqDIlSQnJ9OpUyfGjRsHwKRJk7jnnnu49957GzcwERERaVZqXTRfXDEDIDs7u8pjVDRfX7Zu3UpmZiaZmZmNHUqdpaSk4Orq2mjXr5g9hYqD+xrt+k3Jt40dQBPU1HNiWrimsUMQEWmyal00v/XWWw0Rh0i90BMpRUREpCHUumhuDpKTk+nQoQOtWrVi8+bNmM1mhg0bRnR0NIWFhUyePJnnn3+eTp06AfD9998TGxvLzJkz6dGjB7t372bWrFn88Y9/ZNmyZRw9epTQ0FCmTJnCoUOHWLJkCSdPnuTWW29lwoQJdR75/Oyzz3j77bcpKCjA1dWVkJAQpk+fjpubGwBbtmxhzZo1FBYW4u/vz/Dhw/nlL38JYH89U6dOZcOGDRw4cICgoCB++9vfEhoayu7du3nllVeACw+ugQuPqo6OjuZf//oX69at48SJE3h4eBAWFsbUqVNrnF+j0ciHH36I2Wxm1KhRDBw4kH/+85989tln+Pj4MH78eG699Vb7efn5+bz++uvk5eXh5uZGr169GDt2rL0QPn/+PK+++iqff/457u7uREVFOV37p9Mz1q5dy5YtWygsLMTT05Pbb7+dRx991J67i6PsU6ZMYfHixRQXFxMWFsbEiROdph6JiIhIy9Uii2aADz/8kBEjRvDcc8+xf/9+XnnlFcLCwrBYLDXu4+2332b8+PG4urrywgsv8MILL+Di4sKTTz7J+fPnSU9P5/333+f++++/6ji/++47XnzxRUaPHk2/fv04f/48e/bssbd/8MEH9jhCQkL473//y/z583F1dWXw4MH245YvX05MTAwWi4Xly5fz4osv8tJLL3HzzTczbtw43nrrLV588UUA3NzcOHjwIK+99hqTJ0/m5ptv5uzZsw7Xrc6HH37Ifffdx3PPPce2bdtYuHAh27dvp2/fvowcOZJ169bx8ssv88orr+Dq6sp3333HzJkzGTp0KGPGjKG0tJSlS5fywgsvMHPmTADeeOMNdu/ezfTp0/H19WXZsmUcOnTI/uamKhfn2AcEBFBYWMirr77KG2+8QVxcnP2YH3/8kffee4/JkydjMBiYM2cOr7/+Ok8++eRl+y0rK3N4XLbBYMDd3b3G+RFpigwGQ6Nc71pftylTThwpH86UE2fXKie1LppHjRpV7THXwxSOjh078vDDDwMXnnK4YcMGdu3aVaui+de//jVhYWEADBkyhGXLljFnzhwCAwMB6N+/P7t3765z0VxRUUH//v3tayd36NDB3r5q1SpiYmLo378/cGFJwPz8fD744AOHojkqKorbbrsNuDCinJiYSEFBAW3btsXDwwODwYCvr6/9+OLiYlxdXbn99ttxd3fH39+fkJCQGsfdsWNHHnzwQQBGjhzJ6tWr8fLy4q677gIujGZv3LiRb775htDQUDZu3Ejnzp155JFH7H1MmDCBCRMmcOzYMfz8/Pj3v//N5MmT6dWrFwCTJ0/miSeeuGIcl94QGBAQwKhRo3j11VcdiuaKigp++9vf2r/3kZGRrFy58or9ZmVlORwTEhJCampqTVIj0mQFBQU1ynVr8/9uS6GcOFI+nCknzho6J7Uumh988EGnSr6kpISdO3dSWVnJnXfeWW/BNaRLC0+A1q1bc/r06Vr10bFjR/u/fXx8cHV1tRfMAL6+vhw8eLBOcXbq1ImePXsybdo0evfuTa9evbjjjjvw9PSkpKSEEydOMG/ePObPn28/p7KyEg8PD4d+Ln29F4vj06dP07Zt2yqv26tXL/z9/Zk8eTLh4eGEh4fTr1+/Gk81ufR6RqMRLy8vh30Xnxx48XHchw4dIjc3l5iYGKe+jh8/TmlpKeXl5YSGhtr3e3p6EhwcfMU4cnNzycrKIj8/nx9++IGKigrKyso4f/68fYqGq6urwy9a69atL/uY8ItGjhzJiBEj7Nt6xy/NgdVqvabXMxgMWCwWCgoKsNls1/TaTZVy4kj5cKacOKtLTsxmc40f6FbrovnivNefKi8v59lnn71uHqNsNju/dJvNhtFotP/7ois9Tvoig8HgsH1RZWVlneI0Go386U9/Yt++fezcuZMNGzawfPlynnvuOfujyx9//HG6du3qdN6lLn29Fwu8K/1gubu7k5qayu7du9m5cycrVqzg7bffJiUlhRtuuKHauH+a35/m52IMF/Njs9ns841/ytfXl4KCgmqv+VNFRUWkpKQwbNgwRo0ahaenJ3v37mXevHkO39Oqvm/V/dK5uLjg4uJS65hEmrLG+gNss9n0x/8nlBNHyocz5cRZQ+ek1g83uRyz2UxkZCRr166try4bxcWbzr777jv7vsOHDzdSNBcYDAbCwsKIjo7m+eefx2w2k52dja+vL35+fhw/fhyLxeLwFRAQUOP+zWZzlcW9yWSiV69ePProo6SlpVFUVERubm59vjS7kJAQ8vPz8ff3d3otbm5uWCwWTCYT+/fvt59z9uzZK46MHTx4kMrKSsaMGUNoaCjBwcEO31cRERGRmqrXGwFbtWp13RclrVq1omvXrrz77rsEBARQUlLC8uXLGy2eAwcOsGvXLnr37o2Pjw8HDhygpKTEPq3i4Ycf5rXXXsPDw4Pw8HDKy8s5ePAg33//vcP0gSvx9/fn/Pnz7Nq1i44dO+Lq6kpubi7Hjx+ne/fu3HDDDXz55ZdUVlZWOx3iav3yl79k8+bNvPjii9x33314eXlRUFDAJ598whNPPIGbmxtDhgzhjTfewMvLCx8fH5YvX37FaREWi4WKigo2bNjA7bffzr59+9i0aVODxC8iIiLNW70VzSUlJaxZs6bBiqpracKECcydO5c//OEPBAcH8+ijj/LXv/61UWJxd3dnz549rF+/nh9++IE2bdowZswY+1JtQ4cOxdXVlTVr1vDGG2/g6upKhw4davVEvJtvvplhw4bxj3/8gzNnzvDQQw/Rq1cvsrOzefvttykrKyMoKIjf/e53tG/fvkFep5+fH7Nnz2bp0qU8++yzlJWV4e/vT+/eve2FcUxMDOfPn+f555/Hzc2NqKioKz7OvVOnTowZM4Z3332XZcuW0a1bNx555BFefvnlBnkNAKZn/kHlJatqtFQGg4GgoCCsVqs+PvxfyomIyPXNYKvl/96TJk1yGt0rKyujpKQEg8HAjBkzCA8Pr88YRa4bRUVFDkvRtVQqEJ0pJ86UE2fKiSPlw5ly4qwuOXFxcWm4GwG7d+/uVDRfvGBERESt5tKKiIiIiFwPal00T5o0qSHiaNaKi4v5/e9/X2VbaWkpgH0ljJ964YUXaNOmTYPFVltXei3Q9OIVERERqQ+1LppfeeUVHnrooSpHlIuKinj77beZOHFivQTXXLRu3Zq0tLSrPrcpqe61NLV4RUREROpDrYvmDz/8kLvvvrvKovnMmTN8+OGHKpp/wmQyNZsn9zSn1yIiIiJSU/W2TjNcWDdXD3wQERERkeamRiPNeXl55OXl2bc3b95MTk6OwzGlpaVs376ddu3a1WuAIiIiIiKNrUZF8+7du1m5cqV9+9///neVx7Vp04bHHnusfiITEREREWkialQ0/+pXvyIyMhKbzcZvf/tbnn76aUJCQhyOcXFxwc3NrUGCFGlI0dHRTJs2jX79+tW5r4rZU6g4uK8eorr+fdvYATRBjZUT08I1jXRlEZHmo0Zzmlu1aoWXlxfe3t68/PLLdO/eHS8vL4cvFczNV3JyMpmZmbU6Jzo6muzs7IYJSEREROQaq/XqGTV9aoqIiIiISHNR66IZLtwY+P7773P06FH7wzkuMhgMzJkzp16Ck8aXkZFhvxF0/fr1wIVR5E2bNpGeno6XlxcAqampnDt3jpkzZ5KQkABAeno6cOGNVkZGxlXHcPjwYRYvXszBgwcxGAxYLBbi4+O56aabWLFiBdu3b3dYO3rdunWsX7/e4Zr//ve/Wbt2LQUFBXh6etK/f//Lzr9fuXIl//rXv3j66afp1KnTVcctIiIizUeti+a9e/cye/ZsunfvztGjRwkPD+eHH35g//79BAYGcvPNNzdEnNJIYmNjsVqttG/fnlGjRgHg7e1NTk4O8+bNY/r06WzcuJE9e/aQlpaG0WgkJSWFuLg4Jk6cSHh4OEZj3VY2nDNnDp06dSIuLg6j0cjhw4cxmUw1Pn/jxo0sXryY0aNHEx4ezrlz59i3z3nesc1mIzMzk+3bt/OXv/yFoKCgy/ZZVlZGWVmZfdtgMODu7l67FyZyjRgMhsYOoUoX42qq8TUG5cSR8uFMOXF2rXJS66J5xYoVDB48mN/+9rf85je/YdSoUXTu3JlvvvmG5557rl5uppKmw8PDA7PZjKurK76+vvb9CQkJzJgxg6VLl7Jhwwbi4+PtU3e8vb3t5156ztUqLi4mKiqKtm3bAlyxmK3KqlWriIqK4p577rHv69Kli8MxlZWVvPzyyxw6dIjZs2dz4403XrHPrKwshxVlQkJCSE1NrVVcItdKbX9nrjU9MMmZcuJI+XCmnDhr6JzUumj+9ttviYqKsm9XVlYC0LFjRx588EFWrVpFnz596i9CaZICAwOJiYlhwYIFREREMGjQoAa71r333sv8+fP56KOP6NmzJ3fccUeNfzFOnz7Nd999xy233HLF4xYvXozZbObZZ5+1F/1XMnLkSEaMGGHf1jt+acqsVmtjh1Cli9OtCgoKsNlsjR1Ok6CcOFI+nCknzuqSE7PZXOP79WpdNP/444+4ublhNBoxm82cOXPG3hYcHEx+fn5tu5TrVF5eHkajkaKiIioqKmo1ZaI2oqOjGThwIDt27CAnJ4cVK1YwZcoU+vXrV+XUj4qKCvu/W7VqVaNr9OzZk08++YSvvvqqRm8AXFxc9PRLuW409T+sNputycd4rSknjpQPZ8qJs4bOSa0nm7Zp04bTp08D0K5dO3bs2GFvy8vLw9PTs/6ikybBbDbbP1G4aNu2bWRnZ5OcnExxcTGrVq1yaDeZTE7n1EVwcDAjRozgT3/6E/369WPLli3Ahakgp06dcvglOXz4sP3f7u7u+Pv7k5ube8X++/Tpw+9+9zvmzZvHJ598Um9xi4iISPNQ65Hm7t27s3v3bu644w6GDh3KokWLOHr0KC4uLnz11VcOH1lL8+Dv78+BAwcoLCzEzc2N0tJSFi5cyOjRowkLC2PSpEmkpKQQHh5OaGgoAAEBAeTm5hIWFobZbL7qN1OlpaW8/vrr3HHHHQQEBHDixAkOHjxI//79gQs/jyUlJbz77rvccccd5OTk8OWXX+Lh4WHv4+GHH2bhwoV4e3tz66238sMPP7Bv3z6GDx/ucK1+/fqRkJDAnDlzMJlM3HHHHbWO1/TMP6i85AbBlspgMBAUFITVatVIyP9STkRErm+1Lpqjo6M5e/YsAHfffTelpaV89NFHGAwGHnjgAR544IF6D1IaV1RUFBkZGSQmJlJaWkrPnj3p0qULkZGRAPTq1YvIyEjmzJlDWloabm5uxMTEsGTJEjZv3oyfn99VLzlnNBo5c+YML7/8MqdPn8bLy4v+/fsTHR0NXPi047HHHiMrK4tVq1bRv39/oqKi2Lx5s72PwYMHU1ZWxrp163j99dfx9va2F90/dccdd1BZWcmcOXMwGAyXPU5ERERaFoNNQx4i9aaoqMhhKbqWSqOqzpQTZ8qJM+XEkfLhTDlxVpecuLi41PhGwDotoHvs2DH27t3L+fPn69KNiIiIiEiTdlVPBPzwww958803+e677wBISUmhc+fO/P3vf6dXr17cdddd9RqkXP8SExMpKiqqss3b25uSkpIq2+Lj4xt0OTsRERGRmqh10fzpp5/yyiuvcNttt/HAAw+waNEie1vnzp359NNPVTSLk6SkJIel4C5lMpku2+bj49OQYYmIiIjUSK2L5tWrVzN48GAmTJhAZWWlQ9Hctm1b3n///XoNUJqHms4XEhEREWmKaj2nOT8/n5/97GdVtnl6etpX1hARERERaS5qXTS7urpy7ty5KttOnjzJDTfcUOegRERERESakloXzTfffDMbNmyockmPrVu30r1793oJTGTFihVMnz79isdkZGTw/PPPX6OIREREpKWq9Zzmhx56iD//+c/88Y9/tE/TyM7OZsWKFezZs4fnnnuu3oOUxpGcnEynTp0YN25cjc+Jjo5m2rRp9OvXr+ECa0B1jb9i9hQqDu6r56iuT982dgBN0KU5MS1c02hxiIhI7dVopPmbb76htLQUgJtuuomkpCTOnz/P66+/DkBWVhZWq5WkpCQ6dOjQcNGKiIiIiDSCGo00z5gxg2effZYuXbowa9Ys4uLieOGFFygoKLA/2jg4OLihY5VrKCMjg7y8PPLy8li/fj1wYRR206ZNpKen4+XlBUBqairnzp1j5syZJCQkAJCeng5cWDHjah+ffalNmzbxzjvvcObMGW677TYef/zxy86dz8nJYdWqVXz77bcYjUZCQ0MZN24cFosFgPLychYvXsznn3/O999/j6+vL3fddRcjR45k0qRJDRK/iIiIXP9qVDS7uLhQXl4OQF5eHj/88AMAFovFXoxI8xIbG4vVaqV9+/aMGjUKuPAQkpycHObNm8f06dPZuHEje/bsIS0tDaPRSEpKCnFxcUycOJHw8HCMxjo9cBKAgoICPv30U5566inOnTvHvHnzWLRoEU8++WSVx58/f54RI0bQoUMHfvzxR9566y3S09N5/vnnMRqNrF+/ni+++ILf//73tGnThhMnTlBcXAxQq/jLysocHpdtMBhwd3ev8+uVlsNgMDR2CI3uYg6Ui/+jnDhSPpwpJ86uVU5qVDQHBgby3nvvcerUKQB2797NiRMnLnt8//796yU4aTweHh6YzWZcXV3x9fW1709ISGDGjBksXbqUDRs2EB8fb1+D2dvb237upefURVlZGZMmTeLGG28EYPz48aSkpDBmzJgqr3HHHXc4bE+YMIG4uDjy8/Pp0KEDxcXFBAUFERYWhsFgcFg/ujbxZ2VlsXLlSvt2SEgIqampV/kqpSUKCgpq7BCaDA2+OFNOHCkfzpQTZw2dkxoVzQ8++CAvv/wyX3zxBQDLli274vFvvfVW3SOTJikwMJCYmBgWLFhAREREgz/iuk2bNvaCGSA0NBSbzcaxY8eqLGwLCgp46623OHDgAGfOnKGyshKA4uJiOnTowODBg/nrX//KlClT6N27N7fffju9e/eudVwjR45kxIgR9m2945faslqtjR1CozMYDFgsFgoKCqpckaklUk4cKR/OlBNndcmJ2Wyu8QPYalQ0R0REcMstt3Ds2DFmzpzJY489Rrt27WoVlDQfeXl5GI1GioqKqKiowGQyNXZIdqmpqbRp04bHH3+c1q1bY7PZmDp1qn16UefOnXn55ZfJyclh586dvPDCC/Ts2ZOpU6fW6jouLi64uLg0xEuQFkJ/7P6PzWZTPn5COXGkfDhTTpw1dE5qvOSct7c33t7e/PznPyc8PJyAgIAGC0qaBrPZbB+pvWjbtm1kZ2eTnJzMCy+8wKpVq4iOjra3m0wmp3Pqori4mJMnT+Ln5wfA/v37MRgMVd54eubMGY4ePUp8fDzdunUDYO/evU7HeXh4EBERQUREBHfccQfPPfccZ8+exdPTs97jFxERkeah1ndqTZw4UQVzC+Hv78+BAwcoLCykpKSE4uJiFi5cyOjRowkLC2PSpElkZWWxf/9++zkBAQHk5uZy6tSpenmkuouLCxkZGRw+fJg9e/bw2muvMWDAgCqnZtxwww14eXnxwQcfUFBQQG5uLosXL3Y4Zu3atXzyySccPXqUY8eO8dlnn+Hr64uHh0eDxC8iIiLNQ60fbiItR1RUFBkZGSQmJlJaWkrPnj3p0qULkZGRAPTq1YvIyEjmzJlDWloabm5uxMTEsGTJEjZv3oyfn1+dl2yzWCz079+flJQUzp49y6233kpcXFyVxxqNRn73u9/x2muvMXXqVIKDg4mNjSU5Odl+jJubG++++y5WqxWj0UiXLl1ISkqyr5RR1/hNz/yDyktW1WipDAYDQUFBWK1WfXz4v5QTEZHrm8Gm/71F6k1RUZHDUnQtlQpEZ8qJM+XEmXLiSPlwppw4q0tOXFxcanwjYN0X0hURERERaeY0PUMaVGJiIkVFRVW2eXt7U1JSUmVbfHx8gy9nJyIiIlJTKpqlQSUlJVFRUVFlm8lkumybj49PQ4YlIiIiUisqmqVB1XSekIiIiEhTpjnNIiIiIiLVUNEsIiIiIlINFc0iIiIiItXQnGa55pKTk+nUqRPjxo277DHR0dFMmzaNfv36XbvA6kHF7ClUHNzX2GE0Cd82dgCNyLRwTWOHICIi9UwjzY0oOTmZzMzMWp0THR1NdnZ2wwTUhCxYsIBbb721scMQERERATTSLE2Ur69vg1+jvLwcs1m/AiIiIlI9VQyNJCMjg7y8PPLy8li/fj1wYRR506ZNpKen4+XlBUBqairnzp1j5syZJCQkAJCeng5cWM4tIyPjqmNYsWIF27dvZ/jw4bz99tucPXuWO++8k8cee4z33nuPtWvXYrPZuOeee3jggQfs561du5YtW7ZQWFiIp6cnt99+O48++ihubm72Y/bu3cubb77JwYMHcXFxoUuXLvzud7/D09MTgMrKSt544w02b96M2Wxm2LBhREdH28+/dHpGYWEhkydPZurUqWzYsIEDBw4QFBTEb3/7W0JDQ+3n7Nu3j2XLlvH111/j7e1N3759eeSRR+xxTZo0iSFDhlBQUEB2djZ9+/blxIkTtGvXjscee8zez5kzZ3j88cf54x//yC233HLV+RUREZHmQ0VzI4mNjcVqtdK+fXtGjRoFXHhCXk5ODvPmzWP69Ols3LiRPXv2kJaWhtFoJCUlhbi4OCZOnEh4eDhGY91n1xw/fpycnByefvppCgoK+Pvf/05hYSFBQUHMmjWLffv2MXfuXG655RZ7gWowGIiNjSUgIIDCwkJeffVV3njjDeLi4gA4fPgws2fP5he/+AWxsbGYTCZ2795NZWWl/boffvghI0aM4LnnnmP//v288sorhIWF0atXr8vGunz5cmJiYrBYLCxfvpwXX3yRl156CZPJxJEjR3j22WcZNWoUTzzxBCUlJfzzn//kn//8JxMnTrT3sWbNGh588EEefPBBAL7++msWLVrEmDFjcHFxAeCjjz6idevW9OjR47KxlJWVUVZWZt82GAy4u7tfxXdAmiODwXDZfVW1tVTKiTPlxJHy4Uw5cXatcqKiuZF4eHhgNptxdXV1mIqQkJDAjBkzWLp0KRs2bCA+Pt7+gBBvb2/7ufU1fcFmszFhwgTc3d1p164dPXr04NixYyQlJWE0GgkODubdd98lLy/PXjTfe++99vMDAgIYNWoUr776qr1ofvfdd+ncubN9G6B9+/YO1+3YsSMPP/wwAEFBQWzYsIFdu3ZdsWiOioritttuAy6MRCcmJlJQUEDbtm1Zs2YNAwcOtMcWFBREbGwsM2fOJC4ujlatWgFwyy23cN9999n7vPHGG/nnP//J9u3biYiIAGDr1q0MHjz4ir98WVlZrFy50r4dEhJCamrqZY+XliUoKOiybRaL5RpGcn1QTpwpJ46UD2fKibOGzomK5iYmMDCQmJgYFixYQEREBIMGDWrQ6/n7+zuMkPr4+GA0Gh1GsX18fDh9+rR9Ozc3l6ysLPLz8/nhhx+oqKigrKyM8+fP4+bmxuHDhxkwYMAVr9uhQweH7datWztco7pzLr5pOH36NG3btuXQoUMUFBTw0UcfOZxjs9koLCykXbt2ANx0000O7S4uLgwaNIgtW7YQERHB4cOH+eabb5g+ffoVYxk5ciQjRoywb+sdv1zKarU67TMYDFgsFgoKCrDZbI0QVdOjnDhTThwpH86UE2d1yYnZbK7x04tVNDdBeXl5GI1GioqKqKiowGQyNdi1ftq3wWCoct/FH8KioiJSUlIYNmwYo0aNwtPTk7179zJv3jwqKioA7KO6V1LVDXjV/aBfes7FIvXiOTabjbvuuot77rnH6bw2bdrY/+3q6urUPnToUKZPn86JEyf497//zS233FLtL5CLi4t9OofIT13pZ9lms+kP3U8oJ86UE0fKhzPlxFlD50RLzjUis9nsMM8XYNu2bWRnZ5OcnExxcTGrVq1yaDeZTE7nXEsHDx6ksrKSMWPGEBoaSnBwMN99953DMR07dmTXrl3XNK6QkBDy8/OxWCxOX9WtkNGhQwduuukmNm/ezCeffMIvfvGLaxS1iIiIXC800tyI/P39OXDgAIWFhbi5uVFaWsrChQsZPXo0YWFhTJo0iZSUFMLDw+3ziQMCAsjNzSUsLAyz2WxfjeJasVgsVFRUsGHDBm6//Xb27dvHpk2bHI65//77mTZtGq+++irDhg3DbDaze/du7rjjDvu87Pr2q1/9iqeffppXX32Vu+66C1dXV44ePcrOnTsZP358tecPGTKEf/7zn7i6utbpgSqmZ/5B5SU3CLZUBoOBoKAgrFarRkJERKRZ0EhzI4qKisJoNJKYmEhcXBxz586lS5cuREZGAtCrVy8iIyOZM2cO58+fByAmJoadO3cyYcIEnnrqqWsec6dOnRgzZgzvvvsuU6dO5aOPPuKRRx5xOCY4OJg//elPfPPNN/zxj3/kT3/6E9u3b2/QaSYdO3YkOTmZgoIC/vznPzNjxgzeeuutGt8wOXDgQEwmEz/72c9qNL1EREREWhaDTcNAIhQXF9tH9jt37nzV/RQVFTksRddSaaTZmXLiTDlxppw4Uj6cKSfO6pITFxcX3QgoUhPl5eWcOnWKpUuXEhoaWqeCWURERJovFc3XucTERIqKiqps8/b2pqSkpMq2+Pj4Bl/O7nqwb98+Zs2aRVBQEFOnTm3scERERKSJUtF8nUtKSrIv9fZTJpPpsm0+Pj4NGdZ1o0ePHqxYsaKxwxAREZEmTkXzda6m83BERERE5Opp9QwRERERkWqoaBYRERERqYaK5npUWFhIdHQ0hw8fbuxQWowVK1Ywffp0+3ZGRgbPP/98I0YkIiIizZHmNLdwK1asYPv27aSlpTV2KPUiNja2UdetrJg9hYqD+xrt+k3Jt40dQCMwLVzT2CGIiEgDUdF8HSgvL8dsrt236mrOaQ48PDwaOwQRERFphhq9qkpOTqZDhw60atWKzZs3YzabGTZsGNHR0RQWFjJ58mSef/55OnXqBMD3339PbGwsM2fOpEePHuzevZtZs2bxxz/+kWXLlnH06FFCQ0OZMmUKhw4dYsmSJZw8eZJbb72VCRMm4OrqWqd4KysrWbNmDZs3b+bEiRP4+PgwbNgwHnjgAfsxx48fZ/HixRw4cICgoCB++9vfEhoaCsCZM2dYtGgRe/fu5ezZswQGBjJy5EgGDhzokJP27dtjNpv5z3/+Q7t27Zg1axYrVqxgy5YtnD59Gi8vL/r378/48eMBmDRpEkOGDKGgoIDs7Gz69u3L5MmTeeONN9i+fTsnTpzA19eXgQMH8tBDD2E2m9m6dSsrV64EIDo6GoCJEycyePBgzp07x+uvv8727dspKyujc+fOjB071v59uJKLo9fDhw/n7bff5uzZs9x555089thjvPfee6xduxabzcY999zjkLeaXHP16tWsW7eOH3/8kQEDBuDt7e1w7YyMDL7//ntmzJgBQE5ODqtWreLbb7/FaDQSGhrKuHHjsFgsAPafsalTp7Jhw4Yqv2ciIiIijV40A3z44YeMGDGC5557jv379/PKK68QFhZmL2xq4u2332b8+PG4urrywgsv8MILL+Di4sKTTz7J+fPnSU9P5/333+f++++vU6zLli1j8+bNjB07lrCwME6dOsXRo0cdjlm+fDkxMTFYLBaWL1/Oiy++yEsvvYTJZLIXg/fffz/u7u7s2LGDl19+mcDAQLp27eqQk7vvvpvZs2djs9n47LPPWLduHVOmTKF9+/acOnXKae70mjVrePDBB3nwwQft+9zd3Zk4cSKtW7fmyJEjzJ8/H3d3d371q18RERHBkSNH+Oqrr3jmmWeACyO1NpuNlJQUPD09SUpKwsPDg02bNjF79mxefPFFPD09q83T8ePHycnJ4emnn6agoIC///3vFBYWEhQUxKxZs9i3bx9z587llltuITQ0tEbX3LZtGytWrOCxxx6jW7du/Oc//+H9998nICDgsnGcP3+eESNG0KFDB3788Ufeeust0tPTef755zEa/29K/5W+Z1UpKytzeFy2wWDA3d292rxI82YwGKptu9IxLY1y4kw5caR8OFNOnF2rnDSJorljx448/PDDAAQFBbFhwwZ27dpVq6L517/+NWFhYQAMGTKEZcuWMWfOHAIDAwHo378/u3fvrlPR/MMPP/D+++8zfvx4Bg8eDIDFYrFf96KoqChuu+024MIIbmJiIgUFBbRt2xY/Pz/uu+8++7HDhw8nJyeHTz/91KFotlgsPProo/btL7/8El9fX3r27InZbKZNmzZ06dLF4bq33HKLQ9+AQwEdEBDAsWPH2LZtG7/61a9o1aoVbm5uGI1GfH197cfl5uZy5MgRXn31VVxcXAAYM2YM27dv57PPPuOuu+6qNlc2m40JEybg7u5Ou3bt6NGjB8eOHSMpKQmj0UhwcDDvvvsueXl5hIaGsnv37mqvuX79en7xi18wdOhQ4ML3fNeuXZSWll42jjvuuMNhe8KECcTFxZGfn0+HDh3s+6/0PatKVlaWfZQeICQkhNTU1GrzIs1bUFBQtcfU5v+1lkI5caacOFI+nCknzho6J02iaL60eAFo3bo1p0+frlUfHTt2tP/bx8cHV1dXe8EM4Ovry8GDB+sU59GjRykrK6Nnz55XPO7S13OxGD19+jRt27alsrKS1atXs23bNk6ePElZWRnl5eVO00Y6d+7ssH3HHXewbt06EhIS6N27N7fddhu33367w0joTTfd5BTLxRHqgoICzp8/T2VlZbUjoocOHeL8+fP2qR8XlZaWUlBQcMVzL/L393e4jo+PD0aj0WF018fHx/59rsk1jx49yrBhwxzau3btyu7duy8bR0FBAW+99RYHDhzgzJkzVFZWAlBcXOzwfbrS96wqI0eOZMSIEfZtveMXAKvVetk2g8GAxWKhoKCgUW9WbUqUE2fKiSPlw5ly4qwuOTGbzTV+UFyTKJqrumHNZrPZC6xLE3ClR0ZfZDAYqvxY/WLBdLVatWpVo+MufT0Xi6mLr+G9995j3bp1jB07lg4dOuDm5kZmZibl5eUOfbi5uTlst2nThhdffJGdO3eyc+dOXn31VdasWUNycrL9ej8tvPfv388//vEPoqOj6d27Nx4eHnzyySesXbv2ivFXVlbSunVrkpOTndpqeqPdT/Nf1ffEYDDY81If16xKamoqbdq04fHHH6d169bYbDamTp3qlO8rfc+q4uLiYh8RF7moJv9Z22w2/aH7CeXEmXLiSPlwppw4a+icNImi+XIu3uT13XffERISAtCoayBbLBZatWrFrl277FMEamvPnj306dOHO++8E7hQLFqt1suOaF6qVatW9OnThz59+hAZGcmUKVM4cuSI06j0Rfv27cPf39/hZrvi4mKHY8xms9Obic6dO3Pq1CmMRuMV5wvXp5pcs23bthw4cICf//zn9n0HDhy4bJ9nzpzh6NGjxMfH061bNwD27t1bv4GLiIhIi9Cki+ZWrVrRtWtX3n33XQICAigpKWH58uWNGs+vfvUr3njjDcxmMzfffDMlJSXk5+czZMiQGvVhsVj4/PPP2bdvHzfccANr167l1KlT1RbNW7dupbKyki5duuDq6sp//vMfWrVqdcWPFCwWC8XFxXzyySfcdNNN7Nixg+zsbIdjAgICKCws5PDhw/j5+eHu7k7Pnj0JDQ0lLS2N0aNHExwczHfffceXX35J3759q5wGUlc1ueY999xDRkYGnTt3JiwsjI8//pj8/PzLFtk33HADXl5efPDBB7Ru3Zri4mKWLl1a77GLiIhI89eki2a4cOPW3Llz+cMf/kBwcDCPPvoof/3rXxstngcffBCTycSKFSs4efIkrVu3dppneyUPPfQQhYWFPPvss7i6ujJ06FD69u3LuXPnrnieh4cH7777LosXL6ayspIOHTrw1FNP4eXlddlz+vbty7333ss///lPysrKuO2223jwwQd5++237cf079+fzz//nFmzZvH999/bl5xLSkrizTffZO7cuZSUlODr60u3bt3w8fGp8WutDYPBUO01IyIiKCgoYOnSpZSVldG/f3+GDRvGV199VWWfRqOR3/3ud7z22mtMnTqV4OBgYmNjq5wCUl9Mz/yDyktW1WipDAYDQUFBWK1WfXwoIiLNgsGmv2gi9aaoqMhhKbqWSkWzM+XEmXLiTDlxpHw4U06c1SUnLi4uNb4R0Fj9ISIiIiIiLVuTn55R34qLi/n9739fZdvF9X4vt0rGCy+8QJs2bRostutFYmIiRUVFVbbFx8czaNCgaxyRiIiISMNqcUVz69atSUtLu+pzBZKSki679F9DzXkWERERaUwtrmg2mUx6ik4d1XTuj4iIiEhzoTnNIiIiIiLVUNEsIiIiIlINFc0iIiIiItVocXOapXlbsWIF27dvv+qbPQEKCwuZPHkyzz//PJ06darVuRWzp1BxcN9VX7s5+baxA6gD08I1jR2CiIg0MRppvk4lJyeTmZlZq3Oio6OdHqMtIiIiItVT0SzXRHl5eWOHICIiInLVND3jOpSRkUFeXh55eXmsX78euDCKvGnTJtLT0/Hy8gIgNTWVc+fOMXPmTBISEgBIT08HLiwbl5GRUac4vvjiC1auXMm3336Lm5sb3bp1Y9q0aQBMmjSJIUOGUFBQQHZ2Nn379mXy5Ml89tlnrFixgoKCAlq3bk1kZCRRUVEAvP/++3zwwQf8v//3/wDIzs4mPT2d8ePHExkZCcCzzz5LSEgIjzzyCACrV69m3bp1/PjjjwwYMABvb2+nOLds2cKaNWsoLCzE39+f4cOH88tf/tLe/vXXX7NgwQKOHj1K+/bteeCBB+qUFxEREWl+VDRfh2JjY7FarbRv355Ro0YB4O3tTU5ODvPmzWP69Ols3LiRPXv2kJaWhtFoJCUlhbi4OCZOnEh4eDhGY90+ZNixYwfp6ek88MADTJ48mfLycnbs2OFwzJo1a3jwwQd58MEHATh06BAvvPACDz/8MBEREezfv59XX30VLy8vBg8eTI8ePcjMzKSkpARvb2/y8vLw8vIiLy+PyMhIKioq2LdvH/feey8A27ZtY8WKFTz22GN069aN//znP7z//vsEBATYY/jggw94++23GT9+PCEhIfz3v/9l/vz5uLq6MnjwYM6fP8/f/vY3brnlFhISEigsLKzRtJeysjLKysrs2waDAXd39zrlVJoOg8HQYH02RN/XK+XEmXLiSPlwppw4u1Y5UdF8HfLw8MBsNuPq6oqvr699f0JCAjNmzGDp0qVs2LCB+Ph4+4NILo7Aenh4OJxztd555x0iIiKIjo627/vpTXO33HIL9913n337pZdeomfPnjz00EMABAcHk5+fz5o1axg8eDDt27fH09OTvLw87rjjDvLy8oiKimLdunUAHDx4kLKyMsLCwgBYv349v/jFLxg6dCgAv/71r9m1a5f9cegAq1atIiYmhv79+wMQEBBAfn4+H3zwAYMHD+bjjz+msrKSCRMm4OrqSvv27Tlx4gSvvvrqFV9/VlYWK1eutG+HhISQmppa2zRKExUUFNRgfevhSs6UE2fKiSPlw5ly4qyhc6KiuRkJDAwkJiaGBQsWEBERwaBBgxrsWocPH7YXq5dz0003OWwfPXqUPn36OOy7+eabWbduHZWVlRiNRrp160ZeXh49e/bk22+/ZdiwYbz33nvk5+eze/duQkJCcHNzs/c3bNgwh/66du3K7t27ASgpKeHEiRPMmzeP+fPn24+prKzEw8MDgPz8fDp27Iirq6u9PTQ0tNrXP3LkSEaMGGHf1jv+5sVqtdZ7nwaDAYvFQkFBATabrd77vx4pJ86UE0fKhzPlxFldcmI2m2v8pGMVzc1MXl4eRqORoqIiKioqMJlMDXKdVq1aVXvMpYUogM1mcyouf/rD3aNHDz744AP27NlDp06duOGGG+yFdF5eHt27d69xjJWVlQA8/vjjdO3a1aGtrtNTXFxccHFxqVMf0nQ15B8im82mP3Q/oZw4U04cKR/OlBNnDZ0TrZ5xnTKbzfai8KJt27aRnZ1NcnIyxcXFrFq1yqHdZDI5nXO1OnbsyK5du2p1Trt27di7d6/Dvv379xMcHGwvYnv06EF+fj6fffaZvUDu3r07u3btYt++fQ5Fc9u2bTlw4IBDf5du+/r64ufnx/Hjx7FYLA5fF+c9t2vXjm+++cZhSsdP+xQRERHRSPN1yt/fnwMHDlBYWIibmxulpaUsXLiQ0aNHExYWxqRJk0hJSSE8PNw+3SAgIIDc3FzCwsIwm814enpe9fUfeugh/vKXv2CxWIiIiKCyspIvv/ySX/3qV5c9Z8SIESQlJbFy5Ur7jYAbNmwgLi7OfszFec0ff/wxM2bMAC4UzUuWLAGwz2cGuOeee8jIyKBz586EhYXx8ccfk5+f73Aj4MMPP8xrr72Gh4cH4eHhlJeXc/DgQb7//ntGjBjBwIEDefPNN5k7dy4PPvgghYWFvPfee1edF9Mz/6DykhsEWyqDwUBQUBBWq1UjISIi0iyoaL5ORUVFkZGRQWJiIqWlpfTs2ZMuXbrYl2br1asXkZGRzJkzh7S0NNzc3IiJiWHJkiVs3rwZPz+/Oi0516NHDxITE1m1ahWrV6/G3d2dbt26XfGczp078/vf/54VK1awatUqWrduTXR0NIMHD7YfYzAY6N69O9u3b7cXyB07dsTDw4PAwED7XGSAiIgICgoKWLp0KWVlZfTv359hw4bx1Vdf2Y8ZOnQorq6urFmzhjfeeANXV1c6dOhgX4HDzc2Np556ioULFzJjxgzatWvH6NGj7cveiYiIiAAYbBoGEqk3RUVFDkvRtVQaaXamnDhTTpwpJ46UD2fKibO65MTFxaXGNwJqTrOIiIiISDU0PaMFS0xMpKioqMo2b29vSkpKqmyLj49v0OXsRERERJoaFc0tWFJSEhUVFVW2mUymy7b5+Pg0ZFgiIiIiTY6K5haspnN4RERERFo6zWkWEREREamGimYRERERkWpoeoY0CR988AGrVq3i5MmTjBkzhu+//57t27eTlpZW575XrFjBpk2bOH36NNOmTWP79u18//339oen1KeK2VOoOLiv3vu9Hn3b2AHUkmnhmsYOQUREmjAVzUJycjKdOnVi3LhxNT4nOjqaadOm0a9fvzpf/9y5cyxatIixY8fSv39/PDw8sNlsDB8+vM595+fns3LlSqZNm0ZoaCg33HAD27dvr3O/IiIi0rKoaJZGV1xcTEVFBbfddhutW7e273dzc7vsOeXl5ZjN1f/4Hj9+HIC+fftiMBjqHqyIiIi0SCqaW7iMjAzy8vLIy8tj/fr1wIVR5E2bNpGeno6XlxcAqampnDt3jpkzZ5KQkABAeno6cGEVjqt9JPfWrVt55ZVXAJg8eTIAL7/8Mlu3bnWYnpGRkcH3339P165d2bBhA2azmYyMDE6ePMnixYvZuXMnBoOBsLAwxo0bR0BAACtWrGDlypUAjBo1CrgwVeNSH374IYsXL2b+/Pm4uLjY96enp+Pm5maPSURERFo2Fc0tXGxsLFarlfbt29sLS29vb3Jycpg3bx7Tp09n48aN7Nmzh7S0NIxGIykpKcTFxTFx4kTCw8MxGq/+ftKIiAhuvPFGZs+ezXPPPUebNm3w9vau8tjc3Fw8PDz405/+hM1m48cff2TWrFmEhYUxa9YsjEYj77zzDs899xzp6encd999BAQE8Morr7BgwYIq+xwwYACvvfYaX3zxBQMGDACgpKSEHTt28Mc//vGycZeVlTk8LttgMODu7n7VeZDG19CfRFzsX594/B/lxJly4kj5cKacOLtWOVHR3MJ5eHhgNptxdXXF19fXvj8hIYEZM2awdOlSNmzYQHx8vH1d54tFrYeHh8M5V6NVq1b20Wxvb+8r9ufq6soTTzxhn5bx73//G4PBwBNPPGH/RZk4cSLjxo1j9+7d9O7dGw8PD4DL9tuqVSsGDhzI1q1b7UXzxx9/zI033kiPHj0uG0tWVpZ9FBsgJCSE1NTUGr9uaXqCgoKuyXUsFss1uc71RDlxppw4Uj6cKSfOGjonKpqlSoGBgcTExLBgwQIiIiKaxGOzO3To4DCP+dChQxQUFDBmzBiH48rKyuxzmWti6NChJCUlcfLkSfz8/NiyZQs///nPr/iOdeTIkYwYMcK+rXf81z+r1dqg/RsMBiwWCwUFBdhstga91vVCOXGmnDhSPpwpJ87qkhOz2Vzjh72paJbLysvLw2g0UlRUREVFBSaTqVHjcXV1ddi22Wx07tyZJ5980unYy03xqEpISAgdO3bkww8/pHfv3hw5coSnnnrqiue4uLg4zIGW69+1+uNjs9n0h+4nlBNnyokj5cOZcuKsoXOih5sIZrOZyspKh33btm0jOzub5ORkiouLWbVqlUO7yWRyOudaCwkJwWq14u3tjcVicfi6OC2jpoYOHcrWrVvZsmULvXr1ok2bNg0UtYiIiFyPVDQL/v7+HDhwgMLCQkpKSiguLmbhwoWMHj2asLAwJk2aRFZWFvv377efExAQQG5uLqdOneLs2bONEvegQYPw9vYmLS2NPXv2UFhYSF5eHq+99honTpyoVV8DBw7k5MmTbN68mV/84hcNFLGIiIhcrzQ9Q4iKiiIjI4PExERKS0vp2bMnXbp0ITIyEoBevXoRGRnJnDlzSEtLw83NjZiYGJYsWcLmzZvx8/O76iXn6sLV1ZVZs2bxxhtvkJ6ezvnz5/Hz8+OWW26p9UoWHh4e9O/fnx07dtC3b9+rjsn0zD+ovGRVjZbKYDAQFBSE1WrVx4ciItIsGGz6iyYCwOzZs2nbti3jx4+/6j6KiooclqJrqVQ0O1NOnCknzpQTR8qHM+XEWV1y4uLiUuMbATU9Q1q8s2fP8sknn5Cbm2sfXRcRERG5lKZnSL1ITEykqKioyjZvb29KSkqqbIuPj2/05eyeeuopzp49y+jRowkODm7UWERERKRpUtEs9SIpKYmKiooq20wm02XbfHx8GjKsGmmM+dgiIiJyfVHRLPWipvOBRERERK5HmtMsIiIiIlINFc0iIiIiItVQ0SwiIiIiUg3NaZZmJyMjg++//54ZM2Zc82tXzJ5CxcF91/y6TdG3jR1ANUwL1zR2CCIich3RSLPUSnJyMpmZmbU6Jzo6muzs7IYJqAqxsbFMmjTpml1PREREmj+NNEuz4+Hh0dghiIiISDOjollqLCMjg7y8PPLy8li/fj1wYRR506ZNpKen4+XlBUBqairnzp1j5syZJCQkAJCeng5cWJqurusir1q1ivfff5/S0lIiIiLw8vIiJyeHtLQ0e5wXp2ds2rSJlStXMnfuXIzG//tgJTU1lRtuuIHJkycD8MUXX/D222+Tn59P69at+fnPf84DDzyAyWSqU6wiIiLSPKholhqLjY3FarXSvn17Ro0aBVx42l9OTg7z5s1j+vTpbNy4kT179pCWlobRaCQlJYW4uDgmTpxIeHi4Q+F6NT766CPeeecd4uLiuPnmm9m2bRvvvfceAQEBVR4/YMAAXnvtNXbv3k3Pnj2BC4/N/uqrr3jqqacAyMnJYc6cOcTGxtKtWzeOHz/O/PnzAXj44Yer7LesrIyysjL7tsFgwN3dvU6vTa4tg8HQKNe71tdtypQTZ8qJI+XDmXLi7FrlREWz1JiHhwdmsxlXV1d8fX3t+xMSEpgxYwZLly5lw4YNxMfH2x924u3tbT/30nOu1oYNGxgyZAi/+MUvAHjooYf46quvOH/+fJXHe3p6Eh4ezscff2wvmj/77DM8PT3t21lZWdx///0MHjwYgMDAQEaNGsXSpUsvWzRnZWWxcuVK+3ZISAipqal1fn1y7QQFBTXKdS0WS6NctylTTpwpJ46UD2fKibOGzomKZqmzwMBAYmJiWLBgAREREQwaNKjBrnXs2DHuvvtuh31dunQhNzf3sucMHDiQBQsWEBcXh4uLCx999BERERH2Ue9Dhw7x9ddf884779jPqayspKysjB9//BFXV1enPkeOHMmIESPs23rHf/2xWq3X9HoGgwGLxUJBQQE2m+2aXrupUk6cKSeOlA9nyomzuuTEbDbX+KnGKpqlXuTl5WE0GikqKqKioqJB5wL/tECt7hekT58+zJ8/nx07dnDTTTexd+9exo4da2+vrKwkOjqa/v37O53r4uJSZZ8uLi6XbZPrQ2P9sbHZbPpD9xPKiTPlxJHy4Uw5cdbQOdGSc1IrZrOZyspKh33btm0jOzub5ORkiouLWbVqlUO7yWRyOudqBQcH8/XXXzvsO3To0BXPadWqFf369eOjjz7ik08+ISgoiM6dO9vbO3fuzLFjx7BYLE5fdZ2DLSIiIs2DRpqlVvz9/Tlw4ACFhYW4ublRWlrKwoULGT16NGFhYUyaNImUlBTCw8MJDQ0FICAggNzcXMLCwjCbzXh6el719SMjI5k/fz6dO3e23wj4zTffEBgYeMXzBg0aRGpqKvn5+U7TRx588EFSU1O58cYbGTBgAAaDgSNHjnDkyBF+/etf1yo+0zP/oPKSGwRbKoPBQFBQEFarVSMhIiLSLKhollqJiooiIyODxMRESktL6dmzJ126dCEyMhKAXr16ERkZyZw5c0hLS8PNzY2YmBiWLFnC5s2b8fPzq9OSc4MGDeL48eO8/vrrlJWVMWDAAAYPHuw0+vxTt9xyC56enhw7doyBAwc6tIWHh/PUU0+xatUq1qxZg8lkom3btgwZMuSq4xQREZHmxWDTMJBc52bPno2vr699TejGVFRU5LAUXUulkWZnyokz5cSZcuJI+XCmnDirS05cXFxqfCOgJmzKdeXHH39k7dq1fPvttxw9epQVK1awa9cufv7znzd2aCIiItKMaXqGXHOJiYkUFRVV2ebt7U1JSUmVbfHx8fTv358vv/ySVatWUV5eTnBwMFOnTqVXr14NGbKIiIi0cCqa5ZpLSkqioqKiyjaTyXTZNh8fH1q1asUzzzzTkOGJiIiIOFHRLNdcTecOiYiIiDQVmtMsIiIiIlINFc0iIiIiItXQ9AxpViZNmsQ999zDvffe2yjXr5g9hYqD+xrl2k3Nt40dQDVMC9c0dggiInId0Uiz1EpycjKZmZm1Oic6Oprs7OyGCUhERETkGlDRLCIiIiJSDU3PkBrLyMggLy+PvLw81q9fD1wYRd60aRPp6el4eXkBkJqayrlz55g5c6b9KX3p6enAhZUz6vIYbYAvvviClStX8u233+Lm5ka3bt2YNm2a03GFhYVMnjyZ559/nk6dOgHw/fffExsby8yZM+nRowcA+fn5vP766+Tl5eHm5kavXr0YO3Ys3t7edYpTREREmg8VzVJjsbGxWK1W2rdvz6hRo4ALDyPJyclh3rx5TJ8+nY0bN7Jnzx7S0tIwGo2kpKQQFxfHxIkTCQ8Px2is24cbO3bsID09nQceeIDJkydTXl7Ojh07rrq/7777jpkzZzJ06FDGjBlDaWkpS5cu5YUXXmDmzJmXPa+srMzhcdkGgwF3d/erjkOuPYPB0CjXu9bXbcqUE2fKiSPlw5ly4uxa5URFs9SYh4cHZrMZV1dXfH197fsTEhKYMWMGS5cuZcOGDcTHx9vXYr44Wuvh4eFwztV65513iIiIIDo62r7v4ijy1di4cSOdO3fmkUcese+bMGECEyZM4NixYwQHB1d5XlZWFitXrrRvh4SEkJqaetVxyLUXFBTUKNe1WCyNct2mTDlxppw4Uj6cKSfOGjonKpqlzgIDA4mJiWHBggVEREQwaNCgBrvW4cOHGTp0aL31d+jQIXJzc4mJiXFqO378+GWL5pEjRzJixAj7tt7xX3+sVus1vZ7BYMBisVBQUIDNZrum126qlBNnyokj5cOZcuKsLjkxm801fuiaimapF3l5eRiNRoqKiqioqMBkMjXIdVq1alXjYy9OBbn0F+inj+i22WzcfvvtPProo07nX2lk3MXFBRcXlxrHIk1PY/2xsdls+kP3E8qJM+XEkfLhTDlx1tA50eoZUitms5nKykqHfdu2bSM7O5vk5GSKi4tZtWqVQ7vJZHI652p17NiRXbt21ejYi1NDvvvuO/u+w4cPOxwTEhJCfn4+/v7+WCwWhy83N7d6iVlERESufyqapVb8/f05cOAAhYWFlJSUUFxczMKFCxk9ejRhYWFMmjSJrKws9u/fbz8nICCA3NxcTp06xdmzZ+t0/YceeohPPvmEFStWkJ+fz5EjR3j33XerPLZVq1Z07dqVd999l/z8fPLy8li+fLnDMb/85S85e/YsL774Il9//TXHjx/nq6++4pVXXqm3Ql9ERESuf5qeIbUSFRVFRkYGiYmJlJaW0rNnT7p06UJkZCQAvXr1IjIykjlz5pCWloabmxsxMTEsWbKEzZs34+fnV6cl53r06EFiYiKrVq1i9erVuLu7061bt8seP2HCBObOncsf/vAHgoODefTRR/nrX/9qb/fz82P27NksXbqUZ599lrKyMvz9/endu/dVzVM2PfMPKi9ZVaOlMhgMBAUFYbVa9fGhiIg0Cwab/qKJ1JuioiKHpehaKhXNzpQTZ8qJM+XEkfLhTDlxVpecuLi41PhGQE3PEBERERGphqZnyDWXmJhIUVFRlW3e3t6UlJRU2RYfH9+gy9mJiIiIXI6KZrnmkpKSnJZ+u8hkMl22zcfHpyHDEhEREbksFc1yzdV07pCIiIhIU6E5zSIiIiIi1VDRLCIiIiJSDRXNIiIiIiLV0JxmafEmTZrEPffcw7333lvnvipmT6Hi4L56iOr6921jBwCYFq5p7BBERKSZ0EizOElOTiYzM7NW50RHR5OdnV0v11+xYgXTp0+vl75ERERE6oOKZrlulZeXN3YIIiIi0kJoeoY4yMjIIC8vj7y8PNavXw9cGEXetGkT6enpeHl5AZCamsq5c+eYOXMmCQkJAKSnpwMXlpTLyMi4qutv3bqVlStX2q8LMHHiRAYPHkx0dDRxcXHk5OSwa9cuoqKiCAgIIDMz02FkPDs7m/T0dFasWGHf98UXX7By5Uq+/fZb3Nzc6NatG9OmTasyhi1btpCZmcnUqVPp1avXVb0OERERaV5UNIuD2NhYrFYr7du3Z9SoUcCFp/Tl5OQwb948pk+fzsaNG9mzZw9paWkYjUZSUlKIi4tj4sSJhIeHYzRe/QcYERERHDlyhK+++opnnnkGAA8PD3v722+/zW9+8xvGjh2L0Whk9+7d1fa5Y8cO0tPTeeCBB5g8eTLl5eXs2LGjymPXrFnD6tWrefrppwkNDb1sn2VlZZSVldm3DQYD7u7uNX2Zco0YDIbGDsHuYixNKabGppw4U04cKR/OlBNn1yonKprFgYeHB2azGVdXV3x9fe37ExISmDFjBkuXLmXDhg3Ex8fbH1Li7e1tP/fSc65Gq1atcHNzw2g0VtnXz372M4YMGVKrPt955x0iIiLsI9cAnTp1cjpu2bJlfPjhhyQnJ9OhQ4cr9pmVlWUfEQcICQkhNTW1VnFJwwsKCmrsEJxYLJbGDqHJUU6cKSeOlA9nyomzhs6JimapkcDAQGJiYliwYAEREREMGjSoUeK46aaban3O4cOHGTp06BWPWbt2LefPn+dvf/sbgYGB1fY5cuRIRowYYd/WO/6myWq1NnYIdgaDAYvFQkFBATabrbHDaRKUE2fKiSPlw5ly4qwuOTGbzTV+UrGKZqmxvLw8jEYjRUVFVFRUYDKZrnkMrq6uDtsGg8HpF6SiosJhu1WrVtX2GxYWxo4dO/j000+5//77qz3excUFFxeX6gOWRtUU/6DYbLYmGVdjUk6cKSeOlA9nyomzhs6JVs8QJ2azmcrKSod927ZtIzs7m+TkZIqLi1m1apVDu8lkcjqnPq9/Od7e3pw/f57z58/b9x0+fNjhmI4dO7Jr164r9tOlSxeefvppsrKyWLNGa/uKiIiII400ixN/f38OHDhAYWEhbm5ulJaWsnDhQkaPHk1YWBiTJk0iJSWF8PBw+81yAQEB5ObmEhYWhtlsxtPT86qvHxAQQGFhIYcPH8bPzw93d/fLjup27dqVVq1a8eabbzJ8+HC+/vprtm7d6nDMQw89xF/+8hcsFgsRERFUVlby5Zdf8qtf/crhuNDQUJKSknjuuecwGo0O0y9qyvTMP6i85AbBlspgMBAUFITVatVIiIiINAsaaRYnUVFRGI1GEhMTiYuLY+7cuXTp0oXIyEgAevXqRWRkJHPmzLGP8MbExLBz504mTJjAU089Vafr9+/fn/DwcGbNmkVcXByffPLJZY/19PQkISGBL7/8kqlTp/Lxxx/z8MMPOxzTo0cPEhMT+eKLL5gxYwazZs3iwIEDVfYXFhbGH/7wB9566y37knsiIiIiBpuGgUTqTVFRkcNSdC2VRpqdKSfOlBNnyokj5cOZcuKsLjlxcXGp8Y2AGmkWEREREamG5jRLg0hMTKSoqKjKNm9vb0pKSqpsi4+Pb7Tl7EREREQuR0WzNIikpCSnpd8uMplMl23z8fFpyLBEREREroqKZmkQNZ0fJCIiInI90JxmEREREZFqqGgWEREREamGpmfIdSs7O5vXX3+dwsJChg8fzrhx46rcdy1VzJ5CxcF91/SaTdW3jXht00I91VFEROqXiuYWKjk5mU6dOtWqqIyOjmbatGn069ev4QKrhYULFzJ48GCGDx+Ou7v7ZfeJiIiI1JWKZrkunT9/ntOnT9O7d2/8/Pwuu09ERESkPqhoboEyMjLIy8sjLy/P/qjo6OhoNm3aRHp6Ol5eXgCkpqZy7tw5Zs6cSUJCAgDp6enAhdUxMjIy6hTHF198wcqVK/n2229xc3OjW7duTJs2DYCzZ8+SmZnJ//zP/1BWVkb37t2JjY0lKCiI3bt3M2vWLAD+8pe/ADBz5swq9/Xo0YN9+/axbNkyvv76a7y9venbty+PPPIIbm5uAJSXl7N8+XI++ugjzp07R/v27Rk9ejQ9evSo0+sTERGR5kOP0W6Bzp07x3PPPUf79u0ZNWoUcOGBI3/+85/x8fFh+vTpbNy4kWXLlpGWloa/vz8lJSXExcUxceJEwsPDMRqNeHt7X3UMO/5/e/ceFFXZxwH8y7IgGKIwgMtFQRRQSSwzJ8maSUvNW+kwgiaplaI0mqOjRqNG00wj6Dh5y8m8MllSchG01NG1sTbTihIcRXBkcxcWWIQVNDQWnvePN/Z137O2AXvWRb+fmf2Dc2F/z3cOu789PHtOUREyMzMxY8YMxMfHw2w2o6ioCDNmzAAAZGZmwmAwYOHChfD29saBAwdQU1ODTZs2AQBqa2uxbNkyrFixAjExMfDx8bG5rKqqCmvWrEFiYiJGjBiBxsZG7NmzB+Hh4UhNTQUAbNmyBUajEbNnz4afnx/Onz+P7OxsbNy4EcHBwTbrb2lpsbpdtpubG7y9vVG99DW0cE7zA6fcVfigS5Bwc3ODSqVCdXU1b337N2YixUysMQ8pZiLVlUyUSuW/vkwuzzQ/gnr27AmlUokePXqgT58+luVLlizBqlWrcODAARw7dgwLFy60HEjtDXLPnj2t9ums3NxcxMfHY+bMmZZlERERAACDwYBffvkFH374IWJiYgAAS5cuxeLFi/Hzzz9j9OjRlpug+Pj4WOqxtaygoABjxozB5MmTAQDBwcGYP38+3n//fbz11luor6+HRqPBjh07LFM6pk2bhgsXLuD06dOYPXu2zfrz8vJw6NAhy88DBgxARkZGl3Mhx7jfhx1XoFKpHnQJLoeZSDETa8xDiplIyZ0Jm2ay6Nu3L5KTk7Fz507Ex8fLejtrrVaLcePG2VxXWVkJd3d3REVFWZb16tULISEhqKys7NDzXLt2DdXV1fj++++tlgshUFtbC51OByEE3nnnHav1ZrMZPj4+9/2906dPx5QpUyw/u7m5dagukpfBYHjQJUjw7JAUM5FiJtaYhxQzkeKZZnogLl26BIVCAaPRiNbWVri7u8vyPJ6envddd78DvjMvDkIIvPjii5g0aZJkXUBAAK5fvw6FQoGMjAwoFNaXLW+f82yLh4cHPDw8OlwPOYcrv5EIIVy6vgeBmUgxE2vMQ4qZSMmdCW9u8ohSKpVoa2uzWvbjjz/i/PnzSE9PR11dHXJycqzWu7u7S/bprPDwcJSUlNhcFxYWhtbWVpSXl1uWNTU1wWAwICwsrEPPM2DAAOj1eqhUKslDqVQiIiICbW1tuHnzpmS9I6ahEBER0cOBTfMjKjAwEOXl5aitrUVjYyPq6urw2Wef4bXXXsPgwYPx9ttvIy8vD2VlZZZ9goKCcPHiRZhMJty6datLz5+QkACNRoOvvvoKer0e169fx+HDhwH8dz7qyJEj8emnn6K0tBRarRZbt26Fv78/Ro4c2aHneeWVV1BWVoZdu3ZBq9Va5kvv2bMHABASEoIxY8Zg27ZtOHfuHGpra3H16lXk5+ejqKioS2MkIiKihwenZzyipk6diu3bt2P58uX466+/MGzYMAwaNAgTJ04EAMTFxWHixInYunUrNmzYAC8vLyQnJyMrKwunTp2Cv79/ly45Fxsbi+XLlyMnJwf5+fnw9vbGkCFDLOtTU1Oxb98+rF+/HmazGUOGDEFaWhqUyo4dsuHh4UhPT8fBgwexbt06CCGgUqkwevRoq+fKzc1FVlYW6uvr0atXL0RHR2PEiBEdHpf72o/Rds9VNR5Vbm5uCA4OhsFg4L8PiYjoocBLzhE5kNFotLoU3aOKTbMUM5FiJlLMxBrzkGImUl3JxMPD419/EZDTM4iIiIiI7OD0DOq05cuXw2g02lzn6+uLxsZGm+sWLlwo6+XsHqSOTh952DEPKWYixUykmIk15iHFTKQ6k0lH9uH0DOq09svS2eLu7n7fdb1794a3t7ecpTldS0sLL0FHRET0EOP0DOq0wMBAm5dyU6lU/7juYWuYgf82zZs3b0Zzc/ODLsUlNDc3Y/Xq1czjHsxEiplIMRNrzEOKmUg5KxM2zUQOotFo+KWMvwkhUFFRwTzuwUykmIkUM7HGPKSYiZSzMmHTTERERERkB5tmIiIiIiI72DQTOYCHhwcSEhL4ZcC/MQ8pZiLFTKSYiTXmIcVMpJyVCa+eQURERERkB880ExERERHZwaaZiIiIiMgONs1ERERERHawaSYiIiIisoM3Liey4fjx4ygoKIDJZEJYWBjmzZuHIUOG3Hf7S5cuYf/+/dDr9fDz88O0adMwfvx4q21++uknZGdno6amBn379sWsWbMwatQouYfiMI7ORKfTITs7GxUVFTAajZg7dy4mT57sjKE4hKPzOHnyJM6cOQOdTgcAiIyMxKxZszBo0CDZx+Iojs7k3LlzyMvLQ3V1NVpbW6FSqTB16lQ8//zzzhiOQ8jxWtJOo9Fg8+bNGDlyJFatWiXXEBzO0Zl89913+OSTTyT7ff755/D09JRlDI4kxzFy+/ZtfPnllzh//jxu376NoKAgJCcnY8SIEXIPxyEcnUl6ejouXbok2e/JJ59EWlravy9MEJEVjUYjkpKSxMmTJ4VOpxN79+4Vc+bMEUaj0eb2NTU1Ys6cOWLv3r1Cp9OJkydPiqSkJHH27FnLNleuXBGJiYkiNzdX6PV6kZubK5KSkkRZWZmzhtUlcmRSXl4usrKyxA8//CAWLFggjhw54qzhdJkceWzevFkcO3ZMVFRUCL1eL7Zv3y7mzp0rbty44axhdYkcmVy8eFGcO3dO6HQ6YTAYxNGjR0ViYqL47bffnDSqrpEjk3a1tbUiJSVFrFu3TmRkZMg9FIeRI5PTp0+L119/XTQ0NFg9ugM58mhpaRHvvvuu+Oijj8Tly5dFbW2tuHz5sqioqHDSqLpGjkyampqsjo3r16+LxMREcfr06Q7VxukZRP/nyJEjGDt2LMaNG2f5hBsQEIATJ07Y3P7EiRMICAjAvHnzEBYWhnHjxuGFF15AYWGhZZujR48iLi4O06dPR2hoKKZPn47HH38cR48eddawukSOTAYNGoTk5GQ8++yz3e56o3LksXTpUkyYMAEREREIDQ3FokWLIIRASUmJs4bVJXJkEhsbi1GjRiEsLAwqlQqTJk1CeHg4SktLnTWsLpEjEwBoa2vDli1bMHPmTAQFBTljKA4jVyZubm7o06eP1aM7kCMPtVqNW7duYeXKlRg8eDACAwMxePBgREREOGlUXSNHJj4+PlbHRnFxMXr06IFnnnmmQ7WxaSa6h9lsxrVr1zB8+HCr5XFxcbhy5YrNfcrLyxEXF2e17IknnsC1a9dgNpsBAGVlZZJthg8fjrKyMgdWLw+5MumunJXH3bt3YTab4ePj45jCZeSMTNo/QFRVVWHo0KGOK14mcmZy6NAh+Pr6YuzYsY4vXEZyZnLnzh2kpqZi0aJFWL9+PSoqKhw/AAeTK49ff/0VUVFR2L17NxYsWIAVK1YgNzcXbW1t8gzEgZz1+qpWqxEfHw8vL68O1cc5zUT3aGxsRFtbG3r37m21vHfv3jCZTDb3MZlMNrdvbW1FU1MT/Pz8YDKZJGc++vTpc9/f6UrkyqS7clYeBw4cgL+/P4YNG+aw2uUiZyZ//vknUlJSYDaboVAo8Oabb0reIF2RXJmUlpZCrVYjMzNTrtJlI1cmISEhSE1NRf/+/dHc3IxvvvkGa9euxYYNGxAcHCzXcLpMrjxqampgNBoxZswYpKWlwWAwYPfu3Whra0NCQoJcw3EIZ7y+Xr16FTqdDosXL+5wfWyaiWxwc3P7V8vut078faPNf9pHCPGP612NMzLpTuTM4/Dhw9BoNEhPT+8WX2RqJ0cmXl5e2LBhA+7cuYOSkhJkZWWhb9++iI2NdVDV8nJkJs3Nzdi6dStSUlLg6+vr2EKdyNHHSXR0NKKjoy3rY2JisHr1anz77bd44403HFGyrBydhxACvr6+SElJgUKhQGRkJBoaGlBQUODyTXM7OV9f1Wo1+vXr16kvWbNpJrqHr68vFAqF5BPtzZs3JZ9k29k6Y9zY2Ah3d3fLv9ZtbfNPv9OVyJVJdyV3HgUFBcjLy8PatWsRHh7uyNJlI2cmCoUCKpUKABAREYHKykrk5+e7fNMsRyZ6vR5GoxEZGRmW9e3NQVJSEj7++GNLVq7IWa8lCoUCAwcORHV1tSPKlo2c7zdKpRIKxf9m4IaGhsJkMsFsNkOpdN3WT+5j5O7du9BoNEhMTOxUfZzTTHQPpVKJyMhIFBcXWy0vLi5GTEyMzX2ioqIk21+4cAGRkZGWF6fo6GjJF7qKi4utzo64Krky6a7kzKOgoAA5OTl47733MHDgQMcXLxNnHiNCCLS0tHS9aJnJkUlISAg2btyIzMxMy+Opp55CbGwsMjMzERAQINt4HMFZx4kQAn/88YfLfxlQrjxiYmJQXV1tNYfZYDDAz8/P5V9/5T5Gzp49C7PZjOeee65T9bFpJvo/U6ZMwalTp6BWq6HX67Fv3z7U1dXhpZdeAgB88cUX2LZtm2X78ePHo66uznKNSLVaDbVajalTp1q2mTRpEi5cuID8/HzLmbKSkpJuc11iOTIxm83QarXQarUwm82or6+HVqt1+bNDgDx5HD58GAcPHsTixYsRFBQEk8kEk8mEO3fuOH18nSFHJnl5eSguLkZNTQ0qKytx5MgRnDlzptNveM7m6Ew8PT3Rv39/q8djjz0GLy8v9O/f3+UbIkCe4+Trr7/G77//jpqaGmi1WuzYsQNarfa+17d2JXLkMX78eDQ1NWHfvn2oqqpCUVER8vLyMGHCBKePrzPkyKSdWq3G008/jV69enWqNtf/CyNysvj4eDQ1NSEnJwcNDQ3o168f0tLSEBgYCABoaGhAXV2dZfugoCCkpaVh//79OH78OPz8/DB//nyrS9nExMRg2bJlOHjwILKzs6FSqbBs2TJERUU5fXydIUcm9fX1VjdkKCwsRGFhIYYOHYr09HSnja0z5MjjxIkTMJvN2LRpk9VzJSQkYObMmc4ZWBfIkcndu3exa9cu3LhxA56enggNDcWSJUsQHx/v9PF1hhyZdHdyZHL79m3s3LkTJpMJPXv2xIABA/DBBx90ixsDyZFHQEAA1qxZg/3792PlypXw9/fHyy+/jFdffdXZw+sUuf5uqqqqUFpaijVr1nS6NjfRPiGKiIiIiIhs4vQMIiIiIiI72DQTEREREdnBppmIiIiIyA42zUREREREdrBpJiIiIiKyg00zEREREZEdbJqJiIiIiOxg00xEREREZAebZiIiIiIiO9g0ExERERHZwaaZiIiIiMgONs1ERERERHb8B3Sgmk5VRpGRAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "feature_names = pipeline[:-1].get_feature_names_out()\n",
    "feature_importances = pipeline[-1].feature_importances_\n",
    "fi = pd.DataFrame({'feature':feature_names, 'weight': feature_importances})\n",
    "fi.feature = fi.feature.str[5:]\n",
    "\n",
    "fi.sort_values(by='weight', ascending=False) \\\n",
    "  .head(20).sort_values(by='weight',ascending=True) \\\n",
    "  .plot(kind='barh', y='weight', x='feature', legend=False)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "35033301",
   "metadata": {},
   "source": [
    "## Target correlation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "5db172dd",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "target_corr = movies.corrwith(movies.target)[:-1].round(2).sort_values(ascending=False)\n",
    "plt.figure(figsize=(12,8))\n",
    "target_corr.plot(kind='bar')\n",
    "plt.xticks(rotation=90,size='small')\n",
    "plt.title('Корреляция таргета со всеми числовыми фичами')\n",
    "plt.xticks(range(0,len(target_corr)), labels=target_corr.index)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a825c3fb",
   "metadata": {},
   "source": [
    "## Feature engineering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "8cec0ea1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "oxford_words['length'] = oxford_words.word.str.len()\n",
    "oxford_words.groupby('level').length.median().plot(kind='bar')\n",
    "plt.title('Медианная длина слова в словаре Oxford')\n",
    "plt.ylabel('Количество символов')\n",
    "plt.xlabel('Уровень английского')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "dfe05954",
   "metadata": {},
   "outputs": [],
   "source": [
    "# sns.pairplot(movies[num + ['target']], corner=True, hue='target')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "bce26662",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjgAAAHECAYAAAA0zVAJAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAABMGElEQVR4nO3deVwU9f8H8NcuLHLJJSAiqKiQ4oWopWKimZi3eEvfPKtvXpVpXl/5gqmR+s3jm5oVpuaBIiJmVkICisfPGy0sRQEVUQEBAQ/Yhfn90YP9uu5yLCwszr6ej4ePnJnPzLx3Pku8nPnMjEQQBAFEREREIiLVdwFEREREusaAQ0RERKLDgENERESiw4BDREREosOAQ0RERKLDgENERESiw4BDREREosOAQ0RERKLDgENERESiw4BDREREosOAQ1qTSCSQSCQwMjLCrVu3ym3n6empbPvbb7/VYYVERGToGHCoWoyNjVFaWoqtW7dqXH7y5En8+eefMDY2ruPKiIiIGHComho3bgxvb29s3boVpaWlastDQ0Mhk8nQv39/PVRHRESGjgGHqu3dd9/F7du3ERMTozI/Pz8f+/btw7Bhw+Do6Fju+unp6Zg1axZatmyJBg0aoFGjRhg2bBjOnTtX7jrBwcHKy14v/mnRooVa+z59+kAikajNz8/PR5MmTSCRSDB58mSVZdevX8fChQvRtWtXODg4oEGDBmjevDnee+893L59W21b8fHxkEgkCA4O1ljz5MmTIZFIkJaWppyXlpamcd8vfs74+HiV+VFRUfjHP/4BDw8PWFhYwNLSEt7e3li3bh1KSko0buv06dMYMmQIHBwcYGRkVOkxq0hZ3eX9ebHeFz9PVfpN0/ECAIVCgU6dOkEikaBPnz4at1/e/lu0aKG2n0ePHmH16tV444034OLiAhMTEzg4OGDYsGE4depU1Q/KczVX9Keimrdv347OnTvDzMwMjo6OmDp1Ku7fv69xXzk5OVi0aBHatm0LMzMzWFtbo1+/foiOji63vm3btlVYW3nKfn40/dH03T1//jxGjhwJR0dH5c/N9OnTkZGRodKuqKgI/fr1g5GREfbv36+x1m3btqnMj4iIgJGREfr164eioiKVZU+ePEFISAi8vLyUPxc9evRAWFiYWo26/nkNDw8v9/tfWlqKtWvXwtvbGxYWFmrHsLwaSDd4/YCqLSAgAPPmzUNoaCgGDBignL979248fvwY7777Lvbs2aNx3YsXL8LPzw85OTkYMGAARo4ciezsbERFRaFXr144cOAABg0aVO6+J02apPILa926dVrVHhgYWO4vkMjISGzevBl9+/ZFz549YWJigj/++ANbtmzBjz/+iAsXLsDFxUWr/enKwoULIZVK8dprr6Fp06bIy8vD0aNHMWfOHJw9exa7d+9WaX/lyhX06dMHCoUCAwcORMeOHWFiYgJA+2P2vE6dOmHEiBHK6fj4eBw7dqzS9WrSb1999RWuXLmiRZUV+/PPP/Gvf/0LvXv3xuDBg2Fra4tbt27h4MGD+Pnnn/Hjjz9W+B3U5KOPPoKNjY3a/KVLl5a7ztq1axEdHY1x48bhrbfewokTJ7B161bEx8fjzJkzcHBwULa9desW+vTpg7S0NPTu3RsDBw5EYWEhfvrpJ7z11lvYvHkz3n///XL3NXz4cHh5eSmnt23bVuE4ujJBQUHKv+fl5WH9+vVqbQ4ePIgxY8ZAIpFg9OjRaNasGc6fP4/Nmzfj4MGDOHHiBFq2bAkAaNCgAaKiovDGG28gICAAhw8fxptvvlnu/n/77Te8/fbb8Pb2RlRUFBo0aKBSzxtvvIFLly6hS5cumDp1KkpLS3HkyBEEBAQgKSkJy5cvr/QzVkdhYSE++eSTcpcvXrwYK1euRKNGjfD2228r/1GVlpaG7du310pN9ByBSEsAhKZNmwqCIAgTJ04UTExMhKysLOXyLl26CM2aNRNKSkqESZMmCQCEmJgY5XK5XC60atVKMDU1FRISElS2fffuXcHZ2Vlo3Lix8PTpU7V9L1myRAAgxMfHq8xv3ry50Lx5c7X2vr6+wotf88TERMHIyEjo1q2bAECYNGmSyvL09HTh2bNnatv6+eefBalUKvzzn/9UmR8XFycAEIKCgtTWEQRBeQxSU1OV81JTUzXuu0xQUJAAQIiLi1OZf+PGDbW2JSUlwttvvy0AEE6fPq2ybP78+eXWVt4xq8iNGzcEAMLkyZOrVG+Zf/3rX1XuN03HKyMjQ2jYsKGyz3x9fbXav6b95OXlqXxvy6SlpQmNGzcWXnnlFY3b0kRTzc+rqGaZTCZcvHhRZdnHH38sABCmTp2qMt/X11eQSCRCeHi4yvzc3FyhU6dOgqmpqXDv3j21/YeGhgoAhG3btqltr6JfAz4+PmrLNX13CwoKBDs7O8HIyEg4efKkSvvPP/9cACC8+eabatvPysoS2rRpI1haWgpnz54VBEEQtm7dKgAQtm7dKgiCIJw5c0awtLQU2rZtq7G/yo79f/7zH5X5T58+FQYMGCBIJBKV46vLn9e5c+cKAISuXbtq/P45OjoKUqlUuHnzpsr8ymog3eAlKqqRd999F8XFxfjhhx8AAImJibhw4QKmTJkCqVTz1+vw4cO4efMmZs+ejV69eqksc3Z2xvz58/HgwQMcPXpUbd2nT58CgPIshLYEQcCMGTNgamqKL7/8UmObpk2bqvwLsczAgQPh6elZ4aWA2taqVSu1eVKpFHPmzAEAtdoeP34MAGjfvr1O9l/d4y+XywFA43Gtik8++QSPHz/Gpk2bqrW+JtbW1rC3t1eb37x5c4wZMwbXrl3TeElS19555x107txZZV5wcDCsra2xe/du5eWYy5cv49ixYxg9ejTGjBmj0t7GxgZLly7Fs2fP1C75ADXrt6r0WVRUFHJycjB+/Hj07NlTZdm8efPQvHlz/Pbbb2pni+zt7RETEwM7OzsMHDgQf/31l8ryP//8E4MGDYKdnR2io6PV+uvhw4fYuXMnunXrhrlz56osMzU1xcqVKyEIgtqZTV1ISkrC+vXrMWzYMAwePFhjm8ePH8Pc3Bxubm463z9VjpeoqEZef/11vPLKK9iyZQs++eQTfPfdd5BKpZg6dWq565w+fRrA39e1NV2DTk5OBgD89ddfav/jyM3NBQCYm5tXq97vv/8ep06dwsqVK+Hq6qqxjSAI2LVrF7Zt24bLly8jNzdXZXxLeb8k4uPjNX6exMTEcutJTEzUuE55Y0kePnyI1atX4+eff0ZKSooywJS5e/euyvRrr72GjRs3YsWKFXBzc0OHDh2qHQ6B6h///Px8AH//0tHW0aNHsWfPHkyfPh1du3atsO22bds0Hru8vDyNl45OnjyJ9evX4/Tp08jMzERxcbHK8rt376JZs2Za16wNX19ftXnW1tbw8vLCsWPH8Oeff8LLy0v5c5OXl6fxO5OVlQUAaiEBqFm/VaXPLl26BADo27ev2jKZTAZfX1/88MMPuHTpEpo3b66y3MXFBRMmTMDKlSvh5+eH9957DwBw+/ZtBAYG4uHDh3jvvfc0XhY+d+6c8mdT0zEpC9aajkl1fl6fN2PGDMhkMqxfv15tvFCZ1157DbGxsZg7dy7mzZunvERFdYMBh2ps2rRpmD9/PmJjY7F7927079+/wl8KDx8+BADs27evwu0WFhaqzUtPTwfw95kebeXk5GDhwoVo27Yt5syZoxYGynzyySdYt24dmjRpggEDBqBp06YwMzMDUPGYhWPHjlVpHMrzLl++jMuXL1epbV5eHrp164bU1FS8+uqrmDhxIuzs7GBsbKwcF/Hi4Mt//OMfOHnyJL799luN4eDFXzaVqe7xz87OBoAKB51rUlxcjFmzZsHBwQErVqyotH1F4xpeDDgHDhzA6NGjYWpqiv79+6NVq1awsLCAVCpVjil68XjWhsaNG2uc7+TkBODvwdDA/35uYmJi1Ab2P0+XPzfZ2dlV6rOyGstqflGTJk1U2j3v999/Vw7ETU5OVo5X+uyzz2Bubg5vb2+sWbMGb7/9ttqZyLJjcu7cuQpvTtB0TKrz81rmhx9+wPHjx7F8+fIKB+pv3rwZI0eOxNq1a7F27dpq7Yuqj5eoqMYmTZoEmUyGSZMmIS8vD9OmTauwvbW1NYC/ByUKglDun+cHNpa5fPkybG1tVQZeVtWiRYuQnZ2NjRs3QiaTaWyTmZmJ//73v2jfvj2uXbuGnTt3YuXKlQgODkZwcHCFp+uDgoI0fo5JkyaVu86kSZOq/NlDQ0ORmpqKoKAgnDlzBps2bcLy5csRHByMcePGady+RCLB5s2blb8QX3nlFQQFBSEoKEjZD9oo+9ftK6+8otV6N27cgJmZWbm/zMvz5Zdf4q+//sKqVatga2tbafu4uDiNx1NTkAsMDISJiQnOnz+PqKgofPnll/jss88QHBys9eeriQcPHmicXzYIvqyfyv67fv36Cn9uND2bKjExERKJBO7u7lWu69GjR8jOzq7S5ZWy2sobuH/v3j2VdmUUCgWmTJkCY2NjhIeH4+DBg8pnZxkbG+PgwYMIDw+HsbExJk+eDIVCoXG/c+bMqfCYxMXFqdVUnZ9X4O9/aHz66adwd3fHvHnzKmzr7u6Oy5cvY+LEiQCAcePGISgoqNJ9kG4w4FCNOTo6YsiQIUhPT4e9vT2GDx9eYfvu3bsDABISErTaz++//44HDx7g9ddf17rGs2fPIjQ0FAEBARpPo5dJSUlBaWkp/Pz80LBhQ5Vl6enpSElJ0XrfunLjxg0AwKhRo9SWVfQvUblcjsWLF6NBgwYIDw9XhjVNl2wq89tvv0EqlcLHx6fK6+Tn5yMpKQmdO3eGkZFRlde7ffs2li9fDh8fn1r5hXDjxg14enqibdu2KvNLS0tx4sQJne+vPJr67tGjR0hMTISpqamyvur+3OTk5ODSpUvo0KGDVn1edqt8ZZcFASjHEGm6PKhQKJTH09vbW2XZqlWrcOHCBYSEhKBVq1bo27cvli1bBgBYvnw5+vbti1atWuHzzz/HhQsXsHr1apX1X331VUilUq2PSU0sWbIEmZmZ2LBhQ5XGJ128eBFhYWHo1asXdu3aheDg4HIfD0G6xYBDOvHll1/iwIEDOHz4cKVjPIYPH45WrVph48aN+PnnnzW2OX36NJ48eaIyb82aNQCAt99+W+v6ZsyYAUtLS/znP/+psF3Z6eYTJ06ojLspLCzEe++9p/YvyLpUVtuL/xq9dOkSQkJCyl0vMDAQZ8+exerVq9GxY8dq7z8+Ph4XL17EgAED0KhRoyqvt2zZMsjlcrWBsZX55JNPUFRUhE2bNtXKuIUWLVogOTlZ5VKlIAhYunQprl69qvP9lWfHjh3KMSxlgoOD8ejRI0yYMEH5S7Rr1654/fXXERkZie+//17jtn7//XdkZmaqzPvqq68gl8u1+rl5+vSp8js1evToStuPGDECdnZ2CAsLw//93/+pLFu3bh1SUlLw5ptvqly6TkpKwmeffYZevXph1qxZyvllZ2efH1A8e/Zs+Pj4qPWNo6Mj3n77bZw/fx7Lli3T+PN58+ZNpKamVvGTV+zy5cv4+uuvMWbMGPj5+VXavrCwEBMmTICFhQV27dqlVcCnmuMYHNIJNze3Kt8pIJPJEBkZiQEDBmDw4MHo2bMnvLy8YG5ujjt37uDcuXNISUnBvXv3YG5ujsTERCxfvhz79++HsbExrly5ovYLKC8vD8DfvxhGjBih8qwPALhw4YJyXE1FnJycMH78eOzZswdeXl7w8/PDo0ePEBMTA1NTU3h5eVV5EKKuTZw4EatXr8acOXMQHx8Pd3d3JCcn46effsLIkSOxd+9etXXi4uKwevVqDB48GLNnz672vufPn4/NmzcD+PtOqBcHZ5b9y33btm1IS0vD5MmTceTIESxZsgTnz59Ht27dMH36dK32eeHCBXz88cc1CmUVmTNnDj744AN4e3tj1KhRkMlkOHnyJK5evYqhQ4fi0KFDtbLfFw0aNAg+Pj4YO3YsmjRpghMnTuDEiRNo0aIFvvjiC5W2u3fvxhtvvIFp06bhv//9L1577TXY2NggPT0dV65cwR9//IHTp0/D0dERaWlp+Oyzz5Tjku7du6fWb2UPswsODkafPn3Qp08frFu3DmvXrsXt27fx4Ycfqt3hpYmlpSW+//57jBkzBr6+vhgzZgyaNWuGCxcuIDo6Gk5OTvjmm2+U7UtKSjB16lRIpVJs2bKl3Dsuy0ilUnz//ffo1KkTpkyZglOnTinDwoYNG5CcnIx///vf2LFjB3r16oXGjRsjIyMDf/75J86dO4ewsDCd3MmUmJgIS0tL5T+2KjNr1izcuHED4eHhtT5YnTSolZvPSdTw3HNwKqPpOThlHjx4ICxYsEBo166dYGZmJlhYWAitW7cWRo0aJezYsUOQy+WCIPzvuRhV/VP2/AxB+N9zPjp27CgoFAqV/Zf3bIvHjx8LixcvFlq1aiU0aNBAcHFxEWbMmCFkZ2drfG5IXT4HJykpSRg6dKjg4OAgmJubC97e3sJ3332ncXvZ2dlC06ZNBScnJyEzM1NtH9o8B0eb41/2vJcVK1YInTp1ElasWCE8efJE43Yreg6Ok5OT8OjRI4216OI5OILw93erU6dOgrm5udCoUSNhxIgRwpUrVyrd3otq8hycuLg4Ydu2bcrn2Njb2wuTJ08WMjIyNG4rPz9fWLFiheDt7S1YWFgIpqamQosWLYRBgwYJ33zzjVBYWCgIwv++l1X9U/b9HTp0qODr6yvs3LlT4/4r+u6ePXtWGDFihGBvby/IZDLB1dVV+OCDD4S7d++qtPviiy8EAMLq1avVtvHic3Cet3r1agGA8MUXX6jMLyoqEr766iuhR48egpWVlWBiYiK4uroKb7zxhrB27VohOztb2bYmP68AhFWrVqmto+n7smfPHgGAMG3aNLX2fA5O3ZAIgiDoJCkR1ZJt27ZhypQpqMpXVSKRYOvWrbzGrWNVPa6TJ09GWlpaube50/8EBwdj6dKliIuLU3uNgy7Ex8ejb9++SE1NrfSVHC1atMDkyZP56gASFY7BISIiItHhGByq97y8vDTeNq1JUFCQ2vgbqrmqHtcRI0Yox0ORfrVo0QJBQUFVunPq448/5s8NiQ4vURER6UFtX6IiMnR6DzglJSXYt28fEhISkJeXB1tbW/Tp0wcjR44sd2R9UlKSxrfzrl27Fk2bNq3tkomIiKie0/slqoMHDyImJgYzZ86Ei4sLUlJSsGnTJpibm2PQoEEVrrtu3TqVd6tYWVnVdrlERET0EtB7wLl+/Tq6du2qfMKlo6MjTpw4gZs3b1a6rrW1NSwsLGq7RCIiInrJ6D3gtGnTBjExMcjIyICzszPS0tJw7dq1Kj2aff78+ZDL5XBxccHIkSPVXsRWRi6XK98qW0Ymk5X7PiIiIiJ6uel9DI4gCAgLC8PBgwchlUpRWlqK8ePHw9/fv9x1MjIycPXqVbRs2RIKhQLHjx9HTEwMgoKC4OnpqdY+PDwcERERymkfHx989NFHtfJ5iIiISP/0HnBOnjyJnTt34h//+AdcXV2RlpaGbdu2YeLEiVrdWfDFF19AIpFgwYIFastePIMjkUhgZmaG3Nxcvb5bqK5JJBLY29sjOzu7Sg/No5cb+9uwsL8Ni6H2t7GxMWxtbavWtpZrqdTOnTsxfPhw5duJmzVrhqysLERFRWkVcDw8PMp9o2x5l6MUCoXapSsxK3thoVwuN6gfCEPF/jYs7G/Dwv6unN6fZFxUVKR2O7hUKtW6w1JTU6v0QCsiIiISP72fwenSpQsiIyNhb28PFxcXpKWl4aeffkLfvn2VbXbv3o2cnBzMmjULAHD48GE4ODjA1dUVCoUCCQkJOHPmDObOnauvj0FERET1iN4DztSpU7F3716Ehobi0aNHsLOzQ//+/TF69Ghlm9zcXGRnZyunFQoFduzYgZycHJiYmMDV1RULFy5U3mpOREREhk3vg4z1KSsry+DG4DRp0gT37t3jNVsDwP42LOxvw2Ko/S2TyeDg4FCltnofg0NERESkaww4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6DDhEREQkOgw4REREJDoMOERERCQ6xvouoKSkBPv27UNCQgLy8vJga2uLPn36YOTIkZBKy89fV69exfbt25Geng5bW1sMGzYMfn5+dVg5ERER1Vd6DzgHDx5ETEwMZs6cCRcXF6SkpGDTpk0wNzfHoEGDNK6TmZmJkJAQ9OvXD7Nnz8a1a9cQGhoKKysrdO/evY4/AREREdU3eg84169fR9euXeHt7Q0AcHR0xIkTJ3Dz5s1y14mOjoa9vT0mT54MAHBxccHNmzdx6NAhBhwiIiLSf8Bp06YNYmJikJGRAWdnZ6SlpeHatWuYNGlSueskJyejY8eOKvO8vLwQFxcHhUIBY2PVjyWXyyGXy5XTEokEZmZmyr8birLPakif2ZCxvw0L+9uwsL8rp/eAM3z4cDx58gRz5syBVCpFaWkpxo8fj169epW7Tl5eHqytrVXmWVtbo6SkBAUFBbC1tVVZduDAAURERCin3dzcsHLlSjg4OOj2w7wknJyc9F0C1SH2t2Fhf9etO4O76m/fetsz4Hr4vB73XjV6DzinTp1CQkICPvzwQ7i6uiItLQ3btm1TDjYuz4upVRAEjfMBwN/fH0OGDFFbNysrCwqFQgef4uUgkUjg5OSE+/fvK48XiRf727Cwv6ku3bt3Ty/7NTY2rvLJCb0HnJ07d2L48OHw8fEBADRr1gxZWVmIiooqN+DY2NggLy9PZV5+fj6MjIxgaWmp1l4mk0Emk2ncliH+j0AQBIP83IaK/W1Y2N9UF16G75jen4NTVFSkdju4VCqt8OC5u7vjypUrKvMuX76Mli1bqo2/ISIiIsOj94DTpUsXREZG4uLFi8jMzMTZs2fx008/oVu3bso2u3fvxoYNG5TTfn5+yM7OVj4HJzY2FrGxsRg6dKg+PgIRERHVM3o/3TF16lTs3bsXoaGhePToEezs7NC/f3+MHj1a2SY3NxfZ2dnKaUdHRyxatAjbt2/HkSNHYGtriylTpvAWcSIiIgIASISX4UJaLcnKylK5fVzsJBIJmjRpgnv37r0U10+pZtjfhoX9rR8l7w3Tdwl6YfTdj3rZr0wmq/IgY71foiIiIiLSNQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh1jfRcwc+ZMZGVlqc338/PDu+++qzY/KSkJS5cuVZu/du1aNG3atFZqJCIiopeL3gNOSEgISktLldO3b9/G8uXL0aNHjwrXW7duHczNzZXTVlZWtVYjERERvVz0HnBeDCZRUVFo3LgxPD09K1zP2toaFhYWtVkaERERvaT0HnCep1AokJCQgMGDB0MikVTYdv78+ZDL5XBxccHIkSPRvn37OqqSiIiI6rt6FXDOnj2Lx48fo0+fPuW2sbW1xfvvv4+WLVtCoVDg+PHjWLZsGYKCgso96yOXyyGXy5XTEokEZmZmyr8birLPakif2ZCxvw0L+5vq0svwPatXAScuLg5eXl6ws7Mrt42zszOcnZ2V0x4eHsjOzsahQ4fKDTgHDhxARESEctrNzQ0rV66Eg4OD7op/iTg5Oem7BKpD7G/Dwv6uW3f0XYCeNGnSRN8lVKreBJysrCxcuXIF8+bN03pdDw8PJCQklLvc398fQ4YMUU6XJc+srCwoFArti31JSSQSODk54f79+xAEQd/lUC1jfxsW9jfVpXv37ullv8bGxlU+OVFvAk5cXBysra3h7e2t9bqpqamwsbEpd7lMJoNMJtO4zBD/RyAIgkF+bkPF/jYs7G+qCy/Dd6xeBJzS0lLEx8fD19cXRkZGKst2796NnJwczJo1CwBw+PBhODg4wNXVVTko+cyZM5g7d64+SiciIqJ6qF4EnN9//x3Z2dno27ev2rLc3FxkZ2crpxUKBXbs2IGcnByYmJjA1dUVCxcurNaZHyIiIhInifAynGeqJVlZWSp3V4mdRCJBkyZNcO/evZfi9CLVDPvbsLC/9aPkvWH6LkEvjL77US/7lclkVR6Dw3dRERERkejoJOAIgoD79+8b1B1JREREVH9pPQanqKgIO3fuREpKCjp06IBRo0Zh6dKlSE5ORsOGDREYGIjmzZvXRq1EREREVaL1GZxdu3YhOjoaCoUCP/30E9asWYOCggJMmjQJ5ubmCA8Pr406iYiIiKpM6zM458+fx6hRozB27FicP38eq1evxscff4wePXrAysoKP/zwQ23USURERFRlWp/Byc3NRceOHQFA+V9HR0flfwsKCnRYHhEREZH2tA44paWlMDb++8RP2X+ff8lbaWmpDssjIiIi0l61HvR38eJF3L17V/mshYsXL+LOnTt48OCBTosjIiIiqo5qBZz9+/erTO/bt08nxRARERHpgtYBZ8OGDbVRBxEREZHOaB1wqvqIZCIiIiJ9qfbLNu/fv48//vgDBQUFaNiwIdq3bw8nJydd1kZERERULVoHHEEQ8P333yMmJkblhW4SiQR+fn6YOnWqTgskIiIi0pbWAefw4cOIjo5G//790adPH9jZ2SEnJwfHjh1DdHQ0HB0dMWTIkNqolYiIiKhKtA44R48exVtvvYUpU6Yo59nZ2aF169aQSqU4evQoAw4RERHpldYP+svMzESXLl00LuvSpQsyMzNrXBQRERFRTWgdcMzNzZGVlaVxWVZWFszMzGpcFBEREVFNaB1wOnTogD179iAlJUVlflpaGsLDw9GpUyedFUdERERUHVqPwQkICMCSJUuwaNEiuLi4wNbWFrm5uUhPT4ednR0CAgJqo04iIiKiKtM64Njb22PVqlX46aefkJSUhMzMTDRs2BAjRozA4MGDYWVlVRt1EhEREVVZtR70Z2VlxTM1REREVG9pPQaHiIiIqL7T+gzOrFmzKlwukUjw1VdfVbsgIiIioprSOuBkZWWhc+fOHGtDRERE9Va1xuCMHj0arVu31nUtRERERDrBMThEREQkOtU6g5OamoqSkhLIZDJYWVnBzs4OUimzEhEREdUP1Qo4oaGhKtMmJibw8PDA8OHD0bFjR50URkRERFRdWgecoKAgAIBCoUBxcTHy8/Nx//59XLp0CZ9//jmWLFmC9u3b67xQIiIioqrSOuB4enpqnD9u3Dh88cUXOHDgAAMOERER6ZXOBs4YGRnB39+fbxMnIiIivavWGJzyeHp6lnuGh4iIiKiuaH0G58KFC/j11181Lvv1119x8eLFGhdFREREVBNaB5zIyEg8e/ZM47KioiIcOHCgxkURERER1YTWAScjIwNubm4al7m5uSE9Pb3GRRERERHVhNYBRy6XQ6FQlLusuLi4xkURERER1YTWAcfZ2RkXLlzQuOzChQtwdnaucVFERERENaF1wOnbty9iY2MRHh6OvLw8AEBeXh7Cw8MRGxuLvn376rpGIiIiIq1ofZv4W2+9hZs3b2L//v3Yv38/pFIpSktLAQCvv/46Bg0apPMiiYiIiLShdcCRSCSYNWsW+vXrh8TEROTn58PKygqdO3dGmzZtaqNGIiIiIq1U+0F/bdu2Rdu2bXVZCxEREZFO6OxVDURERET1hU5f1VAdM2fORFZWltp8Pz8/vPvuuxrXuXr1KrZv34709HTY2tpi2LBh8PPzq+1SiYiI6CWh94ATEhKiHKQMALdv38by5cvRo0cPje0zMzMREhKCfv36Yfbs2bh27RpCQ0NhZWWF7t2711XZREREVI/pPeBYWVmpTEdFRaFx48blvrQzOjoa9vb2mDx5MgDAxcUFN2/exKFDhxhwiIiICEA9G4OjUCiQkJCAvn37QiKRaGyTnJyMjh07qszz8vJCSkpKuU9YJiIiIsNS7TM4T58+xfXr11FQUAArKyu4u7vDzMysRsWcPXsWjx8/Rp8+fcptk5eXB2tra5V51tbWKCkpQUFBAWxtbdXWkcvlkMvlymmJRKKstbwgJUZln9WQPrMhY38bFvY31aWX4XtWrYDz448/IiIiAkVFRcp5DRo0wNixYzFkyJBqFxMXFwcvLy/Y2dlV2O7FAysIgsb5ZQ4cOICIiAjltJubG1auXAkHB4dq1/oyc3Jy0ncJVIfY33XrzuCu+tu33vYMuB4+r8e9648+j7k+NWnSRN8lVErrgHPs2DHs2rULXl5e6NOnD2xtbZGbm4tjx45hx44dsLKyQu/evbUuJCsrC1euXMG8efMqbGdjY6N8RUSZ/Px8GBkZwdLSUuM6/v7+KsGrLAhlZWUZ1GUtiUQCJycn3L9/XxkKSbzY31SX7t27p+8SqA7pq7+NjY2rfHJC64Bz+PBh+Pj44MMPP1SZ36NHD/z3v//F4cOHqxVw4uLiYG1tDW9v7wrbubu7q73s8/Lly2jZsiWMjTV/HJlMBplMpnGZIf6PXxAEg/zchor9TXWB3zHD8jL0t9aDjO/evVtugOnduzfS09O1LqK0tBTx8fHw9fWFkZGRyrLdu3djw4YNymk/Pz9kZ2crn4MTGxuL2NhYDB06VOv9EhERkThpfQbHxMQEhYWFGpcVFhbCxMRE6yJ+//13ZGdna3wTeW5uLrKzs5XTjo6OWLRoEbZv344jR47A1tYWU6ZM4S3iREREpKR1wGnbti327dsHT09PlcHAeXl5iIiIqNb7qTp16oTw8HCNy2bOnKk2z9PTEytXrtR6P0RERGQYtA44EyZMwJIlS/Dhhx+iffv2ykHGSUlJMDIyqnSQMBEREVFt0zrguLq6IiQkBOHh4UhKSkJhYSEsLS3RrVs3jB49Gs7OzrVRJxEREVGVVes5OM7Ozvj44491XAoRERGRbujsVQ2G9DwZIiIiqt+0PoMjCAIOHTqElJQUdOzYEX379sW6detw5swZuLi4YMGCBQb7hGAiIiKqH7QOOPv27cP+/fthbm6O06dPIzk5GVevXsVbb72FkydPYs+ePZg9e3Zt1EpERERUJVoHnISEBPj5+WHatGmIjY3FN998gxkzZsDX1xfNmjUr93ZvIiIiorqi9RicnJwc9OjRAwDQs2dPAEDTpk0B/D34+NGjRzosj4iIiEh7WgcchUKhfFpx2X+l0r83Y2RkhNLSUh2WR0RERKS9at0mnpqaiuLiYmWYSU1NxbNnz6r1HioiIiIiXatWwAkNDVWZ/vbbb3VSDBEREZEuaB1wgoKCaqMOIiIiIp3ROuB4enrWRh1EREREOqP1IOOlS5fi7t27tVELERERkU5oHXCuXr2Kp0+f1kYtRERERDqhs3dREREREdUX1bqLKi8vD9nZ2eUut7e3r3ZBRERERDVVrYCzevXqCpfv3bu3WsUQERER6UK1As7IkSPRuHFjXddCREREpBPVCjhdunRB69atdV0LERERkU5wkDERERGJDgMOERERiY7Wl6g4gJiIiIjqO63P4Fy/fh2nTp3SuOzUqVNITk6ucVFERERENaF1wAkLC8Pt27c1LktPT8eePXtqXBQRERFRTWgdcG7fvg0PDw+Ny9zd3XHr1q0aF0VERERUE1oHnGfPnkEq1byaRCLhe6qIiIhI77QOOI6OjkhKStK4LCkpCQ4ODjUuioiIiKgmtA44Pj4+OHz4MOLi4lTmx8fH4+eff4aPj4/OiiMiIiKqDq1vEx8xYgSSkpKwefNmfP/997C1tUVubi6Ki4vRrl07+Pv710adRERERFWmdcAxNjZGYGAgTpw4gUuXLqGgoACtW7eGl5cXevXqVe74HCIiIqK6Uq13UUmlUvTu3Ru9e/fWdT1ERERENVatgAMAd+/exdWrV1FQUIA33ngDNjY2yMnJgaWlJUxMTHRZIxEREZFWtA44paWl+OabbxAfH6+c5+XlBRsbG3z77bdwc3PDuHHjdFkjERERkVa0HjATGRmJEydO4J133sGXX36psqxz585ITEzUVW1ERERE1aL1GZz4+HiMGjUKQ4YMQWlpqcoyR0dHZGZm6qw4IiIiourQ+gxOTk5Oua9qkMlkePbsWY2LIiIiIqoJrQOOtbV1uWdpMjIyYGdnV+OiiIiIiGpC64DTuXNnREZGIicnRzlPIpHgyZMn+OWXX9ClSxedFkhERESkLa3H4IwdOxaXLl3CnDlz0K5dOwBAWFgY7ty5AyMjI4wePVrnRRIRERFpQ+szODY2NggJCYGPjw9SU1MhlUpx69YteHl5Yfny5bC0tKyNOomIiIiqrFoP+rOxscH777+v61qIiIiIdKLaTzLWpZycHOzcuROJiYkoLi5GkyZNMH36dLRs2VJj+6SkJCxdulRt/tq1a9G0adPaLpeIiIjqOa0DzqZNmypcLpFIMH369Cpvr7CwEIGBgWjXrh0WL14MKysrPHjwAObm5pWuu27dOpV2VlZWVd4vERERiZfWAScpKUllOjs7GzY2NjA2/ntTEolEq+0dPHgQjRo1wowZM5TzHB0dq7SutbU1LCwstNofERERiZ/WAWfjxo3Kv5eUlCAgIAALFiwo93JSZc6fP49OnTphzZo1uHr1Kuzs7ODn54c333yz0nXnz58PuVwOFxcXjBw5Eu3bt9fYTi6XQy6XK6clEgnMzMyUfzcUZZ/VkD6zIWN/U13i98ywvAz9XaMxOLr4gJmZmYiJicHgwYPh7++PGzduYOvWrZDJZPD19dW4jq2tLd5//320bNkSCoUCx48fx7JlyxAUFARPT0+19gcOHEBERIRy2s3NDStXroSDg0ON638ZOTk56bsEqkPs77p1R98F6EmTJk30XYJesL/rrxoFnOLiYgCAVKr13eZKpaWlaNWqFQICAgD8HT7u3LmD6OjocgOOs7MznJ2dldMeHh7Izs7GoUOHNAYcf39/DBkyRDldFsyysrKgUCiqXfvLRiKRwMnJCffv34cgCPouh2oZ+5vq0r179/RdAtUhffW3sbFxlU9OVDvgPHv2DFFRUZBKpVUeM6OJra0tXFxcVOa5uLjgzJkzWm3Hw8MDCQkJGpfJZDLIZDKNywzxf/yCIBjk5zZU7G+qC/yOGZaXob+1Djjjxo1TmR40aFCV7ngqzyuvvIKMjAyVeRkZGVpfPkpNTYWNjU216yAiIiLx0DrgjBo1ChKJBA0aNECLFi3QsWPHGhUwePBgBAYGIjIyEj179sSNGzdw9OhRlQcJ7t69Gzk5OZg1axYA4PDhw3BwcICrqysUCgUSEhJw5swZzJ07t0a1EBERkThU611UutS6dWvMmzcPu3fvxv79++Ho6IhJkybh9ddfV7bJzc1Fdna2clqhUGDHjh3IycmBiYkJXF1dsXDhQnh7e+u0NiIiIno51YsnGXfp0qXCt5DPnDlTZXr48OEYPnx4bZdFRERELymtA46mVyQ8TyKR4N///ne1CyIiIiKqKa3v77569Sr++uuvckdQvwwjq4mIiEjctD6DM2bMGBw8eBDPnj3DtGnT4O7uXht1EREREVWb1mdwRo8ejbVr16JRo0YIDAzE5s2bUVhYWBu1EREREVVLtR5BbG9vj08//RQLFy7En3/+iQ8//BDR0dG8PEVERET1QvXfsQDAy8sLX375JYYMGYIdO3Zg8eLFuHHjhq5qIyIiIqoWrcfgPP/Syuf17NkTx48fR2BgIMLCwmpcGBEREVF1aR1w9u3bVxt1EBEREemM1gFn7969tVEHERERkc7UaAwOERERUX2k9Rmc598JVR57e/tqFUNERESkC1oHnBffC6UJL2MRERGRPlXrZZsjR45E48aNdV0LERERkU5UK+B06dIFrVu31nUtRERERDrBQcZEREQkOtU6gxMbG4srV67AxMQEDRs2hJOTE1q2bAmZTKbr+oiIiIi0Vq2Ac/ToUbV5pqamGDRoEMaNG1fjooiIiIhqotoP+lMoFCguLkZBQQHu3buHs2fPIjIyEnZ2dujfv7/OCyUiIiKqqmqdwQEAY2NjGBsbw9zcHI0bN4aXlxekUil+++03BhwiIiLSq2oHHE2GDBmCs2fP6nKTRERERFrT6V1UTk5OGDZsmC43SURERKS1ap3BuX//Pvbt24c//vgDBQUFaNiwITp06IDRo0fDyclJ1zUSERERaUXrgHP37l0sWbIExcXFaN++PWxtbZGbm4vTp0/j4sWLWLZsGZo2bVobtRIRERFVidYBJywsDJaWlggODkajRo2U8x8+fIjPPvsMYWFhmDdvnk6LJCIiItKG1mNwrl69irFjx6qEGwBo1KgRRo0ahaSkJJ0VR0RERFQdWgecoqIiNGzYUOMyKysrFBcX17goIiIioprQOuA4OzsjISFB47ITJ05w/A0RERHpndZjcAYOHIhvvvkGT548ga+vr3KQcUJCAs6fP48PPvigNuokIiIiqjKtA84bb7yBR48eITIyEhcvXlTONzExwYQJE9C3b1+dFkhERESkrWo9B8ff3x8DBgzA9evXlc/B8fDwgLm5ua7rIyIiItJalQJOaWkppFLV4Trm5ubw8vJSa5uZmQlHR0edFEdERERUHVUaZLxu3TqUlpZW2i46OhqffvppjYsiIiIiqokqBZwzZ85UGHJycnKwYsUKbNmyBZ6enjotkIiIiEhbVQo4U6ZMwZkzZ7B27Vq1kHP8+HHMnTsXycnJ+OCDD7BgwYJaKZSIiIioqqo0Buett96CVCrFli1bsGbNGnzyyScoLCzEt99+i3PnzqF9+/aYPn067O3ta7teIiIiokpV+S4qPz8/SKVSfPfdd1i2bBnS09Px7NkzTJ48GQMHDqzNGomIiIi0otVt4m+++SYkEgm+/fZbyGQyhISEwNXVtbZqIyIiIqoWrV/V0K9fP3zwwQdQKBTYtWsXFApFbdRFREREVG1VOoOTkpKiMt28eXMMGDAAv/76K1avXo1x48apLG/ZsqXuKiQiIiLSUpUCzqJFi8pdlpiYiMTERJV5e/furVFRRERERDVRpYAzffr02q6DiIiISGeqFHD69OlTy2UQERER6U61Xrapazk5Odi5cycSExNRXFyMJk2aYPr06RWO5bl69Sq2b9+O9PR02NraYtiwYfDz86vDqomIiKi+0nvAKSwsRGBgINq1a4fFixfDysoKDx48qPDN5JmZmQgJCUG/fv0we/ZsXLt2DaGhobCyskL37t3rsHoiIiKqj/QecA4ePIhGjRphxowZynmVvY08Ojoa9vb2mDx5MgDAxcUFN2/exKFDhxhwiIiISP8B5/z58+jUqRPWrFmDq1evws7ODn5+fnjzzTfLXSc5ORkdO3ZUmefl5YW4uDgoFAoYG6t+LLlcDrlcrpyWSCQwMzNT/t1QlH1WQ/rMhoz9TXWJ3zPD8jL0t94DTmZmJmJiYjB48GD4+/vjxo0b2Lp1K2QyGXx9fTWuk5eXB2tra5V51tbWKCkpQUFBAWxtbVWWHThwABEREcppNzc3rFy5Eg4ODrr/QFVwZ3BXvewXAO7obc+A6+Hzety7/rC/DYs+j7k+NWnSRN8l6AX7u/7Se8ApLS1Fq1atEBAQAODv8HHnzh1ER0eXG3AA9fQoCILG+QDg7++PIUOGqK2blZXFJzHXoXv37um7BKpD7G/Dwv42LPrqb2Nj4yqfnNB7wLG1tYWLi4vKPBcXF5w5c6bcdWxsbJCXl6cyLz8/H0ZGRrC0tFRrL5PJIJPJNG6rLBhR7eOxNizsb8PC/jYsL0N/a/0uKl175ZVXkJGRoTIvIyOjwoTm7u6OK1euqMy7fPkyWrZsqTb+hoiIiAyP3gPO4MGDkZycjMjISNy/fx8nTpzA0aNHMWDAAGWb3bt3Y8OGDcppPz8/ZGdnK5+DExsbi9jYWAwdOlQfH4GIiIjqGb2f7mjdujXmzZuH3bt3Y//+/XB0dMSkSZPw+uuvK9vk5uYiOztbOe3o6IhFixZh+/btOHLkCGxtbTFlyhTeIk5EREQA6kHAAYAuXbqgS5cu5S6fOXOm2jxPT0+sXLmyNssiIiKil5TeL1ERERER6RoDDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJjrG+CwgPD0dERITKPGtra3z33Xca2yclJWHp0qVq89euXYumTZvWSo1ERET0ctF7wAEAV1dXBAYGKqel0spPLK1btw7m5ubKaSsrq1qpjYiIiF4+9SLgSKVS2NjYaLWOtbU1LCwsaqcgIiIieqnVi4Bz//59/POf/4SxsTHc3d0xYcIENG7cuMJ15s+fD7lcDhcXF4wcORLt27cvt61cLodcLldOSyQSmJmZKf9OdYPH2rCwvw0L+9uwvAz9LREEQdBnAZcuXUJRURGcnZ2Rl5eHyMhI3L17F2vWrEHDhg3V2mdkZODq1ato2bIlFAoFjh8/jpiYGAQFBcHT01PjPl4c5+Pm5oaVK1fW2meqzJ3BXfW2b31yPXxe3yXoBfvbsLC/DQv7u/7S+xmczp07K//erFkzeHh4YPbs2Th27BiGDBmi1t7Z2RnOzs7KaQ8PD2RnZ+PQoUPlBhx/f3+VbZUlz6ysLCgUCl19FKrEvXv39F0C1SH2t2FhfxsWffW3sbExHBwcqta2lmvRmqmpKZo1a6bVwfPw8EBCQkK5y2UyGWQymcZlej6BZVB4rA0L+9uwsL8Ny8vQ3/XuOThyuRx3796Fra1tlddJTU3VepAyERERiZfez+D88MMP6Nq1K+zt7fHo0SPs378fT58+ha+vLwBg9+7dyMnJwaxZswAAhw8fhoODA1xdXaFQKJCQkIAzZ85g7ty5+vwYREREVI/oPeDk5ORg/fr1yM/Ph5WVFdzd3bFixQrlNbbc3FxkZ2cr2ysUCuzYsQM5OTkwMTGBq6srFi5cCG9vb319BCIiIqpn9B5wPv744wqXz5w5U2V6+PDhGD58eC1WRERERC+7ejcGh4iIiKimGHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdBhwiIiISHQYcIiIiEh0GHCIiIhIdIz1XUB4eDgiIiJU5llbW+O7774rd52rV69i+/btSE9Ph62tLYYNGwY/P7/aLpWIiIheEnoPOADg6uqKwMBA5bRUWv6JpczMTISEhKBfv36YPXs2rl27htDQUFhZWaF79+51US4RERHVc/Ui4EilUtjY2FSpbXR0NOzt7TF58mQAgIuLC27evIlDhw4x4BARERGAehJw7t+/j3/+858wNjaGu7s7JkyYgMaNG2tsm5ycjI4dO6rM8/LyQlxcHBQKBYyN68VHIiIiIj3Sexpwd3fHzJkz4ezsjLy8PERGRmLJkiVYs2YNGjZsqNY+Ly8P1tbWKvOsra1RUlKCgoIC2Nraqq0jl8shl8uV0xKJBGZmZnoLQ9JWr+hlv/pmJJPpuwS9YH8bFva3YWF/1y1tfm/rPeB07txZ+fdmzZrBw8MDs2fPxrFjxzBkyBCN60gkEpVpQRA0zi9z4MABlYHMPj4++OijjzSGoTrx31362S/pB/vbsLC/DQv7u96qd7eJm5qaolmzZrh3757G5TY2NsjLy1OZl5+fDyMjI1haWmpcx9/fH9u2bVP+ee+991TO6BiKp0+fYsGCBXj69Km+S6E6wP42LOxvw8L+rpzez+C8SC6X4+7du2jbtq3G5e7u7rhw4YLKvMuXL6Nly5blnrqSyWSQGejp0+cJgoDU1FTlGS8SN/a3YWF/Gxb2d+X0fgbnhx9+wNWrV5GZmYnk5GR8+eWXePr0KXx9fQEAu3fvxoYNG5Tt/fz8kJ2drXwOTmxsLGJjYzF06FB9fQQiIiKqZ/R+BicnJwfr169Hfn4+rKys4O7ujhUrVsDBwQEAkJubi+zsbGV7R0dHLFq0CNu3b8eRI0dga2uLKVOm8BZxIiIiUpIIPL9lMORyOQ4cOAB/f39esjMA7G/Dwv42LOzvyjHgEBERkejofQwOERERka4x4BAREZHoMOAQERGR6DDgEBERkegw4BAREb2EeI9QxRhwiIiIXkIBAQFIT0/Xdxn1lt4f9Ef6k52djfDwcMyYMUPfpZAOFBcXIyUlBZaWlnBxcVFbdvr0aeUTwunll56ejuTkZHh4eKBp06a4e/cufv75Z8jlcvTu3Rvt27fXd4mkI9u3b9c4v7S0FFFRUWjYsCEAYNKkSXVZVr3HgGPACgsLcezYMQYcEcjIyMCKFSuUT/1u27YtPvroI9ja2gIAnjx5gk2bNjHgiERiYiJWrVoFU1NTFBUV4dNPP8WGDRvQvHlzAMCKFSvwr3/9iyFHJH7++Wc0b94cFhYWasvu3r0LU1NTPVRV/zHgiNj58+crXP7gwYM6qoRq265du+Dq6oqQkBA8efIE27dvR2BgIIKDg2Fvb6/v8kjHIiIiMGzYMIwfPx4nT57E+vXr4efnhwkTJgAAwsLCEBUVxYAjEuPHj8fRo0cxceJElT6dMGECZs6cqXbGlv7GgCNiq1ev1ncJVEeuX7+OwMBAWFlZwcrKCgsWLEBoaCj+/e9/IygoCA0aNNB3iaRDd+7cwaxZswAAPXr0wIYNG/Daa68pl/v4+CA2NlZf5ZGO+fv7o0OHDvjqq6/QpUsXBAQEwNiYv74rwyMkYjY2Npg2bRpeffVVjcvT0tKwYMGCOq6KakNxcTGkUtV7Bt599118//33CA4Oxocffqinyqi2SaVSyGQylcsXZmZmePLkiR6rIl1r3bo1Vq5cidDQUCxatAizZ8/Wd0n1Hu+iErGWLVsiNTVV32VQHXB2dkZKSora/KlTp6Jr165YtWqVHqqi2uLo6Ij79+8rp5cvX65yKfLhw4fK8VckHqamppg1axZGjBiBZcuWobS0VN8l1Ws8gyNiw4YNQ1FRUbnLnZycEBQUVIcVUW159dVXcfLkSfTu3Vtt2bRp0yAIAmJiYvRQGdWG/v37q/xya9asmcryS5cucfyNiPn4+KBNmzZISUnhGLsK8G3iREREJDq8REVERESiw4BDREREosOAQ0RERKLDgENUyw4ePIixY8dqfPBiXl4eJkyYwLucDMSaNWuwd+9ePHv2DBkZGZg+fTrfJURUSxhwiGpZv379YGJigl9++UVtWXR0NEpKSjBw4EA9VEZ1bejQoYiJicHEiRPx8ccfo127dnwKLVEt4W3iRLXM0tISvXr1QmxsLNLT05W/0BQKBWJiYuDq6ooOHTrouUqqC+7u7vj666/x4MEDWFhY8Fk1RLWIAYeoDgwcOBCxsbH49ddf8e677wIATp06hUePHmHs2LHKdvHx8di0aROWLFmCEydO4Ny5c1AoFGjXrh2mTJmCxo0bq2w3NjYWv/zyCzIyMmBiYgJPT09MmDBB5azAxo0bcezYMeW0mZkZmjVrhtGjR6NTp04q27ty5QqioqJw8+ZNlJSUwM3NDWPHjlUJYOHh4YiIiEBoaCisrKyU82/evIlFixZhxowZ6NOnj3Lf//d//4cdO3Zofczu37+PyMhI/PXXX8jJyYGFhQXc3NwQEBCg8tyXpKQkLF26FJ988gm6d++uso133nkH3bt3x8yZM1WOryYbNmyAo6OjcnrmzJnIyspSaxcUFIR27dop23h6eiq3DwDHjx/Hhg0b4ODggI0bNwIAMjMzMWvWLOWxKeufr7/+GnFxcfD19VXZRk5ODvbt24fExETk5eXBysoKHh4emDZtGmxsbJSf+fla7t+/j6VLl8LBwQGLFy9WvoCxsLAQe/bswblz55Cfn49GjRrBx8cHo0ePhkwmU+7z+e/h80aPHq1cVtVtEdUHDDhEdaB58+Zo27Ytjh07hoCAAJibm+OXX36BhYWFxofzff311+jYsSM++ugjZGdnY+/evQgODsZ//vMf5SP5Dxw4gLCwMPj4+GDChAkoLCzEvn37sGTJEoSEhKBJkybK7dnY2GDevHkAgNzcXERERGDVqlVYv3698kFhx48fx8aNG9G1a1fMnDkTRkZGiImJUb6Zuq7PMuXk5MDS0hIBAQGwsrJCYWEhjh07hsWLF2PVqlVwdnau9rZnzJihtr6msymdO3fGqFGjAACpqanYsmVLhdt98uQJdu7cqfbaDE2Sk5MRHx+v1jYnJweLFi2CQqGAv78/mjdvjoKCAly+fBmPHz+GjY2N2rbKCzfFxcVYunQp7t+/j7Fjx6J58+b4888/ERUVhbS0NCxatEhlO927d8eQIUNU5jVq1Kha2yLSNwYcojoycOBArFmzBnFxcXB3d8fNmzcxdOhQjS/CbNWqFaZPn66cdnV1RWBgII4cOYKRI0fi8ePH2L9/Pzp37oyPPvpI2c7T0xMfffQR9u3bp/L+KZlMBg8PD+W0VCrF6tWrcefOHdjb26OoqAjbtm2Dt7c3Pv30U2W7zp07Y8GCBQgLC6vzgOPp6QlPT0/ldGlpKby9vTF37lzExMRg0qRJ1d62q6srWrVqVWEbhUIBW1tb5XGTy+WVbjc8PBxSqRTdunXT+OqMMqWlpdiyZQu6dOmCW7duqSzbu3cv8vPzsXr1apUzcT179tS4rQcPHmgMNwBw7Ngx3Lp1C3PmzEGPHj0AAB07doSpqSl27dqFK1euoGPHjsr21tbWKt+T52m7LSJ94yBjojrSrVs3NGrUCL/++isOHz4MqVSKAQMGaGzbq1cvlelXXnkFDg4OSEpKAvD328OLi4uVl4LK2Nvbo3379vj999/VtllSUoKSkhJkZ2cjLi4O5ubmaNmyJQDg2rVrKCwshK+vr7JdSUkJBEGAl5cXbt68iWfPnqlsr7S0VKVtRe/FKWujjZKSEkRGRmLOnDmYMGECxo8fj4CAANy7dw93795Va/9iPdru70XFxcVaXXa5ffs2jhw5gokTJ6qEDE1+++03pKenY/LkyWrLEhMT0b59+yoNPn7w4AGCg4NRUFCATz/9VG2/f/zxBxo0aKB26a7se6Ppe1IeXW6LqC7wDA5RHTEyMoKfnx/CwsLw4MEDdOvWTWXMx/M0XYawsbFBQUEBACj/q+myiq2tLQoLC1XmZWVlYcKECcppCwsLzJgxA9bW1gCAR48eAfj7NubyFBYWqvwCff/998tt+7yioiLlvmUyGezt7eHr6wt/f39IJJJy19u+fTuOHDmC4cOHw9PTE5aWlpBIJNi8eTOKi4vV2q9bt65K9VSFQqHAkydP0LBhwyqvs2XLFrRp0wY9e/bEpUuXym2Xn5+PPXv2YMSIERr7Pz8/H3Z2dlXaZ2hoKFq0aIHCwkIcOHAAEydOVFleWFgIGxsbteNsbW0NIyMj5feoKnS5LaK6wIBDVIf69euHiIgIyOXyCm8Nz8vL0zjPyckJAJS/eHNzc9Xa5ebmqv1itrW1xfz58wH8HTjOnTuHNWvWYN68eejatauy/dSpU+Hu7q6xphdDV2BgIMzNzZXTd+/exYYNG9TWMzExwdKlS5X7Pnv2LPbs2YMGDRpg8ODBGvcFAAkJCfD19UVAQIDK/IKCAuU4pOe9/fbbai+YrO7LZB88eABBEJTHuzIJCQm4fv06Vq9eXWnb3bt3w8LCAsOGDdO43MrKCjk5OVXab+vWrbF48WKcOHEC3333Hby8vFQuE1laWiI5ORmCIKgEk0ePHqGkpERlkHhldLktorrAS1REdcjc3BympqZo1qxZhW97PnHihMr0tWvXkJWVpRyT4uHhARMTEyQkJKi0e/jwIf744w+1bRsbG6NVq1Zo1aoVPD09MWnSJJiZmeHkyZMAgDZt2sDCwgLp6enKdi/+MTZW/fdQ8+bNVZY3bdpU42eRSCQq+548eTIsLCxw48aNCo+VRCJR2+fFixfL/eXfuHFjtZorOkNUkXPnzgEA2rZtW2nbp0+fYufOnRg0aFCll5Vu3LiBuLg4TJkyBSYmJhrbeHl54Y8//kBGRkal+x43bhxMTU3x5ptvokuXLti4cSPy8/OVyzt06IBnz54pP0+ZsrvqtHnjuC63RVQXeAaHqA5kZWXh2rVrOHfuHAoKCvDOO+9U2P7mzZvYvHkzunfvjocPH2LPnj2ws7NTjtmxsLDAqFGjEBYWhg0bNsDHxwcFBQWIiIiATCbDmDFjVLYnl8tx/fp1AP87g/P48WO4ubkBAExNTTFlyhRs3LgRhYWF6N69O6ysrJCfn49bt24hPz8f7733XrU/f9mYmaKiIpw/fx6PHz8u90xRGW9vbxw7dgxNmzZF8+bNkZKSgh9//FF5V09tyM3Nxa+//ooff/wRvXr1goODQ6XrnD9/HtbW1mrHXJPffvsNXbp0gbe3d7ltxo0bh8TERAQFBcHf3x/NmjXD48ePkZiYiCFDhpQbJD/44APMmzcPmzdvVp6t6927N44cOYKNGzciMzMTzZo1w19//YUDBw6gc+fOWg0K1uW2iOoCAw5RHUhKSsLmzZthY2ODoUOHwtfXt8L206dPx/Hjx7F+/XrI5XLlc3AsLS2Vbfz9/WFtbY1ffvkFp06dUj4HJyAgQOUWceDvy1tLliwBADRo0AD29vYYP368yi3BvXv3hr29PX788Ud8++23ePr0KaytrdGiRQu1wczaKCoqwpw5c9T2XdnTm6dMmQJjY2NERUXh2bNncHNzw7x587Bnz55q11KZpKQknD17FqNHj8bw4cOrtE5paWmVBhYDf4/D0jSw+Hl2dnb4/PPPER4ejqioKBQUFMDKygpt2rRR6f8XWVlZYcaMGQgJCUF0dDT8/PxgYmKCoKAghIWF4dChQ8rxPUOHDq1SIHueLrdFVBckgiAI+i6CiP5W9iC6kJCQSm9jJiKi8nEMDhEREYkOAw4RERGJDi9RERERkejwDA4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYkOAw4RERGJDgMOERERiQ4DDhEREYnO/wMbruE81WWKuwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "movies.groupby('target').sents_median.median().plot(kind='bar')\n",
    "plt.title('Медианная длина предложения')\n",
    "plt.ylim(5,8.5)\n",
    "plt.ylabel('Количество слов')\n",
    "plt.xlabel('Уровень английского')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "56ea1603",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "movies.groupby('target').len_equal_6ratio.median().plot(kind='bar')\n",
    "plt.title('Шестизнаки')\n",
    "plt.ylim(0.09,0.125)\n",
    "plt.ylabel('Доля от общего количества лемм')\n",
    "plt.xlabel('Уровень английского')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "77be4f21",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjUAAAHECAYAAADBM9u5AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8o6BhiAAAACXBIWXMAAA9hAAAPYQGoP6dpAABC0UlEQVR4nO3de1xUdf7H8ffgDCAgYoKipkJeUjATKzU1Uds0yyyqNTVX0/TXeulimaZlyG7WmtVqablpNzdLUcNsMy8lWqblJbU1rylYqIgohDducn5/9GDWifuIzHDm9Xw8eMg55zvnfGa+M/Hue75zjsUwDEMAAADVnJerCwAAAKgMhBoAAGAKhBoAAGAKhBoAAGAKhBoAAGAKhBoAAGAKhBoAAGAKhBoAAGAKhBoAAGAKhBoAAGAKhBpUCxaLRRaLRTVq1NCRI0dKbBcREWFv++WXX1ZhhQAAVyPUoNqwWq0qKCjQe++9V+z2b7/9Vnv37pXVaq3iygAA7oBQg2qjfv36at++vd577z0VFBQU2T5//nzZbDbddtttLqgOAOBqhBpUKyNGjNAvv/yitWvXOqzPysrSkiVL1K9fP9WrV6/Ex6ekpGjs2LG65ppr5OPjo7p166pfv37aunVriY+ZOnWq/ZTWH3/CwsKKtO/evbssFkuR9VlZWWrQoIEsFoseeughh20HDhzQM888oxtvvFEhISHy8fFR06ZNNXLkSP3yyy9F9rV+/XpZLBZNnTq12JofeughWSwWJScn29clJycXe+w/Ps/169c7rF++fLkGDx6sli1byt/fXwEBAWrfvr1mzpypixcvFruvzZs3q2/fvgoJCVGNGjXKfM1Ks2XLFj3wwANq1KiRfHx81KBBA/Xq1Uvx8fFlPrdz586pW7du8vLy0r/+9a8i+167dq169+6tq666Sr6+vmrRooUmTpyozMzMIm0L+zUnJ0fPPfecwsPD5ePjo2bNmikuLk65ubnF1l+RY0jS6dOn9eyzz6pNmzby8/NT7dq1df311+uZZ57RuXPnJKnE9+Mffy59f1gsFnXv3r3U17o8nH0d9u3bp4ceekiNGzeWj4+P6tevr0GDBmn//v1F2pZUa0nve2c+czAnxulRrQwaNEjjx4/X/Pnz1bt3b/v6jz76SOfOndOIESO0aNGiYh/7ww8/qFevXjp9+rR69+6te++9V+np6Vq+fLm6du2qhIQE3XHHHSUee+jQoQ5/kGfOnFmh2qdMmaLU1NRit33yySeaO3euevTooc6dO8vb21u7d+/WO++8oxUrVmj79u26+uqrK3S8yvLMM8/Iy8tLHTt2VKNGjZSZmamvvvpK48aN05YtW/TRRx85tP/xxx/VvXt35efnq0+fPmrbtq28vb0lVfw1mzdvnkaNGqUaNWqoX79+atGihdLS0rR161a9+eab6t+/f4mPvXDhgu666y598803mj17th555BGH7W+++abGjh0rf39/9e/fXyEhIUpMTNTLL7+sFStWaNOmTapTp06R/fbv319bt27V/fffL5vNpk8//VRTp07Vtm3btGLFCoc/rhU9RlJSknr06KEjR47ohhtu0KhRo1RQUKD9+/frn//8p/7617/K399fsbGxDjUtX75cu3bt0uOPP66goCD7+soIMSWpyOuwatUq3XvvvcrPz1ffvn3VvHlzpaSk6JNPPtHnn3+uxMREtW/fvtJrLO0zB5MygGpAktGoUSPDMAxjyJAhhre3t3Hy5En79htuuMFo0qSJcfHiRWPo0KGGJGPt2rX27Xl5eUazZs0MX19f45tvvnHY99GjR42GDRsa9evXNy5cuFDk2M8995whyVi/fr3D+qZNmxpNmzYt0j46Otr440dr586dRo0aNYybbrrJkGQMHTrUYXtKSoqRnZ1dZF8rV640vLy8jEceecRhfWJioiHJiI2NLfIYwzDsr0FSUpJ9XVJSUrHHLhQbG2tIMhITEx3W//zzz0XaXrx40XjwwQcNScbmzZsdtk2YMKHE2kp6zYrz008/GVar1ahTp46xe/fuItt/+eUX++9/fG4XLlwwbrvtNkOS8c9//rPIY5OSkgybzWYEBgYa+/fvd9j2yCOPGJKMESNGOKwv7NcWLVoYp0+ftq+/cOGC0alTJ0OSsWDBgss6RufOnQ1Jxosvvlik5pMnTxb7/jSM4vv7jyQZ0dHRJW4vr4q+DqdPnzaCgoKM4OBgY+/evQ772r17t+Hv72+0a9euSK3dunUrcuyS3vfOfOZgTpx+QrUzYsQI5ebmasGCBZKknTt3avv27Ro2bJi8vIp/S3/++ec6dOiQHn30UXXt2tVhW8OGDTVhwgSdOHFCX331VZHHXrhwQZLsow0VZRiGRo8eLV9fX7366qvFtik8tfJHffr0UUREhNasWePUsStDs2bNiqzz8vLSuHHjJKlIbYWnSNq0aXNZx33rrbeUn5+vKVOmKDIyssj2xo0bF/u43Nxc3XvvvVq7dq1mzJihJ554okibDz/8UHl5eXr00UfVsmVLh20vvviiAgIC9OGHHyonJ6fIY6dMmeIwuuLr66uXXnpJkvTuu+86fYzt27dr06ZNateunSZOnFjkuMHBwfL19S32ObtCeV+HBQsWKDMzU3FxcWrVqpXDPiIjIzVy5Ejt3LlTP/30k319/fr1S/2WY1nK85mDOXH6CdXOLbfcomuvvVbvvPOOnnzySc2bN09eXl4aPnx4iY/ZvHmzpN/nXhQ3D+XgwYOSfj/vf+eddzpsy8jIkCT5+fk5Ve+7776rTZs2afr06SX+ITYMQwsXLtT777+vXbt2KSMjw2G+SkmBav369cU+n507d5ZYz86dO4t9zB/n0hQ6deqUZsyYoZUrV+rw4cP20FLo6NGjDssdO3bUnDlzNG3aNIWHh+u6665zKhB+9913kn4PduWVl5en+++/X1988YVCQ0M1evToYtvt2LFDktSjR48i26666iq1b99eX3/9tfbu3at27do5bI+Oji7ymFtuuUVWq9W+X2eOUfh8e/fuXWI4v1yXvv+9vb3VoEEDtW/fXtdff32F91Xe16Hws1fS++7AgQOSfv/sFYbXTp066dNPP9VLL72ksWPHqlatWhWqrTyfOZgToQbV0sMPP6wJEyZo3bp1+uijj3TbbbepSZMmJbY/deqUJGnJkiWl7vfs2bNF1qWkpEj6fUSnok6fPq1nnnlGrVu31rhx44oEgEJPPvmkZs6cqQYNGqh3795q1KiRatasKUl6//33S/y/1g0bNmjDhg0VqmnXrl3atWtXudpmZmbqpptuUlJSkjp06KAhQ4boqquuktVqVWZmpmbNmlVkNGPw4MH69ttv9fbbb+vGG28sss+mTZuW+9jS76NY5bV48WJdvHhR0dHR2rBhg5566im99dZbRdr99ttvkqTQ0NBi99OgQQOHdpeqX79+kXU1atRQ3bp1lZaW5vQxnHm+FXXkyBHFxcUVWd+rVy8tWbJEgYGB5d5XeV+Hws/evHnzSt3fpZ+9V155Rbt27dLkyZM1efLkctcklf8zB3Pi9BOqpaFDh8pms2no0KHKzMzUww8/XGr72rVrS5I+/fRTGYZR4s8fJ2BKv4eAOnXqKCQkpMJ1Tpo0Senp6ZozZ45sNluxbdLS0vT666+rTZs22r9/vz788ENNnz5dU6dO1dSpU4s9LVUoNja22OcxdOjQEh8zdOjQcj/3+fPnKykpSbGxsfr+++/15ptv6oUXXtDUqVP1wAMPFLt/i8WiuXPn2r+hdu211yo2NlaxsbH2fiiPwgmvFfmjdPHiRU2aNEmJiYnq3bu35s6dq5UrVxZpV1hHSZNIjx8/7tDuUidOnCj2uKdOnXIIBRU9hjPPt6Kio6Pt/Z2Tk6Ndu3ape/fuWrNmjf7xj39UaF8VfR127dpV6mfv0vds8+bNtXv3bi1atEh///vf7e+f0t7XhcrzmYN5EWpQLdWrV099+/ZVSkqKgoODdffdd5favlOnTpKkb775pkLH+e9//6sTJ07olltuqXCNW7Zs0fz58zVo0KBiT0EUOnz4sAoKCtSrV68iw+wpKSk6fPhwhY9dWX7++WdJ0n333VdkW2kjRHl5eZo8ebJ8fHwUHx9vD2iXfjOnLIV9tnr16nI/pk+fPnrxxRdlsVj03nvvqW7duho+fLhOnjzp0C4qKkpS8afcMjMztXPnTvn6+qp169ZFthf3vL/55hvl5+fb9+vMMQqf79q1a2UYRvme8GXw9vZW27Zt7XNOtmzZUqHHl/d1cPaz5+/vrwceeEDPPfec/f1T1teyy/uZg3kRalBtvfrqq0pISNDnn39e5pyNu+++W82aNdOcOXOK/T936fdz/+fPn3dY99prr0mSHnzwwQrXN3r0aAUEBOiVV14ptV3h18Q3btzoMI/m7NmzGjlypPLz8yt87MpSWFtiYqLD+h07dtgnhRZnypQp2rJli2bMmKG2bds6dexRo0bJarXqb3/7m/bt21dke+FpwUtdeo2iBg0a6F//+pdOnDihkSNHOrQbPHiwbDab3njjDXtwu7T2rKwsDR48uNhRsr///e/2eVaSlJ2drUmTJkmShg0b5vQxbrjhBnXu3Fk//PBDse+ZU6dOKTs7u8j6y1U4n6y4r6+Xpryvw7BhwxQUFKS4uLhig1NBQUGJ87kqqryfOZgXc2pQbYWHhys8PLxcbW02mz755BP17t1bd955pzp37qx27drJz89Pv/76q7Zu3arDhw/r+PHj8vPz086dO/XCCy9o2bJlslqt+vHHH7Vnzx6HfRbOgZg6daruueeeIhNKt2/fbp8nU5rQ0FANGDBAixYtUrt27dSrVy/99ttvWrt2rXx9fdWuXbtSJ/5eSUOGDNGMGTM0btw4rV+/Xi1atNDBgwf1n//8R/fee68WL15c5DGJiYmaMWOG7rzzTj366KNOHzsiIkJvvvmm/vrXv6pdu3b269Skp6dr69atql27dpGw9Uf33Xefhg4dqg8++EDvvPOO/TRlWFiYZs6cqTFjxqh9+/b2a8hs2LBBmzdvVqtWrTR9+vQS64qMjHS4PsuhQ4d055136i9/+Yu9nTPH+PDDD9W9e3dNmDBB8fHx9tNFBw8e1Jo1a7Rv374KX7zwUseOHdPcuXMl/T6atm/fPn3wwQfy8vIqEvzKUt7XoW7dulq6dKliYmLUqVMn3XrrrYqMjJSXl5d++eUXbd68udICW3k/czCxKvjaOHDZdMl1aspS3HVqCp04ccKYOHGiERkZadSsWdPw9/c3mjdvbtx3333Gv//9byMvL88wDMN47733DEnl/nnvvffsxyi8Zkbbtm2N/Px8h+OXdK2Yc+fOGZMnTzaaNWtm+Pj4GFdffbUxevRoIz09vdhrcFTldWp++ukn46677jJCQkIMPz8/o3379sa8efOK3V96errRqFEjIzQ01EhLSytyjIpcp6bQpk2bjHvvvdcICQkxbDab0aBBA6N3797GkiVLyvXcfvvtNyMsLMwICAgocs2d1atXG7fddpsRFBRkeHt7G82aNTOefvppIyMjo8h+CvshOzvbePbZZ42wsDDD29vbCA8PN6ZOnVrsdYYqegzD+P01nDBhgtGyZUvDx8fHqF27tnH99dcbkydPNs6dO1fsY8p7nZpLf6xWq3H11Vcbd999t/H111+X+LjKeh2SkpKMMWPGGM2bNzd8fHyMWrVqGddee60xePBgIyEhoVzHLus6NRX5zMGcLIZRBSdvgWrm/fff17Bhw8o1t6Fw/gaXYTe37t27a8OGDVUy38Wd8TrAnTGnBgAAmAKhBihGu3btiv2Kc3FiY2OLzKcBAFQ9JgoDxWjXrl25g0pJd8oGAFQtt5hTs2fPHq1YsUJJSUnKyMjQ+PHj1aFDB/v2ku7EO3jwYPXr16+qygQAAG7MLUZqcnJyFBYWph49ehR787G3337bYXnHjh2aO3euOnbsWFUlAgAAN+cWoSYqKsrhCpR/9MerkG7dulWRkZHF3nsEAAB4pmo3UTgzM1M7duxQz549S22Xl5en8+fPO/zk5eVVUZUAAKCqucVITUVs2LBBvr6+DnNuipOQkKClS5fal7t06aLHH3/8SpcHAABcpNqFmsTERN1yyy1l3usnJiZGffv2tS9bLBZJUkZGhkvvpVPVLBaLgoODlZ6ezsWyPAD97Vnob8/iqf1ttVrLfW+yahVq9u7dq2PHjumJJ54os63NZiv2tvP5+fkedRqqMMzl5eV51IfAU9HfnoX+9iz0d9mq1ZyadevW6ZprrrmsG7oBAABzcouRmuzsbKWmptqX09LSlJycrICAAAUHB0uSzp8/r++++87h7q8AAACF3CLUHDp0SHFxcfblBQsWSJKio6M1ZswYSdKmTZtkGIa6du3qkhoBAIB7c4srClelkydPetycmgYNGuj48eOcg/UA9Ldnob89i6f2t81mU0hISLnaVqs5NQAAACUh1AAAAFMg1AAAAFMg1AAAAFMg1AAAAFMg1AAAAFMg1AAAAFNwi4vvAQBQXVwc2c9lx/7VZUeWasxb4cKjlw8jNQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBQINQAAwBSsri5Akvbs2aMVK1YoKSlJGRkZGj9+vDp06ODQJiUlRQsXLtSePXtkGIYaN26scePGKTg42EVVAwAAd+IWoSYnJ0dhYWHq0aOHXn311SLbU1NT9fzzz6tnz57q37+//Pz8dPToUdlsNhdUCwAA3JFbhJqoqChFRUWVuH3RokWKiorS4MGD7evq169fFaUBAIBqwi1CTWkKCgr0ww8/qF+/fpo2bZqSkpJUr1493XPPPUVOUV0qLy9PeXl59mWLxaKaNWvaf/cUhc/Vk56zJ6O/PQv9japUHd5nbh9qsrKylJ2drU8//VQPPPCAHnzwQe3cuVOvvvqqYmNjFRERUezjEhIStHTpUvtyeHi4pk+frpCQkKoq3a2Ehoa6ugRUIfrbs9DfVetXVxfgIg0aNHB1CWVy+1BTUFAgSbrxxhvVt29fSVJYWJj279+vNWvWlBhqYmJi7O2l/yXMkydPKj8//wpX7T4sFotCQ0OVmpoqwzBcXQ6uMPrbs9DfqErHjx93yXGtVmu5ByTcPtQEBgaqRo0auvrqqx3WN2rUSPv37y/xcTabrcSJxJ744TcMwyOft6eivz0L/Y2qUB3eY25/nRqr1apmzZrp2LFjDuuPHz/O17kBAICdW4Sa7OxsJScnKzk5WZKUlpam5ORkpaenS5L69eunTZs26csvv1RqaqpWrVql7du3q3fv3i6sGgAAuBO3OP106NAhxcXF2ZcXLFggSYqOjtaYMWPUoUMHjRw5UsuXL9d7772nhg0b6qmnnlKrVq1cVTIAAHAzbhFqIiMjFR8fX2qbnj17qmfPnlVUEQAAqG7c4vQTAADA5SLUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAUyDUAAAAU7C6ugAAqO4ujuznsmP/6rIjSzXmrXDh0YGiGKkBAACmQKgBAACmQKgBAACmQKgBAACmQKgBAACmQKgBAACmQKgBAACmQKgBAACmQKgBAACmQKgBAACmQKgBAACm4Bb3ftqzZ49WrFihpKQkZWRkaPz48erQoYN9+5w5c7RhwwaHx7Ro0ULTpk2r6lIBAICbcotQk5OTo7CwMPXo0UOvvvpqsW3atWun0aNH25etVrcoHQAAuAm3SAZRUVGKiooqtY3ValVQUFDVFAQAAKodtwg15bFnzx6NGDFC/v7+at26tQYOHKjatWu7uiwAAOAmqkWoiYqK0s0336zg4GClpaVp8eLF+tvf/qZ//OMfstlsxT4mLy9PeXl59mWLxaKaNWvaf/cUhc/Vk56zJ6O/UZV4n3mW6tDf1SLUdO7c2f57kyZN1KxZM40ePVo//PCDOnbsWOxjEhIStHTpUvtyeHi4pk+frpCQkCterzsKDQ11dQmoQvR31frV1QW4SIMGDVxdgkvQ3+6rWoSaP6pTp45CQkJ0/PjxEtvExMSob9++9uXChHny5Enl5+df8RrdhcViUWhoqFJTU2UYhqvLwRVGf6MqlfbfYJiPq/rbarWWe0CiWoaaM2fO6NSpU6pTp06JbWw2W4mnpjzxP/aGYXjk8/ZU9DeqAu8xz1Id+tstQk12drZSU1Pty2lpaUpOTlZAQIACAgIUHx+vTp06KSgoSCdPntTHH3+sWrVqOVzLBgAAeDa3CDWHDh1SXFycfXnBggWSpOjoaI0cOVK//vqrvv76a507d0516tRRZGSknnjiCfvEXwAAALcINZGRkYqPjy9x+7PPPluF1QAAgOqIez8BAABTINQAAABTINQAAABTINQAAABTINQAAABTINQAAABTINQAAABTINQAAABTcCrUfPbZZzp79mxl1wIAAOA0p0LNwoULNWrUKM2dO1fJycmVXBIAAEDFOXWbhDfeeEOrV69WYmKiEhMT1bJlS91+++3q1KmTatSoUdk1AgAAlMmpkZqQkBANHjxYc+fO1SOPPKK8vDy9/vrrGj16tOLj45WRkVHZdQIAAJTqsm5oabPZ1LNnT/Xs2VMHDhzQRx99pGXLlmn58uXq0KGD7rnnHoWFhVVSqQAAACWrlG8//fjjj1q+fLn27dungIAAde3aVXv37tWkSZP01VdfVcYhAAAASuX0SM2FCxeUmJioNWvW6Pjx42rcuLFGjhypW265Rd7e3srPz9e8efO0ZMkS3XrrrZVZMwAAQBFOhZp58+Zp48aNysnJUfv27TVixAi1adPGccdWq3r06KH169dXRp0AAAClcirUbNq0Sbfeeqtuv/121atXr8R2DRs21KhRo5wuDgAAoLycCjVvvfWWfH19y2wXGBio7t27O3MIAACACnEq1JQWaLKysrRjxw5Jv4eaqKgo5yoDAACoAKdCzZ49e0rclpKSonfeeUcRERG6+uqrCTUAAKBKOBVq4uLiymwTGxvrzK4BAACc4vRXugcOHKjQ0NAi648fP65FixZdVlEAAAAV5XSoadOmjZo3b15k/cGDBy+rIAAAAGdUyhWFAQAAXM3pkZrPPvtMderUkY+Pj+rUqaMmTZro2muvrczaAAAAys3pUPPdd98VWVerVi1df/31l1UQAACAM5wKNYsXL5Yk5efnKzs7W6dPn1ZycrK2bdumjRs3VmqBAAAA5XFZc2qsVqsCAgLUpEkTdevWTU8++aSGDh0qSRo7dqxeeeWVSikSAACgLE6ffipJ165d5efnJ+n3KwoDAABUhUoPNdzvCQAAuEKl3yahUEREhDO7BgAAcMoVu01C4WRiAACAquBUqBk3bpz994KCAs2aNUsPPvig6tWrV2mFAQAAVIRToaZTp0723wtDTZs2bXTNNddUWmEAAAAVwW0SAACAKRBqAACAKVRaqLFYLJW1KwAAgApzak7NkCFDioSY559/Xl5e/8tIH3zwweVVBgAAUAFOhZqOHTsyMgMAANyKU6FmzJgxlV0HAADAZXFqTs3hw4cruw4AAIDL4tRIzaRJk9S8eXPdfvvtuvnmm2W1VvotpAAAACrEqZGa0aNHyzAMzZ49W6NGjdKiRYt06tQpp4vYs2eP/vGPf+iRRx5R//79tWXLlhLbvv322+rfv78+//xzp48HAADMx6khlujoaEVHR+vnn3/WqlWr9Nlnn+nTTz/VDTfcoD59+igyMrJC+8vJyVFYWJh69OihV199tcR2W7Zs0cGDB1WnTh1nygYAACZ2Wdepad68ucaOHau33npLf/7zn3X48GH97W9/01NPPaW1a9cqNze3XPuJiorSgAED1LFjxxLbnD59Wu+++64ee+wxTncBAIAiKuXie1arVT4+PvawkZOTo/nz5+vxxx/XgQMHLnv/BQUFeuONN9SvXz81btz4svcHAADM57KGPI4cOaLVq1dr48aNys/PV6dOnfTYY4+pefPmOnLkiN5++23NmzdPM2bMuKwiP/30U9WoUUN9+vQp92Py8vKUl5dnX7ZYLKpZs6b9d09R+Fw96Tl7MvobVYn3mWepDv3tVKjZtGmTVq9erX379ikwMFB9+/ZVr169FBQUZG/TtGlTDRw4UNOmTbusAg8fPqyVK1dq+vTpFXpBExIStHTpUvtyeHi4pk+frpCQkMuqp7oKDQ11dQmoQvR31frV1QW4SIMGDVxdgkvQ3+7LqVAza9YshYWFadSoUeratWuJc1xCQkJ0yy23XFaBe/fuVVZWlkaPHm1fV1BQoAULFmjlypWaM2dOsY+LiYlR37597cuFgejkyZPKz8+/rJqqE4vFotDQUKWmpsowDFeXgyuM/kZVOn78uKtLQBVyVX9brdZyD0g4FWri4uLUqlWrMtvVr1/fIYw4o1u3brruuusc1k2bNk3dunVTjx49SnyczWaTzWYrdpsn/sfeMAyPfN6eiv5GVeA95lmqQ387FWrKCjQZGRkV+tp1dna2UlNT7ctpaWlKTk5WQECAgoODVatWLYf2VqtVQUFBatiwYcUKBwAApuXUt59WrlxZ4rYtW7Zo/PjxFdrfoUOHNGHCBE2YMEGStGDBAk2YMEGLFy92pjwAAOCBnBqp+eCDD2SxWBy+jZSbm6v3339fX331lbp06VKh/UVGRio+Pr7c7UuaRwMAADyXU6Hm4Ycf1jvvvCNJ6tOnj5KTkzVr1ixlZGRo7Nixlz05GAAAoKKcCjW9evWSxWLR/PnztX//fm3dulXNmzfX5MmTPfYr0wAAwLWcvvjebbfdJkmaP3++IiMjNWXKlGpxYR4AAGBOl3WbhNtuu00jR47Unj17tGrVqsqqCQAAoMKcGqkZM2aMw6iMl5eX3n//fX3++eeSfr8A2BtvvFE5FQIAAJSDU6EmIiKCU00AAMCtOD1SAwAA4E4ua04NAACAu3A61Bw9elQzZ87U//3f/2ngwIE6fPiwJGnJkiXavXt3pRUIAABQHk6FmuTkZE2aNEl79+5VRESECgoK7Nuys7O1du3aSisQAACgPJyaU7Nw4UI1bdpUzz33nKxWqzZv3mzf1rx5c33//feVViAAAEB5ODVSs3//fvXr108+Pj5FvgVVu3ZtZWZmVkZtAAAA5eZUqDEMQ1Zr8YM8586dk81mu6yiAAAAKsqpUNO0aVNt2bKl2G07d+7UNddcc1lFAQAAVJRTc2ruuOMOzZo1Sz4+PurWrZskKT09Xbt371ZiYqKefPLJSi0SAACgLE6Fms6dOys1NVVLlizRF198IUl69dVXVaNGDfXv31833nhjpRYJAABQFqfv0n3vvfcqOjpau3btUmZmpgIDA3X99dcrJCSkMusDAAAoF6dDjSTVrVtXPXv2rKxaAAAAnObUROHt27dr1apVxW5btWqVfvjhh8sqCgAAoKKcCjWffPKJsrOzi92Wk5OjhISEyyoKAACgopwKNceOHVN4eHix28LDw5WSknJZRQEAAFSUU6EmLy9P+fn5JW7Lzc29rKIAAAAqyqlQ07BhQ23fvr3Ybdu3b1fDhg0vqygAAICKcirU9OjRQ+vWrVN8fLz9Pk+ZmZmKj4/XunXr1KNHj8qsEQAAoExOfaX79ttv16FDh7Rs2TItW7ZMXl5eKigokCTdcsstuuOOOyq1SAAAgLI4FWosFovGjh2rW2+9VTt37lRWVpYCAwMVFRWlVq1aVXaNAAAAZbqsi++1bt1arVu3rqxaAAAAnObUnBoAAAB3Q6gBAACmQKgBAACmQKgBAACmQKgBAACmUGmhxjAMpaamlnj7BAAAgCvJqa905+Tk6MMPP9Thw4d13XXX6b777lNcXJwOHjyoWrVqacqUKWratGll1woAAFAip0ZqFi5cqDVr1ig/P1//+c9/9Nprr+nMmTMaOnSo/Pz8FB8fX9l1AgAAlMqpkZpt27bpvvvuU//+/bVt2zbNmDFDTzzxhG6++WYFBgZqwYIFlV0nAABAqZwaqcnIyFDbtm0lyf5vvXr17P+eOXOmksoDAAAoH6dCTUFBgazW3wd5Cv+1WCz2fwtvbgkAAFBVnL730w8//KCjR4/KMAz78q+//qoTJ05UWnEAAADl5XSoWbZsmcPykiVLLrsYAAAAZzkVambPnl3ZdQAAAFwWp0JNSEhIZdcBAABwWZw+/VSZ9uzZoxUrVigpKUkZGRkaP368OnToYN8eHx+vTZs26dSpU7Jarbrmmms0YMAAtWjRwoVVAyW7OLKfy479q8uOLNWYt8KFRwfg6dzi3k85OTkKCwvT8OHDi93esGFDDR8+XK+88or+9re/KSQkRC+88IKysrKquFIAAOCu3GKkJioqSlFRUSVu79q1q8PykCFDtG7dOh05ckTXXXfdlS4PAABUA24RaioiPz9fX375pfz8/Eq9v1ReXp7y8vLsyxaLRTVr1rT/7ikuvX4QcKXxPvMs9LdnqQ79XW1Czfbt2zVz5kzl5uYqKChIzz33nAIDA0tsn5CQoKVLl9qXw8PDNX36dI+d5BwaGurqEjyKK+e1uFKDBg1cXYJL0N+ehf52X5UaavLz8+1XGK5skZGRmjFjhrKysvTVV1/pn//8p1588UXVrl272PYxMTHq27evfbkwYZ48eVL5+flXpEZ3ZLFYFBoaqtTUVPuFEoEr5fjx464uAVWI/vYsrupvq9Va7gEJpxKIYRj67LPPdPjwYbVt21Y9evTQzJkz9f333+vqq6/WxIkTK31ExNfXV6GhoQoNDVXLli312GOPad26dYqJiSm2vc1mk81mK7F+T2MYhkc+b1Qt3mOehf72LNWhv50KNUuWLNGyZcvk5+enzZs36+DBg9qzZ49uv/12ffvtt1q0aJEeffTRyq7VgWEYDnNmAACAZ3Mq1HzzzTfq1auXHn74Ya1bt07/+te/NHr0aEVHR6tJkyaKj4+v0P6ys7OVmppqX05LS1NycrICAgIUEBCgTz75RDfeeKPq1KmjM2fOaM2aNTp9+rRuvvlmZ8oHAAAm5FSouTRQdO7cWf/617/UqFEjSb9fU+a3336r0P4OHTqkuLg4+/KCBQskSdHR0Ro5cqSOHTumV199VWfOnFGtWrXUrFkzxcXFqXHjxs6UDwAATMipUJOfny9vb29Jsv/r5fX7dfxq1KihgoKCCu0vMjKy1NGd8ePHO1MmAADwIE5/VSkpKUm5ubn2AJOUlKTs7GylpKRUWnEAAADl5XSomT9/vsPy22+/fdnFAAAAOMupUBMbG1vZdQAAAFwWp0JNREREZdcBAABwWZy6S3dZE4F37tzpzG4BAACc5lSomTVrVrHBJj8/X++//77+8Y9/XHZhAAAAFeFUqPnxxx/1+uuvOwSblJQUTZo0SV999ZVGjBhRaQUCAACUh1Oh5tlnn9WuXbvswWbVqlV65plnZLVaNX36dP3pT3+q7DoBAABK5dRE4ebNm+vZZ5/VtGnTNGbMGGVmZqpfv37q37+/atSoUdk1AgAAlMmpkRrpf8EmOztbrVq10gMPPECgAQAALuNUqElPT1d6erqCgoL0yCOP6MCBA3r99dft69PT0yu7TgAAgFI5dfppzJgxRdZt3rxZmzdvti8vXrzY+aoAAAAqyKlQM2rUqMquAwAA4LI4FWq6d+9eyWUAAABcHqfm1Hz22Wc6d+5cZdcCAADgNKdGahYuXKj4+Hh16dJFt99+u8LCwiq5LPO5OLKfy479q8uOLNWYt8KFRwcAeBKnQs0bb7yh1atXKzExUYmJiWrZsqVuv/12derUia91AwAAl3Dq9FNISIgGDx6suXPn6pFHHlFeXp5ef/11jR49WvHx8crIyKjsOgEAAErl1EhNIZvNpp49e6pnz546cOCAPvroIy1btkzLly9Xhw4ddM8993BqCgAAVAmnryh8qR9//FHLly/Xvn37FBAQoK5du2rv3r32G1wCAABcaU6P1Fy4cEGJiYlas2aNjh8/rsaNG2vkyJG65ZZb5O3trfz8fM2bN09LlizRrbfeWpk1AwAAFOFUqJk3b542btyonJwctW/fXiNGjFCbNm0cd2y1qkePHlq/fn1l1AkAAFAqp0LNpk2bdOutt+r2229XvXr1SmzXsGFDrj4MAACqhFOh5q233pKvr2+Z7QIDA7n6MAAAqBJOhZrCQJOXl6fDhw/rzJkzqlWrlq655hrZbLZKLRAAAKA8ygw1ubm5Wrlypfr06SMfHx/7+lWrVmnx4sU6f/68fZ2fn58GDBig3r17X5lqAQAASlBmqCkoKNDHH3+s66+/XuHh4ZKkjRs36r333lPr1q0VHR2tOnXqKCMjQxs2bNC7776rgIAAdenS5YoXDwAAUKjMUOPr6ysvLy8ZhmFf99lnn6ljx4568sknHdr26NFDr732mj777DNCDQAAqFLluvheYGCgzp49a18+evSoevToUWzbHj16KCUlpXKqAwAAKKdyhZomTZrou+++sy/7+/s7zKW51Llz55gsDAAAqly5Qk3Xrl21bt06ff3115Kka6+9VkuXLlVWVpZDu6ysLC1btkytW7eu/EoBAABKUa6vdEdHR2vr1q2aM2eONm3apLp16+rYsWMaM2aM2rZta58o/N///ldWq1VPPfXUla4bAADAQbmvU/PUU0/p888/18qVK7Vjxw5Jv3/de9u2bZKkunXrqmvXroqJiVFISMiVqRYAAKAE5Q41FotFffv2Vd++fXX27Fnl5ubK29tbVqtV3t7e8vKqlBt+AwAAOMWpKwoHBARUdh0AAACXheEVAABgCoQaAABgCoQaAABgCoQaAABgCoQaAABgCoQaAABgCoQaAABgCk5dp6ay7dmzRytWrFBSUpIyMjI0fvx4dejQQZKUn5+vRYsWaceOHUpLS5Ofn5+uu+46DRo0SFdddZWLKwcAAO7CLUZqcnJyFBYWpuHDhxfZlpubq6SkJN13332aPn26nnrqKR0/flwvv/yyCyoFAADuyi1GaqKiohQVFVXsNj8/P02ZMsVh3bBhwzR58mSlp6crODi4KkoEAABuzi1CTUWdP39eFotFfn5+JbbJy8tTXl6efdlisahmzZr231E1eK09C/3tWehvz1Id+rvahZrc3Fx99NFH6tKlS6mhJiEhQUuXLrUvh4eHa/r06S67g/ivLjmq6zVo0MDVJbgE/e1Z6G/PQn+7r2oVavLz8zVz5kwZhqERI0aU2jYmJkZ9+/a1LxcmzJMnTyo/P/+K1on/OX78uKtLQBWivz0L/e1ZXNXfVqu13AMS1SbU5Ofn65///KdOnjyp559/vtRRGkmy2Wyy2WzFbjMM40qUiGLwWnsW+tuz0N+epTr0t1t8+6kshYEmNTVVU6ZMUa1atVxdEgAAcDNuMVKTnZ2t1NRU+3JaWpqSk5MVEBCgOnXq6LXXXlNSUpImTpyogoICZWZmSpICAgJktbrFUwAAAC7mFong0KFDiouLsy8vWLBAkhQdHa0///nP2rZtmyRpwoQJDo+LjY1VZGRk1RUKAADclluEmsjISMXHx5e4vbRtAAAAUjWZUwMAAFAWQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFQg0AADAFq6sLkKQ9e/ZoxYoVSkpKUkZGhsaPH68OHTrYt3///ff68ssvdfjwYZ05c0Yvv/yywsLCXFcwAABwO24xUpOTk6OwsDANHz68xO3XXnutBg0aVMWVAQCA6sItRmqioqIUFRVV4vZu3bpJktLS0qqqJAAAUM24Rai5EvLy8pSXl2dftlgsqlmzpv13VA1ea89Cf3sW+tuzVIf+Nm2oSUhI0NKlS+3L4eHhmj59ukJCQlxSz68uOarrNWjQwNUluAT97Vnob89Cf7sv04aamJgY9e3b175cmDBPnjyp/Px8V5XlcY4fP+7qElCF6G/PQn97Flf1t9VqLfeAhGlDjc1mk81mK3abYRhVXI3n4rX2LPS3Z6G/PUt16G+3+PYTAADA5XKLkZrs7Gylpqbal9PS0pScnKyAgAAFBwfr7NmzSk9P1+nTpyVJx44dkyQFBQUpKCjIFSUDAAA34xah5tChQ4qLi7MvL1iwQJIUHR2tMWPGaNu2bXrzzTft22fOnClJuv/++9W/f/8qrRUAALgntwg1kZGRio+PL3F79+7d1b1796orCAAAVDvMqQEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZAqAEAAKZgdXUBkrRnzx6tWLFCSUlJysjI0Pjx49WhQwf7dsMwtGTJEn311Vc6e/asWrRooYcffliNGzd2YdUAAMCduMVITU5OjsLCwjR8+PBit3/66af6/PPPNXz4cL300ksKCgrSCy+8oAsXLlRxpQAAwF25RaiJiorSgAED1LFjxyLbDMPQypUrFRMTo44dO6pJkyYaM2aMcnJytHHjRhdUCwAA3JFbhJrSpKWlKTMzU9dff719nc1mU0REhPbv3+/CygAAgDtxizk1pcnMzJQk1a5d22F97dq1lZ6eXuLj8vLylJeXZ1+2WCyqWbOmrFbXPGWvZte65LiuVsNmc3UJLkF/exb627PQ31WrIn+33T7UFLJYLA7LhmGU2j4hIUFLly61L3fp0kWPP/646tSpc0XqK9PrC11zXLgG/e1Z6G/PQn+7Lbc//RQUFCTpfyM2hbKysoqM3lwqJiZG77//vv1n5MiRDiM3nuLChQuaOHEik6o9BP3tWehvz0J/l83tQ029evUUFBSkH3/80b4uPz9fe/bs0bXXljwEaLPZ5Ofn5/Bj88ChUsMwlJSUVObIFsyB/vYs9Ldnob/L5hann7Kzs5WammpfTktLU3JysgICAhQcHKw77rhDCQkJatCggUJDQ5WQkCAfHx917drVhVUDAAB34hah5tChQ4qLi7MvL1iwQJIUHR2tMWPG6O6771Zubq7mz5+vc+fOqXnz5nr22WdVs2ZNV5UMAADcjFuEmsjISMXHx5e43WKxqH///urfv38VVmUONptN999/v0eeevNE9Ldnob89C/1dNovByTkAAGACbj9RGAAAoDwINQAAwBQINQAAwBQINQAAwBQINQAAVBN8t6d0hBoAAKqJQYMGKSUlxdVluC23uE4Nqk56erri4+M1evRoV5eCSpCbm6vDhw8rICBAV199dZFtmzdvVnR0tIuqQ2VLSUnRwYMH1bJlSzVq1EhHjx7VypUrlZeXp27duqlNmzauLhGV5IMPPih2fUFBgZYvX65atWpJkoYOHVqVZbk9Qo2HOXv2rDZs2ECoMYFjx45p2rRpSk9PlyS1bt3a4U7058+f15tvvkmoMYmdO3fq5Zdflq+vr3JycvT0009r9uzZatq0qSRp2rRpevbZZwk2JrFy5Uo1bdpU/v7+RbYdPXpUvr6+LqjK/RFqTGbbtm2lbj9x4kQVVYIrbeHChWrcuLFeeuklnT9/Xh988IGmTJmiqVOnKjg42NXloZItXbpU/fr104ABA/Ttt99q1qxZ6tWrlwYOHChJ+vjjj7V8+XJCjUkMGDBAX331lYYMGeLQpwMHDtSYMWOKjMzid4Qak5kxY4arS0AVOXDggKZMmaLAwEAFBgZq4sSJmj9/vp5//nnFxsbKx8fH1SWiEv36668aO3asJOnmm2/W7Nmz1bFjR/v2Ll26aN26da4qD5UsJiZG1113nd544w3dcMMNGjRokKxW/mSXhVfIZIKCgvTwww+rQ4cOxW5PTk7WxIkTq7gqXAm5ubny8nKc6z9ixAi9++67mjp1qh577DEXVYYrzcvLSzabzeHURM2aNXX+/HkXVoXK1rx5c02fPl3z58/XpEmT9Oijj7q6JLfHt59M5pprrlFSUpKry0AVaNiwoQ4fPlxk/fDhw3XjjTfq5ZdfdkFVuFLq1aun1NRU+/ILL7zgcJrx1KlT9vlUMA9fX1+NHTtW99xzj/7+97+roKDA1SW5NUZqTKZfv37KyckpcXtoaKhiY2OrsCJcKR06dNC3336rbt26Fdn28MMPyzAMrV271gWV4Uq47bbbHP6gNWnSxGH7jh07mE9jYl26dFGrVq10+PBh5syVgrt0AwAAU+D0EwAAMAVCDQAAMAVCDQAAMAVCDVDJPv30U/Xv37/YCyFmZmZq4MCBfDPJQ7z22mtavHixsrOzdezYMY0aNYr79gBXEKEGqGS33nqrvL299cUXXxTZtmbNGl28eFF9+vRxQWWoanfddZfWrl2rIUOG6IknnlBkZCRXggWuIL7SDVSygIAAde3aVevWrVNKSor9j1h+fr7Wrl2rxo0b67rrrnNxlagKLVq00FtvvaUTJ07I39+f68gAVxihBrgC+vTpo3Xr1mnVqlUaMWKEJGnTpk367bff1L9/f3u79evX680339Rzzz2njRs3auvWrcrPz1dkZKSGDRum+vXrO+x33bp1+uKLL3Ts2DF5e3srIiJCAwcOdPi//zlz5mjDhg325Zo1a6pJkya6//77df311zvs78cff9Ty5ct16NAhXbx4UeHh4erfv79D6IqPj9fSpUs1f/58BQYG2tcfOnRIkyZN0ujRo9W9e3f7sb/77jv9+9//rvBrlpqaqk8++UT79u3T6dOn5e/vr/DwcA0aNMjhmiw//fST4uLi9OSTT6pTp04O+/jLX/6iTp06acyYMQ6vb3Fmz56tevXq2ZfHjBmjkydPFmkXGxuryMhIe5uIiAj7/iXp66+/1uzZsxUSEqI5c+ZIktLS0jR27Fj7a1PYP2+99ZYSExMVHR3tsI/Tp09ryZIl2rlzpzIzMxUYGKiWLVvq4YcfVlBQkP05X1pLamqq4uLiFBISosmTJ9tvcHj27FktWrRIW7duVVZWlurWrasuXbro/vvvl81msx/z0vfhpe6//377tvLuC3AXhBrgCmjatKlat26tDRs2aNCgQfLz89MXX3whf3//Yi+W99Zbb6lt27Z6/PHHlZ6ersWLF2vq1Kl65ZVX7JfCT0hI0Mcff6wuXbpo4MCBOnv2rJYsWaLnnntOL730kho0aGDfX1BQkMaPHy9JysjI0NKlS/Xyyy9r1qxZ9gt3ff3115ozZ45uvPFGjRkzRjVq1NDatWvtd3uu6tGk06dPKyAgQIMGDVJgYKD9jvKTJ0/Wyy+/rIYNGzq979GjRxd5fHGjJlFRUbrvvvskSUlJSXrnnXdK3e/58+f14YcfFrldRXEOHjyo9evXF2l7+vRpTZo0Sfn5+YqJiVHTpk115swZ7dq1S+fOnVNQUFCRfZUUaHJzcxUXF6fU1FT1799fTZs21d69e7V8+XIlJydr0qRJDvvp1KmT+vbt67Cubt26Tu0LcAeEGuAK6dOnj1577TUlJiaqRYsWOnTokO66665ibzTZrFkzjRo1yr7cuHFjTZkyRatXr9a9996rc+fOadmyZYqKitLjjz9ubxcREaHHH39cS5YscbjXk81mU8uWLe3LXl5emjFjhn799VcFBwcrJydH77//vtq3b6+nn37a3i4qKkoTJ07Uxx9/XOWhJiIiQhEREfblgoICtW/fXk899ZTWrl2roUOHOr3vxo0bq1mzZqW2yc/PV506deyvW15eXpn7jY+Pl5eXl2666aZib1lRqKCgQO+8845uuOEGHTlyxGHb4sWLlZWVpRkzZjiMuHXu3LnYfZ04caLYQCNJGzZs0JEjRzRu3DjdfPPNkqS2bdvK19dXCxcu1I8//qi2bdva29euXdvhfXKpiu4LcAdMFAaukJtuukl169bVqlWr9Pnnn8vLy0u9e/cutm3Xrl0dlq+99lqFhITop59+kvT7Hblzc3Ptp3kKBQcHq02bNvrvf/9bZJ8XL17UxYsXlZ6ersTERPn5+emaa66RJO3fv19nz55VdHS0vd3FixdlGIbatWunQ4cOKTs722F/BQUFDm1LuwdNYZuKuHjxoj755BONGzdOAwcO1IABAzRo0CAdP35cR48eLdL+j/VU9Hh/lJubW6FTKr/88otWr16tIUOGOASL4nz55ZdKSUnRQw89VGTbzp071aZNm3JNID5x4oSmTp2qM2fO6Omnny5y3N27d8vHx6fIabnC901x75OSVOa+gKrCSA1whdSoUUO9evXSxx9/rBMnTuimm25ymMNxqeJOMQQFBenMmTOSZP+3uFMmderU0dmzZx3WnTx5UgMHDrQv+/v7a/To0apdu7Yk6bfffpP0+1eOS3L27FmHP5r/93//V2LbS+Xk5NiPbbPZFBwcrOjoaMXExMhisZT4uA8++ECrV6/W3XffrYiICAUEBMhisWju3LnKzc0t0n7mzJnlqqc88vPzdf78edWqVavcj3nnnXfUqlUrde7cWTt27CixXVZWlhYtWqR77rmn2P7PysrSVVddVa5jzp8/X2FhYTp79qwSEhI0ZMgQh+1nz55VUFBQkde5du3aqlGjhv19VB6VuS+gqhBqgCvo1ltv1dKlS5WXl1fq17gzMzOLXRcaGipJ9j+2GRkZRdplZGQU+WNcp04dTZgwQdLvIWPr1q167bXXNH78eN1444329sOHD1eLFi2KremPQWvKlCny8/OzLx89elSzZ88u8jhvb2/FxcXZj71lyxYtWrRIPj4+uvPOO4s9liR98803io6O1qBBgxzWnzlzxj6v6FIPPvhgkRs4Onuz1hMnTsgwDPvrXZZvvvlGBw4c0IwZM8ps+9FHH8nf31/9+vUrdntgYKBOnz5druM2b95ckydP1saNGzVv3jy1a9fO4RRQQECADh48KMMwHMLIb7/9posXLzpM9C5LZe4LqCqcfgKuID8/P/n6+qpJkyal3kF548aNDsv79+/XyZMn7XNMWrZsKW9vb33zzTcO7U6dOqXdu3cX2bfValWzZs3UrFkzRUREaOjQoapZs6a+/fZbSVKrVq3k7++vlJQUe7s//litjv/P07RpU4ftjRo1Kva5WCwWh2M/9NBD8vf3188//1zqa2WxWIoc84cffijxD379+vWL1FzaSFBptm7dKklq3bp1mW0vXLigDz/8UHfccUeZp4x+/vlnJSYmatiwYfL29i62Tbt27bR7924dO3aszGM/8MAD8vX11Z/+9CfdcMMNmjNnjrKysuzbr7vuOmVnZ9ufT6HCb8NV5C7elbkvoKowUgNcASdPntT+/fu1detWnTlzRn/5y19KbX/o0CHNnTtXnTp10qlTp7Ro0SJdddVV9jk4/v7+uu+++/Txxx9r9uzZ6tKli86cOaOlS5fKZrPpz3/+s8P+8vLydODAAUn/G6k5d+6cwsPDJUm+vr4aNmyY5syZo7Nnz6pTp04KDAxUVlaWjhw5oqysLI0cOdLp5184ByYnJ0fbtm3TuXPnShwRKtS+fXtt2LBBjRo1UtOmTXX48GGtWLHC/m2cKyEjI0OrVq3SihUr1LVrV4WEhJT5mG3btql27dpFXvPifPnll7rhhhvUvn37Ets88MAD2rlzp2JjYxUTE6MmTZro3Llz2rlzp/r27VtiePzrX/+q8ePHa+7cufZRuW7dumn16tWaM2eO0tLS1KRJE+3bt08JCQmKioqq0MTeytwXUFUINcAV8NNPP2nu3LkKCgrSXXfdpejo6FLbjxo1Sl9//bVmzZqlvLw8+3VqAgIC7G1iYmJUu3ZtffHFF9q0aZP9OjWDBg1y+Dq39Pupq+eee06S5OPjo+DgYA0YMMDh67vdunVTcHCwVqxYobffflsXLlxQ7dq1FRYWVmRCckXk5ORo3LhxRY5d1lWUhw0bJqvVquXLlys7O1vh4eEaP368Fi1a5HQtZfnpp5+0ZcsW3X///br77rvL9ZiCgoJyTQ6Wfp9XVdzk4EtdddVVevHFFxUfH6/ly5frzJkzCgwMVKtWrRz6/48CAwM1evRovfTSS1qzZo169eolb29vxcbG6uOPP9Znn31mn69z1113lSuEXaoy9wVUFYthGIariwA8VeHF4V566aUyv3IMACgdc2oAAIApEGoAAIApcPoJAACYAiM1AADAFAg1AADAFAg1AADAFAg1AADAFAg1AADAFAg1AADAFAg1AADAFAg1AADAFAg1AADAFP4fp+Rn/4BhUS0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "movies.groupby('target').charsrate_median.median().plot(kind='bar')\n",
    "plt.title('Медианная скорость речи')\n",
    "# plt.ylim(0.08,0.125)\n",
    "plt.ylim(10,17)\n",
    "plt.ylabel('букв в секунду')\n",
    "plt.xlabel('Уровень английского')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "d2dd3fea",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "movies.groupby('target').B1ratio.median().plot(kind='bar')\n",
    "plt.ylim(0.1,.14)\n",
    "plt.title('Слова из словаря В1')\n",
    "plt.ylabel('доля уникальных лемм')\n",
    "plt.xlabel('Уровень английского')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b9771d92",
   "metadata": {},
   "source": [
    "**Кластеры KMeans**\n",
    "\n",
    "При 4 кластерах есть заметное разделение на группы А2/В1 и В2/С1. Внутри этих групп разделение довольно плохо работает.\n",
    "\n",
    "\n",
    "Но в целом добавление кластеризации в качестве признака несколько стабилизировало модель – при кросс-валидации разброс метрик стал меньше, на тесте немного улучшился скор.\n",
    "\n",
    "В итоге было оставлено деление на три кластера."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "0eddcf6c",
   "metadata": {},
   "source": [
    "<img src='data:image/png;base64,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'>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ea901515",
   "metadata": {},
   "source": [
    "# Заключение\n",
    "\n",
    "Для заказчика необходимо было найти решение определения уровня сложности английского языка в фильмах по субтитрам.\n",
    "\n",
    "Заказчик предоставил данные:\n",
    "- английские словари Oxford, в которых слова распределены по уровню сложности\n",
    "- набор файлов-субтитров, рассортированных по каталогам в соответствии с уровнем сложности\n",
    "- excel файл со список несортированных фильмов и указанием их уровня\n",
    "\n",
    "Цель:\n",
    "- создание модели определения уровня сложности\n",
    "\n",
    "Задачи:\n",
    "- изучение и обработка предоставленного материала\n",
    "- обработка текста субтитров и подготовка для машинного обучения\n",
    "- расчет и добавление дополнительных признаков\n",
    "- тестирование модели и подбор гиперпараметров\n",
    "- выделение наиболее удачных признаков\n",
    "- обучение оптимальной модели на датасете с наиболее полезными признаками\n",
    "- создание приложения на платформе streamlit для демонстрации решения\n",
    "\n",
    "\n",
    "\n",
    "## The Oxford core by CEFR level\n",
    "\n",
    "В словарях Oxford 3000 и 5000 содержатся наиболее важные слова, которые должен знать каждый, кто учит английский.\n",
    "\n",
    "Словарь Oxford соответствует стандартам CEFR – The Common European Framework of Reference for Languages. Это общеевропейские компетенции владения иностранным языком: изучение, преподавание, оценка — система уровней владения иностранным языком, используемая в Европейском Союзе.\n",
    "\n",
    "У нас словари представлены в виде pdf-файлов.\n",
    "- прочитаем их,\n",
    "- разобьем на строки,\n",
    "- выделим слова,\n",
    "- определим уровень\n",
    "- сохраним, из какого словаря взяли слово\n",
    "- удалим дубликаты, оставим те случаи, где слову присвоена более высокая сложность\n",
    "- выгрузим в joblib для дальнейшего использования в приложении Streamlit\n",
    "\n",
    "## Subtitles .srt\n",
    "\n",
    "Субтитры у нас представлены в виде файлов .srt – это текстовые файлы, в которых информация представлены в таком виде:\n",
    "\n",
    "        18\n",
    "        00:04:59,796 --> 00:04:59,921\n",
    "        - ( police radio chatter )\n",
    "        - Man: What's the difference\n",
    "    \n",
    "То есть номер реплики, время и текст.\n",
    "\n",
    "Файлы рассортированы по папкам, соответствующим уровням сложности языка.\n",
    "\n",
    "Просканируем все файлы построчно:\n",
    "- удалим из текста по возможности тэги, скобки, указания говорящего лица\n",
    "- нетекстовые знаки, \n",
    "- лишние пробелы и переводы строки\n",
    "\n",
    "Сохраним временные характеристики реплик:\n",
    "- длительность реплики\n",
    "- скорость речи в знаках в секунду\n",
    "- скорость речи в словах в секунду\n",
    "\n",
    "Соберем все в виде датасета для дальнейшей обработки.\n",
    "    \n",
    "\n",
    "## Excel labels processing\n",
    "\n",
    "Часть файлов лежит в общем каталоге без указания уровня сложности. Для них заказчик предоставил excel-файл со списком фильмов и уровнями.\n",
    " \n",
    "Загрузим файл и изучим его:\n",
    "- удалим дубликаты\n",
    "- поправим несовпадения названий\n",
    "- из указанных уровней оставим более сложный вариант\n",
    "- удалим фильмы, для которых нет информации по уровню сложности\n",
    "\n",
    "## Lemmatization and feature engineering\n",
    "\n",
    "Подготовим текст для машинного обучения:\n",
    "- фильтруем имена собственные, стоп-слова, пунктуацию, цифры, урлы\n",
    "- выполним лемматизацию – приведение слов к нормальной форме\n",
    "- выделим морфологическую разметку\n",
    "\n",
    "Добавим числовые признаки.\n",
    "\n",
    "Слова (леммы):\n",
    "- общее количество\n",
    "- количество уникальных слов\n",
    "- доля уников от общего количества\n",
    "\n",
    "Предложения:\n",
    "- количество предложений\n",
    "- количество слов в предложениях\n",
    "- медианная длина предложения\n",
    "\n",
    "Сложность:\n",
    "- подсчитаем количество уникальных слов каждого уровня\n",
    "- долю слов каждой сложности относительно общего количества уников\n",
    "\n",
    "Целевой признак перекодируем в порядовые числа от 0 до 5.\n",
    "\n",
    "## Feature importances and correlation\n",
    "\n",
    "Наибольший вклад в определение уровня ангийского вносят признаки:\n",
    "- медианная длина предложения, количество слов\n",
    "- медианная скорость речи, букв в секунду. Есть еще похожий признак – количество слов с секунду, но он не отражает длину слов.\n",
    "- леммы, отфильтрованные от стоп-слов, именованых сущностей\n",
    "- из частей речи заметно появление местоимений в числе 20 самых важных признаков.\n",
    "- также проиллюстрировано заметно разное распределение слов уровня В1\n",
    "- выявлена положительная корреляция между уровнем сложности и словами длиной 6 знаков\n",
    "- кластеризация в качестве признака несколько стабилизировала модель – при кросс-валидации разброс метрик стал меньше, на тесте немного улучшился скор.\n",
    "\n",
    "## Итоги\n",
    "\n",
    "Обучена модель регресии XGBRegressor\n",
    "- метрика RMSE на кроссвалидации 0.35\n",
    "- метрика RMSE на тесте 0.61, что свидетельствует о переобучении модели\n",
    "- предполагаю, что имеющего массива данных катастрофически недостаточно для качественной модели\n",
    "- на большем количестве данных предполагаю не стоит использовать узкоспециализированные фичи типа слов равных по длине 6 знакам, искать более общие решения\n",
    "\n",
    "Создано приложение на платформе Streamlit\n",
    "- https://english-level.streamlit.app\n",
    "- есть возможность загрузки пользовательких файлов .srt для определения уровня английского\n",
    "- предусмотрено определение уровня английского в субтитрах на Youtube по ссылке\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "033bc0ab",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}