WineReviews.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "def load_wine_data(wine_path):\n",
    "    return pd.read_csv(wine_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 280901 entries, 0 to 129970\n",
      "Data columns (total 11 columns):\n",
      " #   Column       Non-Null Count   Dtype  \n",
      "---  ------       --------------   -----  \n",
      " 0   Unnamed: 0   280901 non-null  int64  \n",
      " 1   country      280833 non-null  object \n",
      " 2   description  280901 non-null  object \n",
      " 3   designation  197701 non-null  object \n",
      " 4   points       280901 non-null  int64  \n",
      " 5   price        258210 non-null  float64\n",
      " 6   province     280833 non-null  object \n",
      " 7   region_1     234594 non-null  object \n",
      " 8   region_2     111464 non-null  object \n",
      " 9   variety      280900 non-null  object \n",
      " 10  winery       280901 non-null  object \n",
      "dtypes: float64(1), int64(2), object(8)\n",
      "memory usage: 25.7+ MB\n"
     ]
    }
   ],
   "source": [
    "wineReviews1 = load_wine_data(\"winemag-data_first150k.csv\")\n",
    "wineReviews2 = load_wine_data( \"winemag-data-130k-v2.csv\")\n",
    "\n",
    "wineReviews2.drop(['taster_twitter_handle', 'title', 'taster_name'],axis = 1,inplace=True)\n",
    "\n",
    "wineReviews_all = pd.concat([wineReviews1, wineReviews2], axis = 0)\n",
    "wineReviews_all.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x000001234475B988>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x00000123447B4D48>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x00000123477F52C8>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234782D188>]],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 4 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from pandas.plotting import scatter_matrix\n",
    "\n",
    "attributes = [\"points\", \"price\"]\n",
    "scatter_matrix(wineReviews_all[attributes])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Unnamed: 0   -0.063021\n",
       "price         0.437456\n",
       "points        1.000000\n",
       "Name: points, dtype: float64"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_matrix = wineReviews_all.corr()\n",
    "corr_matrix[\"points\"].sort_values(ascending=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 280901 entries, 0 to 129970\n",
      "Data columns (total 11 columns):\n",
      " #   Column       Non-Null Count   Dtype  \n",
      "---  ------       --------------   -----  \n",
      " 0   Unnamed: 0   280901 non-null  int64  \n",
      " 1   country      280833 non-null  object \n",
      " 2   description  280901 non-null  object \n",
      " 3   designation  197701 non-null  object \n",
      " 4   points       280901 non-null  int64  \n",
      " 5   price        258210 non-null  float64\n",
      " 6   province     280901 non-null  object \n",
      " 7   region_1     280901 non-null  object \n",
      " 8   region_2     111464 non-null  object \n",
      " 9   variety      280901 non-null  object \n",
      " 10  winery       280901 non-null  object \n",
      "dtypes: float64(1), int64(2), object(8)\n",
      "memory usage: 25.7+ MB\n"
     ]
    }
   ],
   "source": [
    "fill = \"nowhere\"\n",
    "wineReviews_all[\"province\"].fillna(fill, inplace=True)\n",
    "wineReviews_all[\"region_1\"].fillna(fill, inplace=True)\n",
    "\n",
    "fill = \"mystery\"\n",
    "wineReviews_all[\"variety\"].fillna(fill, inplace=True)\n",
    "\n",
    "wineReviews_all.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import OrdinalEncoder\n",
    "\n",
    "#Ordinal encoder should be fine for this application since there's not much in the way of similarity between categories.\n",
    "#Might try a one-hot encoder later.\n",
    "cat_encoder = OrdinalEncoder()\n",
    "\n",
    "wine_province = wineReviews_all[[\"province\"]]\n",
    "wine_region1 = wineReviews_all[[\"region_1\"]]\n",
    "wine_variety = wineReviews_all[[\"variety\"]]\n",
    "\n",
    "\n",
    "wine_province_ord = cat_encoder.fit_transform(wine_province)\n",
    "wine_region1_ord = cat_encoder.fit_transform(wine_region1)\n",
    "wine_variety_ord = cat_encoder.fit_transform(wine_variety)\n",
    "\n",
    "wineReviews_all[\"province\"] = wine_province_ord\n",
    "wineReviews_all[\"region_1\"] = wine_region1_ord\n",
    "wineReviews_all[\"variety\"] = wine_variety_ord"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x000001234973C788>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234976DF88>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x00000123497A6DC8>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x0000012349C0FC48>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x0000012349C4AB08>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x0000012349C83988>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x0000012349CBD888>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x0000012349CF5588>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x0000012349CF5788>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x0000012349D2F6C8>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x0000012349DA4248>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x0000012349DD8FC8>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234AAE2EC8>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234AD8AD48>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234ADC4C08>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x000001234ADFEA88>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234AE37908>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234C071748>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234C0AC5C8>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234C0E5488>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x000001234C11F308>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234C159188>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234C18CF88>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234C1C7E08>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x000001234C201CC8>]],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 25 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from pandas.plotting import scatter_matrix\n",
    "\n",
    "#See if there's any notiecable correlations using a scatter matrix\n",
    "attributes = [\"points\", \"price\", \"province\", \"region_1\", \"variety\"]\n",
    "scatter_matrix(wineReviews_all[attributes])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "224721\n",
      "56180\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "\n",
    "#Reset back to normal\n",
    "wineReviews1 = load_wine_data(\"winemag-data_first150k.csv\")\n",
    "wineReviews2 = load_wine_data( \"winemag-data-130k-v2.csv\")\n",
    "\n",
    "wineReviews2.drop(['taster_twitter_handle', 'title', 'taster_name'],axis = 1,inplace=True)\n",
    "\n",
    "wineReviews = pd.concat([wineReviews1, wineReviews2], axis = 0)\n",
    "\n",
    "#Split into training set and test set\n",
    "def split_train_test(data, test_ratio):\n",
    "    np.random.seed(42) #To produce the same shuffled indices each time\n",
    "    shuffled_indices = np.random.permutation(len(data))\n",
    "    test_set_size = int(len(data) * test_ratio)\n",
    "    test_indices = shuffled_indices[:test_set_size]\n",
    "    train_indices = shuffled_indices[test_set_size:]\n",
    "    return data.iloc[train_indices], data.iloc[test_indices]\n",
    "\n",
    "train_set, test_set = split_train_test(wineReviews, 0.2)\n",
    "print(len(train_set))\n",
    "print(len(test_set))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Merlot' 'Syrah' 'Pinot Noir' 'Bordeaux-style Red Blend' 'Riesling'\n",
      " 'Chardonnay' 'Cabernet Sauvignon' 'Red Blend' 'Sauvignon Blanc']\n",
      "['Pinot Noir', 'Chardonnay', 'Cabernet Sauvignon', 'Red Blend', 'Bordeaux-style Red Blend', 'Sauvignon Blanc', 'Riesling', 'Syrah', 'Merlot']\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "#get the wine varieties with more than 4000 reviews\n",
    "variety_common = train_set.groupby(\"variety\").filter(lambda x: len(x) > 6000)\n",
    "varieties = variety_common[\"variety\"].value_counts().index.tolist()\n",
    "\n",
    "print(variety_common['variety'].unique())\n",
    "print(varieties)\n",
    "\n",
    "fig, ax = plt.subplots(figsize = (25, 10))\n",
    "sns.countplot(x = variety_common['variety'], order = varieties, ax = ax)\n",
    "plt.xticks(rotation = 90)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['alcohol',\n",
       " 'alcoholic',\n",
       " 'alder',\n",
       " 'ale',\n",
       " 'alejandro',\n",
       " 'alene',\n",
       " 'alentejo',\n",
       " 'alert',\n",
       " 'alessandro',\n",
       " 'alex',\n",
       " 'alexander',\n",
       " 'alexandre',\n",
       " 'alfalfa',\n",
       " 'alfio',\n",
       " 'alfrocheiro',\n",
       " 'alicante',\n",
       " 'alighieri',\n",
       " 'alignment',\n",
       " 'aligoté',\n",
       " 'alike',\n",
       " 'alisos',\n",
       " 'alita',\n",
       " 'alive',\n",
       " 'alkalinity',\n",
       " 'alla']"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.feature_extraction import text\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "\n",
    "train_set = train_set.drop_duplicates('description')\n",
    "#may want to get rid of variety names, years, and punctuations\n",
    "punc = ['.', '. .', '. . .', ',', '\"', \"'\", '?', '!', ':', ';', '(', ')', '[', ']', '{', '}',\"%\", '$', '+', '/', '-', '1', '2', \n",
    "        '3','4','5','6','7','8','9', 'flavor'] \n",
    "stop_words = text.ENGLISH_STOP_WORDS.union(punc)\n",
    "desc = variety_common['description'].values\n",
    "vectorizer = TfidfVectorizer(stop_words = stop_words)\n",
    "X = vectorizer.fit_transform(desc)\n",
    "\n",
    "word_features = vectorizer.get_feature_names()\n",
    "word_features[500:525]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nltk.tokenize import TweetTokenizer\n",
    "from nltk.stem.snowball import SnowballStemmer\n",
    "\n",
    "stemmer = SnowballStemmer('english')\n",
    "tokenizer = TweetTokenizer()\n",
    "\n",
    "def tokenize(text):\n",
    "    return [stemmer.stem(word) for word in tokenizer.tokenize(text.lower())]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Mathias\\anaconda3\\envs\\HOML\\lib\\site-packages\\sklearn\\feature_extraction\\text.py:386: UserWarning: Your stop_words may be inconsistent with your preprocessing. Tokenizing the stop words generated tokens ['abov', 'afterward', 'alon', 'alreadi', 'alway', 'ani', 'anoth', 'anyon', 'anyth', 'anywher', 'becam', 'becaus', 'becom', 'befor', 'besid', 'cri', 'describ', 'dure', 'els', 'elsewher', 'empti', 'everi', 'everyon', 'everyth', 'everywher', 'fifti', 'forti', 'henc', 'hereaft', 'herebi', 'howev', 'hundr', 'inde', 'mani', 'meanwhil', 'moreov', 'nobodi', 'noon', 'noth', 'nowher', 'onc', 'onli', 'otherwis', 'ourselv', 'perhap', 'pleas', 'sever', 'sinc', 'sincer', 'sixti', 'someon', 'someth', 'sometim', 'somewher', 'themselv', 'thenc', 'thereaft', 'therebi', 'therefor', 'togeth', 'twelv', 'twenti', 'veri', 'whatev', 'whenc', 'whenev', 'wherea', 'whereaft', 'wherebi', 'wherev', 'whi', 'yourselv'] not in stop_words.\n",
      "  'stop_words.' % sorted(inconsistent))\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "['10',\n",
       " '100',\n",
       " '15',\n",
       " '20',\n",
       " '2010',\n",
       " '2015',\n",
       " '2016',\n",
       " '2017',\n",
       " '2018',\n",
       " '2019',\n",
       " '2020',\n",
       " '2022',\n",
       " 'accent',\n",
       " 'acid',\n",
       " 'ad',\n",
       " 'add',\n",
       " 'addit',\n",
       " 'aftertast',\n",
       " 'age',\n",
       " 'alcohol',\n",
       " 'almond',\n",
       " 'alongsid',\n",
       " 'alreadi',\n",
       " 'ampl',\n",
       " 'anis',\n",
       " 'anoth',\n",
       " 'appeal',\n",
       " 'appel',\n",
       " 'appl',\n",
       " 'approach',\n",
       " 'apricot',\n",
       " 'aroma',\n",
       " 'aromat',\n",
       " 'astring',\n",
       " 'attract',\n",
       " 'auster',\n",
       " 'background',\n",
       " 'bacon',\n",
       " 'bake',\n",
       " 'balanc',\n",
       " 'barrel',\n",
       " 'basic',\n",
       " 'bean',\n",
       " 'beauti',\n",
       " 'beef',\n",
       " 'befor',\n",
       " 'berri',\n",
       " 'best',\n",
       " 'better',\n",
       " 'big']"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vectorizer2 = TfidfVectorizer(stop_words = stop_words, tokenizer = tokenize, max_features = 500)\n",
    "X2 = vectorizer2.fit_transform(desc)\n",
    "word_features2 = vectorizer2.get_feature_names()\n",
    "word_features2[:50]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Bootleg transformation pipeline to get dataframe into format I need\n",
    "\n",
    "def pipeline(data_set, threshold):\n",
    "    #first extract common varieties based on threshold of how many times the variety appears in the dataset\n",
    "    variety_common = data_set.groupby(\"variety\").filter(lambda x: len(x) > threshold)\n",
    "    varieties = variety_common[\"variety\"].value_counts().index.tolist()\n",
    "    \n",
    "    data_set = train_set.drop_duplicates('description')\n",
    "    drops = ['.', '. .', '. . .', ',', '\"', \"'\", '?', '!', ':', ';', '(', ')', '[', ']', '{', '}',\"%\", '$', '+', '/', '-', '1', '2', \n",
    "        '3','4','5','6','7','8','9', 'flavor'] \n",
    "    stop_words = text.ENGLISH_STOP_WORDS.union(drops)\n",
    "    desc = variety_common['description'].values\n",
    "    vectorizer = TfidfVectorizer(stop_words = stop_words, tokenizer = tokenize, max_features = 500)\n",
    "    usable_set = vectorizer.fit_transform(desc)\n",
    "    word_features = vectorizer.get_feature_names()\n",
    "    return usable_set, word_features, variety_common\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KMeans(n_clusters=15, n_init=5)"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.cluster import KMeans\n",
    "\n",
    "kmeans = KMeans(n_clusters = 15, n_init = 5)\n",
    "kmeans.fit(X2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 : citrus, finish, palat, white, peach, lemon, fresh, grapefruit, lime, wine\n",
      "1 : wine, rich, ripe, fruit, age, wood, acid, textur, drink, charact\n",
      "2 : blackberri, cherri, sweet, currant, soft, dri, tannin, oak, wine, rich\n",
      "3 : chardonnay, pineappl, butter, oak, vanilla, toast, peach, sweet, acid, tropic\n",
      "4 : black, cherri, fruit, wine, pepper, aroma, tannin, finish, palat, plum\n",
      "5 : berri, aroma, finish, plum, herbal, palat, feel, note, nose, oak\n",
      "6 : wine, fruit, finish, cherri, aroma, note, spice, red, tannin, raspberri\n",
      "7 : fruiti, wine, drink, attract, acid, readi, crisp, fresh, soft, light\n",
      "8 : veri, wine, dri, fruit, tannin, ripe, rich, acid, cherri, good\n",
      "9 : pinot, noir, cherri, cola, silki, raspberri, acid, dri, wine, drink\n",
      "10 : wine, structur, tannin, fruit, firm, age, dens, rich, ripe, drink\n",
      "11 : palat, tannin, cherri, alongsid, aroma, offer, black, berri, spice, red\n",
      "12 : cabernet, sauvignon, blend, merlot, franc, petit, verdot, syrah, wine, malbec\n",
      "13 : light, wine, fruit, aroma, fresh, red, finish, color, cherri, strawberri\n",
      "14 : appl, pear, green, finish, lemon, fresh, palat, acid, crisp, wine\n"
     ]
    }
   ],
   "source": [
    "common_words = kmeans.cluster_centers_.argsort()[:,-1:-11:-1]\n",
    "for num, centroid in enumerate(common_words):\n",
    "    print(str(num) + ' : ' + ', '.join(word_features2[word] for word in centroid))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x1080 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Create a heat map to see which clusters above correspond to which varieties\n",
    "variety_common['cluster'] = kmeans.labels_\n",
    "clusters = variety_common.groupby(['cluster', 'variety']).size()\n",
    "fig2, ax2 = plt.subplots(figsize = (30, 15))\n",
    "sns.heatmap(clusters.unstack(level = 'variety'), ax = ax2, cmap = 'Reds')\n",
    "\n",
    "ax2.set_xlabel('variety', fontdict = {'weight': 'bold', 'size': 24})\n",
    "plt.xticks(rotation = 90)\n",
    "ax2.set_ylabel('cluster', fontdict = {'weight': 'bold', 'size': 24})\n",
    "for label in ax2.get_xticklabels():\n",
    "    label.set_size(16)\n",
    "    label.set_weight(\"bold\")\n",
    "for label in ax2.get_yticklabels():\n",
    "    label.set_size(16)\n",
    "    label.set_weight(\"bold\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Ok so that's fun, but let's try to find the optimal number of clusters. I tried\n",
    "#running KMeans to get an elbow graph several times, but it takes a long time so \n",
    "#I'm using mini-batch K-Means instead because it's much faster. \n",
    "from sklearn.cluster import MiniBatchKMeans\n",
    "'''\n",
    "n_clusters: The number of clusters/centroids to form\n",
    "random_state: Determines random number generation for centroid initialization\n",
    "n_jobs: Number of threads used for computation, -1 uses all processors\n",
    "n_init: Number of times algorithm will be run with different centroid seeds\n",
    "'''\n",
    "kmeans_per_k = [MiniBatchKMeans(n_clusters = k, random_state=42, n_init = 5).fit(X2) for k in range(1,25)]\n",
    "intertias = [model.inertia_ for model in kmeans_per_k]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x252 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8,3.5))\n",
    "plt.plot(range(1,25), intertias, \"bo-\")\n",
    "plt.xlabel(\"$k$\", fontsize = 14)\n",
    "plt.ylabel(\"Intertia\", fontsize = 14)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Not particularly insightful, the more clusters there are the less inertia and I don't think it'll dramatically\n",
    "#decrease with more clusters because grouping words is difficult using K-Means. Try a silhouette score just\n",
    "#for fun. WARNING THIS TAKES LIKE 3 HOURS TO RUN COMPUTING THESE ON SUCH A LARGE SET IS ROUGH\n",
    "from sklearn.metrics import silhouette_score\n",
    "\n",
    "silhouette_scores = [silhouette_score(X2, model.labels_) for model in kmeans_per_k[1:]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgcAAADwCAYAAACUqcezAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deZgU1dX48e9hE0ZkkU22GYgiokMURTAxCcZEI8aoWQSVBIx5JcQYk5jNSGJWsrgmvm4v/vQNGoxLQhJ8Q6JmUROJC+4wqCDCzAAKuKAssp7fH6cr9DTdM9UzVV3VPefzPPP0THV31e2enqlT995zrqgqzjnnnHOBDkk3wDnnnHPp4sGBc84555rw4MA555xzTXhw4JxzzrkmPDhwzjnnXBMeHDjnnHOuiU5JNyAJffv21WHDhiXdDOecc64knnjiiQ2q2i/s49tlcDBs2DAWLVqUdDOcc865khCRVcU83ocVnHPOOdeEBwfOOeeca8KDA+ecc8414cGBc865gubOhWHDoEMHu507N+kWuVJolxMSnXPOtWzuXJg+HbZssZ9XrbKfAaZMSa5dLn7ec+Cccy6vmTP3BAaBLVtsu6tsHhw455zLq76+uO3tWaUNv3hw4JxzLq/q6uK2t1fB8MuqVaC6Z/ilnAMEDw6cc87lNWsWdO269/axY+0kWO7CXO2rwsaNUFcH998P//u/8OMfwxe+AB/7GBx5JEydWnnDLz4h0TnnXF5TpsATT8DVV4MIDB1qvQa/+x186Uvwy19Cx45Jt7J18k22/Oxn4a67oFcvaGyE1avta9OmvZ/fpw8MGQKDB8NTT+U/xqpV8PjjFkyJxPda4uDBgXPOuYKGD7fbtWthwAC7kv7Wt+Dyy2H9erj1Vthnn2Tb2Br5Jlvu2AHz51sQNHgwjB4NJ520JwgYPNi+HzSoaY/KsGEWCOQSgXHjrHdhxgw46yzo3j3WlxUZDw6cc84V1NBgJ/9+mSV7ROCyy6B/f/jGN+C11+D3v4f99ku2ncUqNKlSpPgJl7NmNe2FAKiqsp6V7dvhxhvt/q99DT7zGfj85+Hd725920vB5xw455wrqKHBrpY75Jwtvv51mDMHHngAjjsO1q1LonWtF+VkyylTYPZsqKmx4KKmxn7+r/+C88+HZ56Bhx+G00+Hm2+Gww+HY4+F226DrVvb9jri4sGBcxWm0lKqXLIaGqybPZ+pU60bfulSO9mtWFHatrXFrFl2dZ+tqsq2t8aUKbByJezebbfZRaJE4L3vtSGY1avhyithwwZ7/wYPhosughdesMem5e9XtBKmnBZp7Nix6ks2u0qUO8kK7B/e7Nle0c61TnW19QzcemvhxzzyCHz0o9ClC/zlL3ZlXA7mzIFzzrHva2osMCjV34mq9brceCPMmwc7d8KoUfDSSzYUEYjq71dEnlDVsWEf7z0HzlUQr2jnorRrF6xZ03JX+zHHwL/+BZ06wQc+AA8+WJr2tdWECXZ78817X+3HTQQ++EG4807rnfnJT+DFF5sGBpDc368HB85VEK9o56K0dq0FCIWGFbKNGgULF1o3+Uc+YlfDadfYaLdDhiTbjgMOgG9/24Yk8kni77ekwYGInCQiL4jIchG5OM/9IiLXZO5/VkSOzLrvFhFZJyKLC+z76yKiItI3ztfgXJoNGpR/uyp8/ON2RdcORxJdKwUnpTDBQfC4f/4TxoyBM86w7vA0S0twEEhTRcqSBQci0hG4DpgIHAqcJSKH5jxsIjAi8zUduCHrvl8BJxXY91DgBKCk8VVaJo44BzZbPN+VR7ducNpp9k/7uOMs53rOHNi2reRNdGWmocFuizk59ekDf/2r1Qf4/OfhRz9Kb0CatuAg6kmSbVHKnoNxwHJVXaGq24E7gNNyHnMacKuaR4BeIjIQQFUfAl4vsO+rgW8CJfsIVmItbVe+XnsNPvxhK/P63e82Tam66Sb4wx/sH/1NN1mhl3POsX/43/8+vPpq0q13aRUEB2F7DgL77mufualT4dJLrZrirl3Rt6+tGhutPkOPHkm3xBRKiUxiMnEpg4PBQEPWz42ZbcU+pgkRORVYrarPtPC46SKySEQWrV+/PnyrC/CJXy4t3nzTxnhffBH++Ef44Q/zp1R162Z51889Z1d248bBD35gQcI55xQuAevar4YGO3n27Fn8czt3tnUIvv51uO46S3WsqUlXT2tjY3p6DQLNpUSWUimDg3yVpXOv9MM8Zs+DRaqAmcClLR1cVWer6lhVHdsvKPXVBj7xy6XBpk1w8snw7LNW7/7DH275OSLwoQ/BPfdYbvX06fDb39pww4QJVu0uuMrzobP2rb6++F6DbB06WJnls86CRx+1/aWppzWNwUFalDI4aASyP2ZDgDWteEy2A4HhwDMisjLz+CdF5IA2t7YFaZo44tqnLVtsVbjHHoM77rA882IdfDD893/bP8krr7R/2p/4BBx0EHz603DeeT501p41NETzP23hwr23paGn1YODwkoZHDwOjBCR4SLSBTgTmJ/zmPnA1EzWwjHARlVdW2iHqvqcqvZX1WGqOgwLLo5U1Vdieg3/kW/iSLduyUwcce3Ptm17sg9uvdVO6G3Rq5dVaVu+3HogqqstCMgt7ZqGf+iudJqrjliMNPa07txpqZoeHORXsuBAVXcCFwD3AkuBu1R1iYjMEJEZmYctAFYAy4GbgPOD54vIb4B/AyNFpFFEPleqtueTO3EE7B+0V6FzcduxAyZNgvvus+ItZ58d3b47dbLP8YMPFl5itlKGzqIaMqnUoZd33rEMmCiCgzT2tK5da+P6HhwUoKrt7uuoo47SqB19tOrhh0e+W+ea2LFD9YwzVEH1uuviPVZNjR0n96trV9UHH1TdvTve48fp179Wrapq+rqqqmx7EvtJo2XL7PX86ldt31ca36eFC60df/pTcm0oJWCRFnGe9CWbIzJtGlxwga2+VS51xV152b0bzj0X7r7b5gecf37Lz2mLfMvQdu5sXxMmWMnciy+2eQ+5K/alXaFsoy9+0SZp7tzZ9GvHjvw/L1hQeOil3HsRW5vGmE/wXlx8sY3z9+5tc12SfI/SVuMgdYqJJCrlK46egw0bVDt3Vr3oosh37Zzu3q163nl2pfOjH5XuuL/+tfUgiNjtr3+tumWL6vXXqw4fbu0ZNcquLrdtK1272kokf69I8NW5s2q3bqr77afau7dq//6qgwapVlervutdqiNHqh52WOHniyT9Cttuzhx7LS++GN0+d++29/TCC6PbZ2tddZW9vtdeS7olpUGRPQdlFu+nV58+cMopNt64c2fSrXGVRBW+8hUrYHTJJfCd75Tu2Plyrrt1gy98weoq3H679SScc45lOPziF5ZemXbNjYGr2uI3W7bAW2/B669boajVqy1j46WX4PnnYfFim3NUzP7LSTC3JMoraxF7b9Iwb6Wx0T7LvXsn3ZJ08uAgQlOn2j+R++5LuiWuUqjagizXXANf/Sr8+MdJt2iPTp0sf/3pp617ffhwa2NNjVVefO21pFtY2KxZdmLIVlVlK+MVu59KzVpqaIB+/fZ+n9oqTcHBkCGFJ962dx4cROjkk60Hobl1z50rxg9/CD//uV2pX3llOv+RicDEiZbh8PDD8L737am8+JWv2IkgbTP6p0yBr33Nvm9Lmdp8WUunnlr+8w0gujTGXGkLDlx+HhxEqEsXu5L6wx+spK2rfHGe9C67zK7AP/tZuPbadAYGud77XivhvHgxfOpTVjZ32DCbsJu2YkqDM4XZV61qW5na7KGXceNg2bKoWpisuIKDmhrYsGHvCaGl5sFB8zw4iNi0aVag5u67k26Ji1vUi29lBxr77w/f+haceabNNSi3bIDDDrOVH5cvh+7d9150Jw3FlOrqrG1RniAmTYInn7TXXe7q6+OZOxHsM8neg127YM0aDw6aU2b/ctLvqKNg1Cj7x+gq2yWX5E+H+/rXbcLamjWweXO45WpzA4033oCOHa27vmPHeNpfCjU1hScoJt21vHSp/a1G2SMzaZLd3nlndPtMwltv2VdcwwqQ7O9/3TqbOO7BQWFe5yBiItZ7cPHFNqv5wAOTbpHLZ+5cu3INro5mzSrcrawKr7wCS5bY1WZwW+if2yuv2Ekn0LGjLQnbs+ee2+zve/SA//mfvQONXbtsudupU6N5zUmprragJ9/2JNXVwQknRLvPoUNtaOXOO5PvGWmLKGsc5EpDcOA1DlrmwUEMpkyxGea33moTs1y6BFfpwck4GA4AOO64pgFA8H32HJLeva3bvHv3/FfF/frBL38JGzfa1Vf2bfD96tV25Rps27Ejf1uTvrqOQr5iSlVVyc7of/NN69k59NDo9z15Mnz5y3t6JspR8LmLIzgYPNiGyTw4SDcPDmIwZIgtnXvrrfC975XfeHGlK1Qd7zOfaToE0KePBQFnnmknkcMOs9sBA6yHKDfIADvpXX21TUwtRnX1nqu13O3lLuiRueQSOyH06AHXX5/sjP6lS+02juDgU5+yLI277rK//3IUfBbj+Px16mQBQr7epFLx4KBlRQUHIjIWWyb5/1R1s4jsC2xTW1TJZZk61U42Dz8M739/0q1xW7fCI4/AAw8U/qekalkBQSDQr1/z49HByS3s8ERzfvrT9F1dR2nKFPs69FAYMSL5VL+6OruNIzgYNMj+5u+804aFyiHLJFdDg13UDBwYz/6TTmdsbLTssr59k2tD2oUKDkRkALac8tGAAiOw1ROvAt4BvhxXA8vVxz9u3c5z5nhwkITsYOCBB+z77dvtH16XLvZ9rpoaq61fjOCk11ZRBhppVlsLTzyRdCssOOjWrXCFw7aaPNk+S4sXw+jR8RwjTg0NFuR0iqlvuboaHn00nn2H0di4Z3jD5Rf2rbkaeAXoA2R3yN4NnBh1oyrBvvta9+Jdd+29MEu5SFvhmubas3Ur/OMf1o07YQL06gXHH28VBbdsgQsvhP/7PyuFe8ste1e1S8NVer5SxZWmthZWrLAsjiTV1cEhh8SXCfLJT9rntFyzFuJKYwwEw2i7d8d3jOZ4jYOWhQ0OPgTMVNU3cra/BFTAqGg8pk6Ft9+2okjlJuoc/jja87nPwSc+0XIw8PjjcPnl8NGPWnZAblW71lbHc8WrrbXboFs/KXV18U4WHDAAPvhBuzgIk8qaNnEVQArU1Ngk3Fdeie8YzfHgoGVhg4NuQJ6OWPphwwoujwkTLEIux3LKhSbtJZWela8927bB73/fcjCQT3u4Sk+joIt98eLk2rBpk10ZxzHfINvkyVYt8emn4z1O1FTjDw6STGdU9eAgjLDBwUPAOVk/q4h0BL4F/C3qRlWKDh1sUuJ991naVLnYtq3wpL2kJhEVOq5IuGDApcO73gVduyYbHDz/vN3GHRx84hM2bFFuQwvr19v/gEoNDjZssDlHHhw0L2xw8E3gPBG5H9gHuBKoA44Fvh1T2yrC1Kl2dXr77Um3JJyFC2HMmML3J5Va19wSu658dOxoJ+Ukg4M4MxWy9eljKc133lleQwtxpjEGkgwOPI0xnFDBgarWAaOBhcB9QFdsMuIYVX0pvuaVv4MPhmOOsayFNP+DePttuOACW1Fv82b4xjfStRTtJZfsvS0Nkwhd8Wprkw8OOncuTfXSyZNt2Orxx+M/VlTirI4YCKqEenCQXi0GByLSWUQeBXqq6vdU9RRVPVlVv6Oqa0vQxrI3bZr9M0zr2OOf/mRXUddfD1/6klUEvOyyvZeiHTMmubH5xx+3dgwc6JMIy11trQ2zvf56Msevq4ORI+NL08t2+ukWiNx1V/zHikopggMoXFY7bh4chNNicKCqO4DhWH0D1wqTJllufdoWY1q3zir5nXKKRfELF1rZ3+7d7f7sSXvf/a7df999pW/nQw/B//t/tqDRmjU+ibDcBRkLS5Ykc/y6uviHFAK9e8NHPmLBQVJpe8Wqr4d99rEiYHFKqhBSY6MFhgMGlP7Y5STsnIM5wHlxNqSS7b8/nHqqzTsoVEO/lFQtg2LUKJg3D374Q1tm9phjCj/nkktsiGTGjNKuw75tm6UwDh9evqVoXVNBcJDE0MLWrVZnoVTBAdjQQkODFeIqB0GmQtyVHZMMDgYNKu/VTkshbHCwLzBdRJ4WkZtF5Jrsr7AHE5GTROQFEVkuIhfnuV8y+1wuIs+KyJFZ990iIutEZHHOcy4Xkeczj/+9iPQK255SmjrVZgH/5S/JtuPll+1KZto0KwLz1FPWK9ClS/PP69rVVg58+eXSLib1s5/BCy/ADTdYYSlX/oYMsfUVnnuu9Md+4QULjksZHJx6ql2Jl0vWQtxpjIGaGhtaKrSkd1w8jTGcsMHBKOBJ4A3gXdjkxOCrNswOMqmP1wETgUOBs0Qk9090IlaaeQQwHbgh675fASfl2fX9QK2qvht4kZRmT5x0knXTJVXzYNcuuOoqu2r797/huuvgn/8s7p/kccfBuefClVfCM8/E1tT/eP55+MlPbOjjIx+J/3iuNESSm5QYLLhUytUSe/SAiRPh7rvt7zDtShUcJJWx4MFBOGGzFT7YzNfxIY81DliuqitUdTtwB3BazmNOA25V8wjQS0QGZtrwELDXFCZVvS9r4adHgFT+2jt3hrPPhvnzSz8R69ln4T3vga99zaoI1tXB+ee3rq745ZfbMMl558X7j273bvj856234Oqr4zuOS0YQHJQ6g6euzrqTR4wo7XEnT4a1a20htjTbudOWEy9FinASwYEXQAqvqNODiHQVkVoROUxEuhZ5rMFA9qK0jZltxT6mOecCfy6yXSUzbZoV34hz5nL2+gPV1XDaaXDUUTaB7447LDhpy1XB/vvDL35h2QPXXx9Vq/f2v/9rExEvv9wnDlWi2lp44w07YZZSXR0cdJB185fSKadYKnDahxbWrrXAvFJ7Dt580+ZMeXDQslDBQSad8XJsWOEZ4DngDRG5TEQ6hzxWvuktudcNYR5TqI0zgZ1A3ur/IjJdRBaJyKL169eH2WXkjjjC/inGlbWQu/5AQ4MFA+95j3WnTp4czSSjoJv/kkv2pD1F6dVXLTNhwgQbxnCVJ6kyyqXMVMjWvbsFCL/9rV2dp1Wp0hhhz6TAUgYHnsYYXtieg58DnwZmAAdjcwK+AHwG+GnIfTQC2R+5IUBuUeEwj9mLiEwDTgGmqObvqFTV2ao6VlXH9os7R6cAEes9eOQRePHF6Pefb/0BsD++Pn2iO46ITRDctcsKJ0XdNfzVr9rr+J//iX/GtEvGYYfZbSmDg+3bba2DJIIDsJTmdevgwQeTOX4YwYm6FMFBx452ki5lrQMPDsILGxycDXxOVeeo6kuZr18B/wWEzTZ/HBghIsNFpAtwJjA/5zHzgamZrIVjgI0tFVoSkZOwNR5OVdUSJtm1zpQp1uV/223R77tQBB5HZD58uGUtzJ9vix9F5S9/gd/8xgKdkSOj269Ll379bLiolMHBsmUW0CYVHJx8ss2hSfPQQilKJ2crdTqjBwfhhQ0OemLLM+d6CQiVOpiZNHgBcC+wFLhLVZeIyAwRmZF52AJgBbAcuAk4P3i+iPwG+DcwUkQaReRzmbuuBfYD7s+kWt4Y8jUlYuBAOPFEy1qIsijK5s2Fx1Hj+kP/ylfg8MOt92Djxrbvb/Nm+MIXLMXyW99q+/5cupU6Y6FUayoUUlVlaY3z5qWj3kk+DQ2WXdGjR2mOl0Rw0KEDHHBA6Y5ZrsIGB88AF+bZ/mUgdFFgVV2gqger6oGqOiuz7UZVvTHzvarqFzP3j1bVRVnPPUtVB6pqZ1Udoqo3Z7YfpKpDVfWIzNeM/EdPj6lT7Q/ioYei2d9rr9kCL++8s3e9gjjXH+jcGW66ydZkz7f2QbG+/32bODl7duknjLnSq621KomlqhxYV2fDVEn2SE2ebH+vf/97cm1oTqnSGAPV1XbCLlWKZ2OjBQadw86Ua8eKWZVxmoi8KCJzRORXIvICNg/hG/E1rzKdfrpF5lFMTGxogPe/34oZzZsHt9yyZz2EUqw/cPTRth7DDTdYeeXWevppS1k87zx7Pa7y1dba3JKVK0tzvKVLbTisW7fSHC+fk06yv/20Di3U15d2pdOaGpugWaqsFU9jDC9snYOHgJHYSozdgR6Z70eq6r/ia15l6tYNzjjDZi5v3tz6/dTVwXvfa3nJ994LH/940/UQSrX+wI9/DIMHW6bE9u3FP3/XLgsK+vaFn/88+va5dArKKJeqUmJSmQrZ9tnHLg5+//vW/a3ELYmeAyjd0IIHB+GFrnOgqqtVdaaqflJVP5FZlbHFTAKX37RpVja0tZP5/v1vu8LeudOGJyZMiLZ9xdhvP6t5sGQJXHFF8c+/9lpYtMgWferdO/r2uXQKTtSlmHewc6eVTk46OADLWnjzzWQWMWvO1q1W4t2DAwfh6xxcICKfzrP90yJyfr7nuOYde6x1cbamnPKCBfChD1lBoocftkmBSfvYx+CTn7RFnJYtC/+8+nrLTJg40f5puvajRw/rVi5FcLBihV2ppyE4OOEEC4LTNrQQzOSv1ODgrbfsy4ODcML2HHyFppULAyuBr0bWmnakQwebmPjXv+75owzjtttsxvOoUfCvf8G73hVfG4t1zTXWbTpjRrjaB6p76iRcf73XNGiPSpWxkHSmQrYuXWwI8I9/tEnEaVHqNEawXsfevUsTHKxebbceHIQTNjgYAuQrVdFIStcyKAef+YydGOfmrem4tyuvtIBiwgT4xz/SV1Z40CBbRfHvfw/XIzJvHtxzj/U2DBsWe/NcCo0ebQtsxT3+HgQHhxwS73HCmjwZ3n47+VVas5WyOmK26urSFELyGgfFCRscvAIckWf7kcCG6JrTvhx4ILzvfZa10NyVtip885tWUviMM2xYoVR5yMX6/OdtkuRFF9n4ZSEbN1qWw5gx8OUvl659Ll1qa20+QDFDUa1RV2cnof32i/c4YR1/vFUtTdPQQhAclPrkWapaBx4cFCdscHA7cI2InJBZZ6GziJwI/IICaxm4cKZOtRSrJ57If/+OHfDZz9oCROefb9UD01wDoEMHS598+21bBbKQb3/b1lCYPRs6dSpd+1y6BBkLcQ8tpCFTIVunTjZH55578pc8T0J9PfTvD12LXVKvjUodHAwaFP+xKkHY4OB7wMNYdcMtma8/AwuB78bTtPZh0iQ72eerebBli41NzpljpYqvvdbqkafdYYdZT8dtt8H99+99/8KFcOONcOGFMHZs6dvn0mPkSPtMxxkc7N5tQxdpCg7AhhY2b4Y//SnplphSpzEGamose+Ott+I9TmOjBT9pvrhKk7B1Dnao6lnYoktnY+spjFTVM1U1pYVAy0PPnpb3/JvfNB13ff11q3q4YIEVGLr00vKasPed78CIETY5MfvKaPt2q4cwdCj86EfJtc+lQ9eu9jmJMzhYtcrS9EaNiu8YrTFhgs0binMJ92IkFRyUKmPB0xiLE7rOAYCqLlfVu4F52DwEF4GpU62k6oIF9nNjo9UweOIJuPtuO8GWm65drXdgxYqmQcAVV1g9hOuus2VsnYs7YyFNmQrZOnaET33Keg42bUq6NR4cuKaaDQ5E5EMiMiln28XAJuBNEfmLiIRaeMkVduKJNsHw7LNtzH7YMHjpJZvJ/MlPJt261jv+eDjnHKt6OGiQvbaZM2HcOFvb3jmw4OCll+Ibew+Cg7T1HIANLWzdanMPkrRxo3XrlzKNMeDBQTq11HNwMVmpiiIyDvgJcBu23sLhwMzYWtdO3Hmn/WPcutUyE3btsiGENRVQf/KYY+w1rV27JyPjuefCp2+6yldba5+N4CQetbo6Ww01jdU3jz3WAueksxaSSmMEWwipU6d4g4MtW2yo1oOD8FoKDkYDD2b9fAawUFXPU9WrsJUaT42rce3FzJmWzpXtnXdse7n76U/33rZ1a2W8NheNuDMW0papkK1DB0tP/vOfo1n2vLWSDA46drTjxlnrwAsgFa+l4KAXsC7r52OB7LIdjwODo25Ue1MoYi7lOudxqeTX5qJx4IE2gzyO4CDokUhrcAA2tLB9O8yfn1wbgr/HJIYVguPG+T/BaxwUr6XgYC1wIICI7AOMAf6ddf9+wLZ4mtZ+FPqDTOoPNUqV/NpcNDp1svkAcQQHjY022S/NwcExx9jfQ5JDCw0NdgU/cGAyx/fgIH1aCg7+DFwmIscDPwc2A//Muv/dwPKY2tZuzJoFVVVNt1VV2fZyV8mvzUVn9Oh4goO0ZipkE7F6J/fdB2+8kUwbGhps7kNSdVSqq63rP3d4NSpBcDDY+7lDayk4uBR4B/grcC5wnqpmV0E/F8hT5sYVY8oUqxRYU2P/KGpq7OcpU5JuWdtV8mtz0amttZND1CfHpUvtNo2ZCtkmT7ZqqK1dwr2tkkpjDNTU2ETstWvj2X9jo61im3uh4gprtnCtqm4APiAiPYFNqror5yFnYGmNro2mTKncE2YlvzYXjWBS4pIltt5IVOrqoG9f6Ncvun3G4aijbIXVO++Ec88t/fHr6+Hoo0t/3EAwzLhqVTxBiqcxFi9shcSNeQIDVPX1nJ4E55wrWlwZC2mfjBgIhhb+9jfYUOKl7FTt5Jlkz0HctQ48OCheURUSnXMuDkOH2oqJUQYH5ZCpkG3yZOtanzevtMddvx62bfPgwDXlwYFzLnEi0ZdRfvVVm8NQLsHB4YdbQaALL9xTKbUUxcKCE3KSwcG++9oS1nEEB9u2wbp1HhwUq6TBgYicJCIviMjyTBnm3PtFRK7J3P+siByZdd8tIrJORBbnPGd/EblfRJZlblNYB80515LaWqueGVTSbKtyyFTIdvvttsbKtm32HqxaZYuUxR0gBAWQkk4vrq6OpxBSUGnWg4PilCw4EJGOwHXAROBQ4CwRyf2znQiMyHxNB27Iuu9XwEl5dn0x8DdVHQH8LfOzc67M1NZaidtXIlrSrdyCg5kzLWMh25Yt8VcTTbI6Yra4ah14jYPWCR0ciMhoEblWRP4sIgMz204XkTEhdzEOWK6qKzKTGO8ATst5zGnArWoeAXoFx1LVh4DX8+z3NGBO5vs5wOlhX5NzLj2inpRYVwe9ellXfTlIqmGEygkAABuxSURBVJpoQ4Ototq3b7zHaUlcwUEQ/HhwUJxQwYGInMieUsnHA90ydx0IfC/ksQYDDVk/N7J36eUwj8k1QFXXAmRu+4dsj3MuRaIODpYutfoGItHsL25JVROtr7deg6Tfp5oaWxky6jUmvOegdcL2HPwIuEhVPw5kpy4+gPUIhJHvo5c7uhjmMa0iItNFZJGILFq/fn0Uu3TORah/f/uKsuegXIYUILlqokkXQApk1zqIUmMj9Ohh2TAuvLDBwWHAgjzbXwf2D7mPRiD7IzgEyF2UOMxjcr2aNcwxkKYLRf2Hqs5W1bGqOrZf2iuiONdORZWxsGGDzVAvp+AgqCYarG/Qp09pqommLTiIemjB0xhbJ2xw8Ab5u/ePxE7oYTwOjBCR4SLSBTgTyF2HbD4wNZO1cAywMRgyaMZ8YFrm+2nAH0O2xzmXMrW1ViVx9+627Scom1xOwQFYINDQAN27w5lnxh8Y7Nxps/k9OHC5wgYHtwOXi8gQrJu/k4hMAK4Abg2zA1XdCVwA3AssBe5S1SUiMkNEZmQetgBYgS3mdBNwfvB8EfkNtiLkSBFpFJHPZe76GXCCiCwDTsj87JwrQ7W1sHlz27uWyy1TIVvHjjB2LDz6aPzHWrPGArGk0xgBBgyALl08OEiLZtdWyPIdLJVwFTYvoC5zezsQekRMVReQMzyhqjdmfa/AFws896wC218DPhS2Dc659MqelDh8eOv3U1dnV99puCJujfHj4aqr4J13LJMgLmlJYwQr/DR0aLTBwY4dlhrrwUHxwq6tsENVp2D1ByYBZwOHqOpn8q254JxzrXHYYXbb1nkHdXXllamQa/x4O7E99VS8x0lTcADRF0Jau9YKSnlwULywqYyXikhVpkbBb1X1LlVdJiLdROTSuBvpnGsfevSwE0QUwUE5DikExo+328cei/c4aSidnC3qWgeexth6YeccfA/onmd7FeHrHDjnXIuCMsqttXGjjaWXc3AwaJCd0OKed9DQAD17WlCWBtXV9rvLrRTZWh4ctF7Y4EDIX29gDPmrFjrnXKvU1sLzz7f+BBFkKowaFV2bkjB+fGmCg7T0GoAVQtq9e896CG3lwUHrNRsciMjbIvIWFhisEJG3sr42Y5kHd5Wioc659qG21gKDZcta9/xyzlTINm4crFhhSyrHJW3BQdSFkBobrZBUr17R7K89aSlb4QKs1+AWYCaQXdhyO7BSVf8dU9ucc+1QdsZCa07wdXU2w3/YsEibVXLZ8w4++tF4jlFfD0cfHc++WyPqWgdBGmO5TkxNUrPBgarOARCRl4GHM7UKnHMuNqNGWVrb4sUwaVLxz6+rg0MOsXoB5eyoo+x9ePTReIKDrVutkmSaeg6CtkQdHLjihZ1z8HfylEkWkT4i4qmMzrnIdO0KI0a0PmOh3DMVAt27Wy9KXPMOgvH4NAUHVVXQr58HB2lQzITEfPah6UJMzjnXZq1dY2HTJhuvroTgAGxo4bHH2l5OOp/gBJyG6ojZoqp1sGuXTWz04KB1mh1WEJGLMt8qMENENmXd3RF4P/B8TG1zzrVTtbUwb551fXfr1vLjA89n/htVUnBw002wfDkcfHC0+05bAaRAdTW88ELb9/PqqxYgeHDQOi1NSPxS5laA/wKyhxC2AyuBGTjnXIRqa62y3dKlcOSR4Z9XrgsuFRJMSnz00fiCg7SdPKur4f777ffflomEnsbYNs0OK6jqcFUdDjwIHB78nPkaqaofUdUSLA/inGtPsjMWilFXB507w4EHRt+mJIwaZXMP4ph30NAA/fvDPvtEv++2qKmx4aE332zbfjw4aJuwayt8UFXfABCRASISdq6Cc84V7aCDbIW+Yisl1tXZFXansEvKpVycKzTW16dvvgFEl87owUHbhF1boZOIXCYibwOrgWGZ7T8XkfObfbJzzhWpUye7am5Nz0GlDCkExo+HZ56xFRqjlLYCSIGoCiE1NlqvSN++bW9TexS2B+D7wMeATwPbsrY/BpwTbZOcc674jIWtW62iYCUGB1Gv0KhqV+ZpDg6i6DnwAkitFzY4OAuYoap/BLKTahYDEU+Tcc45Cw4aG8OPPb/4oqX8VWJwANEOLWzcaOP6aRxW6NfPrvijCg5c64QNDgYB+Tp5OtFyxoNzzhUtmJS4ZEm4x1fKmgq54lihMa1pjGBVIaNYutmDg7YJGxwsAT6QZ/sk4InomuOcc2b0aLsNO7RQV2cT+EaMiK9NSQmKIUUlzcEBtL0Q0u7dsHq1BwdtEfaq/wfAr0VkKFb86AwROQQ4G4hpSRDnXHtWXW1pfGGDg6VLLcshbal5URg/Hn73O1uhsV+/tu+vHIKDe+9t/fM3bIDt2z04aIuwqYz3YL0EJ2JzDr4HjAA+pqp/ja95zrn2SqS4SYl1dZbhUImyV2iMQn299bIMHBjN/qJWXQ1r19oJvjU8jbHtQtcrUNV7VXWCqnZX1SpVfZ+q3hdn45xz7VttrdU6UG3+cdu3w7JllTffIHDkkXtWaIxCQwMMHpzelStraux3vnp1657vwUHbeTEj51xq1dbCa6/BunXNP275cti5s3KDg6hXaExrjYNAW2sdeHDQdmGLIL0tIm8V+gp7MBE5SUReEJHlInJxnvtFRK7J3P+siBzZ0nNF5AgReUREnhaRRSIyLmx7nHPpFraMcqVmKmSLcoXGcgkOWpux0NhohbT694+uTe1N2J6DC7BFmIKvi4DbgC3AzDA7EJGOwHXAROBQ4CwRyf1TnojNZRgBTAduCPHcy4AfqOoRwKWZn51zFSAIDloqo1xXZ3MURo6Mv01JGT/eaj4sW9a2/ezebcFBGmscBILApS3BweDBNhTjWidUtoKqzsm3XUSeBD4E/HeI3YwDlqvqisxz7wBOA+qyHnMacKuqKvCIiPQSkYFYueZCz1WgR+b5PYE1YV6Tcy79+ve38rdheg6GD4eqqtK0KwnZxZDaEgStX29zNNLcc9C1KwwY0LbgwIcU2qatcdU/sLLKYQwGGrJ+bsxsC/OY5p77FeByEWkArgC+HbI9zrmUC5uxUIlrKuQKVmhsa8ZC2tMYA20phOTBQdu1NTg4E9gQ8rH5KlznzkEu9JjmnvsF4KuqOhT4KnBz3oOLTM/MSVi0fv36kE12ziWtttaqJBYaa9+500onV3pw0LEjHH102yclBifcNA8rQOsLIal6cBCFsBMSn8tMEAy+nhORdcAPgZ+EPFYjkB2rDmHvIYBCj2nuudOAeZnv78aGL/aiqrNVdayqju0XRRUR51xJjB5t6wAUuop8+WXYtq3ygwOIZoXGcus5aCmNNdcbb9giXB4ctE3YCom/zfl5N7AeeEBVnw+5j8eBESIyHFv2+UyswmK2+cAFmTkF44GNqrpWRNY389w1wATgAeB4oI3TdZxzaZKdsTBs2N73B5kKlVoAKdu4cXtWaHzPe1q3j4YGG9Pv0yfatkWtpga2bIHXXy+urZ7GGI2wExJ/0NYDqepOEbkAuBcrwXyLqi4RkRmZ+28EFgAnA8uxTIjPNvfczK7PA34pIp2Ad7AsB+dchTjsMLtdvBhOOWXv+9tTcJA9KbEtwcHQoelfyjg7ndGDg9IrakVFETkeSyVUYImqPlDM81V1ARYAZG+7Met7Bb4Y9rmZ7f8CjiqmHc658tGzp53MCk1KrKuz+/fbr7TtSkIUKzTW16d/vgE0LYQ0Zkz453lwEI1QwYGIDAZ+j52Eg7H+QSKyCPi4qnr6oHMuNs1lLLSHTIVs48e3LThoaIATToiuPXFpbSGkxkarb3DAAdG3qT0Jm61wDbALOEhVh2YyA0Zktl0TV+Occw4sOFi61DITsu3ebdvbW3Dw8stWr6BYO3bYgkZpn4wIVt+iW7fWBQcDB1qFRNd6YYODE4AvqurLwYZMQaILM/c551xsamutcM/y5U23r1plM9PbW3AArat3sGaNBVTlMKwg0rpaB57GGI221jmIoMq3c841r1AZ5fawpkKuo46ymgetGVoolzTGQGtqHXhwEI2wwcHfgGtE5D8fKRGpBn6Zuc8552IzapRdSebOO1i6dM/97cW++7Z+hcZyDA685yAZYYODC4EqYIWIrBKRlcBLmW0XxtQ255wDbOz5oIP2Dg7q6mziWe/eybQrKePGtW6FxnIMDl55xYpchfHWW/D22x4cRCFUcKCqDap6JFaD4ArgKmCiqh6lqo1xNtA55yB/xkJ7y1QItHaFxvp66NWrfNI+a2rstjHkWcbTGKNT1JwDVb1fVf9bVa9R1b/G1SjnnMs1erRNSNy61X5Wbd/BARQ/tBAUQCoXxaYzenAQndDJHiIyHlueuT85QYWq+tCCcy5WtbXWjf7881YUZ/Vq60Juj8FBsELjo4/C1Knhn1euwUHYSYkeHEQnbBGkrwOXYWWN19B0NcUil8VwzrniZa+xMGZM+8xUCLR2hcaGhj29DuUgOMkX23MwaFA87WlPwvYcfBm4UFWvjbMxzjlXyEEHQZcue+YdtOfgAOwkf8UVtkJj164tP37LFtiwobx6DvbZxwoaFRMcDBhgnxPXNmHnHPQgz7oGzjlXKp07wyGHNA0O+vaF9roC+/jxVjHyqafCPT64qi6n4ACKS2f0NMbohA0OfgOcFGdDnHOuJdkZC+2tbHKuYiclllsaY6CYQkgeHESn4LCCiFyU9WMD8AMRORZ4FtiR/VhVvSqe5jnn3B61tXD77bBxIyxZApMmJd2i5AwcWNwKjcHVdzmUTs5WXQ333GPZKS0tM93YCB/4QGnaVemam3PwpZyfNwHvzXxlU6zugXPOxSqYlPj3v8Mbb7TvngMoboXGoOdg8OD42hOHmhqbV7FhQ/NDSJs322fCew6iUTA4UNXhpWyIc861JAgO7rrLbj04gN/9zlZobGnuRUODTdbbZ5/StC0q2bUOmnuNq1fbrQcH0WjrwkvOOVcyNTW2tsA999jPHhzYbZjeg4aG8htSgPC1DrzGQbSam3NwTdideBEk51wpdOiwZ9Ghnj1t3L09y16h8ZRTmn9sfX15LlAVtkqiBwfRam7OweiQ+/AiSM65kgly+jduhOHDYdYsmDIl2TYlJVih8bHHmn+cqvUcnHhiadoVpf33h6qq8MFBuc2pSKvm5hx8sJQNcc65lsydCwsX7vl51SqYPt2+b68BwvjxNgdj927rWcln40bYtKn80hjBMhRqasIFB3362Aqeru18zoFzrmzMnAk7djTdtmWLbW+vwqzQWK5pjIEwhZAaGnxIIUotzTn4tqpubmn+gc85cM6VQqETRNgKepVo3Di7ffRRGDky/2PKtQBSoLq65UqQXgApWs31HIwGOmd9X+irNs4GOudcoNCVb7leEUche4XGQiohOFi3bs9y3fl4cBCtgsGBqn5QVd/M+r7Q1/FhDyYiJ4nICyKyXEQuznO/iMg1mfufFZEjwzxXRL6UuW+JiFwWtj3OufIya5ZNTstWVWXb26swKzQ2NECnTnDAAaVrV5Rqauw2mHSYKyiS5MFBdFo150BEOolI9yKf0xG4DpgIHAqcJSK5WcoTgRGZr+nADS09V0Q+CJwGvFtVDwOuaM1rcs6l35QpMHu2nSyCiWqzZ7ffyYiB8ePhmWcKX1nX19ss/o4dS9uuqLSUzugFkKLXbHAgIh8SkUk52y7GSim/KSJ/EZFeIY81DliuqitUdTtwB3ZSz3YacKuaR4BeIjKwhed+AfiZqm4DUNV1IdvjnCtDU6bAypU2O3/lSg8MoOUVGhsayndIAVouhOQ1DqLXUs/BxcB/3m4RGQf8BLgN+CZwOBB2nvBgbAGnQGNmW5jHNPfcg4H3i8ijIvKgiByd7+AiMl1EFonIovXr14dssnPOpV9QKbFQvYNyDw4GD7aeokI9Bx4cRK+l4GA08GDWz2cAC1X1vMxKjBcCp4Y8Vr71tHILKBV6THPP7QT0Bo4BvgHcJbL32l2qOltVx6rq2H7tdQF451xFGjjQTv755h3s3m0nz3KetNmlCwwa1HJw4AWQotNScNALyO6mPxb4S9bPj7P31X8hjUB27DoEWBPyMc09txGYlxmKeAzYDfQN2SbnnKsIhVZoXLcOtm8v754DaL7WQWOjldPeb7/StqmStRQcrAUOBBCRfYAxwL+z7t8P2BbyWI8DI0RkuIh0Ac4E5uc8Zj4wNZO1cAywUVXXtvDcPwDHZ9p4MNAF2BCyTc45VxHGjYOXX7YVGrOVexpjoLq6+TkHPqQQrZaCgz8Dl4nI8cDPgc3AP7PufzewPMyBVHUncAFwL7AUuEtVl4jIDBGZkXnYAmBFZp83Aec399zMc24B3iUii7GJitNU1dd7cM61K4VWaKyk4KChwYZJcnlwEL3mFl4CuBSYB/wVy1CYlskWCJwL3B/2YKq6AAsAsrfdmPW9Al8M+9zM9u3Ap8O2wTnnKlGhFRrLvXRyoLoatm2znpEBA5re19gIhx+eTLsqVbPBgapuAD4gIj2BTaq6K+chZ2BBg3POuQQFKzTm6zno1s1WNyxnQSGk+vqmwcH27fDqq95zELVQRZBUdWOewABVfT2nJ8E551xCxo+3dMbsrvcgjXHvHK7yUqjWwdq1tiS1BwfR8lUZnXOuQowfb8szZ6/Q2NBQ/kMKULhKotc4iIcHB845VyHyTUqsry//yYgAvXrZAlMeHJSGBwfOOVchDjnEcv2D4GDHDut2r4TgIFhLw4OD0vDgwDnnKkTHjjB27J7gYM0aG4+vhOAA8hdCamy0HoWePZNpU6Xy4MA55ypI9gqNlZLGGMhXCCmocVDuEy7TxoMD55yrINkrNFZKAaRAdTVs2ABbtuzZ5gWQ4uHBgXPOVZDsSYmVFhwEtQ4astbo9eAgHh4cOOdcBcleobGhAXr3tjH5SpCbzrhzp0249OAgeh4cOOdchQmKIVVKGmMgtxDSq6/Crl0eHMTBgwPnnKsw48fbCo1PPVVZwcGgQdChw56eA09jjI8HB845V2HGjbPbxsbKCg46d4bBgz04KAUPDpxzrsIsX77n+zvugLlzk2tL1LJrHXhwEB8PDpxzroLMnQtf+tKen998E6ZPr5wAIbvWQWMjdO1a/itOppEHB845V0FmzmxaBwDs55kzk2lP1KqrLQtj924vgBQnDw6cc66C5JYXbml7uamutjUjXn3VaxzEyYMD55yrIIVKJVdKCeWgEFJ9vQcHcfLgwDnnKsisWVBV1XRbVZVtrwRBkLNyJaxe7cFBXDw4cM65CjJlCsyebVfYwTLHs2fb9koQBAeLFtnwggcH8eiUdAOcc85Fa8qUygkGcvXsCT16wMKF9rMHB/HwngPnnHNlpabGeg7Ag4O4lDQ4EJGTROQFEVkuIhfnuV9E5JrM/c+KyJFFPPfrIqIi0jfu1+Gccy451dWwfbt978FBPEoWHIhIR+A6YCJwKHCWiBya87CJwIjM13TghjDPFZGhwAlAhSTrOOecKySYd9C5M/Trl2xbKlUpew7GActVdYWqbgfuAE7LecxpwK1qHgF6icjAEM+9GvgmoLG/Cuecc4kKgoPBg20hJhe9Ur6tg4GGrJ8bM9vCPKbgc0XkVGC1qj7T3MFFZLqILBKRRevXr2/dK3DOOZe4hszZYOVKGDasckpDp0kpg4N8BS5zr/QLPSbvdhGpAmYCl7Z0cFWdrapjVXVsP++Hcs65sjR3Ltx8856fV62qrLUj0qKUwUEjkL146BBgTcjHFNp+IDAceEZEVma2PykiB0Tacuecc6kwcyZs29Z0WyWtHZEWpQwOHgdGiMhwEekCnAnMz3nMfGBqJmvhGGCjqq4t9FxVfU5V+6vqMFUdhgURR6rqKyV7Vc4550qm0teOSIuSFUFS1Z0icgFwL9ARuEVVl4jIjMz9NwILgJOB5cAW4LPNPbdUbXfOOZcO2Us252530RHV9jfBf+zYsbooqKDhnHOubMyda3MMspelrqqqrBLRcRCRJ1R1bNjHexKIc865slHpa0ekha+t4JxzrqxU8toRaeE9B84555xrwoMD55xzzjXhwYFzzjnnmvDgwDnnnHNNtMtURhFZD+TJlHUF9AU2JN2IdsTf79Ly97u0/P0uvb7Avqoaeu2AdhkcuOKIyKJi8mNd2/j7XVr+fpeWv9+l15r33IcVnHPOOdeEBwfOOeeca8KDAxfG7KQb0M74+11a/n6Xlr/fpVf0e+5zDpxzzjnXhPccOOecc64JDw5cQSKyUkSeE5GnRcSXsYyBiNwiIutEZHHWtv1F5H4RWZa57Z1kGytJgff7+yKyOvM5f1pETk6yjZVERIaKyD9EZKmILBGRL2e2+2c8Bs2830V/xn1YwRUkIiuBsarqOckxEZEPAJuAW1W1NrPtMuB1Vf2ZiFwM9FbVbyXZzkpR4P3+PrBJVa9Ism2VSEQGAgNV9UkR2Q94AjgdOAf/jEeumfd7EkV+xr3nwLkEqepDwOs5m08D5mS+n4P9cbsIFHi/XUxUda2qPpn5/m1gKTAY/4zHopn3u2geHLjmKHCfiDwhItOTbkw7MkBV14L9sQP9E25Pe3CBiDybGXbwLu4YiMgwYAzwKP4Zj13O+w1FfsY9OHDNOVZVjwQmAl/MdMk6V2luAA4EjgDWAlcm25zKIyLdgd8BX1HVt5JuT6XL834X/Rn34MAVpKprMrfrgN8D45JtUbvxambsMBhDXJdweyqaqr6qqrtUdTdwE/45j5SIdMZOVHNVdV5ms3/GY5Lv/W7NZ9yDA5eXiOybmdCCiOwLnAgsbv5ZLiLzgWmZ76cBf0ywLRUvOEllfBz/nEdGRAS4GViqqldl3eWf8RgUer9b8xn3bAWXl4i8C+stAOgE3K6qsxJsUkUSkd8Ax2Grpr0KfA/4A3AXUA3UA2eoqk+ii0CB9/s4rLtVgZXA54PxcNc2IvI+4J/Ac8DuzOZLsHFw/4xHrJn3+yyK/Ix7cOCcc865JnxYwTnnnHNNeHDgnHPOuSY8OHDOOedcEx4cOOecc64JDw6cc84514QHB84555xrwoMD51zsROTnInJ/0u1wzoXjwYFzrhSOAJ5OuhHOuXA8OHDOlcLhwFNJN8I5F44HB865WInIAcAAMj0HmXU77hCRJzPLyjrnUsaDA+dc3MYAW4EXRGQk8BiwE1sSfGWSDXPO5efBgXMubkdgC8GcDiwEblLVT6vq1mSb5ZwrxBdecs7FSkTuBE4AOgKnquqDCTfJOdcC7zlwzsXtCGAe0Bnok3BbnHMheM+Bcy42IlIFvA0cA4wAZgMfUNUnE22Yc65ZnZJugHOuoh0OKLBYVR8XkVHAPSIyTlVXJ9w251wBPqzgnIvT4cCyrMmHlwIPA/MzvQrOuRTyYQXnnHPONeE9B84555xrwoMD55xzzjXhwYFzzjnnmvDgwDnnnHNNeHDgnHPOuSY8OHDOOedcEx4cOOecc64JDw6cc84514QHB84555xr4v8DZI2+o8wa0ucAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 576x252 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8,3.5))\n",
    "plt.plot(range(2,25), silhouette_scores, \"bo-\")\n",
    "plt.xlabel(\"$k$\", fontsize = 14)\n",
    "plt.ylabel(\"Silhouette Score\", fontsize = 14)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KMeans(n_clusters=19, n_init=5)"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#ok so 11, 19, and 23 seem like a good number of clusters (closer to +1 the better) so I'll just go\n",
    "#with 21 and rerun the K-Means with that. Honestly they all suck the silhouette score is really telling\n",
    "#me that all of the instances are close to a cluster boundary.\n",
    "\n",
    "kmeans21 = KMeans(n_clusters = 19, n_init = 5)\n",
    "kmeans21.fit(X2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 : meat, smoke, black, wine, pepper, grill, syrah, pair, red, aroma\n",
      "1 : cabernet, sauvignon, blend, merlot, franc, petit, verdot, wine, syrah, cherri\n",
      "2 : berri, aroma, finish, herbal, plum, palat, feel, note, nose, fruit\n",
      "3 : tast, like, wine, smell, sweet, cherri, fruit, acid, finish, oak\n",
      "4 : import, wine, finish, fruit, aroma, palat, nose, note, citrus, bit\n",
      "5 : blanc, sauvignon, crisp, citrus, grapefruit, acid, lime, green, wine, gooseberri\n",
      "6 : wine, fruit, age, structur, tannin, wood, firm, rich, year, ripe\n",
      "7 : green, appl, wine, finish, fruit, acid, lime, aroma, crisp, citrus\n",
      "8 : simpl, sweet, soft, cherri, wine, finish, fruit, raspberri, candi, light\n",
      "9 : pinot, noir, cherri, cola, silki, raspberri, dri, acid, drink, wine\n",
      "10 : fruit, wine, cherri, finish, red, tannin, aroma, light, spice, oak\n",
      "11 : black, cherri, palat, aroma, pepper, tannin, plum, spice, fruit, nose\n",
      "12 : appl, pear, finish, palat, aroma, lemon, fresh, wine, nose, acid\n",
      "13 : sampl, barrel, wine, fruit, tannin, ripe, structur, juici, wood, veri\n",
      "14 : riesl, finish, peach, lemon, palat, citrus, lime, acid, miner, note\n",
      "15 : blackberri, currant, tannin, dri, cherri, oak, rich, wine, cabernet, drink\n",
      "16 : chardonnay, pineappl, butter, toast, oak, vanilla, acid, rich, fruit, pear\n",
      "17 : wine, acid, drink, fruiti, fruit, ripe, fresh, attract, readi, charact\n",
      "18 : —, fruit, wine, finish, cherri, oak, ripe, acid, tannin, note\n"
     ]
    }
   ],
   "source": [
    "#Get the new clusters\n",
    "common_words2 = kmeans21.cluster_centers_.argsort()[:,-1:-11:-1]\n",
    "for num, centroid in enumerate(common_words2):\n",
    "    print(str(num) + ' : ' + ', '.join(word_features2[word] for word in centroid))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1, 13,  4, ...,  9,  3, 13])"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "kmeans21.labels_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABhgAAARLCAYAAAB2l6nxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzde7Rld1Un+u88FSAhhEBCB3NJeMlLwgWByEPUDqKQVgRUvOLQTmjprgECwqV9gG3TIoPupq/dXlDAW7YxCS1XI60QFRQaWmmUh4UCyksiz5BAIEjIQ0JSmf3HWUVty6pTtc4+dfY6e38+Y6yx9v7ttdeedRKKjPPdc/6quwMAAAAAADDG2qILAAAAAAAAdh4BAwAAAAAAMJqAAQAAAAAAGE3AAAAAAAAAjCZgAAAAAAAARhMwAAAAAAAAox236AK2xXVf7EWXwJK58e8XXQHL5ITbLboCls3NX110BQAb6uu/tOgSWDJ18mmLLgHgsPrqzyy6BJZMnXn/WnQNy+TpdfvJ/+74V/rLk/1nroMBAAAAAAAYTcAAAAAAAACMJmAAAAAAAABGW409GAAAAAAA4CC+gT8fPz8AAAAAAGA0AQMAAAAAADCaEUkAAAAAAKyktapFl7Cj6WAAAAAAAABGEzAAAAAAAACjGZEEAAAAAMBK8g38+fj5AQAAAAAAowkYAAAAAACA0YxIAgAAAABgJa3VoivY2XQwAAAAAAAAowkYAAAAAACA0QQMAAAAAADAaJPdg6GqTk7y4CSnDUtXJfnL7r5mcVUBAAAAALAsfAN/PpMLGKrqzCT/OckT84/ru7mqXpfkJ7r709teHAAAAAAAkGRiAUNVnZHkHUlOT3Ko/btvleTJSR5VVQ/v7s9sZ30AAAAAAMC6SQUMSV6U5P8YHn8syQeSXJP1sOH2Sc5Kcs+sBxAvSvIvF1AjAAAAAABLYK0O9T13jtbUAoZzk3SSH+/uVxzqgqp6dpKXJfln21kYAAAAAABwwNT2sDh1OP/GBte8+qBrD6mqdlfV3qrau+eCi7akOAAAAAAAYN3UOhg+m+TMJD9VVS/s7ptnX6yq45L89PD0yo1u1N17kuxJklz3xd76UgEAAAAA2Mmm9g38nWZqAcPrkvx41kOEZ1fVZVnfgyFJ7pDkXklOyPoYpdctpEIAAAAAAGByAcOLkjwuyX2TnJjkgQe9vn/HjQ8N1wIAAAAAAAswqQ6Q7v67JA9P8otJrs56oDB7XJ3kvyR5ZHd/aVF1AgAAAACw863V9I8pm1oHQ7r7y0n+dZJ/XVX3SHLn4aXPdffHF1cZAAAAAACw3+QChllDoCBUAAAAAACAiZl0wAAAAAAAAMfKpPYQ2IH8/AAAAAAAgNEEDAAAAAAAwGhGJAEAAAAAsJKqatEl7Gg6GAAAAAAAgNEEDAAAAAAAwGgCBgAAAAAAYDR7MAAAAAAAsJJ8A38+fn4AAAAAAMBoAgYAAAAAAGA0I5IAAAAAAFhJa7XoCnY2HQwAAAAAAMBoAgYAAAAAAGA0I5IAAAAAAFhJvoE/n5UIGPa98/cXXQJLZu1eD1p0CSyTa7+QOu3ui66CZXLLLYuugGW05j+72Tpfee7TFl0CS+aEX/+9RZcAcFh16l0WXQLAMbMSAQPA5N325EVXwDL56lcWXQHL6Pb/ZNEVABye/5YCpuyGaxZdAcAxI2AAAAAAAGAlrVUtuoQdTa87AAAAAAAwmoABAAAAAAAYzYgkAAAAAABWkm/gz8fPDwAAAAAAGE3AAAAAAAAAjCZgAAAAAAAARrMHAwAAAAAAK2mtFl3BzqaDAQAAAAAAGE3AAAAAAAAAjGZEEgAAAAAAK8k38Ofj5wcAAAAAAIwmYAAAAAAAAEYzIgkAAAAAgJW0llp0CTuaDgYAAAAAAGA0AQMAAAAAADCaEUkAAAAAAKykNROS5rIjOxiq6tSquqWqbl50LQAAAAAAsIp2ZMAwQ74EAAAAAAALMLkRSVX1Y0dx2YnHvBAAAAAAAJbaTv8G/qJNLmBI8stJet6bVNXuJLuT5FXPfWr+1Xc/et5bAgAAAAAAgykGDPvNNf6ou/ck2ZMk+/7HxXMHFgAAAAAAwAFTDBi+muRWSX4lyecOc81tk/zktlUEAAAAAAD8A1MMGN6b5JuS/M/u/u1DXVBVp0bAAAAAAADAHNbmmqPDFPeweFfWxyM9fNGFAAAAAAAAhzbFDoYXJ7kgyZc2uOaLSe6xPeUAAAAAAAAHm1zA0N1fSPKFI1zTST65PRUBAAAAALCM1mJG0jymOCIJAAAAAACYOAEDAAAAAAAw2uRGJAEAAAAAwHZYMyFpLjoYAAAAAACA0QQMAAAAAADAaEYkAQAAAACwknwDfz5+fgAAAAAAwGgCBgAAAAAAYDQjkgAAAAAAWElrtegKdjYdDAAAAAAAwGgCBgAAAAAAYDQjkgAAAAAAWElrMSNpHjoYAAAAAACA0Vaig2HXN5276BJYMvvecOGiS2DJ7LrTmYsugSXSN3xp0SWwbG65Jbnui4uugiVy6/N+ZNElsGxuuGbRFbBs+pZFV8AS6S9/YdElsGTqticvugT4Gh0MAAu263v+5aJLAAAA9hMuAMBRW4kOBgAAAAAAONiaLRjmooMBAAAAAAAYTcAAAAAAAACMZkQSAAAAAAAryYSk+ehgAAAAAAAARhMwAAAAAAAAoxmRBAAAAADASlozI2kuOhgAAAAAAIDRBAwAAAAAAMBoRiQBAAAAALCS1mJG0jx0MAAAAAAAAKMJGAAAAAAAgNGMSAIAAAAAYCWtmZA0Fx0MAAAAAADAaAIGAAAAAABgNAEDAAAAAADsYFV1h6p6bVV9uKo+VFWPrKpTqurNVfXR4XzHmetfUFWXVdVHqupxM+sPraq/Gl57eVVtOERKwAAAAAAAwEpa2wHHUXpZkj/s7vsleVCSDyV5fpK3dPe9k7xleJ6qun+SpyQ5K8m5SV5ZVbuG+7wqye4k9x6Oc4/08wMAAAAAAHagqrp9km9L8mtJ0t1f7e4vJXlikouGyy5K8qTh8ROT/GZ339jdH09yWZKHVdXpSW7f3e/o7k5y8cx7Dum4Lf/TzKmqTkrytCT3SfLpJBd392cOuuatSbq7H7OAEgEAAAAAYCrumeTzSX69qh6U5D1JnpPkzt19ZZJ095VVddpw/V2SvHPm/ZcPazcNjw9eP6xJBQzDDKg/T3KPmeWfqarzuvt3Z9bOSdLbWRsAAAAAAMtlww0GJqKqdmd9bNF+e7p7z8zz45I8JMmzu/tdVfWyDOOQDnfLQ6z1BuuHNbURST+V9bSlZo4Tk1xSVT845kZVtbuq9lbV3j0XXrz1lQIAAAAAwDHW3Xu6++yZY89Bl1ye5PLuftfw/LVZDxw+N4w9ynC+aub6M2fef0aSK4b1Mw6xflhTCxgeP5wvT/KTSf5tks8l2ZXk4tndrI9k9oe++6nnbX2lAAAAAACwYN392SSfrqr7DkuPSfLBJJcmOX9YOz/J64fHlyZ5SlXdpqrukfXNnN89jFO6tqoeUVWV5LyZ9xzSpEYkJbl71lsuvr+7/zxJquqCJG9Nct8kv11V37G48gAAAAAAWBZrtROGJB2VZyf5jaq6dZKPJfkXWW8wuKSqnpbkU0l+IEm6+wNVdUnWQ4ibkzyzu/cN93lGkguTnJDkjcNxWFMLGG4azu/bvzBsPvHYJH+a9Q0lfn8RhQEAAAAAwBR193uTnH2Ilx5zmOtfkuQlh1jfm+QBR/u5UxuR9NnhfP/Zxe7+dJLHJrk6yanbXRQAAAAAAPAPTS1geHfWN3b+0YNf6O4PJ/muJNdtd1EAAAAAACyf2gHHlE1tRNJvJ7lzkjOr6sTuvn72xe7eW1Xfm+SFWd+rAQAAAAAAWIBJBQzd/QdJ/uAI17w165s+AwAAAAAACzKpgAEAAAAAALbL1EcQTd3U9mAAAAAAAAB2AAEDAAAAAAAwmhFJAAAAAACsJCOS5qODAQAAAAAAGE3AAAAAAAAAjCZgAAAAAAAARrMHAwAAAAAAK6nKLgzz0MEAAAAAAACMJmAAAAAAAABGMyIJAAAAAICVZEDSfHQwAAAAAAAAowkYAAAAAACA0VZjRNLarkVXwJJZ+5bHL7oElkh/+fOLLoElU8eftOgSWDJ9/ZcWXQJLZu1h5y66BJbNLfsWXQHL5OabFl0BS6Zuc+KiSwA24Bv48/HzAwAAAAAARhMwAAAAAAAAo63GiCQAAAAAADhI1aIr2Nl0MAAAAAAAAKMJGAAAAAAAgNGMSAIAAAAAYCVVzEiahw4GAAAAAABgNAEDAAAAAAAwmoABAAAAAAAYzR4MAAAAAACsJDswzEcHAwAAAAAAMJqAAQAAAAAAGM2IJAAAAAAAVpIRSfPRwQAAAAAAAIwmYAAAAAAAAEYzIgkAAAAAgJW0ZkbSXHQwAAAAAAAAowkYAAAAAACA0YxIAgAAAABgJVXMSJrH5DoYqursqvovVfXSqrrfsPaAqvrdqvpAVb2+qh666DoBAAAAAGCVTaqDoaoekeRPcqCu3VX1nUnemOSUYe1+SR5dVQ/u7r9dQJkAAAAAALDyptbB8FNJbpWkhuPkJK9LcurMWiU5McnzNrpRVe2uqr1VtXfPr190TIsGAAAAAGDnqR1wTNmkOhiSPDxJJ3lPkouTnJfk7GHttUn+fZLnJ/nBJI/e6EbdvSfJniTJtVf3MasYAAAAAABW0NQ6GO40nP95d/9ykqfOvPaz3f2+JD8zPL/rdhYGAAAAAAAcMLWA4avD+WPD+SMzr+1fu3w4T637AgAAAAAAVsbUfkn/+SR3S3J6kk91976q+lSSW7p733DNnYfz3y2iQAAAAAAAlkNNfZODiZtaB8NfZX3fim/fv9Ddd+/ue85c85Dh/NHtLAwAAAAAADhgah0ML07yO9k4PDgnySeT/MF2FAQAAAAAAPxjkwoYuntvkr1HuOZ5SZ63PRUBAAAAALCsTEiaz9RGJAEAAAAAADuAgAEAAAAAABhtUiOSAAAAAABgu6wZkjQXHQwAAAAAAMBoAgYAAAAAAGA0I5IAAAAAAFhJBiTNRwcDAAAAAAAwmoABAAAAAAAYzYgkAAAAAABWUpmRNBcdDAAAAAAAwGgCBgAAAAAAYDQjkgAAAAAAWEkmJM1HBwMAAAAAADCaDgbYhLrdKYsugSXSN31l0SWwbG5zwqIrYMnUrl2LLoGl43tibLGbb1p0BSyT7kVXwLK59fGLrgDgmNHBAAAAAAAAjKaDAQAAAACAlVS6a+eigwEAAAAAABhNwAAAAAAAAIxmRBIAAAAAACtpzYSkuehgAAAAAAAARhMwAAAAAAAAoxmRBAAAAADASjIhaT46GAAAAAAAgNEEDAAAAAAAwGhGJAEAAAAAsJKMSJqPDgYAAAAAAGA0AQMAAAAAADCaEUkAAAAAAKykMiRpLjoYAAAAAACA0QQMAAAAAADAaDtmRFJV3THJ9yRJd1+84HIAAAAAAGCl7ZiAIcldk1yY5JYkAgYAAAAAAOZStmCYy6QChqp69wYvn3Dgsnp3ku7uh29DWQAAAAAAwEEmFTAkOTtJJ6nhfLA+6DoAAAAAAGABphYwzDpcc4pgAQAAAACAua0tuoAdbmo/v89kPVi4Lslzk9xj5nj8zHX3SHLPjW5UVburam9V7d3z6xcdo3IBAAAAAGA1Ta2D4RuS/MckT0/yi0m+JcmzuvuqqrrD/ou6+5NHulF370myJ0ly7dW6HgAAAAAAYAtNqoOhu6/r7mdlPVj4YJInJ/lwVT0thx+ZBAAAAAAAo9UOOKZsUgHDft39ziQPTvJvkxyf9U6E/3+hRQEAAAAAAF8zyYAhSbr75u5+SZIHJfmTJPddcEkAAAAAAMBgansw/CPd/dEk315V357kpEXXAwAAAADAcqia+hCiaZt8wLBfd7910TUAAAAAAADrJjsiCQAAAAAAmK4d08EAAAAAAABbyYCk+ehgAAAAAAAARhMwAAAAAAAAoxmRBAAAAADASjIiaT46GAAAAAAAgNEEDAAAAAAAwGgCBgAAAAAAYDR7MAAAAAAAsJKq7MIwDx0MAAAAAADAaAIGAAAAAABgNCOSAAAAAABYSWsmJM1FBwMAAAAAADCagAEAAAAAABhtNUYk7VqNPybbyL9TbKHatWvRJbBs9t3s7ym21q2OX3QFALBt9r36pYsugSW064d/YtElAIdRZiTNxW8fAGAZnXjHRVfAMrnhmkVXAACws93ulEVXAHBMGJEEAAAAAACMpoMBAAAAAICVVCYkzUUHAwAAAAAAMJqAAQAAAAAAGE3AAAAAAADASqqa/nF0f476RFX9VVW9t6r2DmunVNWbq+qjw/mOM9e/oKouq6qPVNXjZtYfOtznsqp6edXGFQgYAAAAAABg53t0d39jd589PH9+krd0972TvGV4nqq6f5KnJDkryblJXllVu4b3vCrJ7iT3Ho5zN/pAAQMAAAAAACyfJya5aHh8UZInzaz/Znff2N0fT3JZkodV1elJbt/d7+juTnLxzHsOScAAAAAAAAA7Wyd5U1W9p6p2D2t37u4rk2Q4nzas3yXJp2fee/mwdpfh8cHrh3XcFhQOAAAAAAA7zhG2GJiEITDYPbO0p7v3HHTZo7r7iqo6Lcmbq+rDG93yEGu9wfphCRgAAAAAAGCihjDh4EDh4GuuGM5XVdXvJnlYks9V1endfeUw/uiq4fLLk5w58/YzklwxrJ9xiPXDMiIJAAAAAAB2qKo6sapO2v84yWOT/HWSS5OcP1x2fpLXD48vTfKUqrpNVd0j65s5v3sYo3RtVT2i1ls7zpt5zyHpYAAAAAAAYCXtgAlJR+POSX53GPd0XJLXdPcfVtWfJ7mkqp6W5FNJfiBJuvsDVXVJkg8muTnJM7t733CvZyS5MMkJSd44HIclYAAAAAAAgB2quz+W5EGHWL86yWMO856XJHnJIdb3JnnA0X62EUkAAAAAAMBoOhgAAAAAAFhJtSQzkhZFBwMAAAAAADCagAEAAAAAABhtUgFDVd2nqu5z0NpDq+pNVfXlqrqmqn6vqr5hUTUCAAAAALAcqqZ/TNnU9mD4cJJbMtRVVWcleVuS45Ps/1F+V5JHVNWDu/vyhVQJAAAAAAArblIdDIPZTOZnk5xw0FolOSXJC7azKAAAAAAA4IApBgyzzknSSd6Q5Izh+MOshwyP2+iNVbW7qvZW1d49F1x4jMsEAAAAAGCnWaua/DFlUxuRdLBTh/NzuvuKJKmq5yT5SJK7bPTG7t6TZE+S5IZr+hjWCAAAAAAAK2fqHQxXD+dPz6x9YjjfuL2lAAAAAAAA+00yYKiqfVW1L8lpw9IZMy/ffThfua1FAQAAAAAAXzPFgKEOcXzHzOtPGs7v2ua6AAAAAABYIlXTP6Zsansw/IvDrF+WJFVVWd/4+X1JLtmmmgAAAAAAgINMKmDo7ouO8Hon+e5tKgcAAAAAADiMSQUMAAAAAACwXWrqM4gmbop7MAAAAAAAABMnYAAAAAAAAEYzIgkAAAAAgJVUvoI/Fz8+AAAAAABgNAEDAAAAAAAwmhFJAAAAAACspKpadAk7mg4GAAAAAABgNAEDAAAAAAAwmhFJAAAAAACsJBOS5qODAQAAAAAAGE3AAAAAAAAAjGZEEgAAAAAAK6nMSJqLDgYAAAAAAGC06u5F13DM/f3/9a3L/4dkW53wq69ddAkskf7CZxZdAkumTrvboktgyex7239fdAksmV2PeuKiS2DZ7Nq16ApYJivwexK2Wfl+L1vspFN95X4Lfeys+0z+L/57fuBvJvvP3N9wAAAAAADAaPZgAAAAAABgJdmCYT46GAAAAAAAgNEEDAAAAAAAwGhGJAEAAAAAsJLWzEiaiw4GAAAAAABgNAEDAAAAAAAwmhFJAAAAAACsJBOS5qODAQAAAAAAGE3AAAAAAAAAjGZEEgAAAAAAK6nMSJqLDgYAAAAAAGA0AQMAAAAAADCaEUkAAAAAAKwkE5Lmo4MBAAAAAAAYTcAAAAAAAACMJmAAAAAAAABG2xF7MFTVnZI8MMkNSf6iu7+64JIAAAAAANjh7MEwn0l1MFTVY6vq56vqO2fW/kOSK5K8OcmfJrmiqn54UTUCAAAAAAATCxiSPD/Jv8lQV1U9M8lPZ73ToobjlCQXVdU5C6oRAAAAAABW3tRGJJ01nPcO52cO5+uS/I/h8XckuV3Wg4c/3rbKAAAAAABYKrVmRtI8ptbBcIfhfONwvmeSTnJOd39fd39fkkcPrz1ioxtV1e6q2ltVe3/tY589NtUCAAAAAMCKmlrA8MXh/K3D+crhfNnMNX87nI/f6Ebdvae7z+7us592z6/bwhIBAAAAAICpBQxvz/o+C6+qqkcn+a3h+W9U1eOr6vFJXj1c+7eHuQcAAAAAABxR1fSPKZvaHgwvTfKkJGdmfc+Fm4b17xqO/TrJhdtaGQAAAAAA8DWT6mDo7r1JfjTJV7PeuXDrrIcJGZ7vP34nyS8uokYAAAAAAGB6HQzp7ldX1TuT/HiSxyS5a5JbJflCkvckuai7//sCSwQAAAAAYAmsTX0G0cRNLmBIku7+aJJnL7oOAAAAAADg0CY1IgkAAAAAANgZJtnBAAAAAAAAx5oJSfPRwQAAAAAAAIwmYAAAAAAAAEYzIgkAAAAAgJVUZiTNRQcDAAAAAAAwmoABAAAAAAAYTcAAAAAAAACMZg8GAAAAAABWki0Y5qODAQAAAAAAGE3AAAAAAAAAjGZEEgAAAAAAK6nMSJqLDgYAAAAAAGA0AQMAAAAAADDaSoxIOuGC1y+6BJZM3/DlRZfAEumbblx0CSyZuvmmRZfAktn1zU9YdAksG13obLH+yg2LLoEl0p/460WXwJJZu9/DFl0CsAETkuajgwEAAAAAABhNwAAAAAAAAIy2EiOSAAAAAADgYGVG0lx0MAAAAAAAAKMJGAAAAAAAgNGMSAIAAAAAYCWVr+DPxY8PAAAAAAAYTcAAAAAAAACMJmAAAAAAAABGswcDAAAAAAArqaoWXcKOpoMBAAAAAAAYTcAAAAAAAACMZkQSAAAAAACrac2IpHnoYAAAAAAAAEYTMAAAAAAAAKMZkQQAAAAAwGoqI5LmoYMBAAAAAAAYTcAAAAAAAACMNqkRSVX1sSTXJrkwya9195cXWxEAAAAAAMuqjEiay9Q6GO6e5AFJfiHJ5VX1S1V138WWBAAAAAAAHGxqAcN+leR2SX4syQeq6g1V9bgF1wQAAAAAAAymGjD8XJIrsx40rCV5XJI3VNWHq+oZVXXbI92gqnZX1d6q2rvngouObbUAAAAAAOw8azX9Y8ImGTB0988nuVuSH0nyrqwHDZXk3kl+OclnquoXjnCPPd19dnefvftHzz/WJQMAAAAAwEqZZMCQJN19c3e/prsfmeRhSX4jyU1ZDxpOTvJ/L7I+AAAAAABYZZMNGGZ1997u/udJ7pr18UmfXWxFAAAAAACw2nZEwLBfd181jE+6a5IfXnQ9AAAAAADsYFXTPybsuEUXcJBPJbnlSBd1981JfvPYlwMAAAAAABzKpAKG7r77omsAAAAAAACObFIBAwAAAAAAbJdam/YIoqnbUXswAAAAAAAA0yBgAAAAAAAARjMiCQAAAACA1VRGJM1DBwMAAAAAADCagAEAAAAAABjNiCQAAAAAAFZSrRmRNA8dDAAAAAAAwGgCBgAAAAAAYDQjkgAAAAAAWE1lRNI8dDAAAAAAAMAOVlW7quovq+r3h+enVNWbq+qjw/mOM9e+oKouq6qPVNXjZtYfWlV/Nbz28qojpy8CBgAAAAAA2Nmek+RDM8+fn+Qt3X3vJG8Znqeq7p/kKUnOSnJukldW1a7hPa9KsjvJvYfj3CN9qIABAAAAAIDVtFbTP46gqs5I8t1J/uvM8hOTXDQ8vijJk2bWf7O7b+zujye5LMnDqur0JLfv7nd0dye5eOY9h//xHe3PGQAAAAAA2F5Vtbuq9s4cuw+65P9N8lNJbplZu3N3X5kkw/m0Yf0uST49c93lw9pdhscHr29oNTZ5tlEHW61kc2ydOvlOiy6BpdOLLoAl0zfftOgSWDK1tuvIF8EIdfyJiy6BJVL3OXvRJbBsbtm36AqAHa679yTZc6jXqurxSa7q7vdU1TlHcbtD/bK8N1jf0GoEDAAAAAAAsHweleQJVfVdSY5Pcvuq+m9JPldVp3f3lcP4o6uG6y9PcubM+89IcsWwfsYh1jfka9gAAAAAAKykqpr8sZHufkF3n9Hdd8/65s1v7e4fSXJpkvOHy85P8vrh8aVJnlJVt6mqe2R9M+d3D2OUrq2qR9T6h543857D0sEAAAAAAADL5T8muaSqnpbkU0l+IEm6+wNVdUmSDya5Ockzu3v/LLdnJLkwyQlJ3jgcG6r1DaGX3PV/twJ/SLZT33Dtoktgmdx846IrYMnU8bdbdAksGXswsNXswcCWu/Xxi66AZWJePlvOr6XYYiff2YazW+j6Jzxy8v8jPfHSd0z2n7kOBgAAAAAAVtPaZH93vyPYgwEAAAAAABhNwAAAAAAAAIxmRBIAAAAAAKupjEiahw4GAAAAAABgNAEDAAAAAAAwmhFJAAAAAACspPIV/Ln48QEAAAAAAKMJGAAAAAAAgNGMSAIAAAAAYDVVLbqCHU0HAwAAAAAAMJqAAQAAAAAAGE3AAAAAAAAAjLaj9mCoqlOSXN/dNy66FgAAAAAAdrZaswfDPCbXwVBV51XVn1XVe6vqJVW1q6rOrKr3JPl8kmuq6mWLrhMAAAAAAFbZ6A6Gqvq2maef7e6/2apiqur7k1yYpIel/zPJbZJ8Q5IHD2u3TvKsqvpwd79qqz4bAAAAAAA4epsZkfTHORAA7E6yZQFDkmcN59m+lKdnPWT46vBZ98l6yHBeEgEDAAAAAACbU0YkzWMzI5KuyYEA4F1bWEuSfGPWw4tnJTklybOT3DbrdZ7b3Q9M8s+Ga79hoxtV1e6q2ltVe/dccOEWlwkAAAAAAKttMx0MH0zyyOHxVu/hcLvhfEF3f6WqLkjyS8PaO4bznw3n2250o+7ek2RPkuT6v+uNrgUAAAAAAMbZTEDwqzOPf2CrCnKtWlIAACAASURBVBlcM5x/oqruk+QnZ177/oM+85oAAAAAAMBmrdX0jwkbHTB094VJXp31MUnPr6oXVtXJW1TPXw7nFyX5UJJ/l+TaYe3VVXVdkouyPkbpL//x2wEAAAAAgO0wekRSVb016+HCviS7sh4CvKCqPp3kiiS3HOJt3d2POYrb/1KS2ev2JfmhJK9JcvscGIvUSV4xtnYAAAAAAGBrbGYPhnOy/gv+DOdKcpsk90ry9Ye4vmau31B3X1pVP5TkqVkPF/6/7n5DVT0hycuS3C/JZ5K8tLtfv4naAQAAAAAgSVI17RFEU7eZgOFgW7qBcnf/VpLfOmjtbUkevJWfAwAAAAAAbN5mAwaxDgAAAAAArLDNBAwv2vIqAAAAAABgu635Lv08RgcM3S1gAAAAAACAFbe26AIAAAAAAICdZ+5Nnqvq+CTflOTrkpyQ5LXdfcO89wUAAAAAgGOqjEiax6YDhqo6LclLkvxIklvPvPTHVfWQJD8+PL+8u8/bfIkAAAAAAMDUbCpgqKr7JXlTkrskmY14eji/LclrkhyfpKvqpd39gXkKBQAAAAAApmP0HgxVdZsklyY5I+vhQudAsJAk6e4vJvmjmaXvmaNGAAAAAABgYjbTwfCvktwr66FC5UDIcLA/SvLE4fG3bKo6AAAAAAA4RsoeDHMZ3cGQ5PtmHr83ydfnH45J2u/9M4/vv4nPAQAAAAAAJmozAcMDZh7/bHd//DDXfX44V5LTNvE5AAAAAADARG1mRNLJM48/usF1J835OQAAAAAAcOysGZE0j810MHx55vHpG1x39szjL23icwAAAAAAgInaTMDwNzOPn3moC6rqLkl+enjaST64ic8BAAAAAAAmajOji96c5JHD4ydX1dsPev0Xk3xbklNm1t60ic/ZOrWZHAUOr048+cgXASxK96IrYMk845T7LroElsyvXPuJRZfAsvH/fWylW91m0RWwbPwdBZNWZUTSPDYTMLwyyU8kOSHrGzg/cua1SvKk4bz/b8/rkvzqHDUCAADA4Z106qIrYJnccM2iKwCAHWP0V/u7+6okT9//dObcBz3fHzI8o7uvnrNOAAAAAABgQjbTwZDu/m9VdWOSVyS502EuuzrJM7v7ks0WBwAAAAAAx8yaEUnz2FTAkCTd/dtV9YYkT07yrUlOz3rXwpVJ/leS13b3dVtSJQAAAAAAMCmbDhiSpLuvT3LRcAAAAAAAACtidMBQVW8dHnaSHxr2ZDjUdScm+cH9z7v7gk1VCAAAAAAAx0IZkTSPzXQwnJMDmzkfv8F1d0ryX2euFTAAAAAAAMCSWNvk+8bEOiIgAAAAAABYMpsNGI7GnY7hvQEAAAAAgAU64oikqnpgkm88zMtPrqovHGL9hCQ/PPN83yZqAwAAAACAY6bWDOCZx9HswfC9SV54iPVK8v8c4b09XPf5kXUBAAAAAAATdrSbPB8uxtko3umZ89uPuiIAAAAAAGDyjjZgSA4EBnWItUPZf911SV48pigAAAAAADjmyoikeRxNwPClJJ+ceX63HAgWrkhy8yHec1PWxyK9M8kruvtj8xQJAAAAAABMyxEDhu5+WZKX7X9eVbfMvPyo7v7UsSgMAAAAAACYrjEjkvb7VA50MByqewEAAAAAAKZvzYikeYwOGLr77segjkOqqlsnuWOSm7r7i9v1uQAAAAAAwMY208FwWFV1xyTnJblvkuuTvLG73zryHscleXaS85OclWRtWL8uyf9M8p+7+39tZd0AAAAAAMA4owOGqnpIkl8Ynu5L8sTuvqGq7pbkz5J83czlz6uql3X3847y3rdN8qYkj9y/NPPySUm+J8njq+qF3f3vx9YOAAAAAAD7VRmRNI+1TbznMUnOSfJPkxzX3TcM6y9OcnoOhAI1HM+pqn96lPf+mSTfPPPe/TrJZ2dqfnFVfecmagcAAAAAALbAZgKGR8w8/qPka2ONvjfrQUDnH4YDSfLUo7z3Dw7v/7Uk3zgcFwz3+50k90vygeH5czdROwAAAAAAsAU2EzDcd+bxXwznByY5cXj8maz/8v8jw/NK8vCjvPeZw/m53f3+7n5/kucMa0/o7o8m+bHh+Yb3rKrdVbW3qvbuueDCo/x4AAAAAABWxlpN/5iwzWzyfNrM408N5/vPrL2iu19eVZ9I8rph7YyjvPeXk5ya9fFLbxjWzhnO/2Q47x3Ot9voRt29J8meJMkN1/RRfj4AAAAAAHAUNhMw3GHm8c3Debar4b3D+cMza8cf5b3fluT7kvxeVX006+OS7jOcPzhcsz9Y+OLRFgwAAAAAAGytzYxI+srM43sN54fMrF02nPf3bnSOPgz4uSR/P7z33lkPF/bf5z8N528ezu8/ynsCAAAAAABbbDMdDJ9Mctbw+Jeq6k1JHjc8vyHJx4bHd555z1VHc+Pu/uuq+o4kL0/y0GH5s0n+XXf/1vD8E0l+KAf2eAAAAAAAgPFq2nscTN1mAoY/zYGA4euTPD3rXQad5E+6e/9+Bw+Yec9lOUrd/Y4k31RVJyc5vrs/d9Dr70vyvk3UDQAAAAAAbJHNjEh6ZZJ9w+ODN09+5czjx848fvvYD+nuaw4OFwAAAAAAgGkYHTB09/uTnJ/kmqx3LlSSG5O8sLvfkCRVdack5w5vqaxv3gwAAAAAANNRNf1jwjYzIind/Zqqem3WxyAdl+SD3X3dzCVfTHLqzPXXz1UlAAAAAAAwKZsKGJKku7+a5C8O89otSYQKAAAAAACwpDYdMAAAAAAAwI428RFEUzc6YKiqF27mg7r75zfzPgAAAAAAYHo208Hwc0l6E+8TMAAAAAAAwJKYZ0TSmN6RzQQSAAAAAABw7KytLbqCHW2r92A4VJBgiBUAAAAAACyZzQQMb8vhOxJOTXK3JCcN13SStye5ZVPVAQAAAAAAkzQ6YOjuczZ6vf43e3cfZetd1gf/e81JjEl4MRze3yQLU1qgylsMVqsRELPUCrZaUpAX5fH4GKn6VKJSKoIpFXmsolKMhxogFaFE4YFS0SIGqC4aCBjkXcAQQOCQk5AQgxBOzvX8sfchk2FmztlnXu577/35rHWve+a3f/ue79mz1yQz1/27flUnJHl8khckuX2SD3X3vuNKBwAAAAAAO6U04NmKbW8w1d2HuvviJD+TSXukp1bVY7b76wAAAAAAAMPZyR0s3rHq45/awa8DAAAAAADssu3e5Hm1fzE9V5KH7ODXAQAAAACA2WmRtCUzFxiq6lmbPLwnySlJ7pfkezLZ5LmSfM1xpQMAAAAAAEbpeFYwPDuTwsHRHCn9dJIPHsfX2TZ97aeG/PIsoJufd/7QEVggK0/4saEjsGD6oP/usb1us2cnu2qyjG7+Df8vxfba87TnDh2BBdJf+vuhI7BgDr/2oqEjsGD2/Ph/GjoCfMVWWiRttnakc+sixO9u4esAAAAAAAAjs1N7MKxevfAb3b1/h74OAAAAAAAcH3swbMnxFBjems1bJB1Kcl2S9ya5pLvffzzBAAAAAACA8Zq5wNDdZ+9ADgAAAAAAYI7sVIskAAAAAAAYt5WVoRPMNa8eAAAAAAAws01XMFTVk7brC3X3xdt1LQAAAAAAYFhHa5H00my+ofMsFBgAAAAAABiPqqETzLVj3YNhq6/ydhUpAAAAAACAETjWAsNWCgRKQAAAAAAAsGCOVmD4eKw+AAAAAABgEWmRtCWbFhi6+z67lAMAAAAAAJgjK0MHAAAAAAAA5s+x7sEAAAAAAACLRYukLZm5wFBV35jkKdNPO8mzu/uGNXNul+RZuWWFxEu6+z1byAkAAAAAAIzI8axgeGKSn8mkuPCatcWFJOnuz1fVPZI8bjrv5iTnbyUoAAAAAAAwHsezB8PZqz7+/U3m/cEGzwEAAAAAAObc8axguNeqj9+/ybwPbfAcAAAAAAAY3srx3IPPEcdTYDht1ccnbjLvyGO15jnHrKoekOTMJHeeDn02yTu6+33Hcz0AAAAAAGB7HE+B4YYkd5h+/M+TbPTH/m9f85xjVlUPTrI/yUM2ePydSfZ19xWzXBcAAAAAANgex7P+42+n50ryS1V1v7UTpmO/mMkGz73qOUdVVf8kyVsyKS7UBsfDkrylqv7xceQHAAAAAICkavzHiB3PCoY3Z/IH/k5ylyTvrqpXJPnr6dg3Jnl8Ji2Sajr2lhmu/5wkt0ly8/R570ty/fRat0vygCTfMZ3z7CTnrneRqtqXZF+SXPi8Z2ffE/71DBEAAAAAAIDNHE+BYX+Sf5dbigdfk+RJa+YceSzT8/4Zrn/29DmP6+5Xrzehqn4wyauSPGKji3T3/iNftz/5gd5oHgAAAAAAMLuZWyR190eS/EomRYRkUgxY28JodXHhV7v7wzN8idtNz5duMufPp+fbznBdAAAAAAC4xdDtj+a8RdLx7MGQJM9K8hu5dUFh9XFk/De7+5kzXvuq6fkFVXX3tQ9Ox14w/fTjs0cHAAAAAAC26rgKDD3xs0nOTPKSTDZx/uL0+Nvp2Dd39787jsu/MpPixA8n+URVfa6qPjY9rkvyiSRPyKSQ8YrjyQ8AAAAAAGzN8ezB8BXd/c4kT92mLEc8L8l3Jvm26ee3zy1tk1avB/nf07kAAAAAADC7kbcgGrvjbZG0Y7r7HzIpMPx0knclOZRbWi4dmo79VJJHdPcXh8oJAAAAAABDq6qvraq3V9W7q+p9VfWc6fgdquqNVfXh6fm0Vc95RlV9pKo+VFXfvWr8oVX1nuljv1W1eQVmdAWGJOnum7v7t7v7YUlOTXL36XFqdz+su1/Y3TcPmxIAAAAAAAb3pUxuyP+mJA9Kck5VPTzJLyR5U3efkeRN089TVfdPcm6SByQ5J8mLqmrP9Fq/k2RfkjOmxzmbfeEttUjaDd395SSfWT1WVXuTXJ3kcHeP/t8AAAAAAMD41Moo78GfSXd3kr+ffnri9Ogkj0ly9nT8ZUnenOTnp+Ov7O4vJbmyqj6S5Jur6mNJbtfdb0uSqro4yWOTvGGjrz3vr54GWQAAAAAALLWq2lNVVyT5bJI3dvdlSe7S3Z9Okun5ztPp90jyiVVP/+R07B7Tj9eOb2h0d/9X1XnHMO3UHQ8CAAAAAAADq6p9mbQtOmJ/d+9fPWe6pcCDqurrkrymqh642SXXGetNxjc0ugJDkhfmKKEBAAAAAGDLNt/DeBSmxYT9R504mXtdVb05k70TDlTV3br701V1t0xWNySTlQn3WvW0eyb51HT8nuuMb2jMLZLqKAcAAAAAACy1qrrTdOVCqurkJI9K8sEkr0vy5Om0Jyd57fTj1yU5t6pOqqrTM9nM+e3TNko3VNXDq6qSPGnVc9Y1xhUMN2WyCcWFSQ5sMOeUJOfvWiIAAAAAABinuyV5WVXtyWRRwau6+/VV9bYkr6qqpyb5eJIfSpLufl9VvSrJ+5McSvKT0xZLSfITSV6a5ORMNnfecIPnZJwFhiuSnJnk0u6+ZL0JVbU3CgwAAAAAACy57v7rJA9eZ/yaJI/c4DnPTfLcdcYvT7LZ/g23MsYWSZdl0gLprKGDAAAAAACwwKrGf4zYGFcwXJDkoiTXbTLn2iSn704cAAAAAABgrdEVGLr7YJKDR5nTSa7anUQAAAAAAMBaoyswAAAAAADArhh5C6KxG+MeDAAAAAAAwMgpMAAAAAAAADNbihZJderXDR2BBVPf9wNDR2CBrHzT2UNHYMH0NX83dAQWzPPf9gdDR2DBHL7kpUNHYNGcdPLQCVggddLJ6c9fPXQMFkj98+8dOgKwmRX34G/FUhQYAAAAAI5V3fUbho7AAulrDwwdAWDHKM8AAAAAAAAzs4IBAAAAAIDlVDV0grlmBQMAAAAAADAzBQYAAAAAAGBmWiQBAAAAALCctEjaEisYAAAAAACAmSkwAAAAAAAAM1NgAAAAAAAAZmYPBgAAAAAAlpM9GLbECgYAAAAAAGBmCgwAAAAAAMDMtEgCAAAAAGA5rbgHfyu8egAAAAAAwMwUGAAAAAAAgJlpkQQAAAAAwHKqGjrBXLOCAQAAAAAAmJkCAwAAAAAAMLO5bJFUVZXkXknS3R8fOA4AAAAAAPNIi6QtmcsCQ5I7JPlYksOZ338DAAAAAADMrXlvkaS8BAAAAAAAAxjd3f9V9fxjmHbyMVxnX5J9SfK7v/787HvKD281GgAAAAAAi2Rl3u/BH9boCgxJnp6kt3qR7t6fZH+S5HOf3vL1AAAAAACAW4yxwHCE9kcAAAAAADBSYywwfCGTFki/lOTjG8y5TZLf3rVEAAAAAADArYyxwPDOJN+W5Mru/v31JlTV3igwAAAAAACwFaWRzlaMcQeLyzJpj3TW0EEAAAAAAID1jXEFwzOTPDfJlzea0N3XZJzFEQAAAAAAWAqjKzB095eTXD90DgAAAAAAFpwWSVsyl6sAqmpvVR2uqkNDZwEAAAAAgGU0lwWGVZSXAAAAAABgAKNrkVRV5x3DtFN3PAgAAAAAAItNi6QtGV2BIckLk/TQIQAAAAAAgI2NscBwhNIRAAAAAACM1BgLDDclOTHJhUkObDDnlCTn71oiAAAAAAAWz8q8b1M8rDEWGK5IcmaSS7v7kvUmVNXeKDAAAAAAAMBgxlieuSyT9khnDR0EAAAAAABY3xhXMFyQ5KIk120y59okp+9OHAAAAAAAFlLZCngrRldg6O6DSQ4eZU4nuWp3EgEAAAAAAGuNsUUSAAAAAAAwcqNbwQAAAAAAALtCi6QtsYIBAAAAAACY2XKsYDjplKETsGD2fPsPDh2BBfL57zt76AgsmNu9/s1DR2DB1B3uNnQEFszKs7916AgsmMMfvnzoCCyQuvsZyY3XDx2DBbJynwcOHQFgx1jBAAAAAAAAzGw5VjAAAAAAAMBa5R78rfDqAQAAAAAAM1NgAAAAAAAAZqZFEgAAAAAAy2mlhk4w16xgAAAAAAAAZqbAAAAAAAAAzEyLJAAAAAAAllO5B38rvHoAAAAAAMDMFBgAAAAAAICZaZEEAAAAAMByqho6wVyzggEAAAAAAJiZAgMAAAAAADAzLZIAAAAAAFhOK+7B3wqvHgAAAAAAMDMFBgAAAAAAYGYKDAAAAAAAwMzswQAAAAAAwHKqGjrBXLOCAQAAAAAAmNkoVzBU1dcn+X+S/KMkn0iyv7vfuWbOlUkOd/d9B4gIAAAAAABLbXQFhqq6R5J3JNm7avhHq+rp3f2bq8a+PknvajgAAAAAABZHafKzFWN89Z6Z5I5JatWxJ8mvV9XTj/UiVbWvqi6vqsv3X/TSHQkKAAAAAADLanQrGJI8IpOVCe9JckGSr03yjCT3T/KrVXWwu196tIt09/4k+5MkX7jeSgcAAAAAANhGYyww3Ht6Pre7P5gkVfX/JfnTJP8syf6qOjhUOAAAAAAAFkTV0Anm2hhbJN04PV95ZKC7b0zyvUnem0lR5JUD5AIAAAAAAKbGWGD41PT8kNWD3X19kkdnUng4ZbdDAQAAAAAAtxhjgeFtmWzsfN7aB7r7M0m+K8lndjsUAAAAAAALZmVl/MeIjXEPhv+WySbPh6vq1Gl7pK/o7r+tqkcnefp0HgAAAAAAsMtGV2Do7r9M8pdHmfPeJE/ZlUAAAAAAAMBXGV2B4VhU1d4kn03S3T2X/wYAAAAAAAZWNXSCuTbPf5z3nQcAAAAAgIGMrsBQVV+1ufM6Tt3xIAAAAAAAwIZGV2BI8sLYvBkAAAAAgJ1WK0MnmGtjLDAcoQUSAAAAAACM1BgLDDclOTHJhUkObDDnlCTn71oiAAAAAADgVsZYYLgiyZlJLu3uS9abUFV7o8AAAAAAAACDGWODqcsyaY901tBBAAAAAABYYCs1/mPExriC4YIkFyW5bpM51yY5fXfiAAAAAAAAa42uwNDdB5McPMqcTnLV7iQCAAAAAADWGl2BAQAAAAAAdkWNcReB+eHVAwAAAAAAZqbAAAAAAAAAzGwpWiTd/M7/NXQEFsyehz566AgskNu9/s1DRwDYnCXDwMitnPGwoSMAAPOqaugEc20pCgwAAMDInHL7oROwSL5w/dAJAACWktvRAAAAAACAmVnBAAAAAADActISdku8egAAAAAAwMwUGAAAAAAAgJlpkQQAAAAAwHJaqaETzDUrGAAAAAAAgJkpMAAAAAAAADNTYAAAAAAAAGZmDwYAAAAAAJZT2YNhK6xgAAAAAAAAZqbAAAAAAAAAzEyLJAAAAAAAllO5B38rvHoAAAAAAMDMFBgAAAAAAICZaZEEAAAAAMByWqmhE8w1KxgAAAAAAGBOVdW9qurSqvpAVb2vqn56On6HqnpjVX14ej5t1XOeUVUfqaoPVdV3rxp/aFW9Z/rYb1XVphUYBQYAAAAAAJhfh5L8bHf/kyQPT/KTVXX/JL+Q5E3dfUaSN00/z/Sxc5M8IMk5SV5UVXum1/qdJPuSnDE9ztnsCyswAAAAAACwnGpl/MdRdPenu/td049vSPKBJPdI8pgkL5tOe1mSx04/fkySV3b3l7r7yiQfSfLNVXW3JLfr7rd1dye5eNVz1jXKAkNVPbaqXl1V/72qvmM69oiq+ququrGq3l1V3zN0TgAAAAAAGIuquk+SBye5LMlduvvTyaQIkeTO02n3SPKJVU/75HTsHtOP145vaHSbPE8LB69O0tOh76+qf5nkD5N8bZJK8k+TvKaqHt7dfzVMUgAAAAAA2FlVtS+TtkVH7O/u/evMu02SP0ryM939+U22T1jvgd5kfENjXMHwb6fnmh4nJfmDJCfn1v/AE5L89EYXqap9VXV5VV3+4tf92U5lBQAAAABgXlWN/uju/d39sFXHesWFEzMpLry8u189HT4wbXuU6fmz0/FPJrnXqqffM8mnpuP3XGd8Q2MsMDwkk6rIHyf5/un59tOxFyY5LclvZVJs+LaNLrL6Rf+x73/UjocGAAAAAIDdVpOlCr+X5APd/eurHnpdkidPP35ykteuGj+3qk6qqtMz2cz57dM2SjdU1cOn13zSquesa3QtkjIpICTJed398ap6X5Ij+y38v919fVU9P8lPJbn7IAkBAAAAAGAcvjXJE5O8p6qumI79+yTPS/Kqqnpqko8n+aEk6e73VdWrkrw/yaEkP9ndN0+f9xNJXppJR6E3TI8NjbHA8A9JbpPkM9PPP7bqsSMbTFw9PW/YRAoAAAAAABZdd/9FNv5b+SM3eM5zkzx3nfHLkzzwWL/2GAsMBzIpMNwryUe7u6vqrUm6u49sKHFk5+qDQwQEAAAAAGAB1Bh3EZgfY3z13p1JteWcIwPdfXZ3f+eqOd8yPX9wN4MBAAAAAAATY1zB8HOZbOZ8YJM5ZyR5S5L/viuJAAAAAACAWxldgaG7r0xy5VHm/HKSX96dRAAAAAAALKQV2/xuxRhbJB1VVe2tqpur6tDQWQAAAAAAYBmNbgXDDJSWAAAAAABgIKMrMFTVeccw7dQdDwIAAAAAwGKruWzyMxqjKzBkssFzDx0CAAAAAADY2BgLDEdogQQAAAAAACM1xgLDTUlOTHJhkgMbzDklyfm7lggAAAAAgMVT7nPfijEWGK5IcmaSS7v7kvUmVNXeKDAAAAAAAMBgxriDxWWZtEc6a+ggAAAAAADA+sa4guGCJBcluW6TOdcmOX134gAAAAAAsJBWxngP/vwYXYGhuw8mOXiUOZ3kqt1JBAAAAAAArKU8AwAAAAAAzGx0KxgAAAAAAGBXVA2dYK4tRYFhz0MfPXQEgA3d/JoXDR2BBbPnB84bOgKL5uZDQydg0azsSW7cbMs1mM07vuEbh47AAnnIhb8wdAQWzJ5HPX7oCAA7RoskAAAAAABgZgoMAAAAAADAzJaiRRIAAAAAAHyVcg/+Vnj1AAAAAACAmSkwAAAAAAAAM9MiCQAAAACA5VQ1dIK5ZgUDAAAAAAAwMwUGAAAAAABgZlokAQAAAACwnFbcg78VXj0AAAAAAGBmCgwAAAAAAMDMtEgCAAAAAGA5VQ2dYK5ZwQAAAAAAAMxMgQEAAAAAAJiZFkkAAAAAACyncg/+Vnj1AAAAAACAmSkwAAAAAAAAM1NgAAAAAAAAZmYPBgAAAAAAllPV0Anm2lwVGKrqLkl+PEm6+5cHjgMAAAAAAEtr3lok3TXJs5P80sA5AAAAAABgqY1uBUNVfXaTh/esmdfdfZcNrrMvyb4k+d3ffkH2/ehTtjMmAAAAAADzrubtHvxxGV2BIckdk3SSI82ves3jvWbeurp7f5L9SZIvXL/hPAAAAAAAYHZjLDCstd4uGwoGAAAAAAAwoDGu//irTIoKf5/kZ7p75ciR5CFHJk3H9mx0EQAAAAAA2NRKjf8YsTEWGM5M8nOZZPuNqnpbVT1g+piVCwAAAAAAMAKjKzB09+Hu/rUkD0zyxiRnJXlXVV2Q5KRBwwEAAAAAAElGWGA4ors/1t3nJHlikuuT/Pskrx82FQAAAAAAC6NWxn+M2LjTJenulye5X5L/luSOA8cBAAAAAAAyBwWGJOnuz3X3U5KckeTBWbXZMwAAAAAAsPtOGDrALLr7o0lSVXur6nCSw909V/8GAAAAAABGomroBHNtLlYwbMJ3HwAAAAAABjC6u/+r6rxjmHbqjgcBAAAAAAA2NLoCQ5IXJumhQwAAAAAAsOBq3pv8DGuMBYYjtD8CAAAAAICRGmOB4aYkJya5MMmBDeackuT8XUsEAAAAAADcyhgLDFckOTPJpd19yXoTqmpvFBgAAAAAAGAwY2wwdVkm7ZHOGjoIAAAAAACLq6pGf4zZGFcwXJDkoiTXbTLn2iSn704cAAAAAABgrdEVGLr7YJKDR5nTSa7anUQAAAAAAMBaoyswAAAAAADArqgx7iIwP7x6AAAAAADAzJZiBUMf/MTQEVgwz7/ftw8dgQXy89dcOXQEFszhD/yfoSOwYOqu9xk6AsCmzvzoe4aOALChmy//k6Ejg/fLJwAAIABJREFUsGD2fPvjho4AX7EUBQYAAAAAAPgqWiRtiVcPAAAAAACYmQIDAAAAAAAwMy2SAAAAAABYTis1dIK5ZgUDAAAAAAAwMwUGAAAAAABgZlokAQAAAACwnMo9+Fvh1QMAAAAAAGamwAAAAAAAAMxMgQEAAAAAAJiZPRgAAAAAAFhOVUMnmGtWMAAAAAAAADNTYAAAAAAAAGamRRIAAAAAAMup3IO/FV49AAAAAABgZgoMAAAAAADAzLRIAgAAAABgOVUNnWCuWcEAAAAAAADMTIEBAAAAAACY2Vy1SKqqOyTp7v7c0FkAAAAAAJhz5R78rRjdq1dVj66qR68Z+96q+pskVyc5WFXvqarvGCYhAAAAAAAwugJDkj9J8sdHPqmqf5bktUnum6SmxwOSvKGq7r/RRapqX1VdXlWX7/+DS3Y4MgAAAAAALJextkhavXX3L+SWQsg/JOkkpyQ5KcnPJXnKehfo7v1J9idJf/y9vVNBAQAAAACYUyt19DlsaIwrGNb6lkyKCi9NcvskpyW5OJMixHcOFwsAAAAAAJbXPBQYbj89/2J3H+ruLyd55nTsLgNlAgAAAACApTbWFkmpqntlskrh2iR3SnLNqoevnp5v3O1cAAAAAADAiAsMST625vP7JPng9ON/ND3/3W6FAQAAAABgwdQ8NPkZr7G+erXOcc6qxx83Pf/lLucCAAAAAAAyzhUMG23cfCBJqqqS3DPJa5O8fLdCAQAAAAAAtxhdgaG733KUxzvJj+xSHAAAAAAAFlXV0Anm2lhbJG2qqvZW1c1VdWjoLAAAAAAAsIxGt4JhBkpLAAAAAAAwkNEVGKrqvGOYduqOBwEAAAAAYLHVXDb5GY3RFRiSvDBJDx0CAAAAAADY2BgLDEdogQQAAAAAACM1xgLDTUlOTHJhkgMbzDklyfm7lggAAAAAgMVT7nPfijEWGK5IcmaSS7v7kvUmVNXeKDAAAAAAAMBgxriDxWWZtEc6a+ggAAAAAADA+sa4guGCJBcluW6TOdcmOX134gAAAAAAsJBqjPfgz4/RFRi6+2CSg0eZ00mu2p1EAAAAAADAWsozAAAAAADAzEa3ggEAAAAAAHbFinvwt0KBAY7Dz19z5dARWCC/faf7Dh2BBfNvr/7o0BFYNN1DJ2DRVA2dgEXTh4dOwCI59OWhE7Bg9jzsnKEjAOyYpSgw1L0fOHQEFs0Xrh86AQAAAADAoKz/AAAAAAAAZrYUKxgAAAAAAGCt0r5zS6xgAAAAAAAAZqbAAAAAAAAAzEyLJAAAAAAAllO5B38rvHoAAAAAADCnquqiqvpsVb131dgdquqNVfXh6fm0VY89o6o+UlUfqqrvXjX+0Kp6z/Sx36pj2KBCgQEAAAAAAObXS5Ocs2bsF5K8qbvPSPKm6eepqvsnOTfJA6bPeVFV7Zk+53eS7EtyxvRYe82vosAAAAAAAMByqhr/cRTd/dYk164ZfkySl00/flmSx64af2V3f6m7r0zykSTfXFV3S3K77n5bd3eSi1c9Z0MKDAAAAAAAsFju0t2fTpLp+c7T8Xsk+cSqeZ+cjt1j+vHa8U0pMAAAAAAAwEhV1b6qunzVsW8rl1tnrDcZ39QJWwgCAAAAAADzq8Z/D35370+yf8anHaiqu3X3p6ftjz47Hf9kknutmnfPJJ+ajt9znfFNjf/VAwAAAAAAZvG6JE+efvzkJK9dNX5uVZ1UVadnspnz26dtlG6oqodXVSV50qrnbMgKBgAAAAAAmFNV9YokZye5Y1V9MskvJXlekldV1VOTfDzJDyVJd7+vql6V5P1JDiX5ye6+eXqpn0jy0iQnJ3nD9NiUAgMAAAAAAMup1tt6YL5097/Z4KFHbjD/uUmeu8745UkeOMvX1iIJAAAAAACYmQIDAAAAAAAwMwUGAAAAAABgZvZgAAAAAABgOa24B38r5qbAUFUPSvKQJF9I8tbu/tTAkQAAAAAAYGmNrjxTVY+vqour6l9OPz+hql6R5J1JXpzk5UmurKpnDJkTAAAAAACW2egKDEmeluQJSa6efv7MJI9LUquOE5P8xyNFiPVU1b6quryqLt+/f/8ORwYAAAAAYO5Ujf8YsTG2SDpjev7r6flJSTrJJ5P80XTsXyW5V5KfSfLq9S7S3fuTHKks9I4kBQAAAACAJTXGAsPtpucjpZm7T8+P6O6PJklVvSjJ3yR50C5nAwAAAAAAMs4WSQem53Om549PzzesmvP30/OeXUkEAAAAAMDiqZXxHyM2xnR/nsnqhf9SVT+S5LXTz/9nVT2tqp6W5H9M535goIwAAAAAALDUxtgi6blJfjDJ1yX5r9OxTvKQ6ZFVYy/a3WgAAAAAAEAywhUM3f3hJI9Nck0mKxdWb5Ndq45f7+6Ldj8hAAAAAAALoWr8x4iNcQVDuvvPqup+SX40ySOT3DvJiUkOJnlnkou7+/IBIwIAAAAAwFIbZYEhSbr7c0n+8/QAAAAAAABGZLQFhs1U1d4kVyc53N1z+W8AAAAAAGBo425BNHaj24NhRr77AAAAAAAwgNHd/V9V5x3DtFN3PAgAAAAAALCh0RUYkrwwSQ8dAgAAAACABVea5GzFGAsMR/jOAgAAAADASI2xwHBTkhOTXJjkwAZzTkly/q4lAgAAAAAAbmWMBYYrkpyZ5NLuvmS9CVW1NwoMAAAAAAAwmJWhA6zjskzaI501dBAAAAAAABZY1fiPERvjCoYLklyU5LpN5lyb5PTdiQMAAAAAAKw1ugJDdx9McvAoczrJVbuTCAAAAAAAWGt0BQYAAAAAANgd425BNHZj3IMBAAAAAAAYuaVYwXD47a8fOgILZuUB3zp0BBbIeS8+f+gILJrDNw+dgAXTBz42dAQWTN3xnkNHYNF8+UtDJ2CB9BdvHDoCC6ZOu8vQEQB2zFIUGAAAAAAA4KuUFklboUUSAAAAAAAwMwUGAAAAAABgZlokAQAAAACwnHRI2hIrGAAAAAAAgJkpMAAAAAAAADPTIgkAAAAAgCWlR9JWWMEAAAAAAADMTIEBAAAAAACYmQIDAAAAAAAwM3swAAAAAACwnMoeDFthBQMAAAAAADAzBQYAAAAAAGBmWiQBAAAAALCctEjaEisYAAAAAACAmSkwAAAAAAAAM9MiCQAAAACAJaVF0lZYwQAAAAAAAMxMgQEAAAAAAJjZ6AoMVXVzVV1fVb9ZVWcMnQcAAAAAgAVVNf5jxEZXYMik6dVtkzwtyQeq6n9U1aMGzgQAAAAAAKwyxgJDknQmhYaVJN+T5E+r6v1V9eNVdfKxXKCq9lXV5VV1+f7X/MlOZgUAAAAAgKUz1gJDkjwlybsyKTRUkvsleVGST1bVr1bVvTd7cnfv7+6HdffD9v3AOTseFgAAAACAeVNzcIzXaAsM3X1xdz8sybcmeVWSQ5m8mqcleXqSj1bVHw4YEQAAAAAAltZoCwxHdPfbuvvcJPdJ8twkn82k0LAnyQ8MGA0AAAAAAJbW6AsMR3T3p7v7F5PcO5P2Se8cNhEAAAAAACyvE4YOMKvuvinJxUkurqqHD50HAAAAAIA5VePe42DsxlhgeGuSPpaJ3f1/djgLAAAAAACwjtEVGLr77KEzAAAAAAAAmxtdgeFYVNXeJFcnOdzdc/lvAAAAAABgYFokbcncbPK8Ad99AAAAAAAYwOju/q+q845h2qk7HgQAAAAAANjQ6AoMSV6YY9zkGQAAAAAAjp8mOVsxxgLDEb6zAAAAAAAwUmMsMNyU5MQkFyY5sMGcU5Kcv2uJAAAAAACAWxljgeGKJGcmubS7L1lvQlXtjQIDAAAAAABbUKWRzlasDB1gHZdl0h7prKGDAAAAAAAA6xvjCoYLklyU5LpN5lyb5PTdiQMAAAAAAKw1ugJDdx9McvAoczrJVbuTCAAAAACAhaRF0paMsUUSAAAAAAAwcgoMAAAAAADAzEbXIgkAAAAAAHaHFklbUZPtDBbcDQeX4B/Jrrrpi0MnYJEsw89hdtdJJw+dgAXTX7xx6AgsmDrhpKEjAGzsBPdiss0OHRo6AYvmtLv6i/h2uubvxv+Hmb33GO33XIskAAAAAABgZgoMAAAAAADAzKz7AwAAAABgOdVouw/NBSsYAAAAAACAmSkwAAAAAAAAM9MiCQAAAACA5aRF0pZYwQAAAAAAAMxMgQEAAAAAAJiZFkkAAAAAACwpLZK2wgoGAAAAAABgZgoMAAAAAADAzLRIAgAAAABgOZUWSVthBQMAAAAAADAzBQYAAAAAAGBmWiQBAAAAALCcdEjaEisYAAAAAACAmSkwAAAAAAAAM1NgAAAAAAAAZjZXezBU1UlJHpDkc9195dB5AAAAAACYZzZh2IpRrmCoqmdV1aeq6rqqenlVnVpV/zTJR5O8I8lHquq1VfU1A0cFAAAAAIClNLoVDFX1Y0menaQzKR+dm+SmJGckufuqqd+X5GeT/MouRwQAAAAAgKU3xhUMT52eK8nnp+cnJPmWJNckef30XEl+aKOLVNW+qrq8qi7f/5KLdzYxAAAAAADzp2r8x4hVdw+d4Vaq6vNJTk3ymO5+fVX9iySvzWRFw4O7+6+r6kFJ3pXkxu6+7VEvesPBcf0jmX83fXHoBCySkf0cZgGcdPLQCVgw/cUbh47AgqkTTho6AsDGThhdswfm3aFDQydg0Zx213H/xXnefP7q8f9h5nZ3Gu33fIwrGI78VeRNa85J8jfT8wenZ7+ZAAAAAADAAMZYlr8myZ2SvKCq/ijJD6567Ker6veS/NiquQAAAAAAMLuRtyAauzG2SHpdJhs4rw52IMld14wlyeu7+zFHvagWSWw3LZLYTiP7OcwC0CKJbaZFEttNiyRg1LRIYrtpkcR20yJpe83D345ve8fRfs/H2CLp+UkOZbKJcyX5YpLHJvn0qrFKcjjJfx4oIwAAAAAALLXRleW7+y+q6uwkT0xyc5KXdPc7q+qRSX4lyT9O8ndJfq273zpcUgAAAAAA5ttoFwfMhdG1SNoR87DMhfmiRRLbaRl+DrO7tEhim2mRxHbTIgkYNS2S2G5aJLHdtEjaXjdcM/4/zNx272i/52NskXRUVbW3qg5XlZ/QAAAAAAAwgHkvy4+2cgMAAAAAwMiVPzFvxegKDFV13jFMO3XHgwAAAAAAABsaXYEhyQuTjL/vFQAAAAAALLExFhiOsDYFAAAAAICdo0XSloyxwHBTkhOTXJjkwAZzTkly/q4lAgAAAAAAbmWMBYYrkpyZ5NLuvmS9CVW1NwoMAAAAAAAwmJWhA6zjskzaI501dBAAAAAAAGB9Y1zBcEGSi5Jct8mca5OcvjtxAAAAAABYTPZg2IrRFRi6+2CSg0eZ00mu2p1EAAAAAADAWmNskQQAAAAAAIzc6FYwAAAAAADArigtkrbCCgYAAAAAAGBmCgwAAAAAAMDMarJfMkxU1b7u3j90DhaD9xPbzXuK7eY9xXbznmK7eU+xnbyf2G7eU2w37ymYP1YwsNa+oQOwULyf2G7eU2w37ym2m/cU2817iu3k/cR2855iu3lPwZxRYAAAAAAAAGamwAAAAAAAAMxMgYG19LljO3k/sd28p9hu3lNsN+8ptpv3FNvJ+4nt5j3FdvOegjljk2cAAAAAAGBmVjAAAAAAAAAzU2AAAAAAAABmdsLQAYDFUVUrSc5M8vVJvnbt49198a6HAoBtVlUnJfmHJIeTfFN3v2/gSMypqnrWLPO7+5d3KgsAABwPezAskar62xmmd3ffd8fCsHCq6huSvC7J/TaY0t2tqMkxq6o3JHlxktd196Gh8wCsVlUHk5yW5JTu/tLQeZhPVXU4yTH/Qtbde3YwDsCmqur3k/ybJM9ZXfCsqv+Q5DlJXtHdPzxUPgCGocCwRNb5Baam57Vvgsrkj8F+geGYVdWfJvmuTaZ4TzGTVT+zrk7ysiS/191/M2wq5l1V/XkmP48euc5jz5o+dsHuJ2PeVNXvJXlKku/u7j8bOA5zavrfumPl/6WAQVXVlUnuneQbuvvKVeP3SfK3Sa7q7tOHSQfAUBQYlohfYNhJVXVdktsm+WCSP05yY9YUr7r7OQNEY05V1U25pZXfkffSX2SyquEPu/uLgwRjrh0pXK3337jpY4ettuJYVNVjM/l5lCS/luSKTNomfUV3v3W3czFfqurJqz49MZM7gCvJf03yyST3TPJ/JdmT5D9094u/6iKwiWlhfSOd5Jokb0zyEitGOZqq+mImP6tu093/sGr85Ex+//tSd588VD7mV1U9PskTs367ZR02YOQUGJZUVd03yVunxzNzyy8w/ynJI5J8e3d/cLiEzJuq+kySOyW5e3cfGDoP86+qTkvyr5Kcm+Q7MvnjSjL5ZfjzSX4/yf7ufs8wCZlHGxUYqurrkly73mOwnmNobaM1IDOpqv+Y5BlJHtLd7141/uAk70zya939c0PlYz7N0Ibrfyd5lCIDm1l1U9l3dfefrxp/RJI/S3J9d582VD7mU1Wdn+R5Gz0c/38Oo6fAsKSq6n8mOSfJ3u6+btX4kT+w/K/uPmeofMyfqvovSf7vTDa7fO/QeVgsVXXnJP86yY8kefB0+Mh/wP4oyVO7+4YhsjF+0zuEj9wlfHYm7523rJn29UlOT3Jtd99x99Ixr45hZahfhplJVX0qyV2S3LG7P7dq/A5JDiY50N13Gyof8+n/Z+++oy0r6/uPvz/MwADSm0gvUlQCNgQEQUANIogmURFBRGKLRkUsiRgpamINqFEsoaPYEKUZ/AUEpNgAQalKF4l0BhhmKPP5/fHs4z0c7p17z3DnPHff83mtNWuf++yN67NWds7Z+/k+pavAoHEuNXCA7S8t+lTRVpJ+DmxLGaD4UeBq4FnApygDFi+0vX29hNFGkv4IbLCAS/JMFTHFZVTV8NqhOW4AXNrV/szmuN1g48Q0cBZlw69TJX0euBZ4tPuCLBURT8HTKS8vGzLyktx5Uf57YB5lSm3EaNZjpLAA5d7Zoeeazv108WAixTSQZf9isq3QHL8p6RBGZhh37rXlaoSK1tsa+CFlffzDgFuAtYFDKIX1dwBvA15NmTWaAkMsyLGUAsOalD3SOkR5zjp28JFiGliDcv/8M3Cc7Ycq54mIPmUGw5DqWs7mLuAERl5g9mna77C9er2E0TZZKiImm6RZlFkL76S8HMNIJ/DPgCMoD6NHAvfaXnngIaMVJB0MHNz82V1k6HY38Avgn23fNKBoERF/JemnwMsY+3nqf23/7QAjxTQg6TvA6+hZxlTSM4DbgO8C7wLuAB7K8jYxHknfpwzw6fV9228YdJ5oP0m/ALYEVrQ9u3aeiOhfCgxDStLngAN58gtMZ+TBF7LGa/QjS0XEZJN0F9B5yRVllsJJwBG2r+i67gZg3dxfMREL2uQ5IqImSZtQ9kdbdZTTdwA72L52sKmi7STdDywDPMv2dV3tzwSuo1kzX9KtwGq2Z1WKGi0i6fXA7pRZxn8BTrX9/bqpoq0k7UhZEeHTwMFOR2VE62Q08fD6KOXl5c2jnDu+OR/RjywVEZNtpeb4F8oshSNt3znKdd8FMuMqJmq/2gGivSSdQylQ7dx8XhDb3nkQuWJ6sH2tpM2ADwA7AitTZhv/DDh8jN/AiPE83hzPlPQlyhJJawLvado7HXlPA+4jYgJsfw/4Xu0c0V6jPEfdDxwE/KOk63nicst5poqY4jKDYcg1I6We8ALTPbIlIqIWSZcCXwROsv1I7TwxvUjaGtgVWI0yMvgM27+smyqmumYGzHzbM8dZGlBkpkxETAGSvgnsz9jfV/8NfBK4GTjX9k6DyhbtJGkxynI26wJL9p63ffzAQ0XrTGCJ5b9eSp6pIqa8FBgiYlJJWgrYhpFOu4ttP1w3VUTECElHAm8f5dTXbL970HmiPbqX2MrSgLEoSVqN0TvubqkQJ1pM0jLAKcBoo3//F/g7YCNgL+A826cPMF60TLO01qnAJmNckn33YkIm8BzVLc9UEVNcCgxDTNKylNGbY408OGzgoaLVJL0R+DIj6+YD3Au8x/Z36qSKNpO0D3AA5SWm93sqLzDRN0lvAY4e47SB/TLyLiJqaDqC/xPYGxhtHfz87sVCk/QySpFhFcrM9bNt/2/dVNE2ks4CXr6AS9IRHBExhFJgGFKStgTOZGSN8yfJg0H0Q9J2wLmUKYzqOf04ZWPCiwadK9qr2TzuO5RO3957CvICEwtB0kXA1pSlIA5vjutQClnrUWZdbVstYEQMra6lbMaS372IqErSfcCywDWU/oSH6Fnmxnb25ouIGDIpMAwpSRdSlrEZS15goi+STgV2Ax4DTmek0243YHHgNNt71EsYbSPpPOAlwBxgacrLyz2UPWPuA+6zvUG9hNFGkmZTNrLcwvbvu9o3A64AHrC9fK180S7NsoDv44n7eZwOfDnLA0a/JN1OuY8eAX7P6B13O1aIFi2XNfNjskj6P2BVYA3bf6mdJ6YXSZtTZq4v1Xsu31MRU1sKDENK0gOUDrvzgJMZ/QXmuArRoqUk3UmZEfNa26d2te8O/Bi42/aqtfJF+0i6F1gO2Ba4iJF1z/8NeA+wk+0ra2aM9pE0l1L0XNn2fV3tK1AKWPNsP+mlJqKXpM5z1PNHOX0JZeZeigwxYV0F0K1t/7p2npgesmZ+TCZJXwHeSc9AjYinolki8EfAWEX0fE9FTHEpMAwpSbcCa9DTwRKxsCQ9AswAlrP9UFf704AHgEdtj7aecMSouu6ppYBOJ90SlHWpHwR+Znu0DQsjxiTpD8AGwDHAB23fJ2l54POUpUn+aHvjmhmjHSQdCvzbGKcNHJZlIqIfzWzQVwGr276zdp6YHrJmfkwmSa8GjqXMJv48cC3waPc1ts8ffLJoM0mfBj68gEvyPRUxxaUCOLyOB/4F2Ay4oHKWmB7uAp4O7AV8s6t9r67zEf2YTdkwXJQi1bLAK4H7m/NbVcoV7XY6ZUmb/YD9JD0ILNOcM3BarWDROv9AuWd+CnyEkaUBPwPsArwOSIEh+vEhYDvgOEkfBK61/XjlTNF+W1G+q8ZcMz+iDz+i3D/LA18e5bxJP1P0bw/KvXMW5RnKlL3S3kyZYfztetEiYiIyg2FISXob5QXYwFGMPvIga9zFhEk6CXgD5Z66AriF0tGyeXPJd23vNcZ/HvEkkq4AnkPZePd7wIu6TwN/sr1OhWjRYpJWpixfM9q9cxPwQtv3DDRUtJKkhymzqp4w2lzS04Hbgbm2l66VL9pH0njFhCwREX3LmvkxmSTNH+eSjDSPvkmaQ5mlvjbwJ0aWxt0GuBB4p+1v1MwYEQuWAsOQah4MFvR//LzARF8k/Q3wK0pnyxNOAXOBF2WdzuiHpOOAfYDXU16Mv9Jzyadsj7U8ScSYmg7gT1I25l0VuJMys+Hj6XyJiWpmvywFbGD75q729YAbgIdsL1snXbRR1/O5xrgkHXfRt6yZH5NJ0sHjXZPlAaNfXYM2Fqf0HcygzGR/BJgDXG97o3oJI2I8KTAMqYw8iEVB0o7A14FndjVfRxlxcG6VUNFazf4dywAP2J4j6V8os2QeA04BPpOlIyKiFkmXAM8FLgM+wcjMvYOAFwKX2n5hvYTRNpLOZZyla2yPtQFmxKiyZn5ETHWSbgNWB1YCrmo+n0gpLrwDeNj20+oljIjxpMAwpCTtO941to8bRJaYfiRtRDMq2PYfaueJiIiYbJIOAL7A2B3CB9o+YoCRIiKeJDPXI2Kqk3QeZQ+i51L2r3oNT/zeuiyDNiKmthQYIiJiSmuWs1kXWLL3XEbcxURMYF3zbuloiQmRNIOyKfguo5w+E3i17fFmjEaMSdJSth+unSPaLTPXY7JJ2gF4H7AJZanAbra94eBTRZtJejewG2Vj57uACxh595sL7G777ErxImICUmAYck3H3Q7AyraPrJ0n2k3SXpQ180frDM7DZvSl+X46Adh5jEvSERwTMoHOlW7paIkJk7QY8EaevJ/Hd1NciIXRzAL9HPByYJbtmZKOAJYDvmD7yqoBo3WyZn5MJkkvA34CLMbIfjGdTiWR56iYBJLWB15NWRr3J7ZvqBwpIsaRAsMQk3QgZZPLJWg66iRdDmwG7Gn7+1UDRqtI+hDw6bFOk4fN6JOk7wN/v4BLck/FhExkXfNuWeM8ImqQtC5wCWVjy78+O0n6HPAB4NO2D6qZMSKGm6T/AV7R1dS7MX2ez6MvkmZRilYG3mX7usqRImIhpMAwpCTtQdkktaPzAvN24GvAt2zvUyddtJGkPwIbLOCSPGxGXyTdRelkuRP4OfAQPZ3EtverEC0ihpSkdfq53vYtiypLTD+SjgH2BR5hZADQDEnbUn4Hf217q5oZI2K4SbqTshHvqynLBBpYnrKB+EuBl9n+U7WA0UqSHqQst7W07Xm180RE/1JgGFLNaM6XUF5WtmfkBebZwO+B62xvWjFitIykOcAs4J+B42w/VDlStJykeygvLBvbvr52nmi/jJCKp2oCm6V2yzJu0RdJtwGrAzsBP2Pk+XwV4A7gDtur18wY7SDpHMr9s3PzeUFse6zlKCOeQNIjwAxgGUYG/ywOLAvcC5xq+zX1EkYbSfopZVnc59i+pnaeiOhfCgxDStL9lIeCtYE/MfICsxTlQeEB28vXzBjtIunXwPOBFW3Prp0n2k/SiZS1zTewfXPtPDE9ZIRUPBXZzyMWpa6OuyWBeYw8ny8DzAYetT2rZsZoh+a7an6zBO6CCqNZxjT60jUAaEngHmBpYBvgfuBq4EHby9VLGG0k6TnAucCNlAGLv81zekS7ZFTV8FqiOd7V075yc1x8gFlievg4ZWPLD0o62KlexlP3FcrGqSdL+jhwLfBo9wVZfiQWwkWUEVLrAxkhFf36FWN31C1GKbR3b3wZ0Y97KJuFr9fTvntzvHugaaLtNMbniKfiL5QCwyrAH4HNKR3DnQIoadWoAAAgAElEQVR8ZrHHwriiOa5EeVZHesLXVmaFRkxx+X/Q4XUzsBGl8w4ASTOBTzV/3lgjVLTah4EHgIOAf5R0PU/sDM706+jXhZSOvOdR1njtZfI7Fv07gPIifLykjJCKvtjeerR2Sa8DDmakuPAQpUga0Y+LKeuaf7vTIOlrlH0ZDFxQKVe0jO3FRvscMQkuBzYBXgB8H9iCskxupzf4h5VyRbuJJ28YHhEtkiWShpSkzwEHAo9ROugMPEhZNgngs7b/tVK8aKFMv47JNoGlSHJPRd8kPT7OJRkhFRMm6bXAIcBmlN+6h4Ejgc/YvrNitGghSVtT9kfr7RAW8Diwne1fDjxYRERD0trAGpQBi3cDXwbeQOlXOAV4v+059RJGGzV7hC6wc9L2joNJExELIwWGISVpBeA3wAaMfJF3qsU3AC+0fV+NbNFO6QyOySbpmPGusb3fILLE9NFVDB1rhFS+q2Jckl5NKSxsQbmX5gJfBz5t+y8Vo0XLSXo98FXKMhEd9wLvtv2dOqmibSRt38/1ts9fVFkiIiJi+kuBYYhJWhX4BLAbsBpwB2UN/Y/bvqNmtoiIiEUhI6TiqZC0K3AoZa8FUTbi/QalsHB7zWwxfUhaCtiWkefzizIiOPoxzsziXpm5FxFVSfoIcEz6oSLaKwWGiIiIiIgJ6JkB8xjwA+BPY11v+8MDihYR8VcTmFncLTP3YoEmsLxktxSsom/Nd9ajwBnAfwM/cTorI1olBYaImBTjTMU2cLftqwaVJ6YHSS8G9gHWBZbsOZ2Nw+Mpk7SU7Ydr54h26HNUMOm0i/FIOrqPy217/0UWJqYNSQf3c73tQxdVlmi/FKxiUeu6xzrPWLcBR1NmNdxcJ1VE9CMFhiGSkQexKE2w0+XPwDttnzGASNFykt4MjLUPQzYOj4UmaSPgc8DLgVm2Z0o6AlgO+ILtK6sGjCkrnSwx2VK0ioipbiLLS3bLUpPRL0m7AXsCrwaWaZrd/Dubshzlj20/VidhRIwnBYYhkpfiWJT6uL8eBba0fcWizBPtJ+kqYNMFXJLvqeibpHWBS4AV6SpUSfoc8AHKWvoH1cwYU5ekffu53vZxiypLTA95Po8aJG0FrANckP1jImKqkLQksDvwBmBXygz2Tqfl/wGfsv3VSvEiYgFSYBgiGXkQi5KkYyijgdcALgJuAdYGXgzcDlwGvAxYAjjB9lvqJI22kDQHmAV8FjgReIie77BMmY1+Nd9V+wKPUL6POgWGbYGfA7+2vVXNjBExPJqi54Tldy/6JekDlJHBJ9k+XNKXgX9qTj8I7Gj70moBIyK6SJoB/D1l4M+WnebmaOAQ25+okS0ixpYCQ0RMCkl7A8cBb7T9va72PYFvUTr07gLOBG62vX6VoNEakn4PPAtYwfYDtfPE9CDpNmB1YCfgZ4wUGFYB7gDusL16zYwRERGTRdJ5wHaUUcG/ogz86Z4Jc6rt19TIFu0i6dmUEeWX235c0lqUTuBNKffVkbZ/UzNjtFdTcH878FZgtU4zMBs4ivL8/kbgz7bXqhIyIsaUAkNETApJ1wAbAcvbfrCrfRnKQ8E1tp8t6V5gSdtLVYoaLSFpH0rR6p22v1E7T0wPkh6hdKwsCcxjpMDQ+a561PasmhkjIgAkbc5Ix90FzotbLARJt1M669YCng+cRhnw8wvgMOBO20+vlzCmOkkrA2cBz2uabqTMijkV6L53HgFekiJD9EvSGcArgMUYma1wPfBl4OhO/0Ln+yzLBUZMPdnEdwhl5EEsIp0p/u+T9O9dL8HvbI6dGQsPkO+emJidgPuAIyX9I3AtZQ+PDtvev0qyaLN7gFWB9Xrad2+Odw80TUQMPUn/DOxDeT7/EaXT91jKSM2OX0t6he3Zg08YLbdic7yD8r5n4NvAyZR7bcUx/ruIjoMoxamO9YFzGNmMt2MW8GHg9QPKFdPHK7s+nwscAZw2SmH9KmDuoEJFxMRlBsMQyciDWJQk/Rb4m+bPe4DbgGcAqzRtvwNeAMwBbrK9ycBDRqs0G18u8Ecqo1eiX5JOAV4NXEr5TjLwTcoybksAJ9vOi3FEDISk/ShLP3T/3l0AvKTnUgOfsf3RQWWL6UHSHcDKwDbAv1J+A7cC/gDcC9xne6V6CWOqk3QVpTg1B/g98GxKccHADygzjt8C/ANZviYWgqS5wEnA4bavqJ0nIvqXAsMQkfSfwPu7mkzZNLV35AHAD9LBEv2QtCvwY8rSI91fLAIeA/ZojmcBx9p+68BDRqs0BYYFcQoM0S9JW1M2c16s9xTwOLCd7V8OPFhEDKVmffzRigmijDg/H9iBMvPqKtubDTZhtJ2k84Ftu5rmUWYtbAL8ltxXMQ5JD1FmWL3Q9mWSnksZqGHKcjV3d+1llaUmo2+SVrN9R/N5aUpBdCXgatu/rxouIiak9+U6prddmuMcygZfneKCge8Du1FGIAC8eODpotVsnwm8HLiYkRdjAxcBL7f9E8qGqstSNm+KGM/64/zboF60aCvbvwDeRFl+S13/7gX2SXEhJkrSOpLWqZ0jWu85lOelT1BGBR/KyDPUK5sBP52lI9Yf9X8hYsGOYOTZXMBRtucx8m74i1rBojWWArB9WXP8beeE7bub411NU5bCjQmRtIekb0j6l67iwhbANcBPge8Al0s6WdLiNbNGxPgyg2GIZORBDIqkpSgjDu6x/XDtPBERvZrvqW0pG1/eAVxke07dVNEmzSyr+baf1Jki6cbm3IaDTxZtIukxSqfvLNuPNZ0o8yjP5zNtW9JilFmgmbkXC0XSi4DtgBttn9K0bQ+sDVxi+5qa+WJq61q2dD1GNuC9aay2fE/FREj6HvD3wGG2D23aLqTMXuhm4F9tf3bAESOiDykwDJHOg0H3D/5E2yIiapG0J7ArIx3BZ9j+bt1U0SaSPt7P9bYPW1RZYvpY0PNSnqViovJ8HhFT3UT2RetcSr6nYoKavT02Abay/RtJ6wE3UO61Ryl7e2zfXHOZ7RdUihoRE5Dpa0NI0tqMjDIYsy1iPJLOoTxE7tx8XhDb3nkQuWJ6kDSDsq/HK3tOvUnS3sAetsfbpyEC4BAm9mLckQJDjErScsAKPW29z1BbNMfHB5Ur2m+056gJPFtFTIikWcC7gFcAK9neWtJelP6An9i+s2rAaIvu3zqP0hbRj9Wa4x+a4/Zd546y/W5JmwFXABsPNFlE9C0FhuF0U9dnj9IWMVEvBeZ3fR6rA08LOBcxlvdSZi6MZtfm/BGDixPTwERegvNdFQtyANA9I0aM/gxl4LZBBIppY4euzx6lLWKhNEsCngu8kCc+k78K2BP4EPCfVcJFW9xCno9i8i3THOc1x+6lkU5rjp3l27IHQ8QUlwLDcMrIg5hMGuNzxFO1D+U76jLKiPKbgXUonXsvaM6nwBD9MKUz+FjKclsRC6P3t26s374TFnWQmDby/BSL0kHAlqO0nwi8kTJoIwWGGJPt9WpniGnpNsoeHi+T9P8oRU8oM0Avaj53Zo3eRURMaSkwDJeMPIhJZXux0T5HTJJNmuPf2765+Xy5pN9R1ufctE6saKEjgTcBy1FeZA4CTgaOtH1BxVzRPr+lrAkMsC/luer4rvMG7gZ+YfvkAWeLdtqxdoCY9l5H+W76MPC5rvbfNceNBp4oIgIuAdYHTgEeoBQTDJxve3ZzzVbNMbNCI6a4bPIcEZNC0kttn1s7R0wfkuYAs4DVu9cGlrQa8H/AXNtL18oX7SJpaWBv4B3A85pmA1cBXwVOtP1ApXjRQtl0NyLaQNJcyvIiTwPm0HxvNfsyPAw8YnvJmhkjYvhI2hz4NU9c/mg+sLPt85prTqAMEjrC9gcGnzIiJiojjiNispwj6Y+SDpK0Vu0wMS3c0ByPlbSFpBUlbQEc3XM+Yly259j+hu0XAFtTRqHPBZ4D/Bcj91XEhNheLMWFiGiBuc1x2Z72FzTHOQPMEhEBgO0rgJ2BM4DrgLOBPbqKC8sCawAXAj+ulTMiJiYFhoiYTOtT1sq/UdJPJP2DpGzIFAvrZMq61LsAl1LW3rwUeCVl5Pn360WLlpvPE5cMFFkDPRaCpNUkfU3SLZLmNsevNjOtIiKmgiua46c7DZL2pOwTY8rSbxERA2f7Atu7297U9sttn9F17gHbO9t+SafoEBFTV5ZIiohJIekzwBsom/DCSOfdPcC3gGNsX14jW7RTs6TNzxlZzqbbJcD2th8ebKpoK0lLAXsB7wSe32kGrgS+BhyfJZKiH5JWBn7DyO+eGPntuwnY0vY9FaJFRPyVpL0oGzr3vvh3vrP2tn3SwINFRETEtJECQ0RMKknbUjrx/gFYtWk2Zb3XbCwffWmKDAcAu1LupzuB04Ev2s6U/pgQSV8E3kzZ5FnAI2ST53iKJB0OvK/508BsRu4xU9YLPrBSvIiIv5L0ZeDdo5z6qu33DDpPRERETC8pMETEIiFpBvAx4CBgJtkIMyIqaTbj7biJsv/CnaNfDba/uqgzRftJ+iNlacBjgANt3y9peeALwFuB621vVDNjRESHpK2A3YHVgDuA023/om6qiIiImA5SYIiISSXpWcAbm38bdJpJgSEmQNI64181wvYtiypLTB9NgWHCDzz5roqJkDQXWBxYyfb9Xe3LA/cCj9hesla+iIiIRUHSi4F9gHWB3t8529558KkiIqKmLFcSEZNC0ocpRYXNO03N8WHKUiRH18gVrXMTE+8INvkdi4mb6CbOGXkRE/UIpcCwNnB/V/vazXHewBNF60javp/rbZ+/qLLE9CHpzQC2j+98XhDbxy/6VDEdNPfTMWOdJs9RERFDKTMYhoikfjp4bXv/RRYmpp2uEcKdTrxfUh4+T8rGqTFRPUvZjCezYmJCJB3cz/W2D11UWWL6kHQBsA1wM/BF4BbKhs/vBdYDLra9XbWA0Qp9zrDKflYxIc19Nd/2zAncY7mvYsIkXQVsuoBL8nweETGEUmAYIn28wGQ5m+hbc3/9BTgRONr21ZUjRQtJOpcnfk89m7K5863AbcCalNHB9wBX2N5p0BkjIgAkvYUyO6/32aozgnO/jAqO8aSwHotC573P9owJ3GO5r2LCJM0BZgGfpbz3PUTP76DtmytEi4iIijJSYfhMdImIiH69lrJZ3OO1g0R72X5p57OkVwCnAwfY/mJX+wGUl5rDBx4wIqJh+1hJ2wBvG+X0N1JciAk6rufvVwCrAxcCfwLWArYF7qb8JkZMxH5jfB5NRhxGP24AngX8e2apR0RER2YwDBFJO3T9uSzwdeA+4AuMvMAcCKwCvM32qQMPGRHRkHQpsAWwnO2HutqXAWYDl9t+Xq18EREATZFhN8psqzspxfaL66aKNpL0JuB44A22f9DV/nrgJODtto+qlS+mH0lbUDqKX1U7S7SDpH0ohdF32v5G7TwRETE1pMAwpCR9FXgH8EzbN3a1bwj8ATjK9mgj8iJGJemGcS6x7Q0HEiamBUkPA0sA+9j+dlf7m4ATgHm2l6qVLyIiYjJJuhrYGFje9oNd7Z3C+nW2F7T2ecRfSVoa2IOyN8wdwGm272rOrQV8GtiT0ieQJZJiQiQdQ7mvlgcuAa4FHu26JHs5RkQMoRQYhpSku4AVgTVt/19X+xqU2Qz32l65Vr5on1E2eYaRKdfZ1yP6Juk6oFOUuoyRmVadWQvX2964RraICABJywK7AusCS/aet33YwENFa3UV1v/V9me72j8C/AcprMcENQWECyj7VnU8CLyKsn7+D4FlyDN69Gki+zrmfoqIGD4pMAwpSQ8ASwO/oLywdDru/gV4MfCg7eXqJYy2kXQTT3zYnAk8vTk+AvzZ9gYVokVLSXor8N+MvoEqwFttHzvQUBERDUlbAmcCK411TTpZoh+SLgc2a/68C7gdeAZl+VKA39veoka2aBdJ/w28dZRTfwSWAtZs/p4NfN72JweVLdotm4ZHRMRoUmAYUpJ+APwdY48++KHt1w0wUkxDzdTsw4D3Ai+3fV7lSNEykt4CfBJYo6v5z8BBtns3xoyIGBhJFwLbLOCSdLJEXyTtBpwCzOCJz+gC5gN72D6jRrZoF0k3UpZGuoUyW0HAaymzrQAeB74CfML23VVCRitJWne8a2zfPIgsERExdaTAMKQkPQM4B9hklNPXADvbvn2wqWI6kiTK6KjLbG9fO0+0T3MPbQKsTBnReZ3z4xURlXXNBj0POBl4iJ6BGymERr8k7UwprG8JLEYpLPwK+Jjtc2pmi/aQNBdYHHi27Wubtk2BqyjfU2+y/Z2KESMiImIaSYFhiElaEtgX2JGRjrufAcfbnlszW0wfkl4I/BKYa/tptfNEO0iaBVxNeQl+le1rKkeKiHgCSbdSZletbPu+2nliemlmga5I2RdtTu080S6ddfJ7Z1F1rZ8/y/ZjVcLFtCBpT8oeRKtRNhE/w/Z366aKiIhaUmCIiEkh6YZRmpek7MMAcJPtDUe5JmJUku4DlgWWsv1I7TwREd0kfYqyd9UOti+onSemD0kzgedTilc/qZ0n2qerkHBoz6lDRmvPhvQxUZJmAD8GXjnK6TMpS7mNt09DRERMMykwDDlJrwFeAaxke09JL6Gs0Xmp7Qfrpos26XqR0RiXfNT2pwcYKVqu2SvmtcDWtn9dO09ERDdJbwM+Q/ntOwq4Fni0+xrbx1eIFi0m6XXAf1E2drbtmZLOBtYH3mn7p1UDRit0PZdPSPaLiYmSdADwhTFOGzjQ9hEDjBQREVNACgxDqhl58ENgN0qHsG3PkHQaZarju21/rWbGaBdJN/HkF5l5wM3At4ATsm5+9EPSdpTNLu8HDgJ+CzzcfY3tWypEi4iYSAeebc8cVJ5ov2agz88oz+bdz+cfAD4P/Lftt9fMGO3QfD9NVDakjwmTdCmwBXAZcBjlXW8d4OPACyj77r2gXsKIiKghBYYhJelA4HNdTZ0XmL8DfgD8xPar6qSLiEjnXURMbRPowEunXfRF0pnALsA1wKaMPJ9vTimyX2n7b2pmjHaQdHA/19vuXUopYlSSHqIsg7uB7Zu72tcDbgAezr57ERHDJx0zw2tfSsfdF4H3d7Vf1hyfNfBEERFPNtaSWxERte1XO0BMO1tTns93B/7Q1d7Z52rNgSeKVkrBIBahzrN57+bzc3rOR0TEEMkMhiElaQ4wC1gemM3ICKlZlCVI5tpeumbGaB9JKwF7A5sAS/Wctu39B58q2krSMeNdYzsdfBERMS1ImkcZALYkZZnJzvP5qsBfgHm2e5+vIiIGRtLvKYMR/wf4KHALZYmkT1GWWr7K9mb1EkZERA2ZwTC8Hm+OvVP3N2mOjxLRB0kbAhcCq452mjIiLwWGmLAUDyIiYsjcBqwLbNPT/sHm+KfBxomIeJKTgX+jLOe2S885A98feKKIiKguBYbhdTVlE6YPdRokbQ18ufnzyhqhotUOAVarHSIiImJRkXQjMN/2hs3n8faJ2XBA0WJ6OAt4B/CjToOka4CNKPfaWZVyRUR0fAbYDXjeKOcuAT472DgRETEVZImkISXp3ZRiwlg3wHtsHznASNFykm6hrA3878BBlHtrD8rU2ZWA99r+ab2E0UaSNqF0toy17NbOg08VEcOqs/l8s2xNZyP6sdabzibP0RdJa1I2c16ZJz6jC7gbeK7t22pki4jokLQ0cABlSaRVgTuB04Ev2u7dmyEiIoZACgxDSpKAU4BXj3L6dGAP5+aIPnStG7wycA8jHTDrAjcCR9j+QM2M0S6SXgCcC4y2H4xI511EDJikmygzGDZoPi/wWcn2+oPIFdOHpI2BLwE7U5YyfRw4G3i/7WtqZouIiIiIGE0KDEOsKTK8HtidsrTNHZTiwndTXIh+SXqA0hG8OPAQsARlHeE5wF3AnbafXi9htI2kk4HXLuCSFBgiImJakrQkZQboPbbn1s4TEdFN0tMp73pL9p6zff7gE0VERE0pMMSoJK1lOxvJxYQ1a1GvAzwduBjYgLKXxzzKfh+zba9QL2G0jaTbKcXPfwKOpIwU3gL4JLAp8Abbl9dLGBHDTNJKALbvqZ0lpj9JmwP/YftVtbNExPBqCgsnUGZZjca2s9dnRMSQWax2gJhaJK0k6QvAdbWzROt0NgbfBPh/lCVsngM8n9IxfEGlXNFeKzfHb3UabP8eeDuwMWXt14iIgZK0h6Q/UtacvlPSHyS9pnauaC9JS0t6o6SPSNpP0ipd59aSdCJwKbBLvZQREQD8F/AyyrveWP8iImLIZAbDkJG0O/AeykjzO4Bv2/56c+59wCHAcgBZeiT6IWknYEvgp8CfgXOAZzWnrwZ2t31DpXjRQpLuB5YBZgH3U6ZgPxt4ELgVuM/2SvUSRsSwkbQdZW+Y3k6U+cBOWRYi+iVpLcogjLW7mh8EXkVZbvIUym9h9h6KiOok3QWsSCmy/5yyNO4TOpVs71chWkREVJSpa0NE0m6Ul5TOC/HGwHaS1gCexshoYAGPDD5htJntcyhFBQAkbQZsDjwGXGP78VrZorXuoHSqrATcRFkW6WeUewpKh15ExCB9mNFnAC8GfBBIgSH6dQhl4E+3ZYGjgaWazwCzgc8PLlZExKg6v4Hb2r6+apKIiJgyskTScHkf5f/mvVMYPwi8m5HCw0mMjDyPWCguLrd9ZYoLsZB+1xw3p2xAL8oeH2tSRkr9tFKuiBheW1O+f75GWcZtVeDrzbltaoWKVtuZck/dDBwOHNF8fibl9+5x4EvAhrY/WStkRETjzOb42AKvioiIoZIlkoZI13TGo4AvUzrr3gu8lfJicxXwZtuXVQsZrSJpTWBd4OHOfSNpSeDfgddRRp5fDXzC9o+rBY1WkrQFZdbCJcBfgO8Dr6B8X50NvMn2nfUSRsSwkfQY5flpedsPNm3LUpZxm5+NLaNfkuYCiwPPtn1t07Yp5bnclN+671SMGBHxV5K2Ac4AbgA+DlwLPNp9je1bKkSLiIiKUmAYImO8FC8H3Ed5gdnU9h8qRoyWkfR14B+BE23v27T9F/Aunrw29d/aPnvwKWM6kbQC8FjnOywiYpAkzWeUdfDHao8Yz3j3FDDLdkYKR8SU0PXdNBan2B4RMXzyxT9cFqP84P+1Y872bEmdzykuRL+2aI7fA5D0NMqMmI45wNKUe+8AyqjziIVm+77aGSIiJH18Iu22DxtMomi7se4p4KOdZ3XIPRURU4LGvyQiIoZJZjAMka7RBjf3nFpvlHbb3nBA0aKlJN0GrA6sZ/tWSbtQ1uU0cAFlXeG9gGOBO20/vVbWaB9JMykbqu5DWYprVs8lGSEVEQM1gZGbT5AZDTGe3FMR0SaSjhnvGtv7DSJLRERMHemYGU7r9vztnnbRx4tODLWVmuPtzbF7g8tjbT8m6QeUAsOKgwwW08KnKJvQQ0ZKRcTUMdHvozxLxUTlnoqIVkjxICIiRpMCw/BJJ11MpgeAlYF1KBt97dR17rzm2HkZnj3AXDE97MnId9bdwEOkcyUi6jq0doCYdnJPRURERESrZYmkISKpd+bCAtnuXUop4gkkXQRsBVwCXAq8ndIBfL3tjZtrng/8BrjG9rNrZY32kfQQsCSwh+3Ta+eJiIiIiBhmko5ewGlTBgX9r+2fDihSRERMAZnBMERSMIhF4GvA1sALmn+diuVRXdfs1hwvH2CumB4uBnYELqwdJCIiIiIieAvjzyg+UNJ3be81gDwRETEFLFY7QES02gmUIgOUpWwE/AD4PICkJYC3NefPGXi6aLsPUpZF+rykZWuHiYiIiIgIYOTdb6x/b5C0b714ERExSFkiKSKeMkmrAhsCt9q+rat9CWDt5s/bbM+tkS/aQ9INPU2rAksDjwN/AR7tOmfbGw4qW0RERETEMJO0G2WA2WzgP4FbKO97BwLLAR+jLJu7NXCu7Z3G+J+KiIhpJEskRcRTZvtO4M5R2h8Brh98omix9SjTrjubO3c+zwTW7Lk2FfKIiIiIiMHZDXgGsIPtv77nSToPuA54EfB3wJ+AzaskjIiIgcsMhoiImDIk3UQfhQPb6y+6NBERERER0SHpLmBFYC3bt3e1PwO4Dbjb9qqSbgNWtb1EpagRETFAmcEQERFThu31ameIiIiIiIhRdQoGP5L0KcoSSWsCH2naZ3Vdd/+As0VERCXZ5DkiIqYMSS+S9ENJR0l60m+UpBnNuZMlvahGxoiIiIiIIfU/lOVLXwicAlwCnApsS5mF/JNmf76VgT/UChkREYOVAsMQk7SKpCMkXSXp/5q2j0j6uKT16qaLiCH1j8AewHW25/eetP04ZV+P1wL7DzhbRERERMQwey9wFaXI0Pvvqub8RsDJwDcqZYyIiAHLHgxDStLKwC+B9SkPA7Y9Q9LXgLcBB9v+ZM2METF8JF0NbAxs3L1xXM81z6RsIneN7WcPMl9ERERExDCTNAt4M7ATsApwF3A2cILteTWzRUREHSkwDClJXwLe09XUKTDsSHk4uMj2dnXSRRtJOppyHz1pVLmkN1NOHj/wYNEqkmYDTwNm2X5sjGtmAo8AD9pebpD5IiIiIiIiIiJiRAoMQ0rSjcA6wBuB7zBSYFgd+DNwu+01a2aMdpE0n+Y+GuPcfNvZWD4WSNIDwNLA2rb/PMY1zwBuA+bYXmaQ+SIiIiIihomkdQBs39L5vCC2b1n0qSIiYipJZ9/wWqM5/qin/f7muMoAs8Q0JmnpzseqQaItbgU2Ad4OHDLGNW/vujYiIiIiIhadm4D5lP6jmyibOY/FpJ8pImLo5It/eD0IrACs1tO+fXO8n4hxSNqDsiFvd9vRPZdt1BxnDyRUtN05wKbAxyQtC3ze9u0AzQyrDwHvo7y8/G+1lBERERERw0NjfI6IiEiBYYj9BngZ8PVOg6R/AT5I6bj7VaVc0S7PBd7CyCgWAfuOcp2B3w4oU7TbEcD+wBLA+4H3N/syAHT2WxAwD/jy4I9KnMkAAB0WSURBVONFRERERAyV4xl53+v+HBERAWQPhqElaRfgTJ78cKCmbVfbZw08WLSKpIOBg5s/u4sMva4E9rZ9+UCCRatJ2g/4JrDYGJfMB95uu3e2TEREREREREREDFAKDENM0oHAvwOLdzU/CnzM9ufqpIo2kbQ8ZaktATdQigwbdF1i4G7bD1WIFy0maSfgP4Ate079CjjI9tmDTxURERERMbwkHQscZfvntbNERMTUkQLDkJO0FrALZS+GO4CzbGfj1OibpEMA2z60dpaYPiStCqzX/HmT7TsrxomIiIiIGFqS5lMGkd0AHA0cZ/vPdVNFRERtKTBExKSS9HRgB2Bl20fWzhMRERERERFPXVNg6DBl6dKzKMWGU20/ViVYRERUlQLDEJH08X6ut33YosoS01Oz7NYnKRv02vZMSZcDmwF72v5+1YARERERERGxUCRtAuwFvBF4ZtPc6VS6GzjR9gdqZIuIiHpSYBgiXdMZJ8T2jEUYJ6YZSXsAp3Q12fYMSW8HvgZ8y/Y+ddJFRERERETEZJG0JaXQ8HpgjabZ6UeIiBg+i9UOEAOnCf6L6NcBlALW+T3tFzTH3s16IyIiIiIiooVs/xr4IXA6kKWRIiKG2MzaAWKgdqwdIKa15zXHvYA/dbXf2ByfMdg4ERERERERMZkkbUF559sTWKvT3BxvqpEpIiLqSoFhiNg+r/NZ0izb82rmiWlnieZ4V0/7ys1x8QFmiYiIiIiIiEkk6Upg086fzfFhylK5R9s+p0qwiIioKkskDa/bJX2pGX0QMRlubo67dhokzQQ+1fx545P+i4iIiIiIiGiLZzGyrPKvgXcBz7C9d4oLERHDK5s8D6meDZ8vAb4JnGT7wXqpos0kfQ44kLL+5kzK/fUgsExzyWdt/2uleBEREREREfEUSLoDOIEyW+HK2nkiImJqSIFhSEm6hJE186F0Bs8BvgscZfviKsGitSStAPwG2ICR4lVn2uwNwAtt31cjW0RERERERDw1kmbazobOERHxBCkwDDFJGwJvBN4APKdp7twQVwNHUooNcyvEixaStCrwCWA3YDXgDuB04OO276iZLSIiIiIiIhaepO3Hu8b2+YPIEhERU0cKDAGApM2ALwM7NE2dG+NWYFfbV1UJFhERERERERHV9Sy1PBrbnjmoPBERMTWkwDDkJC0H7Au8g7JhE5RlbeZR1tGfAZxne8c6CaONJK0GLNnbbvuWCnEiIiIiIiLiKWoKDAti2zMGEiYiIqaMVJaHlKQtgXdSlkdaipG18v8CfJWyPNJqwKXAi2pkjHaRtAzwn8DewKxRLjH5zomIiIiIiGirQ3v+nknZg++1wOPA4QNPFBER1aWzb3j9ktLh2yksXAZ8ETjJ9qNN212SbgE2rJAv2udwYP/aISIiIiIiImLy2e4tMAAgaSvgYuDhwSaKiIipIEskDalmauN84DTg8LE2YpL0GmB528cNMl+0j6TbKbNeHgF+DzxEz/qcWWorIiIiIiJi+pE0G7jTdgYoRkQMmcxgGF5fBL5o+6YFXWT7R4OJE9PA05rj9rZ/XTVJRERERERETCpJ24/SvCSwO7AMsNhgE0VExFSQAsPwWgH4N0ZZ0kbSmwFsHz/oUNFq5wKvAm6qGyMiIiIiIiIWgXPpmaXexcDvBhclIiKmiiyRNKSaJZJse8YY5+bbTgEqJkzSJpR1N38BfBC41vbjdVNFRERERETEZGj6CsZyL7Cr7V8OKk9EREwNKTAMqbEKDJKWBh4c7VzEgkgar5jgFK0iIiIiIiLaSdLBozTPA24GzrR9/4AjRUTEFJDOviEiaQ9gj562o3su26g5zh5IqJhORJkWq9pBIiIiIiIiYnLZPrR2hoiImHpSYBguzwXewsiaiQL2HeU6A78dUKaYPs5n7PU4IyIiIiIiIiIiYppJgWE4dUaadz73uhJ4/+DixHRg+6W1M0RERERERMSiI2kf4ABgE2DJntNZFjciYghlD4YhIml5YAVKUeEGSpFhg65LDNxt+6EK8aLFJM0CfkK5h95l+7rKkSIiIiIiImISSXo98B3GXho3ezlGRAyhVJaHSLPh0v0Akg4rTb65bqqYDmzPk/QiYCnKBl8RERERERExvby7OT4MLE0pNNwDrAzc1/yLiIghs1jtAFGH7UO6N2iStJikN0j6kKTn1swWrXVRc1y/aoqIiIiIiIhYFDanFBVe1mmwvSpwMPAosHulXBERUVGWSBpSkv4D2B/4iu1DJZ0MvKY5PR/YxfbZ1QJG60h6DnAucCPwz8Bvbc+rGioiIiIiIiImhaRHgBmUmesPN81LALOAB4Gf2d65UryIiKgkBYYhJeli4EWUkQfXAbf2XHK27ZcPPFi0lqTHx7kkG35FRERERES0lKS7gBUpyyP9BVgW2IOyFPN5wBzby9RLGBERNWSJpOG1YXO8Etiy+XwisF/z+XkDTxRtp67jWP8iIiIiIiKinf7cHFcDrm4+/5gykx3KfgwRETFkMpp4eC3fHO8BNqWso3ga8CPgGGC5Srmivc6n3EcREREREREx/VwGbAZsBRzfHLsdN/BEERFRXQoMw+seyqiD1wKvaNquo0xxBJhdI1S0l+2X1s4QERERERERi8w/AR8GHrA9R9LywBuAx4BTgM/UDBcREXVkD4YhJel/gO49Fh4AVgGeC/yKskHv82tki+lB0lK2Hx7/yoiIiIiIiIiIiGij7MEwvD4NzGVkbfzP2n4M2K05f1GtYNFekjaS9CNJD1GKVkg6QtLRkp5TOV5ERERERERMAklLS9pZ0uskbVY7T0RE1JMlkoaU7XMlbUrZ4PlG25c1p74L/D/ghmrhopUkrQtcDKxIKVp1pkc9CuwL3A4cVCddRERERERE9EvSHsCrgBtsf7pp24Kyh+OaXdf9CNjT9qNVgkZERDWZwTDEbN9q+4ddxQVsX237Qtu318wWrXQIsBKloNDtR5SCw8sGHSgiIiIiIiKekjcB+wOzutq+CqzFyIoIAl4DHDDwdBERUV0KDAGApBslXV87R7TaKyizFv62p/3a5rjuYONERERERETEU9RZ/ugMAEnrAdtQ3v0eAb5JeecTZcPniIgYMlkiKTrWZWRJm4iFsWpz7N2/Y25zXHGAWSIiIiIiIuKpW605/qE5bt917ijb7272YLgC2HigySIiYkrIDIaImCz3NMf1etp3b453Dy5KRERERERETIJlmuO85rhN17nTmuM1zXHxgSSKiIgpJQWGiJgsFzfHb3caJH0NOJoyO+aCGqEiIiIiIiJiod3WHF8maRZlw2eAxxmZvb5Cc7xrkMEiImJqSIEhIibLZ4D5wPMZWW7rbZTNwOYDX6iUKyIiIiIiIhbOJZT9FU4Bbqds7mzgfNuzm2u2ao63Pfk/j4iI6S4FhgDA9mK2Z9TOEe1l+xfAm4B7KQ+gnX/3AvvY/mXFeBEREREREdG/TwKPAjMYmalg4BNd1+zZHC8cYK6IiJgiUmAYUpJ2H+f8hweVJaYP298D1gZeAezdHNe2/Z2qwSIiIiIiIqJvtq8AdgbOAK4Dzgb2sH0egKRlgTUoxYUf18oZERH1yPb4V8W0I2k+8BXgg7bndbU/AzgReGlmNMTCkDQD2BxYhbIG5xW2H6+bKiIiIiIiIiIiIiZbZjAMt38CfiXp2QCS9gCuAHasmipaS9JbgD8BvwH+pzn+SdK+NXNFRERERERERETE5EuBYXgdR1kf/2+AX0s6FfghsDLwCHBQxWzRQk0R4WhgNZ64B8PTgaMl7VMxXkREREREREREREyyLJE0xCTtBnwDWJ2ySZOA3wFvtn15zWzRPpKuATZu/rwIuAVYB3hx03ad7U1rZIuIiIiIiIiIiIjJN7N2gKjqGuBWyghzUYoMf6B0DEf0az3KPbS37ZM6jZL2ouzrsW6lXBEREREREREREbEIZImkISXpfcBvgRdSigs3NMfXAr+X9MqK8aKdrm+Op/a0/7jnfEREREREREREREwDKTAMr8OBpYH7gdcDGwEfAx4HngGcVi9atNQnKUWq1/S0v6brfEREREREREREREwT2YNhSEmaD5wP7GP71q72rYBvA+vZnlErX7SDpKN7mv6WsqdH7x4MtwNn2d5/sAkjIiIiIiIiIiJiUUmBYUhJ+hjwKY9yA0haFviq7X0GnyzapClUTfhLJEWriIiIiIiIiIiI6SMFhohYaE2BYaKcAkNERERERERERMT0MbN2gKhD0pvHu8b28YPIEq22fu0AERERERERERERUUdmMAypCSxtY9spQEVERERERERERETEqNKBPNxUO0BMP5JWADYCluo9Z/v8wSeKiIiIiIiIiIiIRSEFhuG1X8/fM4ENgP2BpwH/v717j7Wsru4A/l0zvCwiioLg26C22moNKqiAD1q1VVGjCPgIkarE1JjWVm0bQcRE09RiWmPxQYqA+KC1qNG+fEKMimiqQn0UifIwomAAecjwmtU/zr5w5syd4d6Bueeeez6fZGef39q/ffa6k5vJzVnn91tvWfGMmGlVtWOSDyQ5Ksm6RaZ0/J8DAAAAAGuGD/vmVHeftli8qk5N8qPYW5/le1M2L1wBAAAAAGvUYt8yZr5dlOSGJK+YdiLMnCMzWqXw3WHcSc5KsiGj36tFi1oAAAAAwGyygmFOVdVRi4R3SfK8JPfM1htAw2L2Hc6HZVRQSHcfVlXPS/LpJH81rcQAAAAAgLtfdfsceR5V1cZsvYjw5e5+1krlw+yrqpsyKlrulNGqhXUZFas2Jrkxyf929+OmlyEAAAAAcHeygmG+1RbiFyV5/UomwppwdZI9k9wjyVVJ7pfkuCTXD9cfMaW8AAAAAIDtQIFhfi3WjPemJJckOa+7b1vhfJh9P8mowPDAJP+T5Dm5Y1ukTvLTKeUFAAAAAGwHCgxzqrs13OXu9oUk90nyO0n+Psmzckcj+U7yjinlBQAAAABsB3owkKpa2NZmE9196RTSYY2oqqcmeWmSW5N8uru/NuWUAAAAAIC7kQLDHKuq45L8WUbfOp/U3W2FCwAAAAAAi1p351NYi6rqdUlOSLJHRs2eFztgq6rqgKo6q6r+uao2+/+kqtZX1SnDnP2nkSMAAAAAsH0oMMyvhSbPPxvOnVFj3k5yWZJzppEUM+fVSV6Y5MLu3jh5cWgWflGSFw1zAQAAAIA1QoFhfj06o2LCcxcC3f3EJK9Jct8kx08pL2bLwcP5k1uZ8y8TcwEAAACANUAPhjlVVTcnWZ9kpyQ3ZbQl0i5D7DdJzuvuJ08vQ2ZBVV2bZNckO3f3rVuYs0OSm5Nc3933Wsn8AAAAAIDtRxPf+XVNRisVdk7y6yT3zmj1wnXD9cdOKS9my0Kvjr2S/HwLc/YczlZMAQAAAMAa4gO/+XXZcN47yfnD6/clOS2jrZMun0ZSzJyF36NjtjJn4dplW5kDAAAAAMwYBYb5dV5G29bsn+T9GX0Tffx47/RSY4Z8OaPfl2Or6sSq2mfhQlXtXVUnJjkuo6LVF6eUIwAAAACwHejBQJKkqo5MckSSW5N8qrs/NuWUmAFV9YgkF2TUy2PBtcN5od9CZdTn4/e7+8IVTA8AAAAA2I4UGIC7pKqOTnJytrwiamOSY7r7lJXLCgAAAADY3myRNMeqaueq+vOq+o+qOneIvbyqjqqqPe/sfkiS7v5wkmcn+dYil89L8hzFBQAAAABYe6xgmFNVdY8kZyd5YkZb2HR3r6+qjyY5Msmbu/s9U0yRGTQUph42DC/u7iunmA4AAAAAsB1ZwTC/3prkSRkVF8adMcSeu+IZMfO6+8ru/tZwKC4AAAAAwBqmwDC/Xpqkk7x5In7BcH7kyqYDAAAAAMAssUXSnKqqDUl2TLJrkt/kji2Sdk5yY5Kbu3uXaeYIAAAAAMDqZQXD/NownHebiD9hOP9mBXMBAAAAAGDGKDDMr/OH898uBKrqyCQfyWjrpO9OIykAAAAAAGaDLZLmVFW9PKOGzpO/ADXEXtndH1/xxAAAAAAAmAlWMMyp7v5Ykn/KqKAwfiTJ+xUXAAAAAADYGisY5lxVHZDk0CR7Jbkiyee6+9zpZgUAAAAAwGqnwAAAAAAAACzbDtNOgJVTVUctZ353n769cgEAAAAAYLZZwTBHqmpjNm/qvCXd3QpQAAAAAAAsygfI86fu5HovYQ4AAAAAAHNOgWG+nDAxPibJ7knOSvKzJA9K8uIkNyQ5aWVTAwAAAABgligwzJHuvr3AUFVvTLJ3kkO6+5yx+IeTfCnJhpXPEAAAAACAWaEHw5yqqp8meUiSe3f3dWPxeyW5Jskl3f3waeUHAAAAAMDqtm7aCTA1ew/nd1bVbkkynN85cR0AAAAAADZjBcOcqqpvJNl/LHRdkt3Gxt/s7qeubFYAAAAAAMwKKxjm118kuTFJDce9xl7fOFwHAAAAAIBFKTDMqe5eWMFwZpIrktyW5JdJPpHkSd197hTTAwAAAABglbNF0hyqqnVJHjQMr+zuG6eZDwAAAAAAs0eBYQ5V1fokNyfpJI/s7p9OOSUAAAAAAGaMLZLmUHcvbIdUSX4x5XQAAAAAAJhBCgzz6+PD+flTzQIAAAAAgJm0w7QTYGouTHJVktOq6g+SfCfJJr0Yuvv0aSQGAAAAAMDqpwfDnKqqjRn1YKjhPKm7WwEKAAAAAIBF+QB5vtXEGQAAAAAAlkSBYX4dPe0EAAAAAACYXbZIAgAAAAAAls0KhjlXVQ9K8uwkeyW5Isnnu/tn080KAAAAAIDVzgqGOVZVf53khGxaaLo1ybHd/e7pZAUAAAAAwCxQYJhTVXVoks8k6Wze5LmTvLC7P7fiiQEAAAAAMBMUGOZUVX0xySFJrklyapJLkjwkyauS3CfJl7v7D6eVHwAAAAAAq5sCw5yqqquT3CvJ07r7a2PxA5N8Nck13b3HtPIDAAAAAGB1U2CYU1V1U0a9F3bv7uvH4rsl+XWSW7p752nlBwAAAADA6rZu2gkwNb8Yzm+uqnVJUlWV5E1D/PKpZAUAAAAAwExQYJhfX8ioufOxSX5ZVd9LcsUw7iSfn2JuAAAAAACscrZImlNV9dAk30ly74wKCrdfSnJVkv26+9Jp5AYAAAAAwOpnBcOc6u5LkhyY5EtJNmZUWNiY0cqGgxUXAAAAAADYGisYSFXtkmSPJFd194Zp5wMAAAAAwOqnwMDtqmqvJFd39y3TzgUAAAAAgNXNFklzpqr2q6q/q6r3VtUhQ+y1VXVlksuTXFNVfz/dLAEAAAAAWO2sYJgjVXVQRj0XdhgLvzvJWzJq9FxDrJO8vrs/sLIZAgAAAAAwKxQY5khVfSbJoRPh8cLCr5Lcb3j9je4+cKVyAwAAAABgttgiab48MaOCwn8n+dMk/5lRcaGTvKy790ryimHuY6aSIQAAAAAAM8EKhjlSVTdltD3Sfbr72qraPcnVGRUYdunuW6pqpyQbkmzs7h228nYAAAAAAMwxBYY5UlUbk3R3r19uDAAAAAAAxvmG+hyqqrctJQYAAAAAAFtiBcMcWViZsJSpsYIBAAAAAICtsIJh/tS0EwAAAAAAYPYpMMyXE6adAAAAAAAAa4MtkgAAAAAAgGVbN+0EAAAAAACA2aPAAAAAAAAALJsCAwAAAAAAsGwKDAAAcBdU1auqqseOi6edEwAAwErYYdoJAAAAW1dVz0jyjLHQxd196lSSAQAAGCgwAADA6veMJMePjc9JcupUMgEAABgoMAAAwF3zySRnj41vnVIeAAAAK0qBAQAA7oLuvj7J9dPOAwAAYKVp8gwAwKpSVW+eaJr8+S3M26uqbh2bt7GqHlJVu1fVMVV1clV9u6ourarrqurmqvpVVX29qt5VVQ/dwvs+Y+L5PcQPrqrPVNUVw7PePsSX3OS5qg6qqlOq6v+GnDZU1cVV9dGqevoi898+PP/4iUtPn8yxqvatqssmYn+0hTzeMzHv1C3lDAAAsCUKDAAArDZnJLltbHxIVe25yLzDkqwfG3+luy9N8ugkH0zymiRPSPLgJPdMsmOS+yZ5SpK/SfL9qnrRUhKqqtdmtA3SC5LsmaSW8fOkqnatqo8l+WqSo5M8ashp5yQPTfLyJGcPxYedlvPeY25LcvJE7OgtzH3xxPi0bXwmAAAwxxQYAABYVbr78iRfGgutz+YfiCfJ4RPj05f5qF2TfLSqHrCEue/PNv7tXFWV5ONJXraE6UcnOWlbnjM4OcktY+MXVtUeE/k8MaOixoJLs2kPCQAAgCVRYAAAYDWa/Eb9EeODqtonycFjoRuS/NvwemOSbyd5a5LnJ9k/oxUDT0ryhiRXjd33WxmtdLgz65OcleRpSR6T5CVJzlvCfQu5Hzo2vjHJW5Lsl+RxSd6YZMPY9VdX1cLP9g9JHp7kHyfe85tDfPz42VCc+fTYvJ0zWh0x7iUT4490dy/xZwEAALidJs8AAKxGn0pyXZLdhvHTq+r+3f3LYXxYNv2yzFlDs+V093kZFRMW8+1hC6ITx2IHLiGfs5McNvZB/A+XcM+C102M39jdHxwbXzCsMjhuLHZMkq929zVJrqmqaybeY0N3X7yF552U5KVj46OTvG9sPLkaZLkrPwAAAJJYwQAAwCrU3Tcm+eRYaF1GRYUFR2x6x6YrHqpq76p6W1WdU1WXD82UFxo2nzhx71K2SHr3tnzLv6rWZ9TzYdwHFmkifdzEnIOW+6wF3X12kh+MhfarqscN+Tw2o9UcC87t7gu39VkAAMB8U2AAAGC1mvxm/eFJUlUPTPLUsfhlSb6yMKiqZya5MMkJGW1ptHdGWwVtyT2XkMv5S5izmPsm2ZamzXtv4/MWfGBi/CfDeXJ7JM2dAQCAbabAAADAanVOkovHxgcNDZkPT1Jj8TO6e2OSDNsffTR3bK20FHXnU/LzZbzfct97Mbts430LTsuoL8WCV1TVjtm0wHBTkjPv4nMAAIA5psAAAMCqNGxJdMZYaGGbpMMnpo6vdDgwyT4T19+X5MlJ9s2oGfKx25DLxuXeM/hVklsmYq/M5g2aFzu2WXdfm1GhZcH9kvxlkt8bi322u6++K88BAADmmwIDAACr2eQ2Sa9PcsDY+Lzu/tHYeLK48OvufkN3f7O7fzI0Rt5vO+S5qO6+Lck3JsIv6O6Lt3RkVAy498Q9k0WKeyzh8SdNjN8xMbY9EgAAcJcoMAAAsGp194+z6Qf0j8qm2w5Nfkj+q4nx7lV1XFU9pqqeUlUfTvLi7ZDq1kz2Qzi8qj5TVc+tqkdX1e9W1R9X1fFV9Z0k30ry+Il7rpwYP76qXlRV+1bVw6pqs54N3f29bPpvt+PY6yuS/Nc2/jwAAABJkh2mnQAAANyJ05M8ZZH4zUk+MRH7WpKrkuwxFntHNv32/i+T3P/uTPBOnJnkFUmeNxZ7wXAs1bkT452SfGps/LUkBy1y30lZ/N/uY9196zKeDwAAsBkrGAAAWO3OzKgh8aTPdfdV44HuviGjbZS21DPh60lOuHvT27qhf8MR2bQnwtbcnIkVC919frZtxcG/Tr7XwPZIAADAXabAAADAqjY0Iv7sIpcm+zMszP9Ekmcm+XySa5PcmOQHGTV3fuYwXlHdfUN3vzKjZtMfSvL9Ibfbkvw6yQVJPpLkVUn26e5/X+RtXpLkXUl+lGTDEp97U5JTJsIXdPd3t+HHAAAA2ER197RzAAAAtpOqek2Sk8dCb+ruE6eVDwAAsHYoMAAAwBpVVfsk+UqS3x5CNyV5cHcvtm0SAADAsmjyDAAAa0xVXZRkxyQPyKZ/839IcQEAALi7WMEAAABrTFUt9kf+D5Mc0N3XrXQ+AADA2qTJMwAArF23JPlxkncnOVBxAQAAuDtZwQAAAAAAACybFQwAAAAAAMCyKTAAAAAAAADLpsAAAAAAAAAsmwIDAAAAAACwbAoMAAAAAADAsikwAAAAAAAAy/b/D9ki8AwsofUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 2160x1080 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#New heat map for clusters\n",
    "variety_common['cluster'] = kmeans21.labels_\n",
    "clusters = variety_common.groupby(['cluster', 'variety']).size()\n",
    "fig2, ax2 = plt.subplots(figsize = (30, 15))\n",
    "sns.heatmap(clusters.unstack(level = 'variety'), ax = ax2, cmap = 'Reds')\n",
    "\n",
    "ax2.set_xlabel('variety', fontdict = {'weight': 'bold', 'size': 24})\n",
    "plt.xticks(rotation = 90)\n",
    "ax2.set_ylabel('cluster', fontdict = {'weight': 'bold', 'size': 24})\n",
    "for label in ax2.get_xticklabels():\n",
    "    label.set_size(16)\n",
    "    label.set_weight(\"bold\")\n",
    "for label in ax2.get_yticklabels():\n",
    "    label.set_size(16)\n",
    "    label.set_weight(\"bold\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Mathias\\anaconda3\\envs\\HOML\\lib\\site-packages\\sklearn\\feature_extraction\\text.py:386: UserWarning: Your stop_words may be inconsistent with your preprocessing. Tokenizing the stop words generated tokens ['abov', 'afterward', 'alon', 'alreadi', 'alway', 'ani', 'anoth', 'anyon', 'anyth', 'anywher', 'becam', 'becaus', 'becom', 'befor', 'besid', 'cri', 'describ', 'dure', 'els', 'elsewher', 'empti', 'everi', 'everyon', 'everyth', 'everywher', 'fifti', 'forti', 'henc', 'hereaft', 'herebi', 'howev', 'hundr', 'inde', 'mani', 'meanwhil', 'moreov', 'nobodi', 'noon', 'noth', 'nowher', 'onc', 'onli', 'otherwis', 'ourselv', 'perhap', 'pleas', 'sever', 'sinc', 'sincer', 'sixti', 'someon', 'someth', 'sometim', 'somewher', 'themselv', 'thenc', 'thereaft', 'therebi', 'therefor', 'togeth', 'twelv', 'twenti', 'veri', 'whatev', 'whenc', 'whenev', 'wherea', 'whereaft', 'wherebi', 'wherev', 'whi', 'yourselv'] not in stop_words.\n",
      "  'stop_words.' % sorted(inconsistent))\n"
     ]
    }
   ],
   "source": [
    "#I want a small amount to use Birch on\n",
    "birch_train, birch_set = split_train_test(train_set, 0.90)\n",
    "\n",
    "len(birch_train)\n",
    "\n",
    "X3, word_features3, dataframe_common = pipeline(birch_train, 400)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([4, 4, 4, ..., 4, 4, 4], dtype=int64)"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.cluster import Birch\n",
    "\n",
    "birch = Birch(threshold = 0.97, n_clusters = 19)\n",
    "birch.fit(X3)\n",
    "birch.labels_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 : grace, almond, vibrant, steeli, follow, grape, mix, produc, open, lemon\n",
      "1 : ani, compon, leaf, nuanc, sip, bring, plenti, strong, herbal, charact\n",
      "2 : low, off-dri, delic, flower, suggest, alcohol, riesl, littl, clean, mouth\n",
      "3 : aroma, palat, finish, berri, note, cherri, herbal, plum, fruit, nose\n",
      "4 : astring, bring, chocol, bitter, aroma, open, persist, mouth, finish, provid\n",
      "5 : grapefruit, typic, peel, blanc, way, orang, chardonnay, tast, tart, sauvignon\n",
      "6 : leav, remain, bitter, oaki, domin, new, wood, charact, tast, like\n",
      "7 : vanilla, note, brown, surpris, persist, reserv, sugar, provid, high, coffe\n",
      "8 : mix, strawberri, suggest, simpl, bring, best, smooth, fruiti, drink, fruit\n",
      "9 : blackberri, emerg, reveal, wet, element, wineri, 2018, color, licoric, pepper\n",
      "10 : refin, tomato, leaf, aroma, domin, cola, long, nice, pepper, vanilla\n",
      "11 : chunki, citrusi, includ, bitter, mouth, green, citrus, note, aroma, finish\n",
      "12 : pinot, noir, charact, cherri, good, crisp, delic, tangi, price, defin\n",
      "13 : sage, strawberri, skin, mint, wild, start, color, cranberri, scent, quit\n",
      "14 : fulli, short, load, doesn't, bouquet, lean, follow, strawberri, coffe, cola\n",
      "15 : wine, fruit, tannin, cherri, black, ripe, drink, rich, acid, dri\n",
      "16 : given, blossom, mint, medium-bodi, refresh, alcohol, smoke, pear, peach, complex\n",
      "17 : dust, bone, hit, mushroom, black, befor, mark, roast, wild, element\n",
      "18 : bake, zinfandel, sirah, togeth, valley, petit, cranberri, layer, smoki, tight\n",
      "19 : butteri, bordeaux, varieti, classic, domin, toasti, lot, cassi, sauvignon, cabernet\n",
      "20 : butteri, load, modest, turn, oaki, toasti, bake, pear, peach, vanilla\n",
      "21 : pleas, whiff, sugar, depth, floral, linger, just, white, peach, fruiti\n",
      "22 : lemoni, burst, featur, apricot, simpl, orang, big, great, pineappl, sweet\n",
      "23 : low, old, fair, alcohol, cab, style, tannic, note, dri, ripe\n",
      "24 : green, bean, veget, charact, carri, note, palat, feel, bouquet, pepper\n",
      "25 : chard, candi, pineappl, make, tast, peach, like, vanilla, soft, sweet\n",
      "26 : plum, pomegran, appel, strawberri, simpl, lead, great, bit, hint, dark\n",
      "27 : butterscotch, apricot, easi, soft, fair, end, enjoy, entir, espresso, estat\n",
      "28 : varieti, brisk, savori, noir, cola, make, pinot, raspberri, complex, spice\n",
      "29 : like, natur, rang, citrusi, doesn't, food, riesl, make, tast, pinot\n",
      "30 : old, vine, —, citrus, almond, case, aftertast, reserv, lift, support\n",
      "31 : bit, ad, highlight, ferment, soften, lift, subtl, doe, lean, wineri\n",
      "32 : hot, tast, soft, blackberri, cherri, pepperi, like, cab, littl, fruiti\n",
      "33 : broad, hit, turn, quit, open, herbal, rich, black, ripe, cherri\n",
      "34 : brambl, run, excel, pepperi, streak, 100, stone, valley, bring, earth\n",
      "35 : —, short, lift, cut, varieti, classic, tea, lead, littl, nice\n",
      "36 : butterscotch, work, result, play, featur, medium-bodi, bake, make, white, appl\n",
      "37 : valu, length, roast, tangerin, oaki, accent, pineappl, chardonnay, white, peach\n",
      "38 : use, vine, sourc, despit, grape, tight, concentr, age, year, plum\n",
      "39 : distinct, highlight, 2020, combin, stone, cranberri, tight, –, bottl, concentr\n",
      "40 : hot, mint, alcohol, green, blackberri, tannin, espresso, estat, everyday, extract\n",
      "41 : appl, cream, pair, express, butter, full-bodi, pear, chardonnay, white, vanilla\n",
      "42 : line, fair, midpal, realli, lean, end, earthi, tannic, eleg, veri\n",
      "43 : like, tast, smell, soft, butterscotch, pear, fruiti, pineappl, wine, feel\n",
      "44 : sharp, cinnamon, silki, bit, pepper, veri, textur, spice, dri, cherri\n",
      "45 : fig, juic, espresso, anis, combin, mocha, —, complex, offer, nose\n",
      "46 : barrel, spice, effort, ferment, anoth, 100, appeal, provid, french, cinnamon\n",
      "47 : work, smell, color, live, lot, strong, white, tast, like, pinot\n",
      "48 : chard, add, creami, earthi, come, peach, like, herb, crisp, wine\n",
      "49 : grip, intrigu, certain, rustic, depth, appeal, wild, meat, smoke, make\n",
      "50 : thing, continu, chard, sip, pleasant, valley, pear, white, lemon, peach\n",
      "51 : grip, excel, dusti, close, meat, enjoy, produc, smoke, plenti, licoric\n",
      "52 : ani, earthi, tart, like, pinot, —, excel, extra, express, exot\n",
      "53 : nice, highlight, focus, wild, strawberri, core, mouthfeel, pepper, pinot, herb\n",
      "54 : famili, relat, fig, 2016, melon, tangi, grapefruit, need, new, just\n",
      "55 : entir, ferment, sour, bodi, big, lime, lemon, tast, appl, offer\n",
      "56 : addit, doesn't, streak, lean, right, lush, tobacco, develop, make, bottl\n",
      "57 : appl, acid, lemon, wine, finish, crisp, citrus, peach, pear, lime\n",
      "58 : butterscotch, sage, matur, close, apricot, honey, mouthfeel, creami, nice, chardonnay\n",
      "59 : appel, provid, jam, blueberri, live, lush, cedar, need, style, vineyard\n",
      "60 : fresh, close, aromat, zesti, refresh, best, pear, white, peach, nose\n",
      "61 : clean, lemoni, refresh, bodi, fresh, zesti, live, plenti, riesl, lead\n",
      "62 : gooseberri, classic, follow, tangi, live, grapefruit, bright, lemon, herb, veri\n",
      "63 : intrigu, rustic, glass, express, provid, delic, way, live, produc, earthi\n",
      "64 : prune, qualiti, jammi, simpl, cola, feel, raspberri, ripe, cherri, fruit\n",
      "65 : gorgeous, lift, matur, frame, 2020, perfum, beauti, edg, –, pinot\n"
     ]
    }
   ],
   "source": [
    "common_words3 = birch.subcluster_centers_.argsort()[:,-1:-11:-1]\n",
    "for num, centroid in enumerate(common_words3):\n",
    "    print(str(num) + ' : ' + ', '.join(word_features3[word] for word in centroid))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7915"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(birch.labels_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x1080 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#New heat map for clusters\n",
    "dataframe_common['cluster'] = birch.labels_\n",
    "clusters = dataframe_common.groupby(['cluster', 'variety']).size()\n",
    "fig3, ax3 = plt.subplots(figsize = (30, 15))\n",
    "sns.heatmap(clusters.unstack(level = 'variety'), ax = ax3, cmap = 'Reds')\n",
    "\n",
    "ax3.set_xlabel('variety', fontdict = {'weight': 'bold', 'size': 24})\n",
    "plt.xticks(rotation = 90)\n",
    "ax3.set_ylabel('cluster', fontdict = {'weight': 'bold', 'size': 24})\n",
    "for label in ax3.get_xticklabels():\n",
    "    label.set_size(16)\n",
    "    label.set_weight(\"bold\")\n",
    "for label in ax3.get_yticklabels():\n",
    "    label.set_size(16)\n",
    "    label.set_weight(\"bold\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}