2-Naive_Bayes.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Estos dos comandos evitan que haya que hacer reload cada vez que se modifica un paquete\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "from problema_1 import plot_bar_chart\n",
    "# Importo el paquete para graficar\n",
    "from matplotlib import pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Naive bayes - Clasificación\n",
    "\n",
    "### Bibliografía recomendada\n",
    "https://nlp.stanford.edu/IR-book/ - Capitulo 13 -  \n",
    "http://scikit-learn.org/stable/modules/naive_bayes.html#naive-bayes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ejemplos"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](./images/student-admission.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![](./images/chip-test.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Introducción teórica"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Teorema de Bayes\n",
    "$P(y|x_1,x_2,...,x_n) = \\large \\frac{P(x_1, x_2,...,x_n|y) P(y)}{P(x_1,x_2,...,x_n)}$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "El numerador:\n",
    "\n",
    "$ \\large P(x_1, x_2,...,x_n|y) P(y) = P(y, x_1, x_2,...,x_n)$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Por la regla de la cadena:\n",
    "\n",
    "$P(x_1,x_2, x_3, y) = P(x_1 | x_2, x_3,y)P(x_2,x_3, y) = P(x_1 | x_2, x_3,y)P(x_2|x_3, y)P(x_3, y) = P(x_1 | x_2, x_3,y)P(x_2|x_3, y)P(x_3| y)P(y)$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$P(x_1,x_2, x_3, y) = P(x_1 | x_2, x_3,y)P(x_2|x_3, y)P(x_3| y)P(y)$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Ejemplo con Aplicación SPAM"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$P(comprar,oferta,ganaste,SPAM) = P(comprar | oferta, ganaste,SPAM)P(oferta|ganaste, SPAM)P(ganaste| SPAM)P(SPAM)$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Asumiendo independencia\n",
    "\n",
    "$P(x_i|y) = P(x_i|y,x_1,...)$\n",
    "\n",
    "$P(comprar,oferta,ganaste,SPAM) = P(comprar | SPAM)P(oferta|SPAM)P(ganaste| SPAM)P(SPAM)$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Naive Bayes\n",
    "\n",
    "$P(y|x_1,x_2,...,x_n) = \\large \\frac{P(y)\\prod P(x_i/y)} {P(x_1,x_2,...,x_n)} = \\alpha {P(y)\\prod P(x_i/y)}$\n",
    "\n",
    "(Compararlo con teorema de Bayes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$\\large \\hat{y} = arg max_y [{P(y)\\prod P(x_i/y)}]$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ejemplo SPAM: \n",
    "\n",
    "La variable aleatoria $y$ puede ser: [SPAM, NO_SPAM]. Calculo ${P(y)\\prod P(x_i/y)}$ para ambos casos y me quedo con el mayor\n",
    "\n",
    "Como el denominador es el mismo para ambos casos, no hace falta calcularlo"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Detección de SPAM\n",
    "V. Metsis, I. Androutsopoulos and G. Paliouras (2006). Spam filtering with Naive Bayes – Which Naive Bayes? 3rd Conf. on Email and Anti-Spam (CEAS).\n",
    "http://citeseerx.ist.psu.edu/viewdoc/download;jsessionid=1EEA3C5EE5DDF8409BCC0F09BFB7D98B?doi=10.1.1.61.5542&rep=rep1&type=pdf\n",
    "\n",
    "Tenemos un dataset donde ya esta preprocesado el texto\n",
    "- Lemmatization, stemming\n",
    "- Stop words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import CountVectorizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from collections import Counter\n",
    "train_dir = './ling-spam/train-mails'\n",
    "emails = [os.path.join(train_dir,f) for f in os.listdir(train_dir)] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['./ling-spam/train-mails/6-338msg1.txt',\n",
       " './ling-spam/train-mails/5-1300msg3.txt',\n",
       " './ling-spam/train-mails/6-250msg1.txt',\n",
       " './ling-spam/train-mails/6-266msg3.txt',\n",
       " './ling-spam/train-mails/spmsgc10.txt',\n",
       " './ling-spam/train-mails/5-1326msg1.txt',\n",
       " './ling-spam/train-mails/spmsga114.txt',\n",
       " './ling-spam/train-mails/spmsga100.txt',\n",
       " './ling-spam/train-mails/6-110msg3.txt',\n",
       " './ling-spam/train-mails/6-126msg1.txt']"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "emails[0:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<_io.TextIOWrapper name='./ling-spam/train-mails/5-1326msg1.txt' mode='r' encoding='UTF-8'>\n",
      "0 Subject: \n",
      "\n",
      "1 \n",
      "\n",
      "2 n n o u n c n g cunyforum 18 , summer 1994 journal work paper linguistic , cuny graduate center c o n t e n t s development dominican vernacular spanish - - keat green pragmatic verbal abuse - - edgar . gregersen parse japanese relative clause : raise lower ? - - yukus hirose interpretation case - delet sentence japanese - - mana kobuchus - philip is innate functional category ? - - ingeborg lasser language contact sao tome principe ( west africa ) - - gerardo . lorenzino comparative morphosyntactic study restructure portuguese africa brazil - - heliana r . mello artificial cognition language - - john . moyne * * * subscription rate : regular $ 10 ; student $ 5 ; institution $ 20 ; oversea add $ 10 . send check : cunyforum ; linguistic department ; cuny graduate center ; 33 west 42nd street ; york , ny 10036 . inquiry emfgc @ cunyvm . cuny . edu .\n",
      "\n"
     ]
    }
   ],
   "source": [
    "with open(emails[5]) as m:\n",
    "    print(m)\n",
    "    for i,line in enumerate(m):\n",
    "        print(i, line)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def make_Dictionary(train_dir):\n",
    "    # Train dir es la carpeta para parsear\n",
    "    emails_array = []\n",
    "    labels = []\n",
    "    # Emails es una lista con los nombres de los archivos\n",
    "    emails = [os.path.join(train_dir,f) for f in os.listdir(train_dir)] \n",
    "    # All words contendrá todas las palabras del mail\n",
    "    all_words = []       \n",
    "    for mail in emails:\n",
    "        # mail es el nombre del archivo\n",
    "        with open(mail) as m:\n",
    "            # m es el archivo abierto\n",
    "            if 'spmsg' in mail.split('/')[-1]:\n",
    "                # Si el nombre del archivo contiene 'spmsg' es spam\n",
    "                labels.append(True)\n",
    "            else:\n",
    "                labels.append(False)\n",
    "            # email_str va a guardar el mail como string\n",
    "            email_str = ''\n",
    "            for i,line in enumerate(m):\n",
    "                if i == 2:  #Body of email is only 3rd line of text file\n",
    "                    words = line.split()\n",
    "                    all_words += words\n",
    "                    email_str = email_str + ' ' + line\n",
    "            emails_array.append(email_str)\n",
    "    dictionary = Counter(all_words)\n",
    "    return dictionary, emails_array, labels\n",
    "\n",
    "def remove_uncommon_words(dictionary):\n",
    "    # Dictionary es un counter\n",
    "    dict_copy = { k : v for k,v in dictionary.items()}\n",
    "    list_to_remove = dictionary.keys()\n",
    "    for item in list_to_remove:\n",
    "        if item.isalpha() == False: \n",
    "            # Remueve lo que no sean palabras con letras del abcedario\n",
    "            del dict_copy[item]\n",
    "        elif len(item) == 1:\n",
    "            # Remueve las palabras de una letra\n",
    "            del dict_copy[item]\n",
    "    return Counter(dict_copy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def read_emails_and_get_dict(folder, max_features=3000):\n",
    "    dictionary, emails_array, labels = make_Dictionary(folder)\n",
    "    dictionary = remove_uncommon_words(dictionary)\n",
    "    dictionary = dictionary.most_common(max_features)\n",
    "    vocabulary = [key[0] for key in dictionary]\n",
    "    return vocabulary, emails_array, np.array(labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "max_features = 3000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "vocabulary, emails_array, labels = read_emails_and_get_dict('./ling-spam/train-mails', max_features)\n",
    "count_vectorizer = CountVectorizer(vocabulary=vocabulary)\n",
    "feature_matrix = count_vectorizer.fit_transform(emails_array)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[False False False False  True False  True  True False False]\n",
      "['order', 'address', 'report', 'mail', 'send', 'language', 'email', 'program', 'our', 'list']\n",
      "  (0, 19)\t1\n",
      "  (0, 24)\t2\n",
      "  (0, 273)\t1\n",
      "  (0, 305)\t1\n",
      "  (0, 533)\t1\n",
      "  (0, 562)\t1\n",
      "  (0, 701)\t1\n",
      "  (0, 717)\t1\n",
      "  (0, 1285)\t1\n",
      "  (0, 1418)\t1\n",
      "  (0, 1650)\t2\n",
      "  (1, 1)\t2\n",
      "  (1, 4)\t3\n",
      "  (1, 5)\t3\n",
      "  (1, 6)\t2\n",
      "  (1, 11)\t1\n",
      "  (1, 12)\t2\n",
      "  (1, 16)\t1\n",
      "  (1, 18)\t3\n",
      "  (1, 19)\t3\n",
      "  (1, 22)\t1\n",
      "  (1, 27)\t2\n",
      "  (1, 29)\t3\n",
      "  (1, 31)\t3\n",
      "  (1, 32)\t1\n",
      "  :\t:\n",
      "  (701, 748)\t1\n",
      "  (701, 800)\t1\n",
      "  (701, 865)\t1\n",
      "  (701, 873)\t1\n",
      "  (701, 913)\t1\n",
      "  (701, 919)\t1\n",
      "  (701, 952)\t1\n",
      "  (701, 1017)\t1\n",
      "  (701, 1150)\t1\n",
      "  (701, 1232)\t1\n",
      "  (701, 1263)\t1\n",
      "  (701, 1355)\t1\n",
      "  (701, 1485)\t1\n",
      "  (701, 1486)\t1\n",
      "  (701, 1603)\t1\n",
      "  (701, 1637)\t1\n",
      "  (701, 1816)\t1\n",
      "  (701, 1825)\t1\n",
      "  (701, 1903)\t1\n",
      "  (701, 2362)\t1\n",
      "  (701, 2402)\t1\n",
      "  (701, 2495)\t1\n",
      "  (701, 2672)\t2\n",
      "  (701, 2777)\t1\n",
      "  (701, 2986)\t1\n"
     ]
    }
   ],
   "source": [
    "print(labels[0:10])\n",
    "print(vocabulary[0:10])\n",
    "print(feature_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 3000)\n",
      "[[1423 1302 1216 ...,    8    8    8]]\n",
      "[[1424 1303 1217 ...,    9    9    9]]\n",
      "157987\n"
     ]
    }
   ],
   "source": [
    "print(feature_matrix.sum(axis=0).shape)\n",
    "print(feature_matrix.sum(axis=0))\n",
    "print(feature_matrix.sum(axis=0) + 1)\n",
    "print((feature_matrix.sum(axis=0) + 1).sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SPAM: 351 HAM: 351 total: 702\n",
      "\n",
      "Palabras diferentes\n",
      "SPAM: 49290 HAM: 33347\n",
      "\n",
      "Cantidad de palabras\n",
      "SPAM: 98144 HAM: 56843\n"
     ]
    }
   ],
   "source": [
    "def get_histogram_from_matrix(alpha, feature_matrix):\n",
    "    # Alpha es para suavizar y evitar la probabilidad cero. Normalmente 1\n",
    "    # Es como agregar un mail con todas las palabras\n",
    "    words_distribution_un = feature_matrix.sum(axis=0) + alpha\n",
    "    words_distribution = words_distribution_un/(words_distribution_un.sum()) # + alpha*feature_matrix.shape[1])\n",
    "    return words_distribution\n",
    "\n",
    "spam_feature_matrix = feature_matrix[labels,:]\n",
    "ham_feature_matrix = feature_matrix[np.logical_not(labels),:]\n",
    "\n",
    "print('SPAM:',labels.nonzero()[0].shape[0],\n",
    "      'HAM:',np.logical_not(labels).nonzero()[0].shape[0],\n",
    "      'total:',len(labels))\n",
    "\n",
    "print()\n",
    "print('Palabras diferentes')\n",
    "print('SPAM:', spam_feature_matrix.nonzero()[0].shape[0],\n",
    "      'HAM:',ham_feature_matrix.nonzero()[0].shape[0])\n",
    "\n",
    "print()\n",
    "print('Cantidad de palabras')\n",
    "print('SPAM:',spam_feature_matrix.sum(), 'HAM:',ham_feature_matrix.sum())\n",
    "\n",
    "\n",
    "# P(x_i/SPAM)\n",
    "spam_words_distribution = get_histogram_from_matrix(1, spam_feature_matrix)\n",
    "# P(x_i/HAM)\n",
    "ham_words_distribution = get_histogram_from_matrix(1, ham_feature_matrix)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$\\hat{y} = arg max_y [{P(y)\\prod P(x_i/y)}]$\n",
    "\n",
    "$y = [SPAM, HAM]$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_bar_chart(P, vocabulary = None, title='title'):\n",
    "    N = len(P)\n",
    "    plt.figure(figsize=(20, 5)) # Tamaño del gráfico\n",
    "    indexes = list(range(1, len(P) + 1))\n",
    "    plt.bar(indexes, P, width=0.75, color='b') # Grafico\n",
    "    plt.title(title)\n",
    "\n",
    "    # Definición de indices\n",
    "    if not vocabulary:\n",
    "        string_indexes = [str(i) for i in indexes]\n",
    "    else:\n",
    "        string_indexes = [vocabulary[i-1] for i in indexes]\n",
    "    plt.xlim([0,N+1])\n",
    "    plt.xticks(indexes, string_indexes) \n",
    "    plt.xticks(rotation=60)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJEAAAFsCAYAAAB8eijSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nUkoppdSh/wcNBxboyYf/gAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f52c8d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJcAAAFsCAYAAABxZFiVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f5302b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "N = 20\n",
    "plot_bar_chart(np.array(spam_words_distribution)[0][:N],\n",
    "               vocabulary = vocabulary,\n",
    "               title='SPAM')\n",
    "plot_bar_chart(np.array(ham_words_distribution)[0][:N],\n",
    "               vocabulary = vocabulary,\n",
    "               title='HAM')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Primer email codificado:\n",
      "  (0, 19)\t1\n",
      "  (0, 24)\t2\n",
      "  (0, 273)\t1\n",
      "  (0, 305)\t1\n",
      "  (0, 533)\t1\n",
      "  (0, 562)\t1\n",
      "  (0, 701)\t1\n",
      "  (0, 717)\t1\n",
      "  (0, 1285)\t1\n",
      "  (0, 1418)\t1\n",
      "  (0, 1650)\t2\n",
      "Elementos distintos de cero: [  19   24  273  305  533  562  701  717 1285 1418 1650]\n",
      "Producto, no es necesario [P(Y) = P(SPAM) = P(HAM)] por que es para ambos casos igual. Lo mismo el denominador\n",
      "6.44553915496e-47\n",
      "2.35145394257e-43\n",
      "Log likelihood\n",
      "[[-106.35811108]]\n",
      "[[-98.15612516]]\n"
     ]
    }
   ],
   "source": [
    "# Veamos si el primer email es SPAM o no\n",
    "# El primer mail corresponde a la primer fila de la matriz\n",
    "row = feature_matrix[0,:]\n",
    "print('Primer email codificado:')\n",
    "print(row)\n",
    "non_zero_row = row.nonzero()\n",
    "print('Elementos distintos de cero:',non_zero_row[1])\n",
    "print('Producto, no es necesario [P(Y) = P(SPAM) = P(HAM)] por que es para ambos casos igual. Lo mismo el denominador')\n",
    "print(np.power(spam_words_distribution[non_zero_row], feature_matrix[non_zero_row]).prod())\n",
    "print(np.power(ham_words_distribution[non_zero_row], feature_matrix[non_zero_row]).prod())\n",
    "print('Log likelihood')\n",
    "print(np.log(spam_words_distribution[non_zero_row])*feature_matrix[non_zero_row].T)\n",
    "print(np.log(ham_words_distribution[non_zero_row])*feature_matrix[non_zero_row].T)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "¿Por que usamos log likelihood?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def get_accuracy(feature_matrix, labels, spam_words_distribution, ham_words_distribution):\n",
    "    P_SPAM = labels.nonzero()[0].shape[0]/len(labels) # Por que no es necesario en este caso?\n",
    "    P_HAM = np.logical_not(labels).nonzero()[0].shape[0]/len(labels) # Por que no es necesario en este caso?\n",
    "    correct_count = 0\n",
    "    for i, row in enumerate(feature_matrix):\n",
    "        coded_email = row.nonzero()\n",
    "        spam_prod = np.log(spam_words_distribution[coded_email]) * feature_matrix[i][coded_email].T\n",
    "        ham_prod = np.log(ham_words_distribution[coded_email]) * feature_matrix[i][coded_email].T\n",
    "        spam_log_likelihood = spam_prod+ np.log(P_SPAM)\n",
    "        ham_log_likelihood = ham_prod + np.log(P_HAM)\n",
    "        SPAM = spam_log_likelihood>=ham_log_likelihood\n",
    "        if SPAM == labels[i]:\n",
    "            correct_count = correct_count + 1\n",
    "    print('%0.3f'%(correct_count/len(labels)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.990\n"
     ]
    }
   ],
   "source": [
    "get_accuracy(feature_matrix, labels, spam_words_distribution, ham_words_distribution)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "_, emails_array, labels_test = read_emails_and_get_dict('./ling-spam/test-mails', max_features)\n",
    "count_vectorizer_test = CountVectorizer(vocabulary=vocabulary)\n",
    "feature_matrix_test = count_vectorizer.fit_transform(emails_array)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.962\n"
     ]
    }
   ],
   "source": [
    "get_accuracy(feature_matrix_test, labels_test, spam_words_distribution, ham_words_distribution)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sklearn MultinomialNB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.naive_bayes import MultinomialNB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "clf = MultinomialNB(alpha=1, class_prior=[0.5, 0.5], fit_prior=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MultinomialNB(alpha=1, class_prior=[0.5, 0.5], fit_prior=False)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.fit(feature_matrix, labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.99002849002849003"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.score(feature_matrix, labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.96153846153846156"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.score(feature_matrix_test, labels_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([False, False,  True, False,  True,  True,  True,  True, False,\n",
       "       False, False, False,  True,  True, False,  True, False, False,\n",
       "        True, False, False,  True, False,  True,  True,  True, False,\n",
       "        True,  True,  True,  True, False, False, False, False, False,\n",
       "       False,  True,  True,  True,  True,  True, False,  True,  True,\n",
       "       False, False,  True,  True, False], dtype=bool)"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.predict(feature_matrix_test[100:150])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  1.00000000e+000,   4.32577416e-203],\n",
       "       [  1.00000000e+000,   9.25701855e-168],\n",
       "       [  4.22016090e-015,   1.00000000e+000],\n",
       "       [  1.00000000e+000,   9.54021707e-106],\n",
       "       [  9.61964302e-147,   1.00000000e+000],\n",
       "       [  2.47312152e-090,   1.00000000e+000],\n",
       "       [  9.85266748e-006,   9.99990147e-001],\n",
       "       [  9.76629082e-029,   1.00000000e+000],\n",
       "       [  1.00000000e+000,   2.04412793e-077],\n",
       "       [  1.00000000e+000,   7.70903147e-057]])"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf.predict_proba(feature_matrix_test[100:110])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sklearn BernoulliNB\n",
    "Suele ser interesante evaluar Multinomial y Bernoulli para comparar resultados. A veces para textos cortos BernoulliNB puede dar mejores resultados"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.naive_bayes import BernoulliNB"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "clf_Ber = BernoulliNB(alpha=1, class_prior=[0.5, 0.5], fit_prior=False, binarize=0.5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "BernoulliNB(alpha=1, binarize=0.5, class_prior=[0.5, 0.5], fit_prior=False)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_Ber.fit(feature_matrix, labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.86039886039886038"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_Ber.score(feature_matrix, labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.76538461538461533"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf_Ber.score(feature_matrix_test, labels_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ¿Es posible graficar un problema de dimensión 3000 en 2 dimensiones?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.manifold import TSNE\n",
    "from sklearn.decomposition import TruncatedSVD\n",
    "fmd = feature_matrix.todense()\n",
    "svd = TruncatedSVD(n_components=8, n_iter=200, random_state=42)\n",
    "svd.fit(fmd.T)\n",
    "X_embedded = TSNE(n_components=2).fit_transform(svd.components_.T)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIkAAAJCCAYAAABETZcVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJzs3X+MdWd+EPbviV9et27RGxIWWLJY\nsdoFFV63UEa7VPzj7iaT1NvepCiOsqrSLU21RpopvKIVSTxqRanGEJCK1doUR0mnaUsT1gEpA+uW\nN+vFRZVCYFzS2CFCWdhGWRGRlGTfqrL0ujanf5y5fq/nvXPvOfc+55znOefzkax5Z3xn7nPPr3ue\n7/1+v09V13UAAAAAMG9fM/YAAAAAABifIBEAAAAAgkQAAAAACBIBAAAAEIJEAAAAAIQgEQAAAACR\nMEhUVdUjVVX9vaqq/vrl909UVfXTVVX9QlVVf7mqqpupngsAAACAtFJmEv2xiPj5le9/ICL+fF3X\nH42IX4+I70n4XAAAAAAklCRIVFXVRyLiUxHxQ5ffVxHxiYj48cuH/EhEfHuK5wIAAAAgvRuJ/s4L\nEfEnIuI3Xn7/9RHx1bqu3738/isR8Q3b/shv/s2/uf7Gb/zGREMCAAAA4I033vi/67r+0LbH7R0k\nqqrq346IX6nr+o2qqp5a/njNQ+trfv+zEfHZiIjHH388Li4u9h0SAAAAAJeqqvrFNo9LUW72ByNi\nUVXV/xURPxZNmdkLEfG1VVUtg1AfiYh/vO6X67r+wbquD+q6PvjQh7YGtQAAAADowd5Borquv7+u\n64/Udf2NEfFdEfHFuq7/vYj4mxHxHZcP+0xE/MS+zwUAAABAP1KubnbV90bEH6+q6kvR9Cj64R6f\nCwAAAIA9pGpcHRERdV2/HhGvX/77H0XEx1L+fQAAAAD60WcmEQAAAACFECQCAAAAQJAIAAAAAEEi\nAAAAAEKQCAAAAIAQJAIAAAAgBIkAAAAACEEiAAAAAEKQCAAAAIAQJAIAAAAgBIkAAAAACEEiAAAA\nAEKQCAAAAIAQJAIAAAAgBIkAAAAACEEiAAAAAEKQCAAAAIAQJAIAoBDn5xHHx81XACA9QSIAALJ3\nfh7x6U9HvPRS81WgCADSEyQCACB7d+9GvP128++3326+BwDSEiQCACB7h4cRjz3W/Puxx5rvAYC0\nbow9AAAA2GaxiPjRH20yiA4Pm+8BgLQEiQAAGN35+fYA0GIhOAQAfVJuBgDAqDSlBoA8CBIBADAq\nTakBIA+CRAAApHF+HnF83DkVSFNqAMiDnkQAAOxvWTP29tsRZ2dNl+mWDYSWTalffrnnMQIAG8kk\nAgBgfwlqxl5/PeLVV/UlAoCxCBIBALC/NTVjXarP9CUCgPEJEgEAsL9lzdjRUcSP/micx6LTimX6\nEgHA+PQkAgAgjcXi/T5Ed48fzgza1KJoGWO6e7cJELVsZwQAJCRIBABAcoeHTf/qt99unxm0EmMC\nAEYgSAQAQHIygwCgPIJEAAD04rrMoPNzwSMAyJHG1QAAJNFmNbPz8+jU0BoAGI4gEQAAe2sb/Gmz\n1H2bYBMAkJ4gEQAAe7su+HM14LNtqXuZRgAwHkEiAAD2ti74sy7gs2xofXTUfL3ak6hNphEA0A9B\nIgAA9rYu+HNdwGexiHjxxfVNq7dlGgEA/bG6GQAASVxdzezwMOLsrAkQtQ34LINNVj8DgOEJEgEA\nkMzV5e13CfhcDTYBAMMQJAIAIIllD6K3324yiJZlZwI+AFAGPYkAAEhC02kAKJsgEQAASWg6DQBl\nU24GAEASmk4DQNkEiQAASEYPIgAol3IzAAAAAASJAAAAABAkAgAAACAEiQAAuHR+HnF83HxN+VgA\noAwaVwMAzNj5ebMa2a1bES+8EPH22xFnZ80qZdc1oD4/j/j0p9s9FgAohyARAMBMrQZ7Hnkk4r33\nmp+//XYTOLou8HP3bvOYNo8FAMqh3AwAYKZWgz3vvRdx4/Ljw8ceizg8vP73Dg+bx7R5LABQDplE\nAABdLOuzDg+LT585PGzKxd5+uwn23LkTce/e9pe2WDQlZhPZDADApaqu67HH8L6Dg4P64uJi7GEA\nAKy3Wp/12GOTaMYzoZgXAHCNqqreqOv6YNvjZBIBALQ1wWY8i0XxLwEASERPIgCAtjTjAQAmTCYR\nAEBbE2nGo8QMAFhHTyIAgBmZYFslAGCLtj2J9i43q6rqn6uq6u9UVfV/VlX1c1VV/ReXP3+iqqqf\nrqrqF6qq+stVVd3c97kAANjPurZKAAARaXoS3Y+IT9R1/a9FxO+NiG+tquoPRMQPRMSfr+v6oxHx\n6xHxPQmeCwCYovPziOPj5iu90lYJALjO3kGiuvH/Xn77Gy7/qyPiExHx45c//5GI+PZ9nwsAmKBl\n/dNLLzVfBYp6tWyrdHQ0bqmZuCAA5CfJ6mZVVT1SVdXPRMSvRMRPRsQ/jIiv1nX97uVDvhIR33DN\n7362qqqLqqoufvVXfzXFcACAkqh/GtxiEfHii+MGiMQFASA/SYJEdV2/V9f1742Ij0TExyLiX1n3\nsGt+9wfruj6o6/rgQx/6UIrhAAAlUf80O+KCAJCnJEGipbquvxoRr0fEH4iIr62q6sbl//pIRPzj\nlM8FAEzEEPVPPdc2KZ3qRlwQAPJU1fXaBJ/2f6CqPhQR/19d11+tquqfj4i70TSt/kxE/JW6rn+s\nqqq/GBE/W9f1X9j0tw4ODuqLi4u9xgMA8AE9r/luSfndnJ83GUSHh7YXAPStqqo36ro+2Pa4FJlE\nH46Iv1lV1c9GxN+NiJ+s6/qvR8T3RsQfr6rqSxHx9RHxwwmeCwCgm55rm5RO7WbsvkgAwMNSrG72\ns3Vd/766rv/Vuq5v13X9py5//o/quv5YXdf/cl3Xz9R1fX//4QIAdNRzbdMuf75reZpyNgBgCHuX\nm6Wk3AwA6EXPtU1d/nzX8jTlbADAvtqWm93Y9gAAgNI8FLRZLHqNrHT58+vK0zb9btfHAwDsKunq\nZgAAY1tm3rz0UvM1txKtruVpVgIDAIYikwgAmJTcM28Wi6ZkrG15WtfHAwDsSk8iAGBS9PABAPgg\nPYkAgFnalnnTcw9rAIBiySQCAAYzdoBGlhEAMEdtM4k0rgaATJ2fRxwf59d4OWK3sbVpKN33a17X\nrwgAgIYgEQBkKOcVunYd27YAzRCvOeeVwnIOCgIA8yBIBAAZyjnjZdexbQvQDPGal/2Kjo7yKTU7\nP4/41KcivvM78wwKAgDzIUgEABnKOePl1q2IRx5p/t1lbNsCNEO95sUi4sUX8wkQffrTEa++GnH/\nfvOz3IKCAMB8WN0MADK0bYWusZyfR7zwQsR770XcuBFx5063sS0W1z8+19fcp9XsqaW2AbKxm4AD\nANMjSAQAmdoUUBnLalDj3Xcj7t1r/7ttghqrr3kOQZDDw4izs2ab3rwZ8U3fFPHss9tf7+oqbWdn\n+ZTOpTKHfQ8AOVJuBgC0tloSdvNmxJe/3K5/Ttem1Dk37k5ptQTvlVciPv/5dkGRnHtW7Wsu+x4A\nciRIBAC0tgxqPP10RFU1vXTaTOS7BjVyDIL0tfrYLj2Sdu3fVMIKajnuewCYC0EiAKCTxSLiiSe6\nNVruGtTIrXF3btktu6zSlttruE5u+x4A5kSQCADorOtEvmtQI7el6nPMbumagdTmNeSQaZTbvgeA\nOanquh57DO87ODioLy4uxh4GANBCm+bCU2lAvNoo+rHHygxebHsNU3iNAMB6VVW9Udf1wdbHCRIB\nwLwMFbiZWtBhCgGvTa/h+LgpRVs6OmoylVI7OWnGsVhEnJ6m//sAwMPaBoluDDEYAJi7XAIMQy6d\nvq68qdTgSkQz9pLHH7H5NRweNsfEMqjXRy+gk5OI559v/v3WW81XgSIAyIeeRADQs5waBg/ZW2ff\nBsQ59MfJTZ/bZIheQFfHbd8CQF4EiQCgZ/sGZlIGBoZcOWqfoENOgbVcDLFNujbD3uXvb/oeABiX\ncjMASOxqadk+ZTypy8OWgZuhSt92LdGaWqlaClPYJsvSMj2JACBPMokAIKF12R77ZNT0UR7Wd7ZI\nCkNmPJVi122SW9ne6WnEm28KEAFAjmQSAUBC12V77JpRM0Qz4V312Yx76IynEuyyTYZsVA4AlE+Q\nCAASSh3UyTVYMkTwYQqriaXWdZtMoUQNABiOcjMASKiPFaJyLA8bcpU0djeXsr3cSuoAoFRVXddj\nj+F9BwcH9cXFxdjDAAC2WM0keuyx4cqY+ixxm6qpb7PVY/HRRyM++cmIZ5+d5msFgF1VVfVGXdcH\nWx8nSAQA85IqaDB08GGswBR5Oz5uGsWvcnwAwAe1DRIpNwOAGVm3+tquhi6Du67ETanRvK2W1C0p\ngQSA3QgSAcCMlNxLaF1/nZRBLx5WQgBu2Qfs6acjbt5sfjbl/ksA0CdBIoAZKmHiN7apbqPVQMvN\nmxFf/nI5r3FdU/CSg165KykAt1hEfP7zEa+8krZpPADMjZ5EADOjr8t2U99G5+cRL78c8dprEffv\nl/0ap76vxnS118/RUVNeCACUR08iANYaNfOikPScqWenLBYRTzzRBIgiyn6N67KLSOPWrYgbN5p/\nK98CgHkQJAKYmXV9XQZRUO3KaNtoQFN6jUM30J6D8/OIF16IePfdiEceibhzx/YFgDkQJAKYmdEy\nLwpKz5lDdsocXiO7Wz1d33sv4t69cccDU9cl0bbtYwtJ3gUyoycRwMScnzcTvMPDzCb+msdAMZyu\nMJwu51vbxzqHgav0JAKYoawruqSuQDGcrjCcLom2bR9bUPIukBlBIoAJyf6mUPMYKIbTFYbRpUdc\n28dOqe8cMCxBIoACXddnoMtN4VA9DfREACC1lO8tY79Pdcnca/tY2YDArvQkAijMtj4DbXoSDdXT\nQE8EgLwt3zNu3WoalGfXz26NlO8t3qeAudCTCGCitpWUtSkRGaqnQfblb5DI2JkIsIvVPnbPP59p\nP7s1Ur63eJ+aFtdi2J8gEUALOd10pOgzMFRPAz0RmIOsG8bDBqsBkqUSAiUp31u8T01HimtxTvd7\nMJYbYw8AIHerqehnZ+Onoi/7DLRd5n5d+Vnbv9H1uVL/PpRgXSaCY50SHB4272urgaISAiUp31u8\nT03HdVlhXe6Xcrrfg7HoSQSwxfFx86nU0tFRU85VAr0WoH9jn2dt+pDBdUrsSQTrXL0W37kT8cIL\n7a/NJd/vQRt6EgEkkksq+i4p0HPotSA1nLGNuYqQUjf2texjd3q6vZ8d+zs5iXjyyeYraf30T0d8\n6EMRv//3N9fie/e63QPlcr8HY1NuBrBFDqnou6ZAr5YSTPGGR2o4uVgsxjn2lLpBOU5OmgbhERFv\nvdV8PT0dbzxTsrptf/EXm4BR13ugHO73SvDMMxFf/GLEJz4R8corY4+GPsgkAmihzYphfdo1I6iv\nDIddsnf6yPiZQ6YUbOKTb0hjiKzUq397jMy/8/OIT32q+W9KmYfrtu0u90Bj3+/l7plnIn78xyN+\n7dear888M/aI6INMIoAC7JMRlDrDYZfsnb4yfqaeKQXb+OR73jb1o9Krqr2hslIXiwcZRMvvh3R+\n3kzq33mn+f611yI+97lpHB/Xbduxsjyn6otf3Pw90yCTCKAAqTOC9vnEdJfsnb4yfsbsBQO58Ml3\nWVJlrGzqR6VXVTdDZaWenkY891zE7dvN16FLze7efRAgioi4f386Gbh9bVt9Dz/oE5/Y/D3TIEgE\nUIhUE8F9Jw+7lLf0WRJjgswkmInMQsrgzabAhlLcboYs2zw9jXjzzXF6ER0eRty8+eD7Rx+dVgZu\nqm27vByfnAi2XvXKKxHf8R0RX/d1zVc9iaZJuRnAzKRodPvUU83XZ59tX+OvJAauoQP7bKRsNL6p\n3FYpbjdzeY9aLJpJ/csvN9+3fQ+futXSzIgHl+MbNyLefbf5mYUBHhAYmj5BIoCZ6TJ5uNrTYnUu\n+9hjzQ1mW/oCwDUsUTYbKYM3mwIbcwl6pFTye1SX/lMlv84+XI3RP/XUg8vxu+9GPPJIxHvvCbYy\nL4JEADPTdvKwLrnBXBZ6IO1jNlIHbzZN+AUD5kEi4n6u3tdENJfh5eX4zp2Ie/cEW5kXQSKAGWoz\neVgXEDKXJSWrL12S9pG11Mep4M08nJ8PU9blw5v9XL2vefbZ5j+XY+asqut67DG87+DgoL64uBh7\nGACdjTnZ7eu5r5aWLT+dNLEvWy7777rjizLlclyl5jhlF1eXmn/00f6WmneM7m+q1y+4qqqqN+q6\nPtj6OEEigP2MeYPW93PnfOOU89hyldNk4vi4WTFm6eioWaWO8uR0XKXmOH3ANbe9q8dNRL/Hjn0D\ntNE2SPQ1QwwGYMpSLjXcdRXsvpc5TrG8fB8re6dcRnpOcloWe8glp+lXTsdVao7ThmtuN0MvNZ/i\nvRpgae8gUVVVv6Oqqr9ZVdXPV1X1c1VV/bHLn39dVVU/WVXVL1x+/U37DxcgP6kmEbvchOc+gelr\nYjHlSWmfcjpelm14nn66WU2G6CeiOoDVCfHNm/ldh/axPE6PjqaVIdWVa243y6Xmn366+a+vUjOA\nPqRoXP1uRPwndV3/H1VV/caIeKOqqp+MiP8gIl6r6/rPVFX1fRHxfRHxvQmeD2AU16Vzp+o5u0vz\nydz73fbVULP0BtpjlQbkeLy8/nqzH19/fd6T8IeWKCpsSZ2q+uDXKdFouvxr7hgcN0Cpkvckqqrq\nJyLixcv/nqrr+perqvpwRLxe1/Xv2vS7ehIBuRqi58YU+3r0+ZpK7cEwxf28K/1eVlzdGI88EvHe\ne0UcJPbjPJR6zQWgMUpPoqqqvjEifl9E/HRE/Na6rn85IuLy629J+VwAQ7ou1T5ldUgOZQ2pq136\nfE2l9mBQtvFATuVvo1vdGDduNAGiiGwOkk3XBvtxHkq95gLQTbJMoqqq/sWI+N8i4rSu679aVdVX\n67r+2pX//+t1XT/Ul6iqqs9GxGcjIh5//PHf/4u/+ItJxgOQ0rrsj4hxVzVL/YmuDJdh9LGdS/6E\nv+SxJ7fcGLduRbzwQjYnY5tjts1+tK8BYDxtM4lS9CSKqqp+Q0T8lYj4S3Vd/9XLH/+Tqqo+vFJu\n9ivrfreu6x+MiB+MaMrNUowHILV1vVyOj/vpt7PN1dYlqeaPffUP4oNS9wXq63hgBKtNTD7+8Wwi\nKm2uDdv6rzhOAaAMKVY3qyLihyPi5+u6/q9W/td5RHzm8t+fiYif2Pe5AMZ0NdV+rBKLvsqVUrye\nk5OIJ59svs7BruV5Kcs2dj0eclhIy7LaG2RU25Pi2qDMEgDKkKIn0R+MiO+OiE9UVfUzl/89HRF/\nJiK+uaqqX4iIb778HmAyxuoh1Fdwat/Xc3IS8fzzEW+91XydeqAolwDHLsdDLmMXOChDimudvkUA\nUIa9y83quv7fI+K6BU8/ue/fB8jZcrK0nNwOESjqcxnzfZbsvRpoOD+POD3df0y5yqU8b5fjIZex\nW1a7HPteG+7ejbhzJ+LevSwq6Hamr1I6tiVAnpL0JAKYq137bOx7c7zPhK0vi0WTRbT6/ZTlFODo\nejzkMvY+A57kYUoN8Tdd71MFPOYSONGjCiBfKcrNAGZrl3KZMUt9+uxDc3oa8dxzEbdvN1+nnEUU\nMV654Tpd92tOY8+o9Q49mFJJ4XWvJdU1PZcy0CFM6bgAmBpBIoA97NJnY6yb4yEmIKenEW++Of0A\n0VIOAY5N+3VT8CiHsTN9U+pFdN1rSXVNn1PgZErHBcDUCBIBtLVmxr1LRkYpq6LlsPoV2/Wd3QD7\nyClrbV/XvZZU1/SpB05W31OmdFwATE1V1/XYY3jfwcFBfXFxMfYwAB6WuLHGGH0nuryEKfUR6dvY\nPUSu21fHx02AaOnoqMkcAtLTk2gz7ykA46uq6o26rg+2PU7jaoA2Ei8HNUbj6S5NgnNZ/Sp3OTRf\nvW6/5tKcGuYg1TU9x0UJUvCeAlAOQSKANiYy4247AZnIy+1dLhOfq/t1SkuOA+W7dWvz9wDkQ5AI\noI2R1uoeq/TA0uTt5BhMU9YB5Obevc3fk5eplj0C7ehJBJApk/0y5HYzrRcRkBvvZ+Wwr2C62vYk\nsroZQKbmtBxyyXJbSn7qKyQB5bGaWTncewCCRACZMtlnFyZj87a6zDjkJLeAOuu59wCUmwG0NNay\n9TmVMgH5UiYCpODeA6apbbmZxtUALYy11Pl1q1a5cQOuymW1PaBsbVdCBaZJuRlACznU6C8DVS+9\n1HxVTjJvyoq4SpnIcJx/AEyVIBFACzlMvnIIVE1JyZM8AUPW0Y9qP22vCc4/AKZMkAighRwmXzkE\nqqai9Ele6oBhyQEzPkhz4N10uSbkHrB3PtO3k5OIJ59svgLTI0gE0FJfk6+2N/Q5BKqm4uok7+WX\ny5pUpQwYlh4wgxS6BH5yDtg7n+nbyUnE889HvPVW81WgCKZHkAhgRF1v6GUJpLE6ybt5M+K118qa\nVKUMGOaeFQFD6BL4yTlg73ymb1ffI0t4zwS6ESQCGJEb+nGsTvK+6Zsi7t9vfl7SPkgVMMw5KwKG\n0jXwk2vA3vlM364e87mdA8D+qrquxx7D+w4ODuqLi4uxhwEwmGUm0dtvNzf0uX0qPQf2QbMN7t5t\nJpRze+1dLbfVrVsR9+7ZZoxr3bnrfKZvJyfNcbZYRJyejj0aoK2qqt6o6/pg6+MEiQDGNcQNvUnD\nZrYPbawGFJfmGlhkfALcAHTRNkh0Y4jBAHC9xaLfG/vVicTZmYnEOn3vA6ZhtTx0aVmi6PhhaOvK\nlR2HAOxLTyKAidP3qAyWrc7far+XJX1f8jfVc0v/IQD6oNwMYOKUJOTPPiqHnkRlmfq5pVQWgLaU\nmwEQEQ9W7TGRyFefZSM7TSLNPK+lNLFfqQ+9qZdkOR4BSE25GcAM5LpcM41dykbalNAssyheeqn5\n2qrcZqdfYi76LN3q49BTksVVUy0/BEhFkAgAEtplArLM9jo6alcO03YyvVM/Kk2suEbf8cM+Dr2u\n5xbTJgbOJgKI0BAkAoBE9pmAdMn2ajuZ3imLQuoF1+g7ftjXoTeHTEqT23bEwKehj+N9l/dv5x1T\nJUgEAIkMNQFpO5neKYtC6gXX6Dt+6NDbjeyY9sTAy9f1eG8byOn6/u28Y8oEiQAgkaEmIF0m0ztl\nUcwh9YLOhgjiOPS6kx3TXptjWHZIeim3aZfjvUsgp+v79y5BJccVpajquh57DO87ODioLy4uxh4G\nAOzMwmDAkJYT4bffbia3srB2Z1umdX4e8fLLEV/4QsQ776TZpl320fFxEyBaOjpqgtCb/nbb9+8u\n43BckYuqqt6o6/pg2+NuDDEYAJiyqzeWQ9/8CUzBfC2zY1wD9rcuO8T23M1qYGQpxTbtcrwfHkac\nnT0IzmzLDury/t1lHI4rSiOTCAD2MPYnhGM/P8BUuJ6mczWLJ2K898ixA6iOK3LRNpNITyIA2MPY\n/UDGfn7Yl14d5GKOzdNPTiKefLL5mtJqj59HH414+ulxtmkOfc7meFxRNplEALCHsT8hHPv5YR+O\n3/nIIaODDzo5iXj++QffP/dcxOlpur9vn0Ne2mYSCRIBwJ7GvhHu+/nHfn3koY/joGtj2bE4B/Yj\nGJinJ5+MeOutB9/fvh3x5pvjjQfol3IzABjI2OnsfT5/lyWEma59joNN5WRdl50ew9DnwBTL75TF\n5unqe4bAHRAhSAQAbGByR8Tux8G2AEvqXh19BFiGPAemGpQdIxg4xWBbaqenTYnZ7dvpS82AcgkS\nAUCP+moKOpQSMj3o367HQZsAS6pMuL4CLEOeA1MNynYNBu4b4JlqsK0Pp6dNiZkAEbB0Y+wBAMBU\nrTYFXfZ9KO1GfDm5049l3nY9Dg4PI87OHvSiGTrAkuJ4HfIc2Gd75d43abFoN67V/kVnZ7tlmPV1\nLADMgcbVANATTUFhuOBF6ubIYwVddnneKTWGTtHMvK/tkXsgDmCTto2rZRIBQE8Wiw8GiUwqmKO2\nGSQpnidVxk+KbJZd7bK9ppQ5kyL7rI/srzGPCYAhCRIBkK0cP7XtMqZladn5efPY0krNoDSpAlKl\nBV2GLOvrW6oAT+rgZGnHxBzleM8AJVJuBkCWciyfyHFMwPV2nTSWeK6bIPerxGNiTuwf2K5tuZnV\nzQDIUo6r/OQ4pghLPVOmvlf+22eFq66rcQ3puvM91SpxuRvrepfzMUG+789QIkEiALKU49LrOY7J\nUs+UaLny31tvNV/7CBTtO2lMHXRJEdwo9XxPFdgZ+/XPJRBXohzfn6FUgkQAZCnHT22HGFPX7Aqf\nnlKiq5P7fQMn6wIQOU0aUwU3SjzfUwZ2Snz9tLNvIDHHewYolSARANnK8VPbPse0S3ZFThNhaOvq\n+bPr+bQpAJHTpDFVcKPE8z1lYKfr61eK+8Azz0R8/dc3X8e0bp+kCiTmeM8AJRIkAoBM7JJdkdNE\nGNo6PY147rmI27ebr7uu/LctAJHLpDFVcKfE8z1lYKvL6x+7NC0nzzwT8eM/HvFrv9Z8HStQdN0+\nkSEGeREkAoBM7JpdkctEmPzknElxehrx5pu7B4gihsusyakUprTzvc1r77J9275+gYcHvvjFzd8P\n5bp9UmKGHExZVdf12GN438HBQX1xcTH2MABgNCcnzURpsdhv8gxzWRK676Xf57Idx9LX9rXfHlhm\nEi19x3dEvPLK8OPYtE/6Po+BiKqq3qjr+mDr4wSJAADKdd3k6vi4KetYOjpqMjDoxnbsV5/bV+Dh\ngWeeaTKIPvGJcQJES/YJjKdtkEi5GQBAoTb1XemzhGOX8qucS982UQrTrz63b2mleX165ZWIf/pP\nxw0QRdgnUAKZRAAAhdqWhdHHp/a7lPGUXvoj+6Ffti9A/9pmEt0YYjAA0DeTDObo8DDi7OxB8OVq\nFsZikf58WNd8dpcmwiWdp31sx7m7es22fQHyoNwMgOJZ6pi5GmNJ9F3Kg5Rssco1GyBfgkQACZXa\nc6N0Yyx1bF/v7+Qk4sknm6/sbugeH7sEpsYIZpEvy9MD5EtPIoBESu+5UbKht719vb+Tk4jnn3/w\n/XPPRZyejjceoL19y3tdQwGzZ019AAAgAElEQVSGZ3UzgIH5ZHQ8q1kKd+40277PDB/7en9X94+M\nLMYkM7C9FKViy2v2009HPPVU8iECsIckQaKqqv67qqp+paqqt1Z+9nVVVf1kVVW/cPn1N6V4LoAx\nbZpI6LkxrsWi2eYvvNB/n4uS9nWuk9+rWQOyCBiL/jjdpAySv/56xKuv2u4AOUmVSfTfR8S3XvnZ\n90XEa3VdfzQiXrv8HqBY2yYSem6Mb6gMn1L2dc6T39PTpsTs9m2lZowr5XUj16BsSqmC5DIyAfKU\nJEhU1/Xfiohfu/Ljb4uIH7n8949ExLeneC6AsbS5oR26gSwfNGSGTwn7OvdJ2OlpxJtvChAxrlTX\njZyDsimlCpKXlJEJMCc3evzbv7Wu61+OiKjr+perqvot6x5UVdVnI+KzERGPP/54j8MBJm/fTppb\nHB5GnJ09aLTphjY/y8lLj4dBURyzsF2q68a6oGzu16Bd3zYXi/1f2+p2v3XrQRA7920GMHXJVjer\nquobI+Kv13V9+/L7r9Z1/bUr///X67re2JfI6mbAzgZaKqXnOBQjmfJ+nfJrg5zksGJXl/M9h/Hm\nNA6AqcthdbN/UlXVhy8H8+GI+JUenwuYu4HqakooMZqrXXuBTL1EZCrH7Bx6vVC2sXuVdb2W5VKO\nmss4AGj0GSQ6j4jPXP77MxHxEz0+FzB3qZsbmJEWZZ9AT04TFIfdeskDeYVv6JOTiCefbL6SlzGD\nsl2vZbn0BMplHAA0kgSJqqr60Yj4qYj4XVVVfaWqqu+JiD8TEd9cVdUvRMQ3X34P0I+UH+GOmFpS\n+Nx1NPsEepJNUPbceVPPaNpH0kBe4Rv65CTi+ecj3nqr+SpQxFLXa9nYmU+5jQOARqrVzT5d1/WH\n67r+DXVdf6Su6x+u6/qf1nX9ybquP3r59erqZwBpdfkId9OEfqTUkmRz1xlGmvYJ9CSZoCTYeTll\nNOUmaaZB4Rv66qE1o9M8qU2XyVIvobtcy3IpR81lHAzr6rlW6rkHU9NnuRlAnrZN6EfKfU8ydy08\nS2JX+wZ69p6gJNh5Si6ulzTToPANffW1m1R3t+kyWfolVLCFUlw9105Oyj73YEoEiYD52TahHyn3\nPcnctfAsiX2MOjlKsPOUXGyWbP8WvqFPTyOeey7i9u3m6+np2CMqz6bL5IwvoZDUtqygq+fa+blz\nD3JR1XU99hjed3BwUF9cXIw9DGDqMl5vd+/lyjN+bZNnrXkowqbLpEvoOFw+p6XNeXT1MXfuRLzw\ngnMP+lRV1Rt1XR9sfZwgETBLU74jnfJrA0hg02XSJXRYAnPTc3zclI0tHR01maCrzs8jXn65+fez\nzzb7/OSk+fliIUsS+iBIBLALswMAGEybgAJl2Rb4W/f/IwQLoW9tg0R6EgEsldix1FIgAMVwyX5Y\n4b3kWWNb67d1vb/0A4N8CBIBLJV2h1JiUAsGMMeJ+Fxe8xivc91z7jIOl+z1Cu8lzzU2LTawLjAo\nWAj5uDH2AACycXgYcXb2INc59zuUdUEtd9eDUp2Yj+W+uHXrQfPTs7N5TDpXSzem/JrHeJ3rnjNi\nt3G4ZF9vsbAt5mQZGLz6/rnuZ8DwZBIBLJX2caaP3XaSKhNBVkA+VvfFn/2zZSUEplBaEuSuxnid\nKctiXLIbc8l6Y7N1mUabso+A4QgSAbPR6sa0pDuU0oJaGUgZ2JnLxDylviaHq/vi3XcjHnmk+fdc\nJuJzCT6M8TpTlsW4ZAuuA5TA6mbALFhil4i0q+g4prrpc3td/dt37kTcuzevkoW5lD6O8TrXPedc\ntndqVjIDGE/b1c0EiYBZcGNKRPpAxdQniilfX9/n4NT3xVzYj9MmuA4wHkEigBVuTFkyCW2nj4Ca\nc5BNHCPz4BoMMI62QSKrmwHTd34ei7t344t3DuN/vLdwYzpzVtFpZ5+VmNZNAq9bzQaWrP41D67B\nAHkTJAImaTlJ/e5b5/HxF5qPpj/+2Fl83EfT0MrhYbO89zKro21z3k3LlJscssmuxxxlkEEEUAar\nm9GaJUspxerqKX/vB0ZegiqXEyeXcVCMXVdisuobuypl9S+X0+6sara/k5OIJ59svgL0SSYRrWz6\nZBhyszpJffW9w/jDN87i0XdH+Gg6lxOnzTh8xMsau2T+yAZhH7lnm+VyWS+NUsL9nJxEPP988++3\n3mq+np6ONx5g2mQS0YpPhinJ4WEzOY2IeO2xRfzMnxjpo+lcTpxt4/ARLwltywaRhUHJcrmsl2b1\nfVnwuLur10vXT6BPgkS04s2dklydpH78dNGstT30x5a5nDjbxmHWMw0ZRV8W15xy4pGULpfLemlK\nKSXM1dXtZftlJKP3Xkilqut67DG87+DgoL64uBh7GFxDNQrs4PLE+elbI6+stukEtu50+Xrchymv\n/cfHTYBo6eioCSb1/byQkmOTMZycNMfeYqHULBvunyhMVVVv1HV9sPVxgkTswg0StFfEPYSTumxd\noi8dpD522/69Is4ZAOatp/de6EvbIJFyMzpTLgDdFFHNdV19EGXoqQYm9bHbtuSkiHMGgMnbWE2m\n/pSJEiSiMzfv0I17CHrXU8OPPo7dNvHIW7ciHnkk7fMCQBdbPxjXbIuJujH2ACiP5Y2hm+U9hGou\netXD2uFjHLvn5xEvvBDx3nsRN25E3LnjnAFgeOs+GH/o/aiH914YmyARnZnwQnfuISjV0Mfu6k35\nu+9G3Ls33HMDwJIPxpkrQSJ2YsILQB/clAPMQ+5rZvhgnLmyuhmMLfd3SNaz35ixvg9/pxfAtE1x\nFUvvXeSu7epmgkQwphTvkN6RhjfFOxtoyeEPwL6mtnq890ZK0DZIZHUzhrNxDcmZ2nepuK3LLtAL\nS/wxY10P/5OTiCefbL4CQMT0Vn51a8iUCBIxDMGM9fZ9h/SONI6p3dlAB10O/5OTiOefj3jrrear\nQBHA+HL43HZqq8e7NWRKlJsxjKnllG7Qufprn3Ixua3jUebHjLU9/J98sgkQLd2+HfHmm/2PD4D1\n3Dr2x60hudOTiFYGu5jN5B1plJfpHQnI1DKTaOm55yJOT8cbD8Dczehz22RS3Gq7XScHehKx1aAV\nYFPLKb3GKNVfi0Xz7r7HNs0h7Zgtdt1Jdi4jOj1tAkO3bwsQAeRAWVQ3beZL2261dN2gNIJEMzZ4\nQCNBMCO11PPnEt94vXEVYNedZOeSgdPTpsRMgAhgfGN8blvy51Xb5kttbrW0EKU0gkQzVmJAI6U+\n5s8lJkx54xrYLndKu+4kOxcAuGLIz21L/7xq23ypza3W4WHEzZvNv2/enN+ci/IIEs1YiQGNlPqa\nPw/2xpvgY5nz84gvf/nBG9ccg4WD2vVOadeI7twjwZBIyZ+CA4yp9M+rlvOlp5+OeOqph/9/21ut\nqvrgV8iZINHMZVgBNpgh58/JJxgJPpZZ/olXX23esJ5+ep7BwkHteqe0a0R37pFgSGCIT8EFoYCp\nmsrnVa+/3twzr74PLJtR37mz+Vbr7t2I+/ebf9+/X16gjPm5MfYAYCzL+fPduxHffes8Pn73bkSk\nX3JgdcWzs7NEc/V1wYaOf3T1T9y/H/HEE2IIvTs8bA6C5fJ3Xe6UFovddtCuvwdERJLL7UZd3iOs\njgOUZvV+u9Rr13Wf8W1a0Xj1er3P7d9DvBEwAJlEzNpiEfHi4Xl8/IX+PibuJc02wccyU/lkpyjr\nMnukEEDW+r5Wtn2PKL2vBzBfpVcurHsf2HTtvnq9jkiU2O2NgIEIEsG2q/y2CfyWx/QywUhQRqQS\naSSrd0re7CF7fV8r275HlN7XA6BU694HNl27r8tA3TtQ5o2AgVR1XY89hvcdHBzUFxcXYw+DuVnN\n9V/NF73u521+d81TyAzlIcfHTYBo6eiouYOYGycIM9fmFGj5dgPwEG+z/bhuu/Z2vfZGwJ6qqnqj\nruuDrY8TJIJYf5VvM4HfdZK/z7u1d/rp8GZvG0AHLv9AV20/83RtSau3bWpnsQdBItqbw8Vml9eY\nMJNo799J8bvkaQ7n3yayqQCgN9veZt1awny0DRLpSTR3qz1RvvM7Iz71qXH7ovTRxHfXvi9tGlHs\n0qxin3pitcjTU3o3x33poA4Avdn2Ntvm1tIaGzAvgkRzd3Ud9FdfHa+Bbl9NfPcJrLSZwHed5O8z\nKe5jQl3YO39hw2UbHdQBoDfb3ma33VpaYwPmR7nZ3K3mmK4ao+Sjr7KTXUvC+uwZlEtPosJyjAsb\nLgBA9jbdWl69PX/66YgnnphvlTyUTE8i2js/j3j55YgvfCHinXfGm333GQHoEliZU8+gwvrBFDZc\nAICird7a3rwZUVVN8UEJt7nAB+lJNKLiymEWi4jPfz7ilVfGLfnos+ykS0nYnHoGFdYPprDhMpbi\nLsIAkKfV2/Nv+qYmQBSx/TZ3Km/FU3kd0IVMosRKSyQpWl+rQs0pkyiiuNW1ChsuQ7vuHFw9cCIc\nRADQUdvb3BJvh9eZyuuApbaZRDeGGMycrEskyfpiUuqMe/WqfXaW9qq9/Mhk3XbZtr02/e4OBtk9\ni0VR+76w4TK067L5lteLH/qhiLpuSmtTXzsAYMLa3uYWNx+6Rp+vo9QpGPOg3CyxosphSl6uYJ+y\nrjZ5o+vK09pur0RLmpe8e2A06y7CV1dxfOed5t8llIQCQEba3OYWNR/aoK/X4R6f3AkSJbZPW53B\na15L65+zater9j5X5YG3V8m7B0az7iK8er149NGm82ZE2XeuAJCpPtuMDqmv1+Een9zpSZSJUWpe\nSy+03SVPc5/lsQbeXqXvHshKXz2J5IsDAB24x2csbXsSCRJlYrSlvec2wdl2Vd62Pbpurz237767\nZ267FwblLg8AtnI/+jDbhDEIEhVm8LnGnK9M17321Dth5Amk+Sv0bLToPgCUwf0o5KNtkEhPokwM\nWrs7925p13XcS10gvO/f27NJ1SD1zoM30sqPTTBjU+nMCQA90X8HyiNIlJFEi2Jtl8vVOrfZdeoJ\n3z5/L0Egr/f569yDjWETzN5UOnMCQE98ngLlESSaoxyu1mPNrjcFplJP+Pb5ewkCeb3PX0fOlMpB\nLvFWRrQpuj+BYxwA9uHzFCiPnkRzcbUPz9g9icbo5bFPUfTQ26uEAu59t2fur6+FibwM+tD24Bj7\nWgxAb1zigZxk05OoqqpvrarqH1RV9aWqqr6v7+djjXVZO4PVtl1jjGymXdM+xsh6KuFjl5EzpXJQ\nwm5iJG2OcfWKAJN0fh7xqU9FfOd3PnyJLynJdPk6PvWpMsYLpHGjzz9eVdUjEfFSRHxzRHwlIv5u\nVVXndV3//T6flyvWTVbGns0uZ9dDfrxyeBhxdvbgk/22gamxtt9iMf5+2mbXMe66L/qw58d8Jewm\nRtDmGM/x2gzAXlYTSZfefjvi5Zebfy//39lZ3h8wnZ83Qa7795vvv/CFiFdeyXe8QDp9ZxJ9LCK+\nVNf1P6rr+p2I+LGI+Laen5OrcuhBtM7Q2Uy7pn3kuv1KlksKjkwO+tLmGHdtAZic1fj/qtdeawJF\npSRS3737IEAUEfHOO3mPF0in10yiiPiGiPille+/EhEfX31AVVWfjYjPRkQ8/vjjPQ9npsbI2snV\nLmkftl8/ckjBkclBn7Yd464tAJOzmkj6NV8T8c/+WfPzZcDlscfySKTe5vAw4od+6MG4b97Me7xA\nOr02rq6q6pmI+Ja6rv+jy++/OyI+Vtf1f7zu8RpXMzk6FuZndZ9E6DwNACS1vNW4dSvihRc+eJsR\nUc6t4fn5gzK5Z5/Nf7zAZm0bV/cdJPo3IuJP1nX9LZfff39ERF3Xf3rd4wWJmJShlr7KLRA1xkpw\nbZ9v3T6J+GDQKKdtCQAULbfbNGC+clnd7O9GxEerqnqiqqqbEfFdEaHpB/MwxApeufXUGXo8XZ/v\nuvKyF19sfpZ67CUtYQIAJDf2gsIAXfUaJKrr+t2IOI6IvxERPx8Rn6vr+uf6fE62MGkdzhBNaXNb\nSn7o8XR9vk37JPXYcwvgAQAAbNF3JlHUdf1qXde/s67rf6mu69O+n48NTFqHNcQKXrmtjjT0eLo+\n36Z9knrsuQXwAAAAtui1J1FXehL17Pi4CRAtHR09KLPJmFruLXLbQDn3JBr6b2mKDQAAZCCLxtVd\nCRL17Oqk9c6diHv38gkurDHIPDu3IAvt5b7vch8fAAAwC4JErLdpTc4MJ7G9Jz/J9ihXqn0nkAMA\nAExcLqubkZvlEgv37hXRL6X3Fjf6xpQrxb7bpU+X5u8AAMBECRLNVW4Nj6/Re+/nQrYDa6TYd10D\nTZq/AwAAE3Zj7AEwkmX0pYAym8Wix+EVtB2yM3aZVop9d3gYcXb2oGRtW6BpXVDJMQMAAEyEnkRA\nd1Pq5dQl2DWl1w0AAMyGnkSQQu79Z8Ya3z79gPYZcx+vd9mnq02wp/f6RwAAgPHIJILr5J41Mub4\ndn3ufcac4vWOXSLXpym/NgAAYC8yiWBf+66e1XeWz5grs+2aUbPPmFPsj6k2nZ7yawMAAAYjSATX\n2Wf1rCEm7WOvzNalTGtpnzHv+3p3Wcks51LDVWMGDAEAgMkQJILr7NN/ZohJe4n9cfYZ876vt0uQ\nKVWQb6hA09gBQwAAYBL0JII+5N7PaGra9uNp+7jj4yZAtHR01GRNdR3TkMeAnkTTc3LS7NfFIuL0\ndOzRAABQsLY9iW4MMRiYnWXWi0l7/1aDMWdnm4Mxi0W7fXF42PytZYBnl8ycddlkfR4HbV8bZTg5\niXj++ebfb73VfBUoAgCgZ8rNyF9JvWFW7dKzh+76KO1LUcqnBIx9XL3elXb9AwCgSIJE5M2qTWzT\nVzBm3yBfiT2jdlBqDDd7V4+XiR4/AADkRbkZeRu6ZKcrfWDGl3NpX24lYImP1y6VfnS0LC3TkwgA\ngAHJJCJvfWSJpEp9kOWUD6V92/VwvA6xiN+snZ5GvPlmxMc/Ll0LAIBBCBKRt9QlOyknymbIlKSH\n41XbpQEIRgMAMCBBIvKXMksk5UT5uhlyykwl2QOk0kNEZyZtl8YlGA0AwID0JGJeUixtvrSuF06q\nJi2avdCHp55qvj77bLLjKbe2S5OT8poFAABbCBIxL6mbHF+dIadqtJ17w+6uNPge12rQ8bHHmiAR\nZci5MTsAAJOj3AxSSlXSU0qzlzYlcXqqjC9lyZIyyOFpzA4AwEAEiZiXvgMWqZq0jNHspevkv+22\n1FNlfKmCjgJ+lEyAEwBgK0Ei5mWIgEWqT/2HzB7YZfLfdluWkhU1ZamCjgJ+lEqAEwCgFUEi5kXA\nYr1dJv9tt+W6AIVP9IeXIujo/KFUApwAAK1UdV2PPYb3HRwc1BcXF2MPg6nTRPlhVxsbt8022WVb\n7vpc5MH5Q4lcdwCAmauq6o26rg+2Pk6QCIiI4Sb/x8dNycfS0VGT4YIADPTJ+QUAzJggEbCfviZU\nPtFfb8ztYvIMAACT1jZIpCcR8LA+m7yOsXJbCcbqmaKhLwAAcEmQCHhY3wGLIVduK8VYTaE19AUA\nAC4JEgEPK3kVq1JXThsrw6rkfQ0AACSlJxGwXol9avQ7aufqvi1xX8M2juvx2QcAkA2Nq4HhjT0h\nsHLadgJpTNG6wKfjfFz2AQBkReNqYFg5NEBWOrWdHkRMzbprj+N8fPYBABRJkIj8DdFjptQ+NjnJ\nYUJg5bTtBNKYmnXXHsf5+Lrsg67vwd6zAaA3ys3I2xDp6lLi07Ady3F+HvHyy82/n33WfqJs1117\nxi5/pd0+6Pre4b0GAHai3IxpGCI7JYcMmCmQxVOW11+PePXV/koDp/BJ/xRewxxcd+1ZLJqeZK5F\n1+v7GG+zD7q+B+/ynu1cBoDWBInI2xAlA3MsS+jrhtmkrAx9B0Zz6E+1rym8hjlx7emur2O86/tL\n1/fgro93LgNAJ4JE5C1Fdsq2G9a5ZcC4YSZVYPS6c6uvINSQ2QAyDMlN6uP/6jF+crL/397l/aXr\ne3DXxzuXAaATPYmYtox7F1zbqqHvPhqWiSdi/+Ns07nVx3k39Lmc8bWDGer7nFra92/n+P7iXAaA\niNCTCBqZfoJ47Yet+2T5tP2UeajyOj0g8rZvec6mc6uP7Lyhz+W5ZRiStz6O/+Uxfvv2g5/t+7dz\nLN92LgNAJ4JETFuON6yx4X5/14lAl+DSEDfMStqmb9u5lbpHzBjnsj435GLT8b9PQH6xiDg9TXdu\n5RqQcS4DQGvKzZi+DJdBvjb7fde0+NxS/IcaT4b7dlaG3v72N3O27vhPVUrl3AKAyWtbbiZIBCNJ\n2pMot54LQ4wnt9dMOUyImYrcPiAAALIlSAQ9OD+PePnl5t/PPpvZ/DK3ia8G3ORIcJEpcTwDAC21\nDRLdGGIwMAXn5xHPPBPxzjvN96+9FvG5z2V0P75YZDSY6H88h4cRZ2cPJkeZ9JsaRQ4BwhzG0Ma6\nvl85jxc2WfYAKuHcAwCKoHE1tHT37oMAUUTE/fvZLJY2T7k2SB1aDk3CcxhDW5k2s4edacoMACQk\nSAQtHR5G3Lz54PtHH+0wv7QcfD9MjoZfGj7XMbQluAgA0Ltdpz+mTePTkwiifaXMTj2J9IygTzkc\nXzmMAQCALOx6a+iWsl9texLJJGL2ulTKLBYRn/9881/rC1ZJWRaUJ4fMmBzGcJWPoQAARrHr9GfT\n77m1G44gEbPXewxHDxT6lkPZXZcx9P0u30ePJHcmAACt7Dr9ue73Smp/OQWCRMxe7zGcHLMs6I9g\nwmZDvMunjvyWemfiWAQAenTdrUab6c+6373u9xRmDEtPIohyVu/+gCEHXeQGGoFC6u2Oj5tgy9LR\nUZOBlFLq/TDEmFNzLAIAPdrnVqPr73Z5vGnL9fQkgg5yqNbpZMjMhlKzKMbgY47thii/TJ29V2LJ\nqGMRAOjRPrcaXX+37a2daUsagkRQoiEngCab2y3zZW/dKi+YMLShyi9TRn5LLBkdObCl0g0Apm2f\nW43V3715M+LLX95+z9Dm1s60JQ3lZlCiIUtJUj/X1HJAr26fO3ci7t2bzuujXCOdayrdAGAett1q\nbPr/5+cRL78c8dprEffvb75naHtL4x5ks7blZoJEUKoSexJN8co9dr+aqQXdKN7YpwQAML42t/1t\n7hl26V/k1ni9QXoSVVX1TFVVP1dV1T+rqurgyv/7/qqqvlRV1T+oqupb9nkeYI0hGymleq4p5oCO\nWdaj8Hq9vmud1FJtVGILJwBgN9fdFrW57W9zz7BL/6Kies1maN+eRG9FxB+KiL+1+sOqqn53RHxX\nRPyeiPjWiPgLVVU9sudzAaWb0uxx+Y4YMV6/mikG3fbVd+BMYG6rEls4AQDdbbotanPb3+aeoe30\nwWd46dzY55fruv75iIiqqq7+r2+LiB+r6/p+RHy5qqovRcTHIuKn9nk+IAP75HAu3wlKzwFdzXs9\nO2te0xj1NIeHzfMv829LDrqlsi5wlvI46/vvT8RisbJZ5H0DwCRtui1qe9v/gXuGa/7/tr+z7tbc\nLcfu+lrd7Bsi4pdWvv/K5c+AsaQIr6fIophCDmguGTxSNh7Wd7balLLhhiDzCgAma9tt0brb/jZT\nkquP2TZ9yOXWfCq2ZhJVVfWFiPhta/7XSV3XP3Hdr6352doO2VVVfTYiPhsR8fjjj28bDrCLVOF1\nWRSNnDJ4tn38Mjd9ZqstM2KsYNeeawYATFbX2642U5Jdpi053ZpPwdYgUV3X37TD3/1KRPyOle8/\nEhH/+Jq//4MR8YMRzepmOzwXcNXV8o5UEzVX4MZUyuamqo/AWeqV+XItwUo9LtcMAJi0LrddbaYk\nu0xb3JqnVdX1/nGZqqpej4j/tK7ri8vvf09E/M/R9CH67RHxWkR8tK7r9zb9nYODg/ri4mLv8cCs\nrZvMRqSb4OY6uYU+pVzXPXXAKZW+xuWaAQBEu1uNXG+TpqCqqjfquj7Y9ri9GldXVfXvRsR/ExEf\niojPV1X1M3Vdf0td1z9XVdXnIuLvR8S7EXG0LUAEJLIu/P7ii+nC68qbmKOUGTG5lmD1NS7XDAAg\n2mX8yAoaX5JMolRkEjE7fXzCLvwO/Uh1vuZ6juY6LgCgKH0mEUtQ3l3bTCJBIhhLnxMyV0/IW67n\naK7jAgCK0PcUx+dZu2sbJPqaIQYDrNHnWo1TWGYepizFOdpmDdkxxpVCH68NAOhdn1OcoZa6n/tt\niCARjOXwsAmBR1j1B+hm+VHaSy81X0u5i2lz11XqawMAep3iDDF9chsiSATjWXZlOzqSKwl0M9RH\naSm1vesq8bUBABHR7xRniOmT2xBBIhiP3h+UYu45tzkqMROx7V1Xia8NAHhfn9XrfVfGr96G3LgR\n8df+WsTJST/PlSuNq2EMuq5RCsdqvkoLNHc5lkp7bQBA9treXpyfR/ypPxXxxhsPfvbccxGnp/2P\nsU8aV0PO5DFSCsfqZmNmWeXSZLqtLjnifb822XEAMCtdeg0tFhH37z/8+3MhSARjUE5BKRyr1+ur\ns2GOAYxUY8ohsKUjJQDMTtfPPa/eqpTymVwKgkQwBk2rKcU+x+rQwY6hn6+PLKscAxg5jmkfsuMA\nYHa6fu55etqUmN2+PY1Ssy4EiWAsOXyiDm3scqwOHVgYI5DRR5ZVjgGMHMfUxnVBQ9lxADA7u3zu\neXoa8eab8woQRQgSAdCHoQMLYwQyUmcEnp9HfPnLEY8+2nyfSwCjxKDKpqChTE4AmCWf0bdzY+wB\nADBBh4cRZ2cPVrLqO7Aw9PMtLRZp7jRWV/66eTPi6acjnn02j7uYZVClpNXG1gUNV8edar8BAEyM\nIBEA3bRZP3TowEKJgYxVq0GNd96JeOKJvF5DaUGVsYKGAACFq+q6HnsM7zs4OKgvLi7GHgYA11nN\neHnsMeU6qVy3XdsE5FjPtgMAeF9VVW/UdX2w7XEyiQBob1sZT65yDxisy4RaDRydnZUTkMtlW5eW\n/QQAkAGNqwFob2pNjCgrmpgAABzvSURBVHNytZtiiauKlbKtAQBYS5AIgPZKXBkqVbDluiXV+1Ji\nQG6XbT30dgUA4FqCRAB0U9r6oSmCLSkyZLoGQ0oMyHXd1jKPAACyoicRAOMZon9NipXP9u3FtGt/\nodL66nTd1qX2uAIAmCiZRACMY8gskn2zn/bNRiqxv9CuumzrnEvqlMEBADMkSATAOEoKnOxb+pVz\nMGRMuZbUKYMDAGZKuRkA4zg8bEqv3n67jMDJPqVfKUrecllaPrUcS+qUwQEAM1XVdT32GN53cHBQ\nX1xcjD0MAIYy1cBHaqs9jR57LK+smymyvQGAiamq6o26rg+2PU4mEQDjyTGLJEdzzWwZI4i4fM47\ndyLu3RPABABmRZAIAHJXWmleCruuCJfqOWUQAQAzpHE1AMMactWoqaxQlWuD5z6N0di8pGbqAAA9\nECQCYDhDrho1tRWqFosmg+ju3fJfSxtjrAhnFbqiTCUGDAA5ESQCYDhDZmq0fa5SZppTC3ptM0b2\n1Bwztgo1t9MBAIYiSATAcIbM1GjzXCXNNOdYCrVYRLz44rDBmjGek862nQ6lxH4BIDeCRAAMZ8hM\njTbPVVLgRSkUvG/T6VBS7BcAcmN1MwCGNeSy99ueq6RVw5ZBr6GXhIcMbTod1sV+nS4A0E5V1/XY\nY3jfwcFBfXFxMfYwAJiT83OBlz7ZvgxsmUm0jP1qLwUAEVVVvVHX9cHWxwkSATB5AhXjMFtnJE55\nAPigtkEiPYkA2F0J3WH3aVBSwuvLWUk9n5gU/ccBYDeCRADsZtfgS4rAS5e/sWugQvfbh3Xdd5pt\nAwAURZAIgN3sEnxJEXjp+jd2DVTIgvmgXfbdkKvZLcco8wsAYGeCRADsZpfgS4rAS9e/sWugQhbM\nB+2674aq+5H5lQeBOgAomiARALu5LviyaZKYIvCyy9/YJVAxdBZM7nIPmsn8Gp9AHQAUz+pmAKTT\nZjWrFMsOWbpoHDlv99xWUst5W/Xl+LgJEC0dHTXBWQBgdG1XNxMkAiAdk8QyTSWgkcvryC1glUKb\nbTvF1w0AE9E2SKTcDIB0ci9J4mFTKhHKZd3zXErfUvUHanuMKNEEgOIJEgGQjklieXIJaExJDsHS\nlMG/LsdILoE6AGAngkQAbNclI8EksSw5BDT6MtZKWzkES1MG/6Z8jAAAH6AnEQCb6TMyfbn08klp\n7sftpte/y/6e4jECADPStifRjSEGA0DB1mUkmCROy2IxvX069+N2mc10NbCzGjw6O2sfPJviMQIA\nPES5GQDrLUt1bt1SakJ5UpVIjVWylsK60s/UPahK3j4AwENkEgHwsKulKnfuRNy7N51SE6Uz5Wq7\n767LpOn6XLtk3eTs8LB5Lctze5+g7xS3DwDMnCARAA+7mm1w716TkTAFJrab5RxA67rv9i2RmmLJ\nWorg2dK+2yfnYw0AZkq5GQAPm3KpjiXfr5dy2fQ+DL3vprqqV6oVCPfZPrkfawAwU4JEADwsxRLe\nuU4CpzrxX7VrcC73ANrQ+y6Hpexzts/2yf1YA4CZquq6HnsM7zs4OKgvLi7GHgYAKRwfNwGipaOj\n4UvWritnmXKZyz5Lv5ewbPyU992clHCsAcCEVFX1Rl3XB1sfJ0gEQC/GngSO/fxj2Tc4ty0II0gz\nLWPuT8cSAAxGkAiA8Y05Ccwhk2kMfQbH5hp4myr7EwBmo22QSE8igLnrs7l0qga5u8ih99AYjbtT\n9ZNaN259ZKbF/gQArpBJBDBnU88kGLuUpsRtu2ncfbwmJUfX63vblHqMAgCdySQCYLupZxKMmclU\n6rbdNO7Uq30NvQLevpldQ2aGDbFt+sw6AwCKJEgEkIOxJlo5lGRNVdttm9ske9u4Uwbehgyk7Rt0\nGTqgNdS22Wd/Dr1NAIDeCRIBjG3MiVbqzBAeaLNtc5xkD3lMDBmk3DfoMnRmWAkB3FKz5QCAawkS\nAYxt7InWmCVZU7dt23bd90NlHQ11TJQUkLru9/vaJyUEcEsIZAEAnezVuLqqqj8XEf9ORLwTEf8w\nIv5wXddfvfx/3x8R3xMR70XEH63r+m9s+3saVwOz1PeS5ZoC56vLvs+pyXCpx9W+4776+zntk12k\n2I+pj4VSjy0AyFzbxtX7BokOI+KLdV2/W1XVD0RE1HX9vVVV/e6I+NGI+FhE/PaI+EJE/M66rt/b\n9PcEiYDZ6mOyVtoEdq6Tw7av+/i4KUtbOjpqsn2GVtpx1adc9skuctyPOY4JACZikNXN6rq+W9f1\nu5ff/u2I+Mjlv78tIn6sruv7dV1/OSK+FE3ACIB19i3vWdfbZuwyti5y7M0zlLb7PpfSnpKOq77l\nsk92keN+zHFMADAzKXsS/YcR8b9c/vsbIuKXVv7fVy5/9pCqqj5bVdVFVVUXv/qrv5pwOAAzsm5y\nVdIE1uRwu1x61JR0XPUtl32yixz3Y45jAoCZ2VpuVlXVFyLit635Xyd1Xf/E5WNOIuIgIv5QXdd1\nVVUvRcRP1XX9P13+/x+OiFfruv4rm55LuRnAjq4r0yilhEuZyW7G2r+lHFcRZY11aKm2TcptbH8B\nQC8G6Ul0+USfiYg/EhGfrOv67cuffX9ERF3Xf/ry+78REX+yruuf2vS3BIkA9lD65Kr08Q9NYG07\n26h/tjEAFGGQnkRVVX1rRHxvRCyWAaJL5xHxXVVVPVpV1RMR8dGI+Dv7PBcAW5S+lH3p4x+aEr3t\nprCNzs+bBtm59umawjYGAN63b0+iFyPiN0bET1ZV9TNVVf3FiIi6rn8uIj4XEX8/Iv7XiDjatrIZ\nAIXJffI6dfq3bFf6NiqhoXvp2xgA+IC9y81SUm4GUAglJv1qW3qnRG+7krfR8XETIFo6Omqy7XJT\n8jYemm0FwEgG60mUkiARQCFKmbzmoOukUACOJcfCtJS+wAAARRukJxEAM6XEpJ1dyoX0eGFpsWgC\nCUdHAkRTsO7cLqGkEIBZESQCoDuT13Z2CfgIwLFKQ/fpWHdu7xsU1hsOgMQEiQB4oMuEw+R1u10C\nPgJwsJ8cAifrxrDu3N4nKCwLCYAe6EkEQCN1/xN9Nhpz3Q5zfd1zlNO+zqGPU9cx7Lr99IYDoAM9\niQDoJmUvHJ9wPzDHjKux938OmSRzMfa+viqHnl5dx7DrNUJpKgA9ECQCoJFywpFiomaiX64xJ+q5\nBS2mLoegzKocAidDjUFpKgA9ECQCoJFywrHvJMlEv2xjTtRzC1pMXQ5BmVU5BE6GHMMcMxUB6JWe\nRAD0Y58+JXptlK/N/u+jl00OPWnmJqeeRADAWm17EgkSAZAfE/3p63MfC1rkYWr7YWqvB4BZESQC\noGwmZNMmW6w8Xc7JqQV6p/Z6AJgdq5sBTN3QjZ2Hfj69NqYtdS+b3Bud9zG+IV9z1z5hU+sNNbXX\nAwDXECQCKNGmCVtfk9HUjaRzn9TTr5TNfXNvdN7X+TPka+4aJMmtofW+dn09rnMAFEaQCKBE103Y\n+po4pv4UPfdJPcNIlS2We5ZHH+Mb+jV3DZLksMpYSru8Htc5AAokSARQousmbH1NHFNnBeQ+qacs\nQ2St7JMR0sf4hs7U2SVIMrWS0a6vx3UOgAJpXA1QqnVNZEtZMUoT2P2dnDTbcbGIOD0dezTj67PR\neYrjtY/xldrcvdRxd+U6B0BGrG4GMFelTMBKGWeOTk4inn/+wffPPSdQ1CcrsaUzt8CJ6xwAmRAk\nAoCpevLJiLfeevD97dsRb7453nimbm6BjV20DYYIuAHAKNoGifQkAoDSXJ2EC1j0a2pNmFPr0qB5\nWy+lUlYDK2WcANCRTCIAKJGeRI0u5TxKf/rRNTvouv1QSsZWH+N0bALQM5lEADBlp6dNidncA0Rt\nM1gsR96friutXbdKWCmrgaUep2MTgIwIEgHQUD5BabpM1ksJQJQoVTle12DTWFKP07EJQEYEiQDw\nSTZl6jJZLyUAUarrsoO6/o0Sej+lHucux6agPgA90ZMIgGmvOKTXx7TpSZQn27qbrsdxCb2bAMhK\n255EN4YYDACZOzyMODt7MOmYSpbF6mTq7MxkaooWi/b7tMtj2Z3zrl/rytNsXwASUW4GQDllHl3p\n9UEqynvac95107XcV+kkAD0SJAKgkaKnSGr7Tsw3TaZM+ml7DOjZ1Y0gRjddg2pTDeoDkAU9iQDI\nU6q+G+t6fejpQZdjYMo9u/pSYk+iTWPu8/W4HgEwgLY9iWQSAZCnVCUr6zKklMPQ5RiQGdNdjpmJ\nm2zKFus7k0xmEAAZESQCIE99TsxN+qetTRlZl2PAJH76NgUNhwgqdwmqKZUFoEfKzQDIV98lHqWV\nw7Bdl9KdKR8DU35tfdh03PRRDrbr/lGaBsCO2pabCRIBQGom6OPRP0ggYVdD9STaZ/84vgHYkZ5E\nADAGK2GNa+xSwhxKgfTc2s2mkq+UPZb22T+HhxE3bzb/vnlTqSwAyQkSAUBKJujjGrN/UC4BwrED\nZaUaKsC37/6pqg9+BYCEBIkAICUT9PG1yfroIyCQS4BQo+3uhgzw7bN/7t6NuH+/+ff9+4LQACR3\nY+wBAMCkLCeAd+9G3Lr1YBI35Yn6vv1ahu7htNoT5uwsXSDl8LD5e8teM2MGCBeLaR9zqa0L8PW5\n/XbdPzkdYwBMkkwiAEhtsWgmby+8MH7pUd/2zcAYo0Srr4yfxSLizp2I27ebr4I05SglA1CWGAA9\nEyQCgD50CUTk0Ox4V/sGXMYo0eorIHB+3gQG33qr+Vri/pyrkoIvKZtoA8AVgkQAsCpVwKZtICKX\nZse72jfgMkYGR18BgVx6ErEbwRcAiKqu67HH8L6Dg4P64uJi7GEAULpde9ys9qp57LH9AwhtxnF8\n3ASIlo6OmolqSUrrSdSX1McP/ZrKcQcALVRV9UZd1wdbHydIBEByfU++Nv39fSbqYwRsdhmvyW2+\n7JsyCOgBMDNtg0TKzQBIq+/yqW1/f5+SnxJKn0ovT5s6JUtluHqdePnlcvuCAUBCgkQApNV3X5Zt\nf3+fQM9YzWu7BBb0vYH9rV4nHn004gtfEHgFgBAkAiC1vrNxtv39fQM9uWeClLJUN+Rs9TrxyU9G\nvPNO83OBVwBmTk8iAPazrgfLmD2J5iCX15/LOGAf+hMBMAMaVwPQv7lNrgRFHpjbvmfanNsATJzG\n1QD0b079cTSM/qA57XumL/cy05TOzzXpBuBagkQA7G5O/XG6BEXmMAmb076HqRDsBmALQSIAdjfW\namBjaBsUmcskbE77HqbiumD3HALbALSiJxEAtNWmb8nxcRMgWjo6aspYAMa2rpdYhP5iADOgJxEA\n89P3p+Ft+pYowwJytS4DUH8xAFbIJAJgGnJabctKSd3YXjCe666dzkuASWmbSSRIBMA0KPMqUw7B\nPZNh5u7qOZDDeQlAUsrNAJgXZV5lGrvUZS6NxmGTq6W0Y5+XAIxGkAiAabDaVjpDrnS0b3Bv37F2\nnQxbBYoSdT1uBd0BZku5GQDwwBhlJruWe6UYa5e/oQSHEu163CrDBJgU5WYAQHdjlJm0WTVunRRj\n7ZKBpgSHErU5btdlGl13XsqmA5g0QSIA4IFdy0zGmDimKolpG6RSgkOJth23Xfpy6eEFMHl7BYmq\nqvovq6r62aqqfqaqqrtVVf32y59XVVX911VVfeny///raYYLAPRql95OY00ch+5Dpe/VeGSv7G7b\ncdslQ042HcDk3djz9/9cXdf/WUREVVV/NCL+84j4IxHxb0XERy//+3hE/LeXXwGA3C0W3QIg6yaO\nQwVQuo61tOfjgz11zs4E6HaxumrZ6vcRTWbR2dmDnkWbMuS6PBaAIu2VSVTX9f+z8u2/EBHLLtjf\nFhH/Q9342xHxtVVVfXif5wIAMqUMiz7JXtnfpmy/rhlyTz0V8fTTgnUAE7VvJlFUVXUaEf9+RNyL\niH/z8sffEBG/tPKwr1z+7Jf3fT4AIDPLSaaVkOiD7JX9bcv2a5Mhd3WVtGef7W+8AIxmayZRVVVf\nqKrqrTX/fVtERF3XJ3Vd/46I+EsRcfz/t3e3sZKeZR3A/1e6gKIfKmnV2laoyYaAuxXJSe0HPxCs\npVSz6xsJBkNTTRqSNtFEo8BJNGq20ZCIqa3EJlIhqUUiiicCSYFA4EvBU2zpmra6vpQCDV3UomaT\nmtrLD/McOmxP98zumbed+f2SyTxzP888c/fk2jk5V6/7vnbetsutepexVNVNVbVdVdsnT5481/8O\nAGCRzrVD2dmyN836sRfU/k2j2k9FF8BaqO5dczdnf6Oqlyf5SHcfqqo/SfLp7r5nOPdoktd19xkr\niTY2Nnp7e3sq8wEAVszplQwSBjC5ra0zV/tNct6/P4DzVlXd390be1233+5mB8deHknyyHC8leSt\nQ5ezq5N8Y68EEQDAGalkgN1NUmF3pmq/SToUqugCWAv73ZPo96rqlUmeTfJYRp3NkuSjSa5PciLJ\nqSQ37vNzAIB1Z28aeL5pdH+btEOh7n4AK29fSaLu/tkXGO8kN+/n3gAA38IG2fB8kyZ4zkQCFoDB\nvpabAQBraHMzOXx49Dxv89ogG84X09iU2lIyAAb7XW4GAExir01hzxebm8mtt46Ojx8fPR87trj5\nwLqbVoWdpWQAZIrdzaZBdzMAVtIqdQU6fPi55FCSHDqUPPTQ4uYDAMCe5tLdDACYwCp15To9uTWv\nZNck3ZsAANgXy80AYNZWaVPYnaVlW1ujBNE8lppNo3sTAAB7UkkEALO2apvCHjs2WmI2SYJoGhVA\nq1SJBQCwxFQSAcA8rOOmsNOqAFqlSiwAgCWmkggAmI1pVQCtWiUWAMCSUkkEAMzGNCuA1rESCwBg\nziSJAIDZ2KkAuvfeUYJo1kmera35fRYAwAqq7l70HL5pY2Ojt7e3Fz0NAFh+kyZE1iVxMr7/0Utf\nalkaAMCYqrq/uzf2us6eRABwvtlJiNxxx+j5hTqHTXrdKtABDQBg3ySJAOB8M2lCZJ0SJ9deO6og\nSnRAAwA4R5JEAHC+mTQhsk6JEx3QAAD2zZ5EALAI+90ryJ5EAABMaNI9iSSJAGDebLIMAMAc2bga\nAJbR1layuTm9vYK2tpJbblntTanPlZ8NAMBZObDoCQDA2hivINqxn72Cxu93110qksb52QAAnDWV\nRAAwL+PdxpLk0KH9JS/WqXvZ2fKzAQA4a5JEADAvp3cbO3Zsf9Ut69S97Gz52QAAnDUbVwPAPE27\n25juZS/MzwYAIInuZgAAAABEdzMAWC06dQEAMGOSRACw7HY6dd1xx+hZoggAgBmQJAKAZadTFwAA\ncyBJBADLaHx5mU5dAADMwYFFTwAAOM3O8rJTp5K77kruuWf00KkLAIAZkiQCgGWz2/Ky22+XHAIA\nYKYsNwOAZWN5GQAAC6CSCACWzZEjlpcBADB3kkQAsIyOHJEcAgBgriw3AwAAAECSCAAAAABJIgAA\nAAAiSQQAAABAJIkAAAAAiCQRAAAAAJEkAgDGbW4mhw+PngEAWCsHFj0BAGBJbG4mt946Oj5+fPR8\n7Nji5gMAwFypJAIARra2zvwaAICVJkkEAIwcOXLm1wAArDTLzQCAkZ2lZVtbowSRpWYAAGulunvR\nc/imjY2N3t7eXvQ0AAAAAFZGVd3f3Rt7XWe5GQAAAACSRAAAAABIEgEAAAAQSSIAAAAAIkkEAAAA\nQCSJAAAAAIgkEQAAAACRJAIAAAAgkkQAAAAARJIIAAAAgEgSAQAAABBJIgAAAAAiSQQAAABAJIkA\nAAAAiCQRAAAAAJEkAgAAACBTShJV1a9VVVfVRcPrqqrbqupEVX2xql47jc8BAAAAYDb2nSSqqsuT\n/HiSL40NvzHJweFxU5L37PdzAAAAAJidaVQSvTvJryfpsbGjSd7fI/clubCqLpnCZwEAAAAwA/tK\nElXVkSRf6e4HTzt1aZLHx15/eRjb7R43VdV2VW2fPHlyP9MBAAAA4Bwd2OuCqvpEku/d5dRmkncm\nuXa3t+0y1ruMpbvvTHJnkmxsbOx6DQAAAACztWeSqLuv2W28qg4nuSLJg1WVJJcl+UJVXZVR5dDl\nY5dfluSr+54tAAAAADNR3dMp3qmqf0uy0d1fr6qfSHJLkuuT/EiS27r7qgnucTLJY1OZ0PNdlOTr\nM7o3TEocsmhikGUgDlkG4pBFE4MsA3G4Pl7e3RfvddGelUTn6KMZJYhOJDmV5MZJ3jTJhM9VVW13\n98as7g+TEIcsmhhkGYhDloE4ZNHEIMtAHHK6qSWJuvsVY8ed5OZp3RsAAACA2dpXdzMAAAAAVsM6\nJYnuXPQEIOKQxRODLANxyDIQhyyaGGQZiEO+xdQ2rgYAAADg/LVOlUQAAAAAvICVTxJV1e9W1Rer\n6oGqureqvm8Yr6q6rapODOdfu+i5srqq6l1V9cgQa39dVReOnXvHEIePVtUbFjlPVldVvamq/qGq\nnq2qjdPOiUHmpqquG2LtRFW9fdHzYfVV1Xur6smqOj429rKq+nhV/dPw/F2LnCOrr6our6pPVdXD\nw+/jXx7GxSJzUVXfVlWfr6oHhxj87WH8iqr63BCDf1FVL170XFmslU8SJXlXd1/Z3a9J8rdJfnMY\nf2OSg8PjpiTvWdD8WA8fT3Kou69M8o9J3pEkVfXqJG9O8oNJrkvyx1V1wcJmySo7nuRnknxmfFAM\nMk9DbN2R0e/gVyf5+SEGYZb+LKPvt3FvT/LJ7j6Y5JPDa5ilZ5L8ane/KsnVSW4evv/EIvPydJLX\nd/cPJXlNkuuq6uokv5/k3UMM/meSX1rgHFkCK58k6u7/Gnv5HUl2NmE6muT9PXJfkgur6pK5T5C1\n0N33dvczw8v7klw2HB9N8oHufrq7/zXJiSRXLWKOrLbufri7H93llBhknq5KcqK7/6W7/zfJBzKK\nQZiZ7v5Mkv84bfhokvcNx+9L8lNznRRrp7uf6O4vDMf/neThJJdGLDInw9+9/zO8fNHw6CSvT/KX\nw7gYZPWTRElSVceq6vEkb8lzlUSXJnl87LIvD2Mwa7+Y5GPDsThk0cQg8yTeWBbf091PJKM/3pN8\n94Lnwxqpqlck+eEkn4tYZI6q6oKqeiDJkxmtdPjnJE+N/c9sv5dZjSRRVX2iqo7v8jiaJN292d2X\nJ7k7yS07b9vlVlq9cc72isPhms2Myo3v3hna5VbikHMySQzu9rZdxsQgsyLegLVWVd+Z5ENJfuW0\nFQ8wc939f8M2LJdlVN37qt0um++sWDYHFj2Baejuaya89M+TfCTJb2WUJb187NxlSb465amxRvaK\nw6q6IclPJvmx7t758hWHTM1ZfBeOE4PMk3hjWXytqi7p7ieG7QaeXPSEWH1V9aKMEkR3d/dfDcNi\nkbnr7qeq6tMZ7Y91YVUdGKqJ/F5mNSqJzqSqDo69PJLkkeF4K8lbhy5nVyf5xk6pJ0xbVV2X5DeS\nHOnuU2OntpK8uapeUlVXZLSR+ucXMUfWlhhknv4uycGhk8qLM9o0fWvBc2I9bSW5YTi+IcnfLHAu\nrIGqqiR/muTh7v6DsVNikbmoqot3OixX1bcnuSajvbE+leTnhsvEIKnnChpWU1V9KMkrkzyb5LEk\nb+vurwxf1Ldn1O3iVJIbu3t7cTNllVXViSQvSfLvw9B93f224dxmRvsUPZNR6fHHdr8LnLuq+ukk\nf5Tk4iRPJXmgu98wnBODzE1VXZ/kD5NckOS93X1swVNixVXVPUlel+SiJF/LqKL8w0k+mOT7k3wp\nyZu6+/TNrWFqqupHk3w2yUMZ/V2SJO/MaF8iscjMVdWVGW1MfUFGxSIf7O7fqaofyKiRxMuS/H2S\nX+jupxc3UxZt5ZNEAAAAAOxt5ZebAQAAALA3SSIAAAAAJIkAAAAAkCQCAAAAIJJEAAAAAESSCAAA\nAIBIEgEAAAAQSSIAAAAAkvw/543lV5kTyKEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f5d1b70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "plt.scatter(X_embedded[labels,0], X_embedded[labels,1], color='b', s=10)\n",
    "plt.scatter(X_embedded[np.logical_not(labels),0], X_embedded[np.logical_not(labels),1], color='r', s=10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(702, 3000)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fmd.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "pca = PCA(n_components=2)\n",
    "X_new = pca.fit_transform(fmd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABJQAAAJDCAYAAABdQU7oAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10f7af9e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(20,10))\n",
    "plt.scatter(X_new[labels,0], X_new[labels,1], color='b', s=10)\n",
    "plt.scatter(X_new[np.logical_not(labels),0], X_new[np.logical_not(labels),1], color='r', s=10)\n",
    "plt.ylim([-2,2])\n",
    "plt.xlim([-10,10])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "eps=0.100000 and n_samples=702 lead to a target dimension of 5617 which is larger than the original space with n_features=3000",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-36-d33bf3c000a4>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0msklearn\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mrandom_projection\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      2\u001b[0m \u001b[0mtransformer\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mrandom_projection\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mSparseRandomProjection\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 3\u001b[0;31m \u001b[0mX_new\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtransformer\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit_transform\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfmd\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m~/anaconda3/envs/mllab/lib/python3.6/site-packages/sklearn/base.py\u001b[0m in \u001b[0;36mfit_transform\u001b[0;34m(self, X, y, **fit_params)\u001b[0m\n\u001b[1;32m    516\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0my\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    517\u001b[0m             \u001b[0;31m# fit method of arity 1 (unsupervised transformation)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 518\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfit_params\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtransform\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    519\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    520\u001b[0m             \u001b[0;31m# fit method of arity 2 (supervised transformation)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/anaconda3/envs/mllab/lib/python3.6/site-packages/sklearn/random_projection.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, X, y)\u001b[0m\n\u001b[1;32m    364\u001b[0m                     \u001b[0;34m'%d which is larger than the original space with '\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    365\u001b[0m                     'n_features=%d' % (self.eps, n_samples, self.n_components_,\n\u001b[0;32m--> 366\u001b[0;31m                                        n_features))\n\u001b[0m\u001b[1;32m    367\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    368\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mn_components\u001b[0m \u001b[0;34m<=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: eps=0.100000 and n_samples=702 lead to a target dimension of 5617 which is larger than the original space with n_features=3000"
     ]
    }
   ],
   "source": [
    "from sklearn import random_projection\n",
    "transformer = random_projection.SparseRandomProjection()\n",
    "X_new = transformer.fit_transform(fmd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Gaussian NB\n",
    "https://en.wikipedia.org/wiki/Naive_Bayes_classifier#Sex_classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Datos originales de wikipedia\n",
    "data = [\n",
    "    {\"genero\":'hombre', 'altura': 6, 'peso': 180,'pie':12}, \n",
    "    {\"genero\":'hombre', 'altura': 5.92, 'peso': 190,'pie':11}, \n",
    "    {\"genero\":'hombre', 'altura': 5.58, 'peso': 170,'pie':12}, \n",
    "    {\"genero\":'hombre', 'altura': 5.92, 'peso': 165,'pie':10}, \n",
    "    {\"genero\":'mujer', 'altura': 5, 'peso': 100,'pie':6}, \n",
    "    {\"genero\":'mujer', 'altura': 5.5, 'peso': 150,'pie':8}, \n",
    "    {\"genero\":'mujer', 'altura': 5.42, 'peso': 130,'pie':7}, \n",
    "    {\"genero\":'mujer', 'altura': 5.75, 'peso': 150,'pie':9}, \n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Altura en cm y peso en kg\n",
    "data = [{'altura': 183, 'genero': 'hombre', 'peso': 82, 'pie': 12},\n",
    "         {'altura': 180, 'genero': 'hombre', 'peso': 86, 'pie': 11},\n",
    "         {'altura': 170,'genero': 'hombre','peso': 77,'pie': 12},\n",
    "         {'altura': 180, 'genero': 'hombre', 'peso': 75, 'pie': 10},\n",
    "         {'altura': 152, 'genero': 'mujer', 'peso': 45, 'pie': 6},\n",
    "         {'altura': 167, 'genero': 'mujer', 'peso': 68, 'pie': 8},\n",
    "         {'altura': 165, 'genero': 'mujer', 'peso': 59, 'pie': 7},\n",
    "         {'altura': 175, 'genero': 'mujer', 'peso': 68, 'pie': 9}]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Matriz y lista antes de llenar\n",
      "[[ 0.  0.  0.]\n",
      " [ 0.  0.  0.]\n",
      " [ 0.  0.  0.]\n",
      " [ 0.  0.  0.]\n",
      " [ 0.  0.  0.]\n",
      " [ 0.  0.  0.]\n",
      " [ 0.  0.  0.]\n",
      " [ 0.  0.  0.]]\n",
      "[None None None None None None None None]\n",
      "\n",
      "Matriz y lista con los datos:\n",
      "[[ 183.   82.   12.]\n",
      " [ 180.   86.   11.]\n",
      " [ 170.   77.   12.]\n",
      " [ 180.   75.   10.]\n",
      " [ 152.   45.    6.]\n",
      " [ 167.   68.    8.]\n",
      " [ 165.   59.    7.]\n",
      " [ 175.   68.    9.]]\n",
      "['hombre' 'hombre' 'hombre' 'hombre' 'mujer' 'mujer' 'mujer' 'mujer']\n"
     ]
    }
   ],
   "source": [
    "GNB_features_matrix = np.zeros((len(data),3))\n",
    "GNB_labels = np.empty((len(data)), dtype=object)\n",
    "print('Matriz y lista antes de llenar')\n",
    "print(GNB_features_matrix)\n",
    "print(GNB_labels)\n",
    "\n",
    "for i, person in enumerate(data):\n",
    "    GNB_features_matrix[i, 0] = person['altura']\n",
    "    GNB_features_matrix[i, 1] = person['peso']\n",
    "    GNB_features_matrix[i, 2] = person['pie']\n",
    "    GNB_labels[i] = person['genero']\n",
    "print()\n",
    "print('Matriz y lista con los datos:')\n",
    "print(GNB_features_matrix)\n",
    "print(GNB_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "medias de los hombres: [ 178.25   80.     11.25]\n",
      "varianzas de los hombres: [ 32.25        24.66666667   0.91666667]\n",
      "medias de las mujeres: [ 164.75   60.      7.5 ]\n",
      "varianzas de las mujeres: [  90.91666667  118.            1.66666667]\n"
     ]
    }
   ],
   "source": [
    "men_means = GNB_features_matrix[GNB_labels=='hombre',:].mean(axis=0)\n",
    "women_means = GNB_features_matrix[GNB_labels=='mujer',:].mean(axis=0)\n",
    "men_vars = GNB_features_matrix[GNB_labels=='hombre',:].var(axis=0, ddof=1)\n",
    "women_vars = GNB_features_matrix[GNB_labels=='mujer',:].var(axis=0, ddof=1)\n",
    "print('medias de los hombres:', men_means)\n",
    "print('varianzas de los hombres:', men_vars)\n",
    "print('medias de las mujeres:', women_means)\n",
    "print('varianzas de las mujeres:', women_vars)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def gaussian(x, mu, sig):\n",
    "    return np.exp(-np.power((x - mu)/sig, 2.)/2)/(np.sqrt(2.*np.pi)*sig)\n",
    "\n",
    "def plot_gaussian(mu=0, sig= 1, points = 200, N = 2, marker=None, label = None, x = None, color = None, \n",
    "                  x_label='',y_label='',ax = None, figsize = (20,10), center_label_tick = None):\n",
    "    if ax is None:\n",
    "        f, ax = plt.subplots(1, 1, sharey=True, sharex=True, figsize = figsize)\n",
    "    if x is None:\n",
    "        x = np.linspace(mu-N*sig, mu+N*sig, points)\n",
    "    y = gaussian(x, mu, sig)\n",
    "    ax.plot(x,y, marker = marker, label = label, color = color)\n",
    "    ax.set_ylabel(y_label)\n",
    "    ax.set_xlabel(x_label)\n",
    "\n",
    "    if center_label_tick is not None:\n",
    "        labels = [item.get_text() for item in ax.get_xticklabels()]\n",
    "        labels[int((len(labels)-1)/2)] = center_label_tick\n",
    "        ax.set_xticklabels(labels)\n",
    "\n",
    "    return x, y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABIsAAAHiCAYAAABoTExbAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkiRJktT5F18CpYbhlmoEAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10b82c160>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sample = np.array([180,90,10])\n",
    "\n",
    "f, axs = plt.subplots(3, 1, sharey=False, sharex=False, figsize = (20,8))\n",
    "keys = ['altura', 'peso','pie']\n",
    "for i, ax in enumerate(axs):\n",
    "    plot_gaussian(men_means[i], np.sqrt(men_vars[i]), ax=ax, N=10, color='b', label='hombre')\n",
    "    plot_gaussian(women_means[i], np.sqrt(women_vars[i]), ax=ax, N=4, color='g', label='mujer')\n",
    "    ax.scatter(sample[i], gaussian(sample[i], men_means[i], np.sqrt(men_vars[i])), color='b')\n",
    "    ax.scatter(sample[i], gaussian(sample[i], women_means[i], np.sqrt(women_vars[i])), color='g')\n",
    "    ax.title.set_text(keys[i])\n",
    "    ax.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Posteriori de hombre: 6.29781391114e-05\n"
     ]
    }
   ],
   "source": [
    "p_hombre = len(GNB_features_matrix[GNB_labels=='hombre',:])/len(GNB_features_matrix)\n",
    "p_altura_hombre = gaussian(sample[0], men_means[0], np.sqrt(men_vars[0]))\n",
    "p_peso_hombre = gaussian(sample[1], men_means[1], np.sqrt(men_vars[1]))\n",
    "p_pie_hombre = gaussian(sample[2], men_means[2], np.sqrt(men_vars[2]))\n",
    "posterior_hombre = p_hombre*p_altura_hombre*p_peso_hombre*p_pie_hombre\n",
    "print('Posteriori de hombre:', posterior_hombre)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Posteriori de Mujer: 2.2363824638e-07\n"
     ]
    }
   ],
   "source": [
    "p_mujer = len(GNB_features_matrix[GNB_labels=='hombre',:])/len(GNB_features_matrix)\n",
    "p_altura_mujer = gaussian(sample[0], women_means[0], np.sqrt(women_vars[0]))\n",
    "p_peso_mujer = gaussian(sample[1], women_means[1], np.sqrt(women_vars[1]))\n",
    "p_pie_mujer = gaussian(sample[2], women_means[2], np.sqrt(women_vars[2]))\n",
    "posterior_mujer = p_mujer*p_altura_mujer*p_peso_mujer*p_pie_mujer\n",
    "print('Posteriori de Mujer:', posterior_mujer)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "HOMBRE\n"
     ]
    }
   ],
   "source": [
    "if posterior_mujer>posterior_hombre:\n",
    "    print('MUJER')\n",
    "else:\n",
    "    print('HOMBRE')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Notar que no hace falta calcular el denominador"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Utilizando sklearn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.naive_bayes import GaussianNB\n",
    "gnb = GaussianNB()\n",
    "y_pred = gnb.fit(GNB_features_matrix, GNB_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['hombre' 'mujer' 'mujer' 'hombre']\n",
      "[[  9.99077440e-01   9.22560042e-04]\n",
      " [  3.59206251e-07   9.99999641e-01]\n",
      " [  4.34003950e-01   5.65996050e-01]\n",
      " [  5.59309521e-01   4.40690479e-01]]\n"
     ]
    }
   ],
   "source": [
    "samples = np.array([[180,90,10], [160, 70, 8], [178, 73, 9], [178, 74, 9]])\n",
    "print(y_pred.predict(samples))\n",
    "print(y_pred.predict_proba(samples))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Problema 2\n",
    "Utilizando las siguientes funciones/paquetes resolver los siguientes problemas de clasificación en **problema_2.py**:\n",
    "\n",
    "**genfromtxt** de **numpy** para leer los dos datasets:\n",
    "- ./datasets/student_admission.txt\n",
    "- ./datasets/chip_tests.txt\n",
    "\n",
    "**train_test_split** de **sklearn** para dividir entre test set y train set. Recomendamos un 40% de datos para test set\n",
    "\n",
    "**GaussianNB** de **sklearn** como modelo de ML.\n",
    "\n",
    "Y la función **plot_boundaries(X_train, X_test, y_train, y_test, score, predict_proba, ax=ax, h=h)** incluida en problema_2.py para graficar los resultados. X_train es un np.array con los features de entrada, y_train es la etiqueta. Lo mismo con X_test e y_test, Score es el 'accuracy' del modelo, predict_proba es la función que dada una entrada de la probabilidad de clasificar correcto y h es el paso para la grafica del 'boundary'"
   ]
  }
 ],
 "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": "3.6.4"
  },
  "toc": {
   "colors": {
    "hover_highlight": "#DAA520",
    "navigate_num": "#000000",
    "navigate_text": "#333333",
    "running_highlight": "#FF0000",
    "selected_highlight": "#FFD700",
    "sidebar_border": "#EEEEEE",
    "wrapper_background": "#FFFFFF"
   },
   "moveMenuLeft": true,
   "nav_menu": {
    "height": "48px",
    "width": "252px"
   },
   "navigate_menu": true,
   "number_sections": true,
   "sideBar": false,
   "threshold": 4,
   "toc_cell": false,
   "toc_position": {
    "height": "295px",
    "left": "1px",
    "right": "20px",
    "top": "67px",
    "width": "212px"
   },
   "toc_section_display": "block",
   "toc_window_display": false,
   "widenNotebook": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}