KyoungHa-Park/Word2Vec_v2(설문자료 추가).ipynb

## Word2Vec_v2(설문자료 추가).ipynb
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Obzw_9t92vTC"
   },
   "outputs": [],
   "source": [
    "# code 및 내용출처 : \n",
    "# https://www.slideshare.net/lucypark/nltk-gensim\n",
    "# https://www.slideshare.net/YBkim2/1-word-cloud-108912512"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "XnkFask12vTJ"
   },
   "outputs": [],
   "source": [
    "\n",
    "# 그래프를 노트북 안에 그리기 위해 설정\n",
    "%matplotlib inline\n",
    "\n",
    "from matplotlib import font_manager, rc\n",
    "import matplotlib as mpl\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.font_manager as fm\n",
    "\n",
    "# 한글 폰트 지정\n",
    "font_name = font_manager.FontProperties(fname=\"c:/Windows/Fonts/malgun.ttf\").get_name()\n",
    "rc('font', family=font_name)\n",
    "\n",
    "mpl.rc('figure', figsize=(8, 4))\n",
    "mpl.rc('axes', grid=True)\n",
    "\n",
    "# 그래프에서 마이너스 폰트 깨지는 문제에 대한 대처\n",
    "# mpl.rcParams['axes.unicode_minus'] = False"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "xEedMdbm2vTO"
   },
   "source": [
    "## 1.Data proprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "4ZTHI-3l2vTR"
   },
   "outputs": [],
   "source": [
    "# 데이터 읽기 : txt 자료\n",
    "def read_data(filename):\n",
    "    with open(filename, 'r') as f:\n",
    "        data = [line.split('\\t') for line in f.read().splitlines()]\n",
    "        data = data[1:]   # header 제외\n",
    "    return data\n",
    "\n",
    "train_data = read_data('./ratings_train.txt')  # 네이버 영화평가\n",
    "test_data = read_data('./NPS_2016.txt')        # 회사 설문조사"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ZbWjwsif2vTX",
    "outputId": "9a44ce21-0934-4b02-a0ff-825d0877ad91"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "150000\n",
      "1694\n"
     ]
    }
   ],
   "source": [
    "# 데이터 확인\n",
    "print(len(train_data))\n",
    "print(len(test_data))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "1uoId8dI2vTf",
    "outputId": "4dc0608d-4f53-4a8b-f3a9-95333a669ce8"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[['9976970', '아 더빙.. 진짜 짜증나네요 목소리', '0'],\n",
       " ['3819312', '흠...포스터보고 초딩영화줄....오버연기조차 가볍지 않구나', '1']]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### 중간 확인1\n",
    "train_data[:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ZX2nJiT42vTn",
    "outputId": "e3cf1ce3-7ea2-4c85-e954-ba0fd5b8d161"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[['3', '믿음직스럽다', '9'],\n",
       " ['4', '첫째아기도 매일에서나온거안에서만 먹였었고 괜찮았기에 두달뒤 둘째태어나면 매일로 먹일생각이에요', '7']]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### 중간 확인2\n",
    "test_data[:2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "MlClLZGv2vTw",
    "outputId": "cae37c87-ddb6-4e23-e4e6-64b0574c85d9"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[['104', '\"아직은 출산전이라 직접 아이에게 먹여본 적 없으나, 엄마들 입소문으로 익히 알고 있습니다.'],\n",
       " ['맞는 분유 찾아서 먹일 생각인데, 앱솔루트도 꼭 시도해 볼거에요.\"', '8'],\n",
       " ['105', '아이의 영양과 건강에 좋아서', '8']]"
      ]
     },
     "execution_count": 109,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#### 중간 확인3 : 오류 Case\n",
    "test_data[110:113]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "cZDfO20R2vUF"
   },
   "outputs": [],
   "source": [
    "# <--- data cleansing(노가다 중...)  --->"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "61e_pYWz2vUJ",
    "outputId": "0ae72066-dd14-4e3b-dac3-abd4ede2fcec"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 0 ns\n"
     ]
    }
   ],
   "source": [
    "# 형태소로 토크나이징\n",
    "%%time\n",
    "\n",
    "from konlpy.tag import Okt\n",
    "pos_tagger = Okt()\n",
    "\n",
    "def tokenize(doc):\n",
    "    result = ['/'.join(t) for t in pos_tagger.pos(doc, norm=True, stem=True)] # ex '더빙/Norm', '나다/Verb'\n",
    "    return result \n",
    "\n",
    "train_docs = [(tokenize(row[1]), row[2])    for row in train_data ]  \n",
    "test_docs  = [(tokenize(row[1]), row[2])    for row in test_data  ]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "g59KKCpU2vUQ",
    "outputId": "a4a926e6-18f2-40b9-e42b-d9f14e6a0fdc"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<원본>\n",
      "['8062501', '공유 존잘!!!ㅎㅎㅎ', '1']\n",
      "---------------------------------------------------------------\n",
      "<정제 후>\n",
      "(['공유/Noun', '존잘/Noun', '!!!/Punctuation', 'ㅎㅎㅎ/KoreanParticle'], '1')\n"
     ]
    }
   ],
   "source": [
    "#### 중간 확인4 : 형태소 토크나이징(train) \n",
    "from pprint import pprint\n",
    "print(\"<원본>\")\n",
    "print(train_data[166])\n",
    "print(\"---------------------------------------------------------------\")\n",
    "print(\"<정제 후>\")\n",
    "pprint(train_docs[166])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "dq9HD92n2vUU",
    "outputId": "b37b05e7-ccd3-4f81-ed5c-07f3861d7b1d",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<원본>\n",
      "['269', '가격대비 괜찮아요..', '6']\n",
      "---------------------------------------------------------------\n",
      "<정제 후>\n",
      "(['가격/Noun', '대비/Noun', '괜찮다/Adjective', '../Punctuation'], '6')\n"
     ]
    }
   ],
   "source": [
    "#### 중간 확인4 : 형태소 토크나이징(test) \n",
    "from pprint import pprint\n",
    "print(\"<원본>\")\n",
    "print(test_data[266])\n",
    "print(\"---------------------------------------------------------------\")\n",
    "print(\"<정제 후>\")\n",
    "pprint(test_docs[266])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "5mcvzIMo2vUY"
   },
   "source": [
    "## 2.Data exploration\n",
    "+ 주요 단어들에 대한 histogram\n",
    "+ 주요 단어들에 대한 world cloud"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "XTxLt0_z2vUZ",
    "outputId": "0abaebb3-dbd5-4574-9a6f-b4a80d563bd5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2159924\n",
      "['아/Exclamation', '더빙/Noun', '../Punctuation', '진짜/Noun', '짜증나다/Adjective']\n"
     ]
    }
   ],
   "source": [
    "tokens = [t   for d in train_docs \n",
    "              for t in d[0]]\n",
    "print(len(tokens))\n",
    "print(tokens[:5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "QtNIVDf62vUc",
    "outputId": "fddc644b-71d1-4377-a2ea-da65b511fb73"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Ak5Tlf-42vUh",
    "outputId": "e22fe224-a8a1-4e50-a22c-71d412f6661a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18890\n",
      "['믿음/Noun', '직/Noun', '스럽다/Adjective', '첫째/Modifier', '아기/Noun']\n"
     ]
    }
   ],
   "source": [
    "token_test = [t   for d in test_docs \n",
    "                  for t in d[0]]\n",
    "print(len(token_test))\n",
    "print(token_test[:5])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "uTcz66kN2vUk",
    "outputId": "32ec3cb7-1041-40d5-eafc-9a27e994c016"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "COFrAu-S2vUm",
    "outputId": "4cb3471a-24b8-4db6-8c38-f7918091c0ce"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of Token : 2159924 \n",
      "unique Token    : 49895\n",
      "\n",
      "[('./Punctuation', 67777),\n",
      " ('영화/Noun', 50818),\n",
      " ('하다/Verb', 41209),\n",
      " ('이/Josa', 38540),\n",
      " ('보다/Verb', 38538)]\n",
      "Wall time: 3.16 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "import nltk\n",
    "text = nltk.Text(tokens, name='NMSC')\n",
    "print(\"number of Token : {} \\nunique Token    : {}\\n\".format(\n",
    "    len(text.tokens), len(set(text.tokens))))\n",
    "pprint(text.vocab().most_common(5)) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "HR12Kis_2vUq",
    "outputId": "f893863a-f30c-4680-dfb5-b57824b201fe"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "number of Token(test) : 18890 \n",
      "unique Token(test)    : 1702\n",
      "\n",
      "[('가/Josa', 567),\n",
      " ('먹이다/Verb', 529),\n",
      " ('분유/Noun', 452),\n",
      " ('먹다/Verb', 451),\n",
      " ('./Punctuation', 421)]\n",
      "Wall time: 23 ms\n"
     ]
    }
   ],
   "source": [
    "text_test = nltk.Text(token_test, name='NMSC')\n",
    "print(\"number of Token(test) : {} \\nunique Token(test)    : {}\\n\".format(\n",
    "    len(text_test.tokens), len(set(text_test.tokens))))\n",
    "pprint(text_test.vocab().most_common(5)) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "iNm4TH3E2vUu",
    "outputId": "21e9c019-c2ca-49bb-a3a2-ddf34a2593d3"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x3bded630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(14,5))\n",
    "text.plot(50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "FeiyFZA42vUy",
    "outputId": "916d1571-d4ef-4e18-db2a-7cbc280c4f1e"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0MAAAF7CAYAAAAQdlAGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzs3Xd4XOWZ9/HvM9KoWpYs917ANgaMwTImEFp6IZAsISSkbyCkv1uS3WzaJtmUzabsZpPsbgLJpkDKBkgILY0SugEJMM0F425j2ZasYqtLz/vHjI1swJZGGo1kfT/XNZdGZ84z9z2jM2fmp3PmnBBjRJIkSZJGm0SuG5AkSZKkXDAMSZIkSRqVDEOSJEmSRiXDkCRJkqRRyTAkSZIkaVQyDEmSJEkalQxDkiRJkkYlw5AkSZKkUckwJEmSJGlUMgxJkiRJGpXyc91Af02YMCHOmTMn120A0NraSnFx8ZCOtaY1rWlNa1rTmta0pjWPpprZUFNTszvGOPGIM8YYR9SlqqoqDhfV1dVDPtaa1rSmNa1pTWta05rWPJpqZgNQHfuQLdxNTpIkSdKoZBiSJEmSNCoZhiRJkiSNSoYhSZIkSaOSYUiSJEnSqGQYkiRJkjQqGYYkSZIkjUqGIUmSJEmjkmEoQ13dPVRvb8t1G5IkSZIyZBjKQE9P5D0/fpB/vbeB3z/+bK7bkSRJkpQBw1AGEonAK46bDMA/XvsYG3fvy3FHkiRJkvrLMJShv37pHE6bXkhzexcf/vnDtHV257olSZIkSf1gGMpQCIGPnFrOrMoSnnq2iX+56alctyRJkiSpHwxDA1CaTPDf71hKQX6CXzywmesf2ZbrliRJkiT1kWFogE6cXs7nzz8egE//9nHW7WzOcUeSJEmS+sIwNAjevnwWbzx5Gi0d3Xz45w/T0tGV65YkSZIkHYFhaBCEEPjqXy3mmImlrK3dy2evf4IYY67bkiRJknQYhqFBUlqYz3+/o4qiZILfPLyNa6q35rolSZIkSYdhGBpEC6eU8eU3LQbgc797glXPNuW4I0mSJEkvxjA0yC6qmsHFy2bQ3tXDh3/+MM1tnbluSZIkSdILMAxlwRcvOJHjppSxYfc+PvWbx/3+kCRJkjQMGYayoLggj/96x1JKC/K46bFnuXrFply3JEmSJOkQhqEsOWbiGL725pMA+NJNq3hsa0OOO5IkSZLUm2Eoi85fMo13vWQ2Hd2p7w81tvj9IUmSJGm4MAxl2WffsIjF08vZuqeVT1y70u8PSZIkScOEYSjLCvPz+K+3L6WsKJ8/P1XLj+7ZkOuWJEmSJGEYGhKzxpfwzbcsAeBrv19Nzab6HHckSZIkyTA0RF5zwhQuO3MuXT2Rj/7iEZrae3LdkiRJkjSqGYaG0CdfdxxLZ1XwbGMbVzzcmOt2JEmSpFHNMDSEknkJvvv2pSTzAg9sbWdHY1uuW5IkSZJGLcPQEJteUcwrjptMD/DbR7bluh1JkiRp1DIM5cBFVTMAuLZmi4faliRJknLEMJQD5yycSHlhgmd27WPlVr87JEmSJOWCYSgHknkJzp5dBKS2DkmSJEkaeoahHDl3djEANzy6nbbO7hx3I0mSJI0+hqEcmVOR5IRpY2lq6+LWVbW5bkeSJEkadQxDObT/QArX1WzNcSeSJEnS6GMYyqE3njydZF7gzrW72NnkOYckSZKkoWQYyqHK0gJetnASPdFzDkmSJElDzTCUY8+dc2ir5xySJEmShlDWwlAIoSmE8Jf05a9CCAtDCLeFEO4NIXyj13xfCiHcmZ5+Qrb6Ga5edtwkxpcW8PTOvTy+zXMOSZIkSUMlm1uGnooxnpu+/Bb4NnBpjPGlwJwQwmkhhLOAyTHGc4APAN843B0ejZJ5Cd548nQgtXVIkiRJ0tAYkt3kQgj5QFGMcWN60nXA6cCrgV8CxBifACqHop/h5s1VqTD0u0e3097lOYckSZKkoZDNMDQuhHBPCOH/gMlAXa/b6oBxwCRgV6/pXSGEUfc9phOmlbNo6lgaWzu5fdXOXLcjSZIkjQoh21/aDyFcALwLGBdjfGV62sWkgtAs4MYY493p6XfFGM9+gfu4HLgcYOrUqVU33nhjVnvuq5aWFkpKSgZl7E1r9/Hjlc1UTS3k02eOG5Ka2R5nTWta05rWtKY1rWlNa+bCsmXLamKMy444Y4xx0C9AXq/rLwGuAu4Gpqen/QpYBJwP/Gd62vHAL45031VVVXG4qK6uHrSxu5rb4jGfujnO+9TNsbapdUhqZnucNa1pTWta05rWtKY1rZkLQHXsQ27Jz0IQA1gYQvgh0A60Ah8ktavctSGEduCGGOOqEMIa4PUhhLuBZlIHURiVJowp5NyFk7h1VS2/e2Q77z97Xq5bkiRJko5qWQlDMcangDMOmbyZ1EETes/XA3woGz2MRBdVzeDWVbVcW7OVy86aSwgh1y1JkiRJR61Rd7CC4ezlx01iXEmSNbXNPLm9KdftSJIkSUc1w9AwUpDvOYckSZKkoWIYGmYuqpoBwO8e3UZHV0+Ou5EkSZKOXoahYeaEaWM5bkoZe1o6uX215xySJEmSssUwNMyEEA5sHXJXOUmSJCl7DEPD0BtPnk5eInDHmp3sam7PdTuSJEnSUckwNAxNLCvk3AUT6e6J/O7RbbluR5IkSToqGYaGqf27yl33sGFIkiRJygbD0DD18kWTqChJsurZJp7c3pjrdiRJkqSjjmFomCrMz+ONS6YBHkhBkiRJygbD0DD25gPnHNruOYckSZKkQWYYGsYWTy9nweQx1O/r4C9rPOeQJEmSNJgMQ8OY5xySJEmSsscwNMy9KX3OodtX76Rur+cckiRJkgaLYWiYmzS2iLPnT6CrJ/K7R7fnuh1JkiTpqGEYGgEuqpoJwHUPu6ucJEmSNFgMQyPAKxZNorw4yZPbm9jY0JnrdiRJkqSjgmFoBChK5nFB+pxDd2xszXE3kiRJ0tEhP9cNqG/eXDWDq1Zs4q7NbTy4oZ6iZIKiZB5F+XkUJRMUJlM/C/IShBBy3a4kSZI07BmGRoglM8o5dtIY1u3cy8U/uP9F5wsBCvMPDkpFyTyKknmcOqGHqqohbFqSJEkaxgxDI0QIgS+cfwLfuOkR8otKaevspr2rh7bObto6e2jv7Katq5vO7khbZw9tnT3Awd8venQLjJ/yDB8855jcPAhJkiRpGDEMjSBnzp9A8dmVVB1m8053T0wHpIPDUvWmej7/uyf52u9XU5zM4z1nzBm6xiVJkqRhyDB0lMlLBEoL8yktPPhPe/y0sWzbuoUf1DTx+RuepDiZx8WnzsxRl5IkSVLueTS5UeTV80r43BuOB+CTv3mM3z26LccdSZIkSbljGBplLj1zLv/wmoXECH//65X84YkduW5JkiRJygnD0Cj0kZcdy0dfdizdPZGP/fJh7lizM9ctSZIkSUPOMDRKffzVC7j0zLl0dkc+eFUN9z2zO9ctSZIkSUPKMDRKhRD47HmLeMdps2jv6uGyn1ZTs6k+121JkiRJQ8YwNIqFEPjSG0/kwqXTaeno5r3/+xCPb23MdVuSJEnSkDAMjXKJRODrbz6J8xZPpbm9i3f97wOs3tGU67YkSZKkrDMMify8BP/x1pN55aJJNLR08s4fPsgzu/bmui1JkiQpqwxDAqAgP8H33r6Us+ZPYPfedt5x5QNsqW/JdVuSJElS1hiGdEBRMo8r3rWM5XMq2dHUxiVXruDZxtZctyVJkiRlhWFIBykuyONH713GkpkVbN3TyjuufICGtu5ctyVJkiQNOsOQnqesKMnP/no5x08dy/rd+/jinXtobOnMdVuSJEnSoDIM6QWVlyS56tLlHDtpDJubuvjILx6mq7sn121JkiRJg8YwpBc1fkwhP33fcsoLE9yzbjdfvnlVrluSJEmSBo1hSIc1vaKYfzyjgmRe4Cf3beRXD27OdUuSJEnSoDAM6YiOm1DAV960GIDP/e4JHtpYn+OOJEmSpIEzDKlPLj51Ju976Vw6uyMfvKqGrXs8B5EkSZJGNsOQ+uzTrz+OsxdMpG5fB5f9tJp97V25bkmSJEnKmGFIfZafl+C7l5zCvAmlrN7RzMd/vZKenpjrtiRJkqSMGIbUL+XFSa58zzLKivL5w5M7+PZtT+e6JUmSJCkjWQtDIYTZIYRtIYTXhhCmhBBuCiHcHUL4SQghmZ7nQyGEu0IID4QQzslWLxpcx0wcw/fevpREgO/c9jQ3P/ZsrluSJEmS+i0rYSiEkA98A7guPekrwFdjjGcBu4ALQwizgfOBc4AL0vNrhDhnwUQ+/fpFAHz8mkd5YltjjjuSJEmS+idbW4a+AHwP2H8M5oUxxvvS168DTgdeCVwTU2qB+hBCRZb6URZceuZcLqqaQVtnD5f/rJpdze25bkmSJEnqsxDj4H4BPoTwauC0GOOXQghfAFYA/xxjPCN9+3zgs8Bq4PEY403p6T8HPhtj3PAC93k5cDnA1KlTq2688cZB7TlTLS0tlJSUDOnY4Vazszvy+TvrWVPXycLxSb54TiXJvJDVmtkaa01rWtOa1rSmNa1pzaGvmQ3Lli2riTEuO+KMMcZBvQC/Aq4Evg9UAzcDz/a6fTnwTeDDwLt6Tb8FGHOk+6+qqorDRXV19ZCPHY41dza1xZd89dY4+5M3xY//+tHY09OT9ZrZGGtNa1rTmta0pjWtac2hr5kNQHXsQ3YZ9N3kYoxvizG+P8b4QeAm4LvAvSGEpelZ3gzcCtydvk4IYRKQH2PcO9j9KPsmlhVy5buXUZRMcG3NVn50z/M27kmSJEnDzlAdWvuTwH+EEO4AAvDHGOPjwCMhhPuAXwIfH6JelAUnTi/nW285GYCv3rKKO9fuynFHkiRJ0uHlZ/POY4xf6PXr8w6dHWP8IvDFbPagoXPeSVNZUzuf79z2NB/9xcNc/5GX5rolSZIk6UV50lUNqr99xXxec8Jkmtu6eP9Pq9nb0ZPrliRJkqQXZBjSoEokAv9+8ckcN6WM9bv38e8rGujqNhBJkiRp+DEMadCVFubzw/csY3xpAStrO/jCjU/uP2KgJEmSNGwYhpQVM8aVcMW7l5FMwNUrNvPjezfmuiVJkiTpIIYhZU3V7HF85NRyAL5881Pcvro2xx1JkiRJzzEMKavOmlXM375yPj0RPvaLR1j1bFOuW5IkSZIAw5CGwN+8Yj4XLJnGvo5uLv3JQ+xsbst1S5IkSZJhSNkXQuDrF53E0lkVbG9s4/0/q6GtszvXbUmSJGmUMwxpSBQl87ji3cuYMa6YlVsa+PivV9LT4xHmJEmSlDuGIQ2ZCWMK+d/3nkpZYT43P/4s//7ntbluSZIkSaOYYUhDasHkMr73jqXkJQLfu2Md19VszXVLkiRJGqUMQxpy5yyYyBfOPx6Af/rNYzy4oT7HHUmSJGk0MgwpJ951+hzee8YcOrsjH7iqmk11+3LdkiRJkkYZw5By5rPnLeLchRPZ09LJ+37yEI2tnbluSZIkSaOIYUg5k5+X4LuXnMLCyWU8s2sfH/55DZ3dPbluS5IkSaOEYUg5VVaU5EfvXcaEMYXcu66Of/7dk8ToIbclSZKUfYYh5dyMcSVc+e4qCvIT/PLBzfzong25bkmSJEmjgGFIw8Ips8bxrbcsAeArt6zioe1tOe5IkiRJR7v8XDcg7Xf+kmls2L2Pf//zWr5xXwNXP3U7FSVJKooLKC9OUl6SpKI4SUVJMvV7cUHq9l7zFCXN95IkSeobw5CGlY+9/Fhqm9r4+QOb2bqnla17Wvs1viA/wYwxCc7d/hSnzavk1DmVVJYWZKlbSZIkjWSGIQ0rIQS+8leLec3UNubMP4GG1g4aWjppaO2ksaWDxtbOA783tHTSmL59//SOrh7WN/Sw/t4N/O+9qe8eLZg8huVzK1k+dzynza1k8tiiHD9KSZIkDQd9CkMhhOUxxgdDCIXAZcAfYozPZLc1jWalyQSzxpcwi5I+j4kxsre9i+vueIj6/Ak8uKGORzY3sLZ2L2tr93L1is0AzBlfclA4mjGumBBCth6KJEmShqm+bhn6d+BM4G+ABuB/gFdnqykpEyEEyoqSLJ5USFXVAgDau7p5fGsjD2yo58EN9VRvrGdjXQsb61r4dfVWAKaVF7F8biWTEy0UTmlk4ZQyknl+90iSJOlo19cwFEIIRcDUGOPXQwiXZLMpabAU5uexbE4ly+ZU8pGXQVd3D08928SDG+p5YEM9D22sZ3tjG9c/uh2AHzx8D4X5CU6YNpaTZlRw8swKTppRzpzxpSQSbj2SJEk6mvQ1DP0YuB14fzoUNWSvJSl78vMSnDSjgpNmVHDZWfPo6Yk8vXMvD2yo408PP8PWlgQb61p4eHMDD29+bjEvK8pnyYxUMNofkqaU+90jSZKkkayvYWhdjPGM/b+EEL6VpX6kIZVIBBZOKWPhlDJOKKijqqqKxpZOHtvWwMotDazc2sjKLQ3sbG7nnnW7uWfd7gNjJ5UVpoNROcWt7RzT0kFFiUeukyRJGikOG4ZC6lvlBcAXQgivAgJQBHwXOCX77UlDr7wkyVnzJ3LW/IkHpu1obGPl1gYe29rAyi2NPLY1FZBuXVXLratqAfjS3X9mzviS9Janck6eWcEJ08opLsjL1UORJEnSYRxpy9AZwJeBJcAfSYWhLuAn2W1LGl6mlBcxpXwKrzlhCgA9PZGNdft4bGsjK7c2cN/qbWxs7D5wcIYbVqa+g5SXCMyfNCb93aNUSPIADZIkScPDYcNQjPFe4GUhhE/HGL86RD1Jw14iEZg3cQzzJo7hTadMp2ZaGyedfApra5tTASm9i93a2mZW70hdfvXQFoCDDtBQ2dPKiUu6Kcx365EkSdJQ6+t3hr4RQjgPqCS1dYgY48+y1pU0AiXzEpwwrZwTppVzyfJZALR2dPPk9kZWbm1M72LX8LwDNFz15B2894w5vOO0WX7nSJIkaQj1NQzdDDwFrAZi9tqRji7FBc8d2nu/3gdouOaB9WxqbOcbf1zD925fx1tPncn7XjqXWeP7frJZSZIkZaavYagoxvi3We1EGiV6H6DhJWMbaS2fzRV3refup3fzk/s28rP7N/LaE6fw/rPmccqscbluV5Ik6ajV1zD0lxDCghjj2qx2I40yIYQDwWjVs0388O4N3LByG7c8voNbHt/BstnjuOysebzq+MnkedJXSZKkQdXXMPQq4O0hhF3p32Pv8w5JGrhFU8fyrYuX8I+vXchP7tvIz1dsonrTHqo31TBnfAmXnjWPi5bO8FDdkiRJg6RPYSjGeHq2G5GUMnlsEZ987XF89GXH8uvqLfzong1srGvhc9c/wb//aQ3vesls3nX6nFy3KUmSNOL1KQyFEN596DSPJidlV2lhPn/90rm86yWz+eOTtVxx1zOs3NrId25fx/fvWs9LZxTyd5MbOGlGRa5blSRJGpH6uptcca/rxwLTAcOQNATy8xKcd9JUXr94Cg9t3MOVd6/n1lW13LGxlTu+dy9LZlbw7pfM5ryTplKUdBc6SZKkvurrbnI/6P17COEz2WlH0osJIbB8biXL51aycfc+vnXDg9y5uYOVWxr4+JYGvnzzU1x86kzeedpsZlZ6aG5JkqQjSfR3QAihEFichV4k9dGcCaW8d8lYHvj0K/m3Ny/mhGlj2dPSyQ/uXM/Z37iDS3/yEHes2UlPj6cFkyRJejF9/c7Q/aROthqALuCb2WxKUt8UF+Tx1lNncfGymTyypYGr7t/EzY89y22rd3Lb6p3MHl/CO0+bzVuWzaCipCDX7UqSJA0rHk1OOgqEEFg6axxLZ43jM+ct4tfVW/j5is1sqmvhK7es4pt/WsMFS6bxbo9CJ0mSdEBftwzNAb4BTAOeAT4RY9yZvbYkZWrCmEI+fO6xfODsY7h99U6uWrGJu9bu4pqarVxTs5VjxuVz/p61nL1gIktmVHgyV0mSNGr19WhyPwA+E2OsDiFUAf8NXPRiM4cQCoDrgDJSu9a9HRiTHlcE3Bdj/If0vF8Czk73cnmM8ckMH4ukXvISgVcdP5lXHT+ZDbv3cfWKTVxTvYVn9nTx7Vuf5tu3Pk15cZIz50/gnPkTOWvBBKaWFx/5jiVJko4SfQ1DBTHGaoAYY00I4UgnNukC3hpjbAkhvBN4D3AWcGmMcWMI4ZoQwmlAATA5xnhOCOFEUlufXp/ZQ5H0YuZOKOVzbzieT7x6IT/94wq2dZdz59pdbK5v4ebHnuXmx54FYMHkMZw9fyJnL5jI8rmVHqpbkiQd1foahmIIoTLGWB9CqAQKDztzjD1AS/rX+UA18KoY48b0tOuA04HxwC/TY55I37ekLCkuyOPUaUV8sOpEADbu3sddT+/irrW7uO+ZOtbW7mVt7V5+eM8GCvMTnDZvPGfPn8A5CyZy7KQxOe5ekiRpcIUYj3zo3RDC6cB3gG3ATOBvY4x3H2HMPwCXA2vTP/8zxnhR+rZXAWcCU4DvxhifSE+/Bzg7HaZ639fl6ftg6tSpVTfeeGN/HmPWtLS0UFKS2flcMh1rTWtmq2ZnT2TN7g4e2dHBytp2NjR0HXT7hOIE08ckmDgmyfjiPMYVJ6gszmN8+mdZQSCEF/7+0XB6nNa0pjWtaU1rWnN41cyGZcuW1cQYlx1xxhjji16AHwLJXr9PAEqBrx1u3CH38Trg/4Bbe027GPgo8HXgrF7T7zrS/VVVVcXhorq6esjHWtOaQ1Wztqk1XlezJf7NLx+OS//lT3H2J2867GX+p2+JL/3abfHN/31v/PDVNfGLNzwZf3Dnunj9I1vjb267P6Ne+9PvYI2zpjWtaU1rWtOaQ1szG4Dq2IescqTd5ObHGDt7BafdACGEw6asEEIZsDfdyGYgDygMIUyPMW4DLgS+CBxL6kAMd4cQjge2HjG9SRoSk8qKuHDpDC5cOoOensia2mbueOhxyiZOZ0dTG7VN7dQ2tbGjsY3apjaa2rrYuqeVrXtaX/D+tvE0H3nZsS+69UiSJGmoHSkMvdhZGo807jjg2yGEdqCV1FagCcC16Wk3xBhXhRDWAK8PIdwNNAMf6HvrkoZKIhFYNHUsLdOLqKqa84LztHR0sbOpPR2U2tJBqZ3tDa388ckdfPNPa9nZ3M7nzz/Bw3lLkqRh4UihZk0I4azY6/tBIYQTgMbDDYoxPgS89JDJG0gdNKH3fD3Ah/rerqThqqQgnzkT8pkzofR5t33vd/fynQeb+Nn9m9i9t53/eOvJFOZ7pDpJkpRbRwpDnwCuDyHcCjxKare2S4C3ZbsxSUeP02cUUXXicVz+s2pueXwH9fse5Ip3L2NsUTLXrUmSpFEscbgb098ROhd4nNSub9uBc2OMT2e/NUlHk9OPGc//feB0JpUVsmJ9PW/9wQp2NrXlui1JkjSKHTYMAcQYu2KM18UYvxZj/EWMsXkoGpN09Dl+2liu+9AZzJtQyqpnm7jwf+5j/a69uW5LkiSNUkcMQ5I0mGZWlnDNB09nycwKtu5p5aLv38/KLQ25bkuSJI1ChiFJQ278mEJ++f7TOGfBROr3dXDJlSu4c+2uXLclSZJGGcOQpJwoKcjnh+9ZxoVLp9PS0c2lP3mI3z7iqcYkSdLQMQxJyplkXoJvvWUJHzhnHl09kb/7v5Vcedf6XLclSZJGCcOQpJwKIfCp1y3is+ctAuArt6ziKzc/RU9PzHFnkiTpaGcYkjQsXHbWPP7zbSeTzAtcefcG/v7Xj9LR1ZPrtiRJ0lHsSCddlaQh88aTp1NZWsAHr6rh+ke3U7evgw+c6P9sJElSdvgpQ9Kwctb8ifzq8tMZX1rA3U/v5mv3NNDV7RYiSZI0+AxDkoadxTPKue5DZzCxrJAndnXw1VtW57olSZJ0FDIMSRqW5kwo5fvvXEp+gP+9dwO/edjDbkuSpMFlGJI0bFXNruTSU8YC8KnfPM4T2xpz3JEkSTqaGIYkDWuvPqaES5bPpL2rhw9cVUPd3vZctyRJko4ShiFJw94XLjiBU2ZVsK2hlY/84mEPqCBJkgaFYUjSsFeYn8f331nFxLJCVqyv94AKkiRpUBiGJI0Ik8cW8f13LiWZFzyggiRJGhSGIUkjRtXsSr5wwQmAB1SQJEkDZxiSNKK847TZHlBBkiQNCsOQpBHn0AMqdHpABUmSlAHDkKQR59ADKvyrB1SQJEkZMAxJGpE8oIIkSRoow5CkEcsDKkiSpIEwDEka0d6+fBZvO9UDKkiSpP4zDEka0UIIfPGNHlBBkiT1n2FI0ojnARUkSVImDEOSjgqHHlDhL5tac92SJEka5gxDko4avQ+o8D8PNfKJa1by2NaGHHclSZKGq/xcNyBJg+nty2excfc+rrx7A9fWbOXamq0smVnBu14ymzecNJWiZF6uW5QkScOEYUjSUSWEwGfOO54lY5p5tLmMX1dvYeWWBlZuaeArNz/FxafO5J2nzWZmZUmuW5UkSTlmGJJ0VJo6Jp83nHM8H3/1Qm5cuZ2frdjIE9ua+MGd67nirvW8bOEk3nX6bM6ZP5FEIuS6XUmSlAOGIUlHteKCPC4+dSZvWTaDR7c0cNX9m7jpsWe5ffVObl+9k1mVJbzzJbN4S9VMxpUW5LpdSZI0hAxDkkaFEAKnzBrHKbPG8ZnzFvHr6q1cvWITm+tb+Ootq/nWn9Zy/pJpvPv02bluVZIkDRHDkKRRZ/yYQj507jFcfvY87li9k6tWbOLOtbsOHHChvDDBiY+sYMHkMo6bUsaCyalLaaGrTEmSjia+s0satfISgVceP5lXHj+Zjbv38fMHNnFtzVb2tHRy77o67l1Xd9D8MyuLWTh5LAunjGHhlLEsnFzGvImlJPM8S4EkSSORYUiSgDkTSvnMecfzqdct4o/3PEjBxDms3tHM2tpm1uxo5plde9lS38qW+lZuXVV7YFwyLzBvwhgWTiljal4Lp5wSPSCDJEkjhGFIknpJJAKTSvOpWjSZVyyafGB6Z3cPG3bvY006IO0PSpvrW1hT28ya2ubUjGWr+dTrFuWoe0mS1B+GIUnqg2Re4sB3h3pr6eji6dq91Gzaw1dufoof3LmeeRNKeeups3LUqSRJ6it3dJekASgpyGfJzApmU1RZAAAgAElEQVTed+Zc3r90LACf+e0T3PfM7hx3JkmSjsQwJEmD5FXzSrjszLl09UQ+dPXDrN+1N9ctSZKkwzAMSdIg+tTrF/HKRZNpbO3kfT95iD37OnLdkiRJehFZCUMhhOIQwhUhhNtDCA+FEN4QQpgSQrgphHB3COEnIYRket4PhRDuCiE8EEI4Jxv9SNJQyUsE/vNtJ3P81LFsrGvhg1fX0NHVk+u2JEnSC8jWlqEC4JsxxpcDrwI+D3wF+GqM8SxgF3BhCGE2cD5wDnAB8I0s9SNJQ6a0MJ8fvXcZk8oKeWBDPZ/57ePEGHPdliRJOkRWwlCMsTHGuDb9ayfQBCyMMd6XnnYdcDrwSuCamFIL1IcQKrLRkyQNpanlxfzoPadSlExwTc1Wvn/n+ly3JEmSDpHV7wyFEAKprT2fP6RWHTAOmERqK9Gh0yVpxFs8o5xvv/VkAP7tD6v5wxPP5rgjSZLUW8jWrhshhATwn8BtMcbrQwj3xxhPT9+2HLgYWA80xxivSk+/Bbg4xrj3kPu6HLgcYOrUqVU33nhjVnrur5aWFkpKSoZ0rDWtac2RV/O3q/dy9eN7KciDL507nmMrk1mvmc2x1rSmNa1pTWsO9tjBtmzZspoY47IjzhhjHPQLkAR+BLym17RrgaXp6/8GvBZYDFyfnjYJ+NOR7ruqqioOF9XV1UM+1prWtObIq9nT0xM/8etH4+xP3hRP/fKf4/aGlqzXzOZYa1rTmta0pjUHe+xgA6pjH3JLtnaT+zvgNcCnQgh/CSH8DPgk8B8hhDuAAPwxxvg48EgI4T7gl8DHs9SPJOVMCIGv/NViTptbyc7mdi79STX72rty3ZYkSaNefjbuNMb4deDrL3DT8w6dHWP8IvDFbPQhScNFQX6C77+zigv/5z6eeraJv/nVo/zgXVXkJUKuW5MkadTypKuSNETGlRbwo/cso7w4ya2ravna71fluiVJkkY1w5AkDaF5E8fwP+9cSn4icOXdG/jlg5tz3ZIkSaOWYUiShtgZx0zgq3+1GIDPXf8E967bneOOJEkanQxDkpQDF586kw+cM4+unsgHr65ha5MHVJAkaagZhiQpRz75muN49fGTaW7r4l/uqmdT3b5ctyRJ0qhiGJKkHEkkAt9+28mcOmccda09XHLFCjbXteS6LUmSRg3DkCTlUElBPj/+6+UsHJ9ke2Mbl1y5gi31BiJJkoaCYUiScmxMYT6fPWscS2dVsK2hlbddYSCSJGkoGIYkaRgoSSb46fuWc0o6EF1y5Qq27jEQSZKUTYYhSRomyoqS/PR9yzl5ZgVb96QC0baG1ly3JUnSUcswJEnDyNiiJD+7dDlLZlawpb6VS65YwXYDkSRJWWEYkqRhZmxRkp+9bzlLZpSzub6Ft12xgmcbDUSSJA02w5AkDUPlxUl+dulpnNQrEO1obMt1W5IkHVUMQ5I0TJUXJ7nqfadx4vSxbKpr4ZIrDUSSJA0mw5AkDWPlJUmuvvQ0Tpg2lg279/H2K1dQ22QgkiRpMBiGJGmYqygp4OeXncbxU8eyfvc+LrliBTsNRJIkDZhhSJJGgP2BaFE6EL3tyhXsbDYQSZI0EIYhSRohxpWmAtFxU8pYvyu1haihrTvXbUmSNGLl57oBSVLfVZYW8Iv3v4S3X7mC1Tua+cSfWzn28fsoSuZRmJ+gMJlHUX4eRckERcn0z/y81O3p64XJBMXJPDr3dhFjJISQ64clSVJOGIYkaYSpTG8hescPH2D1jmYe2rgn4/uafO9tLJ87ntPmVnLa3EqOnTTGcCRJGjUMQ5I0Ao0fU8iNHzuT39z+IHOOmU9bVw9tnd207//Z2U1bZ+p6W1ev6509tHd109zWxcMbd1Pb1M6NK7dz48rtQCponTpnHKfNHc/yuZUsmjqWvIThSJJ0dDIMSdIIlcxLcGxlkqp54zMa/1B1NRUzF7JiQz0PbqjngfV17Gxu549P1vLHJ2sBKCvMZ9mccamtR/MqWTy9fDAfgiRJOWUYkqRRKhEC8yeXMX9yGe96yWxijGyub+GB9fU8sKGeBzfWsaW+lTvW7OKONbsAKE7mMWtsgplPVFNRkqS8OElFcTJ1vaTgoN8rigsoK8on4ZYlSdIwZRiSJAEQQmD2+FJmjy/l4lNnArC9oTW11WhDPQ9uqOOZXftYU9fNmrraPt4njC3aH46SJLvbWLD5caaMLWLK2CImjS1kSnnqenlx0u8rSZKGlGFIkvSiplUU86ZTpvOmU6YDsKu5nd/f+zCTZ86lsaWThtYOGlo6aWztpKG18+BpLZ00t3fR2Jq6fVP6Pquf3fyCtQrzE0zuHZLGFjGlvIhJY4to3NXBxLoWJo0tpCiZN0SPXpJ0tDMMSZL6bGJZISdMLKDqhCl9mr+ru4emti4aWjrY09LJikefonTCNGqb26ltbGNHUxu1TW3UNrWzt72LzfUtbK5veeE7+8sdAFSUJNOBqYgp6dA0KR2iUuGpkAmlhe6eJ0k6IsOQJClr8vMSVJYWUFlakJqwu4iqqrkvOO/e9q5UMDoQktqpbWpjR2Mbzzxbx77uPHY2t9PQ0klDSyerdzS/eN1EYFJZIZPLiyiJbSytW5PeBbCE2eNLmDim0F3yJEmGIUnS8DCmMJ8xE8dwzMQxz7utpqaGqqoqenoidfs6DoSk2uYXCE9NbTS0dLK9sY3tjW0A3Ltl3UH3V1KQx6zKVDCaM76UWemfs8eXMLW82MOJS9IoYRiSJI0YiURgYlkhE8sKOfEwh/lu6+xmZ1M7zza2cvcjqwhlE9lY18Lmun1srGuhsTW1ZemFti4V5CWYUVlMaehk0uMPUZTMozCZoCiZR1F+HkX7r/eaVphMUNjrto31HZTuaErP/9z0wvyEW6QkaRgxDEmSjjpFyTxmjS9h1vgS8vcUU1W18KDbG1o62FTXwqb6Fjbt3pf6WbePTXUt7GxuZ/2ufakZd+7MvInb7n7ByYX5CQrz9weqg0MVHftYvnsNC6aUcdyUMuZOKCWZl8i8B0nSYRmGJEmjTkVJARUlBSyZWfG821o6UgdyuLfmCWbOmUdbVw9tnd20d3bTnr7e1pn+2dVNe2fPgXlS8/VQ19hEXkHRc/N1dtPW1UNHVw/t6UtTW9cL9vbg9ud26UvmBeZNGMPCKWWpy+TUz+kVxR4gQpIGgWFIkqReSgryOW7KWPZNKezzUfMOtf87Tofq6Yl0dD8/UO2//sDKVXSWTGT1jmbW1jazub6FNbXNrKlthpXP3U9pQR7zJ6e2Hi1IB6RdTV3MbG6joriAgny3JklSXxiGJEkaIolEoCiR96LnSkoeskvfvvYunt65l7Xp7zetrU393L23nUe3NPDoloaD7+CPtwGpA0RUFCcZW7z/hLcFVJQkKS9JUl783O8VxUm2N3QyeU8LFSUFlBbk+Z0mSaOKYUiSpGGqtDCfk2dWcPIhu/PV7W1nbe1e1uxoYk3tXtbtbGZbXRPtPXk0tHbS0tFNS0f3gaPpHdGfU+dwyk8EyotToamiOElFSUHq9wOhav9tqTDV2R0H+yFL0pAyDEmSNMKMH1PI6WMKOf2Y8Qem7d81L8bIvo5uGlo6aGjppLG187mfrR00ps/T1NDaQWNrJ9vrmugkn4aWTlo7u6nb10Hdvo4+9VGSDJy3aSUXnDyN0+eNJ9+DPUgaYQxDkiQdRUIIqXM2FeYzY9yR5+/9/ab2rm4aWztTgal3iGrpOCRUdbK9oZV1O/dyTc1WrqnZyoQxBZy3eCoXnDyNpbPGubudpBHBMCRJkgAozM9jUlkek8qK+jT/7+54gGe6xnHjyu1s2L2Pn96/iZ/ev4npFcWcv2Qa5y+ZyvFTxxqMJA1bhiFJkpSRGWPzeWPVAv7ulfN5YlsTN6zcxk2PPcu2hla+f+czfP/OZzhmYikXLJnOBSdPY+6E0ly3LEkHMQxJkqQBCSGweEY5i2eU86nXLaJ60x5uWLmNWx7fwTO79vEft67lP25dy+Lp5VywZBrTYhcxRrcYSco5w5AkSRo0iURg+dxKls+t5PPnn8C963Zzw8rt/OnJWh7f1sjj2xoBGHP7n1gwOXVC2f3nSlo4uYzxYwpz/AgkjSaGIUmSlBXJvATnLpzEuQsn0dbZzR2rd3LjY9u5b+1OGtq7eHhzAw9vPvhcSRPGFLJwyhgWTh7LwiljWDA5FZZKC/3IImnwuWaRJElZV5TM43WLp/K6xVOpqalhzsITWVPbzJpeJ5Ndmz6h7O517dy7ru6g8TMri1k4uYzQ3syxu1anTx6bOv9ReXHBc+dCKklSnPTksZL6JuthKISQF2PsznYdSZI0cowfU8gZYwo545gJB6bFGNm6p/W5cJQOS8/s2suW+la21LcC8Of1zxz2vgvyEpSXJA8KTK17mxi/9pGMeq2vb6Ayg7FNDQ0c8+xTB4JaefpEthW9TmRbVpQkL2Fwk3IlK2EohFABfB54JfAPwB9CCAuB/waKgPtijP+QnvdLwNnpXi6PMT6ZjZ4kSdLwFkJgZmUJMytLeMWiyQemd3b3sGH3Ptbt3Mujq9ZRPnHqgZPKNrR29jo3Umpae1cPu5rb2dXcfnCBLdszby7DsX/ZtOGwt4cAY4vSW7WKk4xNB6WmhgYqn+5/AEu0NTNjfhuTx/bt8OjSaJetLUM9wPeBxl7Tvg1cGmPcGEK4JoRwGlAATI4xnhNCOBH4BvD6LPUkSZJGoGRe4sB3hyZ3bKeq6tjDzt/W2X3gJLH7Txj71Np1zJ07N6P6GzZsyGjsmnXrGTdp+oGQ1tDaSdP+vtLTmtu6UmGutZNNh97B5swC2E1fv4O3LpvJB86Zx4xxJRndhzRaZCUMxRibgKb9++uGEPKBohjjxvQs1wGnA+OBX6bHPBFCqMxGP5IkafQoSuZRlMw7aOvI+LZtVJ08PaP7q+nekdHYmu4dVFXNO+w8Xd09NLd10dCaCm77A9Mz6zcwd+6cftXr6YFr7lvNim3tXLViE798cDMXLp3Oh889ljme40l6QSHGmL07D+ELwApgJfDdGONF6emvAs4EpqSnP5Gefg9wdoyx55D7uRy4HGDq1KlVN954Y9Z67o+WlhZKSjL7j0umY61pTWta05rWtKY1Dzdud2eS36zex72b2+gBEsCZs4p486IxzBj74v8HH2mP05pHT81sWLZsWU2McdkRZ4wxZu0CfAF4LVAM3Npr+sXAR4GvA2f1mn7Xke6zqqoqDhfV1dVDPtaa1rSmNa1pTWtasy/j1u/aGz/x60fjMZ+6Oc7+5E1xzj/dFD90dXV8cltj1moO1VhrHl01swGojn3IK4msRLHnB65WoDCEsH8b84XAbcDdwP6tRccDW4eiH0mSpKPd3AmlfOMtS7jjE+fyzpfMIplIcMvjO3j9d+7msp8+xKNbGo58J9JRLltHk5sC/AqYA1wYQvgQ8PfAtSGEduCGGOOqEMIa4PUhhLuBZuAD2ehHkiRptJpZWcKX37SYj75sPlfctZ5fPLiJW1ft5NZVOzlr/gQ+9vL5LJ/r17Y1OmXrAAo7gHNf4KbTD5mvB/hQNnqQJEnSc6aUF/HP5x/Ph192DD+8ewNX3b+Ru5/ezd1P72b53EqOKe1gW/525owvYXZlKeUlyVy3LGVd1k+6KkmSpOFjwphC/ul1x/HBc+bx43s38uN7N/DghnoeBH75xHPnNqooSTK7soTZ40uZM76EWQd+ljBxTCH7jxosjWSGIUmSpFGooqSAv3vVAi49ay5/eHwH9z7xDG35Y9hU18Kmupb0eZoaWbm18XljSwrymFVZwpzxpYT2Jv5Uu4rCZB5FyQRF+XkUpn8W7Z+W/lmYnlaYn5rW2N5D3d72F+ju8Fo6e4gxGsg0YIYhSZKkUWxsUZKLT53JMYmdVFVVAamjDe9qbmdTfQsbd+9jc30LG+ta2Fy3j411LTS2drJ6RzOrdzSn7mTd+swbuOHWjIbl3fB7youTVBQnKS9JHrheUVKQur5/WkmS8uKCA7+3dvXQ3RPJSxikZBiSJEnSIUIITBpbxKSxRZw65/kHV2ho6WBTXQsb6/axcvUzTJwynbbObtq6umnv7KG9q5u2zp7UtM709fS09vS09q4eOjs7yU/2/7tJLW2dtHVH6vd1UL+vo/8P8Le3kMwL6a1Yz2292r/FqqjXlq3C9G176pq4cduT/a8F1O9u4s66NelaB9//C9XevwWtpbOHzu4eknlDcgDoUckwJEmSpH6pKCmgoqSAJTMrmNG9g6qqYzK6n5qamgNbo/o7bvGSU2hs7aSxtYPG1s70bn2dNLR20tiSntb6/Gn72jvp6IbO7khndxfN7V19L7xuY797PWDtuszGXf978hKBovyDQ1PhQaFqf5B6Lkzt2tXETdv7H94a65p5smMjk8cWMXlsEVPGFjFhTAH5R2kgMwxJkiRpxCnITzCxrJCJZYX9GldTU8PSpUtp7+pJXQ7acnXwFq32rvT1rh42bNzEzJkz+91njLB+02bGT5p64P76suWsrauHlvZOOnuguyeyr6ObfR3d/Sv+9MZ+9wvwm9UHh6hESB14Y0p5UTokFTJlbNFzgSk9PXWu05HFMCRJkqRRJYRwYEsKxX3bTa8muZuqqrkZ1asprqeqakH/x6W3nHV29xwU1HqHqQOBbX+4Su+quGXLFmZkEN7WbthMorSSnU1t7Ghqo7apjd17O9jZ3M7O5nbg+QfU2O8N80tYtqzfJXPKMCRJkiQNY8m8BMm8BGVFfR9TU1OfUXhLBbfFB03r7O5hZ3M7tU1t1DbuD0np3/eHpsY2xhWNvF3pDEOSJEmSXlQyL8H0imKmVxQfdr4Hq6uHqKPBM/LimyRJkqRhJ28EnvfJMCRJkiRpVDIMSZIkSRqVDEOSJEmSRiXDkCRJkqRRyTAkSZIkaVQyDEmSJEkalQxDkiRJkkYlw5AkSZKkUckwJEmSJGlUCjHGXPfQLyGEXcCmXPeRNgHYPcRjrWlNa1rTmta0pjWtac2jqWY2zI4xTjziXDFGLxlegOqhHmtNa1rTmta0pjWtaU1rHk01c3lxNzlJkiRJo5JhSJIkSdKoZBgamCtyMNaa1rSmNa1pTWta05rWPJpq5syIO4CCJEmSJA0GtwxJkiRJGpUMQ5IkSZJGJcOQJEmSpFEpP9cNaPgKIZwAVAIBIMZ4V247Ui6FEBIxxp5c9zHchBBKY4z7ev1eFGNs6+d9+NxKWTKS3stCCCfHGB/t9fuiGOOqXPYkHe0MQxkIIZQDrwfK9k+LMWbt6BkhhDOAvwfGkdqaF2OML89WvXTNnwOlwGogpi9ZfwMJIZwC/Fu69jbgH2KMm/p5H/36YDkYNYdKCOFjwIeAxvSkGGM8ow/jMl5mQwivAf4JmBhCWAZcGmP8ryzXHPBrLIPlINOaN4QQvhRj/EsIYTnwDeCcPtbM6LlNj81ouc3F+mQgBvA4M/p7DuT5Ger3hoHIVa85eP/M+L1sgMtCpmP/LoTwDPCvwMdIrUvemM2aOXqtDPnYkbTuy9XnktHyvnIow1BmfgfcDpwCbKEfuxtmuKD9AHg3sIrUirxfMqw5L8Z4en9r9aqZ6Rve94G3xxifCSGcClwJvLqPNTP9YDmQmkO9Ur4MODnG2NGX/nrJeJkFPgu8HPhTjLEthPAGoC/P60BqDuQ1lulykGnN84BPhRD+HthMHz+4pGX63ELmy23G65MBfoDJ9M0908eZ6d9zIOvbgSzzA1mfvBr4KtAGdAF/23vrwnDqdSB1c/BeNpBlIaOxMcb3hBAuBLYDn48x9md9kmm/uXit5GLskK/7MnxtQg4+lwyw7oA+p+ZcjNFLPy/A7emfX03//G0/xj4AHJO+fiqpD0FHGnPLAPvNpOZ/A+MHUPMvwD8DvwW+A3yvj+NuPeT3I/baa967gSRwR/r33w9BzcdJvYEUAYVAYZbHXZfh32Mgy+xdh9zHfUNQcyBjM10OMqoJLAT+j9R/cn8FzMn2c5ueN6PldiDrk0yX2/TYfq+HBvg4M/17DuT5yXi5HcjzC9QAlenrs/cvV8O016F+/8z4vWyAy0JGY0n9w+sG4DXAdcBbh6BmLl4ruXhuh3zdl8lrMz3vkH8uGUjdgTy3w+HilqHMNIUQSoGxIYTTgAX9GNscY3wGIMb4UAihL2OeCCF8hdSm/Zge+6cs1zwVWBdCWJ3+PcY+7I7VS0+M8V9CCF+NMX46hPDbw80cQihIX/1D+r/jfwaWA/f1o2aMMXaGEPb/V6J8CGpuiTE+0o/5BzquJYRwJ6ke9y8Ln+7DuIEss1eFEK4BZoQQrgB+38dxA6k5kLH9Wg4Goea/Ah+LMW4LIUwn9Z+18/s4tt/P7SAstwNZn2S63EI/10OD8Dgz/XsO5PkZyHILmT+/u2OM9QAxxk0hhJY+jMlVr0P9/jmQ97KBLAuZjq0A3hRj7Akh/Bn4XB/rDaRmLl4ruRibi3Vfv16bufpckuP3lZzzpKsZCCGMBdqBmcD/A26IMd56hDH7F7T/R2rf5f0L2itijF84wtjPHzIpxhj/pQ99ZlxzoEII1wPvILVLw1XA/8YYTzjM/HeQegHtf3fbfz3Gvu+G835Sm3OXkNoytSXG+KUs1/w60Ek/VwADGHfOodNijHf2YVy/l9lDxi8g9byujTGu7OOYjGsOcGy/loPBqDkQ/X1uB7rcZro+SY/t93Kb6XpoEB5nRn/PAT4/A32d9ev5TS87AG8D6nnuuV0cY/zH4dTrQOrm6r1sgMtCv8aGEKbEGHf0+pv2Hrg2GzV7jcvFa2XIxw7lui/T12auPpfk8n1lODAMZSCEMCPGuDWk/i11HnB/jLHuCGMGvIBn0GfGNUMISeBy4DjgEeDHsR8LywBWrmfEGPvzH5BDx2fyoT3jmjlaKZ9Iatesx2KMT/dxTL+X2V5j7weuJRVo9/RlzCDUzHhsenwmy0G/aoYQvh5j/Mf087P/tbH/9dWn/zxn+tymxw7otZKJTJbbQXiTzehxDnQZysQgLLf9/QD94xe5KcYY3zeceh1I3Vy+lw2VEMLfxhi//QJ/0yP+LQeh9pC/VkaaoXxtpscP+eeSgdYdyQxDGQgh3BdjPCOE8CFSH/YXxxj7tFtMJgtarzeCA/oToDKs+TNgJXAbqS94z44x/k0/xme0cg0h3BZjfEV/eu01NtMP7RnXHGohhE8AZ5La4nEWcFOM8cVWur3HDWSZzQfeAFwCNAFXxBgfynLNgYzNdDnoV80QwkUxxmtDCCfGGJ/oa51D7iOj5zY9NqPldqDrk0wNINRk+jgzWoYG8vwMZLkdiBDCm2OM1/VzTK56Her3z4zfywa4LPRrbAjhKzHGz4QQ3hljvLovNQar3xy9VoZ8bC7WfZm8NtPjcvK5ZKS9rwwWvzOUmf1/8EUxxv8XQri9H2O/BPR3QXttr+vzSYWL/sik5qwY47vT1x8NIfR3d6FfA2cAHyS1cv0AffsexZMhhKtJ7afaBRD7ftjVs0h9sPx+CKE/HywzrpmDlfKFwEtjjDGE8B1Sb/BHDEMMYJmNMXYB14cQ7iV1qNdfAsdms+YAx2a6HPS35j+FEG4DvhdCeBXP/dea2Mej/Q3guYXMl9uM1ycDfMPLZD0EmT/OTJehgaxvB7LcDuT5vYzUl+37I1e9DvX750DeywayLPR37KtDCNXAJ0IIO3vfEPv+3YtM+83FayUXY3Ox7svktQk5+FwywLoD/ZyaU4ahzPwphPAI8NEQwv6jdfRVvxe0GGN7r1+fCCFc1M9+M1m480IIIf2hO0HqUKb9kenKtaafdZ4rmPkHy4xrMvQr5fb9u3jE1Bds8/o4LuNlNoTwNuDtpJadnwJ93Q94IK+TjMcOYDnob81vA9eQ2h3vjxy8C09f/8OZ6XMLGS63A1yfDOQNL9M32UxfnxktQwN8fgayzEPmz++zIXVglQd47rk90oFVctXrkL5/MoD3soEsCxmMfR+pw/KXAy/h4PVJn8LQAPod8tdKLsbmaN2XyWsTcvO5JOO6g/A5NafcTW4Q7F/R9nHe9xw6Lcb40yOM6X2M9+nAG2OMb+pHf5nUvAx4E3Arqf/E/SH28WSQ6fGfT4//KKkX120xxpf2cexMYEaM8f6+1kuPO/SD5c3pD8ZZq/kC9/OFmMEXevs6LoTwXaAWuAl4HTA3xnh5BvX6s8x+mtQuZzv6W+f/t3fmcbdOZR///hxDIlPRoNKEJqVJA5U5aXylFBqkOdFkqKjkpQxJ7ytNiETJm5Ap4phVNEmZS0QSSoVw+L1/XOs+z/3ss4d72MPznLO+n8/+PHvv517rWnvd617jNTSV2SZtm3bQRKakXW3vWzf/lLZV3TZpt237k468Krf3Jv1QKW3r57PG/Rxm/TRu8yl91X6hkWOVjjzGUtY2csc9lrVpC3XTStrB9sGSXm375CoyhlnejnxG/qxMIu0k+r42z+ak5iWTHlcmQT4ZaoAi8NuXgTmEnv+OQCVDdttHNmhoRcA4A7cD29cpbxOZtg+VdB7wLGA327+rKXNPYM/is6T1qqRLk8O1gDUVgcMOsv2+imKfBLyn7sSyjcwuHcDao0wH7ETsIL4L+B3wgYryardZSRvZPotoc69VyY1tlR39Ns9Jm7Q0bwdNZX5D0u6ES9xPEXFQfj9AVqu6TXk0bbeN+5MW7bZp39f4d7a4n23qp027bVy/ts+VtC7wGMJebskBSSZW1nGPny3HsjZjb92071ScZu8u6Sqmq91W8ibXQCYwmWdlQmnH3vc1eTaTvLHPS1rKbTVPnTT5ZKgBks4B3mb7htQxf9X2qyumnd/QCJuaqgP7SsATgWts/7NmeSvLVDIG73iYgFp6y107V1fwfCbpHNvrS5prewNJZ9juG/24mFhKei8L6slWmbTXlllKW3htKTqA77mao4ha6SQtafs+TbmYnY8r2JBGrPIAACAASURBVKc0abOSNrV9RpfdWNv+9ihktixv63bQpLySTgK+CnzC9ssknegBUePb1m1R3hbttlF/0rS9p7RN+75Gv7Nl+2taP41lpvRN+5ODgbuADYB1gSNtbz1Dy9pY7oTGsjZjb+W0kjYiHKm8iog3Nl9NzjW8yTUp7ySelUmlHXff1+TZTOnGPi8ZgtzG93PS5JOhZjxo+wYA2zdKWrpG2k1LDe0/kp40KIGkbYgTgEuAZ0k6wPapI5L5HOByplb5BZX1lhNfArYtd66EUfsgLGnZ9Hdx4GEV0hR2M//pzKtiWZvIjIT2nk06gAbpdgc+TdimTHPjTDX7lNpttjRhuNf294rvJW1RQV4jmS3Ttm0HTcv7UNunSdo5fR7YfoZQt9Cw3bbpT5q290Ttvq8Q2/D5bHQ/W/a3bdp8m/p9mu2NUt3eL2mVGVzWcY2frceyNm2hblrHSfFZkt7uiuqjQyzv2J+VSaSdUN/X5NmECcxL2sgdwjx1stjOr5ovYtfmyen9kwl7mKpp5wLLAmcTi9GLK6S5GFgyvV8SOLtmeZvIXL/j83o1ZZ7d8blSHRGGoxcAtwLnAq+uIfPNHZ+3GIPMbYALCWP6s4HNR5xuxY7PK4yqzQLLETuwPyUikq9BqJr8blQyh5S2aTtoJBM4GnhtajvrEjZKI63bNu22TX/StN2mtLX7oZa/s+n9bFM/jdttm/olNklWTWmWA86fwWUd9/i5fsfnymNZy7bQKC3wduBt5dcYZE7iWZlE3Y6972vybKZ0Y5+XtJHbpm5nwmviBZiNL2IScybh0WYu8MwaaWs3tM5GBZxTs7yVZRKnDUsRcWyWSI16OeBXNWXW6lyBlTo+r0xS46wgq9HEso3MUpqxdspd2sK5o2qzqQ6/BfwFODy9PxR4w6hktixv28Vbo/IS3qn2Ak4BDgRWHmXdtm23bfqTpu02XV+r7xvC72x6P9vUT+M236Z+U1s/DbiFOPV44Qwu61jGT4YwlrVsC43SEmEoitf+wDFjkDmJZ2USdTv2vq/us9m232ta1rZy29TtTHhlNbkG2L4K2KROGkkr2b7D9k+B9SStDNzm1GoGcJOkbZnyIHbrgOvbyHwJ8N9Mdxc8DziiiswSHybivCwD3Eu4OO7HjyTdREwKT7f9txqyngB8BFgN2C2V+QFCrWxUMgvucbLZcdj0LDaKdAqPNF8Ani7pIuI3LgH8oYqwJm3W9mXAdpLeafvwOmmbymyZ9gk0awdtZELYCu1eJ0HLum3bbhv1J4na7b1F39fqd7a4n43rp02bTzTtT1ax/co6giZV1jGOn8MYy9o8K43S2v56+bOkT41B5tiflQmlHWvfl6j7bE5qXjLJcWXiZAcKNZD0XabbH9wHXEUYG/5jQNoLgXJDq1zxSX9zN8IjyOXA523fWSFdG5mftL1P1euHgaSnESoCG5CCibqC04VS+toTyyHIPIoYaIsO4L9sv2mE6b5f5brS9Y3bbCmPbzsFLUw6xAfZ3mEUModU3lrtoK1MSd8EdrV9R1WZpbS16raUrnG7bdqfpLS1223Lfqj27xzC/axdP8Notymfpv3CQcRp43HAUbb/PdPKOsHxs/FY1vJZaZy2lMdShMH9m0chcxLPyiTTjrvvS+kqP5ulNJOal0xkXJkJ5MVQDSSt1vHV4kQj397VvE61auBNaCpT0vLA5pSM51zNI9cwJrMCNk7lfowrRk5uOrFsKXMSnfKjmX5ferpdbdtmUx5nl+uj8/MwZQ6pvHUXb22f658RaiZXEydRtv2SimWtVbdd0jdqt01p0d7bDu6Vf+cw2lBdhiWzZb+wOPAa4L+Av9n+2Ewq66TGz6Zj2aSQdDExhhYnWV+0fcKIZI39WZmtjOPZ7JJurPOStnJnM1lNrga2/9Tl6+skvb9i+iuA3UoN7TOSejY0SXNZ0BOWU14bjUJmiRMJw7vnADcCVY+Ed+v4XHSu36GaNzmA9YA3EbZGdQaBxxZvbM+T9PQaaRvJTLs8tVSk2qSTdCihDrZi+urPRNTyXnJatdnEXZLWsv1bSU9mymvb0GUOqby12sEQnusX1ihbJ7XqtguV2+2Q+pOm7b1pP1RQ+Xc2vZ9t6mdI7bZx/SYeTniPWoUIzNxLxkTKOu7xs0TtsaxNW2j7nNnu9H43kKYyJ/GsTCLtJPu+RKVnswtjnZc0kTuMup0J5MVQDSSdThx1fsf2vcX3ruiDvUTVhrZZx+eVCZfVf64pr47Mggdtf07SPrY/KemHVYS06FzXJDznvIow/jvMoR9eh1oTyzYyJ9gpr+GIZbMPsAcD9N+H1GY/BBwiaQXi5KOv/VcbmUMqb9120EimpIcRNkqPIRyEHFejjAW16jbJbdpuG/cnQxzw6gyyjX5nizbUpn5atdu29SvpFKL9HAG8xvb9M62sExg/C5qMZW3G3kZpJX0W2NvhfnkJ4EjC/nEP22ePSObYn5UJpZ1Y31fn2UzXj31e0lLuMOepk8MzwIvDbHkRi8e3EHqY+xBRsKumXRPYG/g1EXPnRTVlv4VwlbjuOGQSg8wywMHAC6nuket0IvLwUjV/36nAVnXTdeTxhJTPRcD5wNqjkkl4KSq/HksMKl8aRbpS+jPS3y8TO5y/GVWb7ZLXnIrXtXlOWpe3QTtoJBM4nhg8nkY4t9hx1HXbtt2W8qjVn7Rpt037oaa/c0htqG79tJI5hH5hbM9Yi75vIuMnDceypm2haVrgwtL7PYFXpHL/ZIQyx/6sTDrtOPu+lL7u8zX2eUlbucO4n5N+TbwAs/UFPJ8w5jyqYqfTdGB/FPBD4ADgITXTtnmolksP01OA/wE2rpiu6cSy0SKqR15VJ+1DkTmuTjmleR6wEmEU+VPCcH8kbbaU7hWEu9XLgYcAHxy1zLZp67SDpjLpcB1KzUlL07pt027b9CelPOpOJpr2fa2fzwb99DDqp227rTOZvZhY9JdfFwLnzbSytrwvkxjL2oy9tdMW/QfwCOC40vdzR13ehvdkrPXTNu24+76mz+Yw+r26ZW0rdxh1O+nXxAswW1+poT2C0AH9DHDygOtrNzRgu/RA1TpFaiNzyHVUZ2I5jB2qWhPLtjIn2Sk3vB+PBh5ftc2W0p1PuPGemz6f1kD2I1NbriSzI23d8rZZvD2T2I0dKBP4GVPxjNYof64hr3bdNm23Q+hPmrb3pqfFwzzVrHI/W9VPl/xqtfk2/QKwL/D89H5D6m2SvKLBMzaUPqyq3KZtqEW5GreFpmmJmEKfJxZ+z0jfLQH8cpTlbXJPJlE/Let2In1fSlvr2Wzb77XopycyrsyUV/Ym15Bks/E12zekz4vbntfn+sWBNxKRgS8DDrE9SJ/3QeBXhEe24kaJit6qmsgspS17tFkN+IftOg4JynmtArwfeIGreQ16PvBB4uH8mu0LK8o5n+hszrC9gaTTXNG/f12ZkrYD3gN8xDVsm5qmK6WfS4dOsCsYoUt6PKHOtYzt49N3iwOPs/3HAWnPc9gpnW17Q0kXVWl/pfRL2b43lX2Tfs/JgHz6PmOl69q0g08S+vPzn2tgNdvXdbn2Wz2yse13VpTXtm4rt9s2/UmbdtumHyrl0ahP6FaWXm2obX/bI89KbX4I/cKZtjcpfT7Lg20EVrD9D5U8GKZ7tbzt20dV1h55Dn38LKWtPZa1fFYapU2OIV4J/NHhLAJJjyQWRn1thkbUdl9t++Rhy5tQ3U6k70vpaz+bpWvHMi9pI3cUbW8S5MVQRSS9iggkWnSqiwEPsuDEtIqR+FAG9jq0kZmMOXezvdeA655DdOZd/51eD7pizIcGi6hWE8s6MifRKaf0S5U+rg68yva+Pa49CVge+Bqxe3MA4V72c8D9wJbEpGKQo4B3A5sSwQvPAW7s1xYkHWt7q9Ln4n7Mtb1Bn3TrE648y/UCCz5jfRcaddqBpMcC3yee5TcC7yQ8H14A/I1wyXu47c37yWxK3brtk0+tZ6VB/kMZ8Nr2fYN+p6SXAW+FBYyIp2H7PXXkVijXz4A7iXYEYTC9DXB8xc2Ktv3CqcDnbP9U0lrAN22/aECars+lBrvOr13WAWPDfKqMDeMYyyZBL2cGI5ZZjMu2bUnPt31p+l8tN/8LK5N4NrvkMdJ5SVu5CwPZm1xFbJ8CnDKkvC4los+vArxf0id6DOxD6xyryuyR9n6F56xB3ELoqLZC0la2j7V9K7Bn2hWswlGSjgMeK+kbwGmjkmm7qqvxoaQrpS+3g8slbdnn8qWBHYlYAyp9fwNxjD6n4/teMr8p6Vxiwn617d8MSLLKoDx78HPgGnpPZAeWNVGnHbwf2Dnl/T6gqN8rqVhHaaK3L6FedxOws7t7VVyABnXbKbtyu23Tn7Rtt6V8GvVDNX7nL4E/0aANtexv7+62qZAmmwMZQv2+C9hX0oHAHcC7W+TVt8wNy1oeG9YjNhPPa5DPWMayNm2hRdpXExsyP5BU9/Srqcw/AFcQoRpeDOxHnKrDiJ6VSaSdcN/X+Nkc17ykjdxJLOJHQV4M1UDSGwkjzAtsn98in6oNrXHn2EJmcf3nmZpQrErsNPTF9l+AvzQpXwevA44t5VtJrarlxLKWzEl0yil9ecK1KhE7ox/u+Nvr/91kPdP25SWZdwKPlLQh8Afb11fM89GS3kbYUPTj8cD/MnX6+lRioFYpz0uAT/XLpGY7WJOIxzAH2InYVVsgywHl/hqwte3rJL0A+CZx2tOTFnXbSZ12O4nJVmc+TTc6qv7OVYGDmGpDjwZu7rjmIuC/u6Rt09/Wfr7KtK1f2zcTJ2JFfm3ie/Qtc5OylscGhRv5ObbPlbQMsGKdcW0cYxnt2kKjtKlNfxf4bjr92iv9tiqnX03Le63tzSWdmT6XF0D92sHY66dl2on1fS2fzbHMS1rKHdo8dZLkxVA9XkIYZm8qaXfgo7Z/1yCfSg2tZefYSGaJ04tLgdur/k5JixHxb15EqIwsDvzAfSJ+S3olcYJRTGBWTkfL8y+JIndXVWoysWwrkwl0yokiIJ+B2wm1rlHxHMIJQWcQwMWIHdmrbfddmCRMqA0NUl26EtgEQNIc4BiX1O0G0XCBUfW0qR//crIpsn1JxcOARnXbpt1OaLLVSaV+qOnvtH0VJXUsSd+1/ZYqBRtWfyvpq4RtioiA01VoXb/ppGQ7Isr9pcBZddLXoG1Zf0ccmhWxdL5GvZgkIx/L2rSFIbWjFWxXPv1qIXOBzbK0cTXodHAi9dM07aT7vqrP5qTmJRMcV2YM2WaoBpK+BOxl+w5JjyHiHew36IZ3a2jAreVL6N/Ay3lV1R1tJVPSF2zvVvr8OdufrlC+zwBX2P5++rwYsfN+te3vDUrfkdfKwD2OaMr9rnur7aOS7DKLAU9PsqtM2ivL7EjT1OFD7XSS3lyuR0lbODlE6HLtmcDHiIn3NkzZDL2LUAE7BzjY9YMeFvmfa/vlXb6fpmuuijZD6drHEzGUAO4ibJ5uA3awfdeAtLXbgaS9gZOIe/AK4D/A0cChRB2dABza7TmRtGR6uyOhVncmsA6wke3P9ivrIHrVbZ/ra7fblK62TniddjvMvi/lN/B3SvoooRp6CfAwV7RR7MhjU9tn1Ohvu9pXVGnzXdLUNZp+BeFp7aGEh9OXekBgx3KZO8vY67cMo6yldJsAHwW+aPsnA66dyFjWJZ/G9hM12lERIPq0cr+sis5u6sqUdIbtTUt/5xIn2wK2r9oOOuStA2xp+55hlnXYaUfd96XrGz2bXfIZ27ykrdyUbvbZGnkGuLSbLS8iqu5Kpc8PI3adVmmY38rAsjWu37T0fvFRyCRcA7+XmNy9J712JBY4VfI/ust3DwcOGpBuW+ChHd+9DXj5EO7bueOQSU33tHXSEbEy1iRiC62e7tOz6BNAkJjMP5PYibqaUN16A3Ai0UGfO0Bm3zgJ5fbYWd/AGel1JnBs+v7sCnVxJPCYju/WAz49inZAqM2cn8r8aOCTxK7+SYRu91nEyWa3/OYCZ6e/5fdVfmejuh1WuwUeV+63mvQnTdt7Sjuw72v6O4lNqkcTi9uTgV0alO/s0nsNqh96xIKp0hba1C9hY/clYqIM8MMBeb42PZNnAt8rykiosJ5GuHU+fERl3ZcYL28hToKe2bRuqrQhWo5lpXw27fjcdOzdtFda4MT0d2tgI6aCa19PLOpXK76rIGc54MlVyluSU/wtt/uubRfYmOgnF3il/z+fCv018Jou3zWq26Zp+92TAemG/myW0rXq34EN65Z1GHKHdU8m9cpqcvW4lvDCBYDtf0naDdiN2OXqiaRtCc9Cd5e+fiVh7Htun3SL2S48FO1GTDBxNRfDTWTeSeyOO/0VcDdx/FqFSyTtTEQ8nidpWcJ72UkD0q1DDMTlst5DdOx90ZTr1DIGHrD9MiJy+VBldsP17SDqpHsC8BFiUPwEcV8eAHrucNr+r/T2+TB/l8q2X5e+P2xAuV6c0u1LBAG8NKmcvSD9/4we6XqdaFTRH5tH7KKVWY4pxwY9adIObN8EvLSUR7HL/Nr01Td7yXPaTZf0EtsXDSpfR9pGdZuo3W4lvcjJ1aqk1xDxS24j7bBX6U9S2tfZPjGlqWqz0ajvo93zeaftHwM/lvQ2SUfZfmuviyWdRjxPD3Z8vznRbv4lacv0m7uxRa+sK5S1KxXrdzfi1Hcv9Xb1Xs7zJBbsi2X7r1Tw9taN4hRtUFlt71pK8wJgD0mn2j6ygoyxj2VpZ30ZSuNu+h3zJD3J9h96pJtj+4HS59UcDlV26/NcF31eZ3u5hrABWbzL/7rJXoHY4HkhcF2aP/R7tos8nyTpR1RT6/wjU/OgTxKL3AdK//8IoQ0yiA8BPyp/0a+sCm+jnR7RLiH1memaP9k+uoLsgn73BE15EZ7/FdFHaOqSOFXrljc1ns0Sbecl7yI2OIDp/Yikp9v+/bDkaroH2LJt7/z3knDFUBOTJC+GamD7K12++7Wkyyskr9XQJK1EqOgsJukY24fE1/NdYVJaJA1FZsrzr8CRkv5o+7xUFhGeyQZi+yCFr/vjkorcPcARaWLSj24d/TwqtNEWE8tGMlXBTWzqAPbp+G4NwpZqULpvd35v+zLCg9I7bR/eL48+/IaYBNflucVExvbZkj5FDIB16bUoLbMrcEAa1AsX9lcTA0tfWi4wCk4gXGrXYS9iN7cJTeq2Sbv9nKT/Smk/RNhpSOF+vZho3WO716S+YCfiZHE+FRZSTQf3pn3C1ZQmZ7a/LeluSbvb7uY4ASI44TRVsfR2Z2Bd4nTyg8QO6wLY/nuPfAe2+ab9SZL7A8I+4HGE/eDqkvYg2v+Vg2RXLeMApm3SSXqU7Vv6JbB9CbCVpL0lvdf21wfIGPtYBqxP8kQpqTgtfpjtJxFqtL1UyE4j7IoPtP1RIvD4hvRfzFSxV6hyzfHADlFkvZToSx8A3pg2fjop2t2zupSva3kd9pHXEULeBZxj+0EFewAXu4c6n6TXEV47DTygsEv5AdHn/TzJ/K3tXbokL7xEHgB8PP39PuEBb6eU9h895O5L2NJcm677V+rrBtlGTfMiLOmbRBDvKg6lmj6bTeclRxEnvHMUKvImnpktgVVtP1HSQfR27tNE7k8JJ0fFYmhPwgnSbf3KOhPJi6EhUHFXtW5D+wChWna8pG+XdhZOSXlZ0pG2j+2RvonMMvsQE4CCH1Bx59D2t4gBoC0PUK+NDmPSXkVmUxfi86hwwjGA9YHDYb4e+UG2d6iScNAEpQ/3FycLijgJyzTJxANsA9I1twHvaJJ/icbtwM0covxO0ncIVbd5KZ+eDkM6GErdMrjdioiZtBVxuliwrKcCbp7ZNeF0m4170wTmt8Qu7b/SZZe5ZJfRRXYnVfuhTgY+n7a/1OW7/0sbTD2TdXwuPCDea/tuRSDf99YtbJU2zxBCEti+kdj9/RyhHvjfxCSoStoqZQRA0v8RpxgCnmj7qSx4f7/NAG+KJdmfSguNgaK7fDeysUzSQ4Hn2D4gLaDmAMcwtelR5cRv7RrXDouyrI8RKm3PJeYTC9jNFnOXjhM3FE5oei6SJe1PLIAPLxZChGOM37qP05t0qnxi5/eStvUAu8G0wPqjpH84PBH+w/a16W9fN+22d5W0TnGCU9rs6IukpwL/U/4KOFnTneRc4j42yW2ezRJV+r2yxzql63ckFoqfLZV/mHJXY7r3zlWAA5l+UngpfTRYZgp5MTRZ+jW0tYlGBTFQrk4NQ+OGMsvM3/mw7TQ4jJIrgWMldR6z7lkjj7oTy06ZBX1luqEL8aRW0VW1ogaPLeU3T1LfSOpDYpgxTMbBsBYYVflFi7RN6rbJs2LbxxEntsWiZzPgUeVreiQ8jVKsJiXVXYXhfdP+qEo/NIw+YT6276hzOVHG8uRhKDE8FhDUIiSBOlzp2jZhl3N6/5SNebin1EOLdtTZbmpNuFwxJlcXRjKWSXoJcSL4KUUct0EhAQZR5URn0LVV83gU4W3vVsJ9+b8UAYF36pUgTZxX9nQV0N3c33nC2oSDm7UkXZE22p4uaRdJu9jer4+8twA/sV0+gR/4+xQOa/YhTuVgSn2tat3UuQ+RwL5SoZlzREdZngc81X1U8lo8m436PYUq6anASsTp58uZ+s1Vfnsht/xdX7m2r5J0MGFzdn8qx1LE4vv3FTSCZgx5MTQ+2gzsg9ThqsgsWnhVmddJei/hwvSVxERtZDhUEL8C8zvLH9r+T81sak0sO2Q+lDCUPaGKIDVwIZ7SbQ9sQBjaf7dzR64Cd0lay/ZvJT2Z2LEcKR5uDJNxMNbFm+0jkyrEY21fXDNt7bott9symh6DaoF/d/luoLvzUt5lm5sDJe1TNS0N+75ev3NAOddnSoe9LGvaXw/WYb/F9tGS3ivpIYSq3GV1ylKHpv0J44/x4S7vHyXpA8CfklrRAu1iCPel9kStRN2x7EFCnXgp4vSp9iS6Abcp7NaWA8oqYv8mFnOH0kMFrINiEd9J14W8pGcRtqPXS1oOeIPDc9igBe3fbH9W0tOAwyS9xfY/be8n6UhJa9i+ukfaPYHNJV3sUP+vyhKEV9AHJW1GOAGqQ9MTutdL6pyDPY5Q0+xH03hTvfr3jQckfRKxCPoo8EXiBLeSLWg/uRV4NfDc1E8+hQj5cZztuQ3ymhh5MTQmGjS0XxGqBicQA/HRJGM9pgaQb9n+7hBllvkQsTt2FHAVo41n08mLCS9QtRZDLSftDwFeRtR3FfYgvBLtmWQtBuyuDtfXXXgKcAjwDOAnkj5dR02FuC+HKGxqHkifR47GF8OkNeNevEn6JLAWsGbaVT7I9vtqpB9W3U4z9O4iZyvgTURwWwiPYp0Gyb1YNeWxA3C97VulagGVbH9F0rcJr0q9jHd7lXld4Fc1Ng1+Thid95vAdit3L5uffYELCHXAN1UsQxMa9SeeGTE+7iW8nhU7/d3qt+l9AcY7lqUT5a0IW9c3Sup5qjIsbG9d/lw8W55ygnNExaz+SoydLyYWDcsRwad7qf/uBWxh+0ZJWxM7+j1PdcpFTOW7QtJnCfud96T/HQi8mXCc1I2bbL9V0vskbW+7ryOfDv5sexdJLyfUs3ak+iKn6aL2aBbckP47A7Q8RvBs7gJUmSuIcHCxJBUXQ+mk61WDrrPd656enuQ9nnjeVpJ0v+0LqsifCeTF0MzlEEKl5WPAUUlv3a4YE0ah797JFYRXssKA9Bb39gj2H6KjHDma8thS7BAuBhxfjAdM7RoO/O1VJ5YdMinkasrwcJDMNYqJC3Hhg5K+QkxqBsVTut72RZKOIRY2K/db1JZxBA5tqypZGQ0pTsK4GfPibVPb6ye1sf9IelLFMjaND3MY8BjCbfrlhLe8VzN4UnAS4b2prLP/T4UNkAjvW714lqRrgRNsfzx9V2dysTJh7/b79BuWtf3vNFlbzfZve6R7HeG96m5JyxDG4d8kDOp/nFRPyjyeMOAtnuGnEv1e2dPRJXTYT9h+Q0c+xWTvFIXHs1GfDrTpT4o0lxJOVioF6hwi/7BdDtDYra4a3Reg8DzXaeNzDrG4Pjpd0zN2UJOxzPZdki5JJyfFact/iJP505iylevGsumZWiNd+5A6shOfr3phKuMXUzlvKH4CsD8xeb6P3gv5hzhsWiD6hoGe/RLzY9s5Ak1vL2lF23+3/RtJVwzKwPbXJB0iqbCDrkLxXJ4r6emSNqCCSqjCgcLjS5vJdTRdXmN72nxK0nOpvpFU69mUdBFxCrgkoaJ5PXE/n11FFFPP02JE33kncHlqiz/vke7P9B4jlyYWV/0WVq8k7IVWJ5xafIdw458XQ5kpVMFbEDDNW5DDO1HnsWidY97CUH8PpgaCeYQXli8ytXvQFUk7El5figmSbVd++OvgDo8tHeVY1hUCftWdWBYyVXI5XJOmLsTLZbhL0jsJ9Y9f2x44gKSTh48QesHFwFA5MF4dJF1D7DJ+LO0c/nCmL4QmtHhzuv9OO38PG5SgTd3a3j7lMdf2KzVlDNxvwv4ppyCICo9CD4ms/PoqMgkVsc2Bj0l6t+0iMGNfJL2diF+RPur1xA79W4mT7w2I+vptR7rtmTIyPkyhYvdm4mT1KMIz4nPomDQ6vDRtkvKYAxzjPgbdfZhvPD6GhRAMoT8pcEMX/xW5rbR4LuwXrQiWXOzwLzBpanlfbmJB99l/IozCC5uN9ehBi7HsbEJtTQ67pi8zFRC6J8MYJ21XMvBP115GqlsASc9I31+Qxrd+avaN7L0c9odlPlCW4/7e1sqqxPsS9kxVnDPdR9jEFHwD2N/hsa8vDoc6uw66rgdKbb7MEsSpTy0qPpv32t5cEfD0ANtvT4Xo6uCmxHVMPYNfI2yVfpg+f3JAuf6qMBW42fa9SfaKtq+W9H7gUocHyF6cTqwnVifcun+CMW2mD4u8GBoPrb0FJb5Y9UInL3MKt6XzGHMc4QAAGAxJREFUPc5J+qft71fIYntg7QGd2tCQtBewb3nhI2l3Ygenr15xy0n7jsBP087RWsDNwA4eYK/k5i7Ef0UYnhb5PCDpI8SA8PGeqab4OhEIrXBnOUqaxkmYCBNcvH2CGAzWIHbXqrgqHkbdFve/8Hz2qJ4X2j8vvT9F0mOp59lQ6ZnYW9KnJK3tkivqPnKPpMtus6S3pknxuwi9+s50h1GKg5VOkE4kXOMuRQy0C7igT9c+nqlJ612K+Cm3Ec/1Xd3SdJFfR3W1NR39yRzChXTf/qTKJpt6uORuUc5uJwyyfQOlCXmPsjS6Lw6115u75FeeuPebxDcay9JuPpLauh4f6J1tyNxIsmcZsBCCcDbzmFTHmxOukmtTQU752k+W3v9J0j6Dxtt07f1EIPGCNxEOFdow8J64T3yyXqi72tm1xCbdY9I117tLGA2m+vW7gWW7fN+rnEcztTnQhO0Idcw/EGPZs4kwBXfT3wnRgcC1ti1paYcN7ZPpbrs2Y8mLoTHgFt6COvgoMdkDQNJj3cMgL+3Irg88TdLhUQzXMTi8dlwLocQjSEaektYk4nlc6moGlm0mllIEobzB4X5zY0Lndf9BCZ1ciCs8Dp1csUOfvxCV9IaU7kaqLYQAbrT9q4rXtsLDiWEyTsa6eJO0ku070sniemk37bYqJwlDrttCjaenXEkvI05jymqhaEFV1Pd0S9+x8NmfGo47JJUNwu93uL5ehTBo/0qv5yZNID9ATIT3AF5P2CaYsE/odZK6FxELZP4EWtJ6hKvhXjrvE8f1QxIMa5OtLb1iN3XS+L5Iuhn4NdFO13TE+inT75mrNZYp3EY/vPhInHz9iqm2B/A72wM3J0tj9CDvbEPD9j+Bf1a8fA/gREl/BFYACnXRkWy0pbH2tUzvh26WtB/RBm6wPfD0LbEW7V3Sj2rT42YWPM28AziYKTfTgxxi3EMFLYMh0rmh8EJJ/ybsz3qG5bB9jSKgq4ln+SxHLKrIVHqie8SdmknkxdCYUE1vQemk5FXEA7Oq7TVZsLF+m96B39Yjdk12T+n6xSPqxt2SziVipxim7+iMiB8oYk6sCrzFES19IE0mlpqyGZpHeKj5fsrrJwpvdnXYEPgxNR0+EDErzqTeDv3laafyPKbuS5Vgoo3xcOIkjJwJLN5+JOkmYgJ7uqe7iq3EkOq28Hz2rj7X/JKpoIW96Lq7nnY5X9Hl+0OI3cQbbf9fn3zfRGzkiDDO/hIx2fw3PRY0kp5AqNe9hVC9OJAIenoboR61DdBL5jxiB7bMcrSP8TUS1NC+c4ibbJUp9ZvzvyIWC+UYU3Z3W8s29+UqJzfu6u3Suxd1x7IDiPF5X6Kt3U1MZD9B2G9tRahqLrAYUjhb2IJQMXoWU2P0yGINqUIwb4BupxCOoPEvBFbx9Dh0oyrveYSKq4g63JZoF/sT7vtfKulN3bRX1N22+AXq8FLZ2fZ6nNIsgHs7B6iN7b8obIEvS2V4GqHu+aDtzkVSrzweVHhomxRlW7muwdolnWj7dYS2SnkBdD3RV29JaLNUsnWfJHkxND5qeQuyvYekeUTn+9UeeQ7qsG5y8sIkqe5E4NDBlwwVA1vavlNhnPhxSc+wXdmQtM7E0lM2Q98lJorbEicKbyJ2IHvSo1Me6PChR7ofDErXQaFOUgx+po8HsTZo/DFMhsK4Fm+2102D3NuBz0o6i/DweM2gtG3qVtIRxIbBmgqD4MLFa7/J4apMBcdbkthxvJ3pBuy9guP9hZjAvJioz+IUc09CjeJVku5Jz1Q3/u4UFFFSYbdxJRFc92DCyUUnLyQcx9wDXJbUvf5NTEiRdDExyezm7npX4ACFx8XiObuaODmciTS276y7ydYWd9h3SvomcdpT5eSlzX3p5tK7UA8V/cfCumPZewjj+DUJFe15aSF2i6T7HPYVveQ9hFC9LjQLyiEtRkWrYN5Jxa1z539UKn1PI7Q+IOz/vkY4eXii7R9JuhT4AmlzsqOcPW2LB9DPOcAoOQjYUBEfaX+mu5bvR7ltLTmKgvVhC0m3EhtQP0+bbP+mtyZAsbnR+TxcQ4xvc7r8b0aSF0Pjo4m3oM4HZ46kQo3lO13+38njJd1NNMZih+E3VQpbdfdiiDyYXtj+JbC1pPdL2tt9ojtD+0m77TMlPVNhJPlr4p70u75Rp9yiMy/nUSWuxrAYdwyTVkxi8eZwerFbmhxtDHxGoYM/SCWmcd3afkePf/VzTXyVwpD8D4Qnr9fb3hti19J9nHfYvjkNiCsA6xR9g6Tdbe+ksAX5LL3bdtd+yvZNks6StHEXdZXrCBuGMyUtDSwj6YnAa9L/VyMWc93yvY1YaM0K3M6+s6mL/0ZIeirwP+WvgJM71gaXdOuzW96XR6aNKhEq1RBtrqDnrn7dscz2Xgq7rScQrpzrLDQ623rnwq3gGIfr5dZ4OMG8O/McifqYQ6V4vp2bwvPmDUxpuNzKlIriAkh6I7GIusD2+RVl/lXSPNvT+ou0SNnK9lH1fkVlHqoIDbAL8Hnbt/deQ09jWjspvR/1ouI4QvUQYrPqZ6X/9Qo43dnel+xy0j0OBzStyYuh8dHEW1Bn4zdTE4BBBovnMD3q9G8APOUWty9p57U4rViNcJ/69Cppm2B7hy7ffVXhinIQbSbtn0uyvkSo71RGYfD8euI5ugf4pitEXG6aLqWdS0fnUmHi3QjPjBgmdZjk4m09Qh3syVSIVTWiuh20m/sRYhf+FmCupEcRKkEn0dv+puA84lRyeUmXAJsx1QfdRLjO7sUjNOWBrNCBL058TwKe15nA9qWSXqUpV/efINTqrkqXXEXsPs561M6+s7VL7jrYvlLSMbaPKH+fVJGe6uTqegTsx5TK4IGpLJVUv+uOZZIeQ3gqu5y4J19wzWDKXTBxgjOSkyK1D+Y9dlL/czDhJbI4eXgkoV7Vi5cQv3FThYOlj9ruFUOpzOclfdgRomQX2/vZvk/ShoS63ihYlQgFsDTRlqDCgsZTrs6xXd54qKwl04Sk0tftpP0MBs83y8wqxwkFeTE0JtzAWxBM8xkPoW/6g+JDv10Gt/QsYvvFJTlLMCEVE4crykHXNJ5Y9tsR70fSDV8aeGPqVFcA9pf0oO2eLjCbpiuxWen96lTQhR4GnlwMk8qMe/GmcPTxNuIeXAwc5gZu2odVtxV2c0Woac4hgv7eBuxl+6q+qYK/OzlRkPRDYuwoJjCr0DtoKbaf2+W7s9PfG5iKjdJ5TbdT0EqbBrOMNvadQ3PJXYPXS+qcHD2OKXXNoePwStg0bd2x7EPAjrb/kE4PDpN0OeHM4wma7hCkKn+1fczgyxrTNpj3WJC0IrA1cer2VODdtv8p6VqFQ6F1Gdx2L7B9Ulq0flXSfhX69/Jk6Zml96OcA19r+y1pnDhCEci38aLaNdytDxOnkAyDLkt/70uqde/o+H7GkxdDY8Q1vAVJ2oeYZG1EcsVIGKouyZQ6ylhW4LbvlzROryaNGeOk/cW231yS+4+0U7Uz4RRh2OmK68sqG5crPNmNHElb2T7Wo41hMjTG1A6+RLiN/lxNVZppjLluf5H+3kvssj6DqdOWfnTabBi4WhG1fh0iYGOmOY3sOxtusrXlaBbcKf47Q1bVGgUVx7L/MOXSeAnit95LuJ0uNjt6qd7eQ5wAPaAIcjk/tlHjQlfnejcM5j1G7iNcTN9HOC5anThZ3pUIuHu17eN7J58iqe9uS7T9tw3YOO1V/3VOPOpyMcxXUf4S8AqP3gnVuLkttfMVCM2Dgn8T9/hQBnvNmxHM6AnNwkIXHUoY4C0oPTSdD44chqp94zkMA0mfZ6oDWZVo2DOeMU4sr9GCXm/exXQ922GmA+a7Gy5YFVi7Umnb8zpKO9bD0ncfFWNqB4sxPQ5EU8ZZt+c7XO+ekuxLPi5pf9s7D0i3osI9d9lm4+OE+trvvWAQxkw92th39txkk/Rkl9zcDonX2J42pimc3owkKHdbGoxlXwYOl3Qfoda5cxp3B9oe2f7f9LYz/tPYjMgdwby3B76nisG8x4UjntSPAZJK6KGS/pZU3fraBieupeRMxPa/FJ4MdyM8VvZCwFbpnj4pbeKIHl7ShkHHwmdlWtoLz0Rsb13+rKSuZLtwinPEuMvUFHkswbUXbdLxKEz3FnQ14Uml8BZ0q+25A/LpZmg8EiS9PL01cHtFvdyJk/TZtx58ZWs5ixPetJ5P7ATOAX5o++ujSFdKX3jiKezHvucOw9BhIOmVhFekQtd+ZcK4df4lhG3D5sOWPQzG0Q7SvXwj4ea5so3STKtb9YlXVrqmc6JyYEX1icwAJG1DqMrNx/b7K6Tbjo44JIWNQWqbnyEWqkM9HZB0FAsauS9B2KocPkxZw6DpWCZp8WFtTIx67FY4l/ix7TtL3z0O2MkV7YQngaRlYP4iqU0+fe+VpBcTgUQ7uWHQvGsYKMIQ7Gr7X6OWNUkkbTCO+hwFeTE0RiSd7ZKxu6QzbY/8lKcOabd4G2IH7WyXItfPRGbaxHKUKGIwPRr4bdvBo0UZVgbucbg5njFMuh0kG6UPEqftjWyUZmrdZmYmkjYgXNvuzJQr56sI9c11gF3a2NrMdmbbWDZMJG3qEcefy/RG3cNoFCp5VcNoZMZIXgyNAU15C9qMCCxm29vP0MXQ4cD1wCWEoeMpHoGL1lGzsE0sJX2IcCv8K+AFhBedvvGQhiBzW+B4lzwTJZXPP3n8rtcbMep2oHD3vFihilTYKAEv6GejtDDUbWZm0O0UNBmX7020sUXSnmthGcuqImkxR8ygBTZeM5lMf7LN0Hho4y1o3DzF9jsBJP0EOJURuGgdJt0mloTO9p+ooOc9S9gGeInDde4jgMOB145Y5jrE/S/X6z1E1PgZx4TawXOIfvQ6mO/9cE9Ji0t6tu1edh+zqm4zM4+08D6MiO1xIrA9Yej/0nTJCUSg3EVyMcQsHMuaIGkl4l4vlhbGh8TXmh+MtlgkLYpIeg4L2nAtgO19RiC7cRiNzHjJi6Hx0chb0AR4UOF+tDD4XKL47GpRxifBojCxvLsY0GzfpnChO2q6Gf3OY+b2G2NrB5L2BZ5V+vwOwtPUxsAytp8naX+glxrEbKvbzMzjk8QJ8TWSViecWexFqIh+lIjHMy4HKzOR2TiWNeEDwEG2j5f0bUmFM41TSOpYko50xbhMCyG3AIXa8nqEytp5oxaq9mE0MmMkD7zjo7G3oDFjIshW52czFSV6prEoTCyfJ+mi9F7A04vPtsfpxekBZm69jq0d2N51muDYhf008G6g8ChV14PUTK7bzMxjedvXAKQF0Uq2/wEcmbwpHinpzQPyWJiZjWNZE9YmBaIlJv2rs5DYyQ4D238B/gKQFiRzbJ+bnDesWMXpTUNahdHIjJc88I6Hc4CdSp9/AzBDvby82fZfJ12IIbBQTSxtLz8BsVcCx0oqDAtFTCS6BcOcqYykHSTXrJfbviw5T1iXerFEFoa6zUyWyyS9EzgG2Bb4taQnEA4UninpeGZJSIQRsbCMZXVYZNXhKvI7Yu9qCSJG3NeAUS2GWoXRyIyX7EAhMw1JRxMD6CG2L5l0eaog6YOEXm7hpYX0fs8mXr1mIpJOB44DvuMWAT5byN9kph/tj7MdJFmXAHcRC5htiFgX3wP+1/amgxykKOL2XGh7LMGTMwsXyVvahwkboXOB/1mUbUM6mY1jWROS2/vf2T5B0hGE2twpRDyeYpPlW8N2sT5bkbQJoUb6xRG7O28VRiMzXvJiKLMAkh5LuAl+JhFt/LjZOmGTtBYxUMzqSULTmDYt5G1k+yxJzybc9+6ZJvib2z51VHJHiYYYgDIthnYElgfWtn2LIgbU94CvAFsAPxiwGPoGEQfkHklyqTOWtKbtq4ZR1kxmUWVhGst6IWlFYqNsKeAo29/I3uSmk2w8n02oFM4DNrN9+WRLlZlJ5MVQpieSliLUL95A6CJ/3fZtky3VdNKktLMRf8/2Hcnt8ZeBN9i+f8HUs5NhxLSpIONU25tL+jBwNnAA4Rr+y8SEfyciiNwfhy27CRpzAMrU7n5JOGf4MPGcbAhckGStRMSD2qtL2tMJI945hFqLCW9QfwK2tL2dpDNyDIpMZjiUxrItgIuYgWPZMJE01/YGky7HTETSC4CPA6cuynG4MtNZbNIFyMxcbN9r+7BkiHkdcNCky9SFPwM3dbyel1zNHg28Y2FYCElaK+1yYvtS29sBHwM2lnTyiGQuRbjzvqzjX7sQi6KPjEJuQ64n2ugW6e91kp4i6UdEm7h2BGoiDyQ3qR8Gvmz7ONt/sf0+22/qthACsL2Z7U1tb2R7k7ToKdT6Vk1/6zpfyGQyPSjGMiL20jyi/1qY2XvSBZip2L7E9lbAGpLeO+nyZGYGC42BeWbkXGd720kXohPbJwJIWtr2Pen9UsCNwDuBD7FwGKW/H7hN0hdt3wnTY9qMQJ6IE6B9FUGDH8pUXKMHbF8oabcRyG2E7bkAkraxfUrpX68pAlBKumNYAShtf6X0/ipJO1ZNm3YmZfvnwyhLJpOpzC2ELc1CeyoEMEpbmIUF25+StNqky5GZGeTFUKYSM9UANQUePBy4S9LywHts3wBcKWkf4CuSHmX7lokWtD3/JuLnLGD7ZHveiGQW6odF/oWu/Yw7tZh0AErbd9S4/G3AQyV9GnjXQtA2M5lZge0/MOWGOrMIIOnjwDOYcqwz/6+ngvIem06LMosoeTGUme18DNjR9h/SCcDukj4BvIhwY7nLQjLZPAG4nfBeNg4MvAU40vabU1DRU4ggooslT2gzwl4oMdsCUH6EUIk7TFJx4poNODOZTGa4fItwLlGmc0NvlTGVJTNDyYuhzGxnpbTbh+2bJS1NTDJfRbTvhcJzkO2LBl81VGT7Xkk/k/Q8pk/UDyTUD2dSnKxZF4DS9hXJA91+xEJzByIGxknMzIDMmUwmM9tYh/D8+XBiEfS39P1fge3S+7wRtYiTF0OZ2c6lkj5A7P5sBVxh+/fAjpIeDhwq6R2FnU2mMvulv3OBpYnF0YOSTrd9LhHXZCYxmwJQXkNSPbR9afJMd1mOGJ/JZDLDxfZpwGmSXkfMeU8A1rB9xWRLlplJZNfamVmNJBGnFBsQLlMPLMeRkPRk4IaFwaPcJJG0oe2zJ12OXszmAJSSHrEwu/nNZDKZSSPp0UQ4g48Bc22fVPpfjsu0iJMXQ5lMJpPJZDKZhRJJuwLLAC8BDrJ9sqQnAVsSqnOb5bhMizZZTS6TyWQymUwms7DyU2BZQuX7NZJ+AtwJ/CL9/8xJFSwzM8gnQ5lMJpPJZDKZhZ4U5+2Dtt8x6bJkZg55MZTJZDKZTCaTWSSQtHx2qpQpkxdDmUwmk8lkMplMZpFksUkXIJPJZDKZTCaTyWQmQV4MZTKZTCaTyWQymUWSvBjKZDKZzESQNEfSAZLOknSxpM+NQMYRkp467HwzmUwms3CQXWtnMplMZlJsBjxgeyMASUtNuDyZTCaTWcTIJ0OZTCaTmRR/BJ4taWUA2/dK2i6dFP1C0nsAJH1W0n6STpB0vqSNJJ0h6XJJL0vXHCFpT0lnSvqlpM06haV8zpV0nqTnpe++KunCdDK1xBh/eyaTyWRmAHkxlMlkMpmJYPv3wM7AVyXtlU6GTk4nRS8D3l+6/C7brweOTWleAWwHfKB0ze22NwE2AfYuy5K0MbCC7ZcDrwf2lLQi8HTb6wIvsX3/SH5oJpPJZGYsWU0uk8lkMhPD9m+BLdNJzhHALyStAswDHlK69Ofp77XAz2xb0h+BFUrXnJnyvF3SvZJU+t9zgY0knZM+z7H9d0lflHQwcDFw9JB/XiaTyWRmOPlkKJPJZDITQdKjJC2ZPp4PrA1sYnsX4GBg6dLl7vG+zDop39WAeZ4eSO9q4Pu217e9PvCKpBZ3qu0dgFdLWqv1j8pkMpnMrCKfDGUymUxmUjwDOEDSP4EHgB2BD0i6ELgQuLVmfs+RtDXwUGCnjv+dCGwm6QLgX8C3gPOAEyXdBdwGXNP4l2QymUxmVqLpG2eZTCaTycw+JB0BfMH2lZMuSyaTyWRmD1lNLpPJZDKZTCaTySyS5JOhTCaTyWQymUwms0iST4YymUwmk8lkMpnMIkleDGUymUwmk8lkMplFkrwYymQymUwmk8lkMoskeTGUyWQymUwmk8lkFknyYiiTyWQymUwmk8kskuTFUCaTyWQymUwmk1kk+X97BVriaCg5lgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x3beca320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(14,5))\n",
    "text_test.plot(50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "cKB3TBdW2vU2",
    "outputId": "1bafa281-1ce0-43da-a35f-222df4047513"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 0 ns\n",
      "이/Determiner 것/Noun; 적/Suffix 인/Josa; 이/Determiner 거/Noun; 것/Noun\n",
      "은/Josa; 10/Number 점/Noun; 배우/Noun 들/Suffix; 이/Noun 게/Josa; 수/Noun\n",
      "있다/Adjective; 내/Noun 가/Josa; 최고/Noun 의/Josa; 네/Suffix 요/Josa; 이/Noun\n",
      "영화/Noun; 들/Suffix 이/Josa; 끝/Noun 까지/Josa; 때문/Noun 에/Josa; 적/Suffix\n",
      "으로/Josa; 못/VerbPrefix 하다/Verb; 사람/Noun 들/Suffix; 1/Number 점/Noun;\n",
      "영화/Noun 를/Josa\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "text.collocations()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "kq7xQxlU2vU4",
    "outputId": "c123dd5e-f1d2-4087-a2ab-68557ef0b7a0",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 0 ns\n",
      "앱솔/Noun 루트/Noun; 먹이/Noun 고/Josa; 분유/Noun 를/Josa; 유기농/Noun 궁/Noun;\n",
      "조리/Noun 원/Suffix; 잘/VerbPrefix 먹다/Verb; 아기/Noun 가/Josa; 매/Modifier\n",
      "일/Modifier; 때문/Noun 에/Josa; 자다/Verb 먹다/Verb; 아이/Noun 가/Josa; 자다/Verb\n",
      "맞다/Verb; 탈/Noun 없이/Adverb; 것/Noun 같다/Adjective; 먹이다/Verb 보다/Verb;\n",
      "아기/Noun 에게/Josa; 아이/Noun 에게/Josa; 고/Josa 있다/Adjective; 수/Noun\n",
      "있다/Adjective; 이/Josa 없다/Adjective\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "text_test.collocations()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "tFZ43HZp2vU7"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "JRZexc3p2vU9"
   },
   "source": [
    "## 3.Sentiment classification with term-existance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "XVdOq5Fl2vU-"
   },
   "outputs": [],
   "source": [
    "selected_words = [f[0] for f in text.vocab().most_common(2000)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "rQXyfdBg2vVA"
   },
   "outputs": [],
   "source": [
    "def term_exists(doc):\n",
    "    return {'exists({})'.format(word):(word in set(doc)) for word in selected_words}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "fDh8qt_R2vVC"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 26.7 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "train_docs = train_docs[:10000]\n",
    "\n",
    "train_xy = [(term_exists(d), c) for d, c in train_docs]\n",
    "test_xy  = [(term_exists(d), c) for d, c in test_docs ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "YgK9PN5m2vXM"
   },
   "outputs": [],
   "source": [
    "# test_xy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Naïve bayes 적용"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 31.6 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "classifiers = nltk.NaiveBayesClassifier.train(train_xy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "네이버 긍부정 모델의 Accuracy : 0.021841794569067298\n",
      "Most Informative Features\n",
      "         exists(수작/Noun) = True                1 : 0      =     38.0 : 1.0\n",
      "     exists(이딴/Modifier) = True                0 : 1      =     32.1 : 1.0\n",
      "         exists(최악/Noun) = True                0 : 1      =     30.1 : 1.0\n",
      "       exists(♥/Foreign) = True                1 : 0      =     24.5 : 1.0\n",
      "         exists(노잼/Noun) = True                0 : 1      =     22.1 : 1.0\n",
      "         exists(짜증/Noun) = True                0 : 1      =     19.5 : 1.0\n",
      "        exists(쓰레기/Noun) = True                0 : 1      =     19.4 : 1.0\n",
      "         exists(여운/Noun) = True                1 : 0      =     18.9 : 1.0\n",
      "          exists(굿/Noun) = True                1 : 0      =     17.1 : 1.0\n",
      "        exists(발연기/Noun) = True                0 : 1      =     16.9 : 1.0\n",
      "Wall time: 12.5 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "print('네이버 긍부정 모델의 Accuracy : {}'.format(\n",
    "    nltk.classify.accuracy(classifiers, test_xy)))\n",
    "classifiers.show_most_informative_features(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'1'"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "review = \"\"\"나는 정말 재미있었음 정말 재미있게 봤고 액션신도 멋짐 왜 재미없는지는 모르겠다\"\"\"\n",
    "review = tokenize(review)     # 문법 Tag 추가한 객체로 변환\n",
    "review = term_exists(review)  # 기준 용어들이 포함여부 판단\n",
    "classifiers.classify(review)  # 분류모델로 평가"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Wall time: 6.11 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "classifiers = nltk.NaiveBayesClassifier.train(test_xy)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "네이버 긍부정 모델의 Accuracy : 0.1693\n",
      "Most Informative Features\n",
      "exists(ㅜ/KoreanParticle) = True                1 : 9      =     71.8 : 1.0\n",
      "        exists(어리다/Verb) = True                1 : 9      =     71.8 : 1.0\n",
      "        exists(모르다/Verb) = True                4 : 9      =     49.7 : 1.0\n",
      "          exists(접/Noun) = True                2 : 9      =     43.1 : 1.0\n",
      "         exists(가지/Noun) = True                1 : 9      =     43.1 : 1.0\n",
      "        exists(2/Number) = True                1 : 9      =     43.1 : 1.0\n",
      "          exists(알/Noun) = True                1 : 9      =     43.1 : 1.0\n",
      "      exists(너무나/Adverb) = True                1 : 9      =     43.1 : 1.0\n",
      "         exists(경험/Noun) = True                1 : 9      =     39.9 : 1.0\n",
      "        exists(스럽게/Josa) = True                1 : 8      =     34.3 : 1.0\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "print('네이버 긍부정 모델의 Accuracy : {}'.format(\n",
    "    nltk.classify.accuracy(classifiers, train_xy)))\n",
    "classifiers.show_most_informative_features(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Word2Vec : 유사단어 분류"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import namedtuple\n",
    "\n",
    "TaggedDocument = namedtuple('TaggedDocument', 'words tags')\n",
    "tagged_train_docs = [TaggedDocument(d,[c]) for d,c in train_docs]\n",
    "tagged_test_docs  = [TaggedDocument(d,[c]) for d,c in test_docs ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\khpark\\Anaconda3\\lib\\site-packages\\gensim\\utils.py:1197: UserWarning: detected Windows; aliasing chunkize to chunkize_serial\n",
      "  warnings.warn(\"detected Windows; aliasing chunkize to chunkize_serial\")\n"
     ]
    }
   ],
   "source": [
    "from gensim.models import doc2vec\n",
    "\n",
    "# 사전 만들기\n",
    "doc_vectorizer = doc2vec.Doc2Vec(vector_size=300, alpha=0.025, min_alpha=0.025, seed=1234)\n",
    "doc_vectorizer.build_vocab(tagged_train_docs)\n",
    "\n",
    "# Train document vectors\n",
    "for epoch in range(10):\n",
    "    doc_vectorizer.train(tagged_train_docs, \n",
    "                         total_examples = doc_vectorizer.corpus_count, \n",
    "                         epochs = doc_vectorizer.epochs)\n",
    "    doc_vectorizer.alpha -= 0.002\n",
    "    doc_vectorizer.min_alpha = doc_vectorizer.alpha"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('중학교/Noun', 0.5307413935661316),\n",
      " ('심심하다/Adjective', 0.45919501781463623),\n",
      " ('초등학교/Noun', 0.4558684527873993),\n",
      " ('어이없다/Adjective', 0.44689100980758667),\n",
      " ('학교/Noun', 0.42081159353256226),\n",
      " ('그때/Noun', 0.41403624415397644),\n",
      " ('새록새록/Noun', 0.40217041969299316),\n",
      " ('애틋하다/Adjective', 0.3943653404712677),\n",
      " ('생각나다/Verb', 0.3895311951637268),\n",
      " ('초딩/Noun', 0.38764917850494385)]\n"
     ]
    }
   ],
   "source": [
    "pprint(doc_vectorizer.wv.most_similar('어리다/Verb'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.5307413619916911"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc_vectorizer.wv.similarity('어리다/Verb', '중학교/Noun')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from konlpy.tag import Okt\n",
    "twitter = Okt()\n",
    "\n",
    "# txt 자료 로드\n",
    "f = open('./sample1.txt', 'r')\n",
    "texts_org = f.read()\n",
    "f.close()\n",
    "\n",
    "#단어 추출\n",
    "text = twitter.nouns(texts_org)\n",
    "\n",
    "# 말뭉치 만들기\n",
    "result_nouns = ''\n",
    "for txt in text:\n",
    "    result_nouns += \" \" + txt\n",
    "\n",
    "results, lines = [], result_nouns\n",
    " \n",
    "    \n",
    "for line in lines:\n",
    "    malist = twitter.pos(line, norm=True, stem=True)\n",
    "    result = [ word[0]   for word in malist    # 어미/조사/구두점 제외\n",
    "               if not word[1] in [\"Eomi\", \"Josa\", \"Punctuation\"] ]\n",
    "    rl = (\" \".join(result)).strip()\n",
    "    results.append(rl)\n",
    "\n",
    "from gensim.models import word2vec\n",
    "\n",
    "data  = word2vec.LineSentence(results)\n",
    "model = word2vec.Word2Vec(data, size=200, window=10, hs=1, min_count=3, sg=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "celltoolbar": "Slideshow",
  "colab": {
   "name": "NPS_Word2Vec_v2.ipynb",
   "provenance": [],
   "version": "0.3.2"
  },
  "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"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}