clickbait.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import json\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy\n",
    "import math\n",
    "import seaborn\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from collections import OrderedDict\n",
    "%matplotlib inline\n",
    "\n",
    "plt.style.use('ggplot')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "instances_file_train = \"clickbait17-validation-170630/instances.jsonl\"\n",
    "truth_file_train     = \"clickbait17-validation-170630/truth.jsonl\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# read thr data from files\n",
    "import json\n",
    "instances = []\n",
    "truth = []\n",
    "\n",
    "with open(instances_file_train, \"r\") as inf:\n",
    "    instances = [json.loads(x) for x in inf.readlines()]\n",
    "with open(truth_file_train, \"r\") as inf:\n",
    "    truth = [json.loads(x) for x in inf.readlines()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# compact relevant data into one list of dicts\n",
    "dataset = {}\n",
    "\n",
    "# lists: postText, targetParagraphs, targetCaptions\n",
    "for i in instances:\n",
    "    dataset[i['id']] = {'postText': i['postText'], 'targetTitle': i['targetTitle'],\n",
    "                        'targetDescription': i['targetDescription'], 'targetKeywords': i['targetKeywords'], \n",
    "                        'targetParagraphs': i['targetParagraphs'], 'targetCaptions': i['targetCaptions']}\n",
    "\n",
    "for t in truth:\n",
    "    dataset[t['id']]['truthMean'] = t['truthMean']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# get list of scores and a list of the postTexts\n",
    "cb_id_list = numpy.fromiter(iter(dataset.keys()), dtype=numpy.int64)\n",
    "cb_scores_list = []\n",
    "cb_feat_postText = []\n",
    "\n",
    "for i in cb_id_list:\n",
    "    cb_feat_postText.append(dataset[str(i)]['postText'][0])\n",
    "    cb_scores_list.append(dataset[str(i)]['truthMean'])\n",
    "\n",
    "cb_scores = numpy.asarray(cb_scores_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "showing info https://raw.githubusercontent.com/nltk/nltk_data/gh-pages/index.xml\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from nltk import download\n",
    "download()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# sanitize text with nltk\n",
    "from nltk.stem.wordnet import WordNetLemmatizer\n",
    "from nltk import word_tokenize\n",
    "from nltk.tokenize import TweetTokenizer\n",
    "import string\n",
    "\n",
    "lemmatizer = WordNetLemmatizer()\n",
    "tknzr = TweetTokenizer(strip_handles=True, reduce_len=True)\n",
    "punctuation = string.punctuation\n",
    "\n",
    "def preprocess_tokenize(text):\n",
    "    for token in tokenize(text):\n",
    "        token = preprocess(token)\n",
    "        yield token\n",
    "\n",
    "def tokenize(text):\n",
    "    return [token for token in tknzr.tokenize(text)] # if token not in string.punctuation]\n",
    "\n",
    "def pos_tokenize(text):\n",
    "    token_list = [[token] for token in tknzr.tokenize(text) if token not in string.punctuation]\n",
    "    return token_list\n",
    "            \n",
    "def preprocess(string):\n",
    "    string = string.lower()\n",
    "    string = lemmatizer.lemmatize(string)\n",
    "    if string.endswith(\"'s\"):\n",
    "        string = string[:-2]\n",
    "    return string"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__More possible features:__\n",
    "* avarage text/word length in characters per Tweet\n",
    "* known entitiy count\n",
    "* noun frequency\n",
    "* POS Tag frequency\n",
    "* stop word frequency\n",
    "* punctuation frequency\n",
    "* tweet sentiment\n",
    "* you word list"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__POS TAG NGRAMS__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(19538, 87888)\n"
     ]
    }
   ],
   "source": [
    "# POS Tag frequencies\n",
    "from nltk.tag import pos_tag_sents\n",
    "\n",
    "all_pos_tags = [pos_tag_sents(pos_tokenize(tokens)) for tokens in cb_feat_postText]\n",
    "tag_list = []\n",
    "for tweets in all_pos_tags:\n",
    "    tweet_tokens=\"\"\n",
    "    for elements in tweets:\n",
    "        tweet_tokens += elements[0][1] + \" \"\n",
    "    tag_list.append(tweet_tokens)\n",
    "\n",
    "pos_tag_cv = CountVectorizer(ngram_range=(2, 5))\n",
    "cb_feat_pos_ngrams = pos_tag_cv.fit_transform(tag_list)\n",
    "print(cb_feat_pos_ngrams.shape) # 2459, 2649"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__POS TAG Frequencies__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from collections import OrderedDict\n",
    "tag_frequency_list = []\n",
    "for tweets in all_pos_tags:\n",
    "    tag_freq=OrderedDict({'NN': 0, 'NNS': 0, 'VB': 0, 'IN': 0, 'VBG': 0, 'RB': 0, 'JJ': 0, 'NNP': 0, 'JJS': 0, \n",
    "                          'DT': 0, 'CC': 0, 'VBD': 0, 'PRP$': 0, 'MD': 0, 'PRP': 0, 'CD': 0, 'TO': 0, 'VBN': 0, \n",
    "                          'WP': 0, 'VBZ': 0, 'JJR': 0, 'WRB': 0, 'VBP': 0, 'WDT': 0, 'RBR': 0, 'EX': 0, 'RBS': 0, \n",
    "                          'WP$': 0, ':': 0, 'NNPS': 0, 'UH': 0, 'SYM': 0, 'FW': 0, 'LS': 0})\n",
    "    for elements in tweets:\n",
    "        tag_freq[str(elements[0][1])] += 1\n",
    "    tag_frequency_list.append(list(tag_freq.values()))\n",
    "cb_feat_pos_frequency = numpy.asarray(tag_frequency_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(\"'\", 'O'), ('Jurassic', 'O'), ('Park', 'O'), (\"'\", 'O'), ('in', 'O'), ('8', 'O'), ('-', 'O'), ('bit', 'O'), ('is', 'O'), ('just', 'O'), ('as', 'O'), ('cool', 'O'), ('as', 'O'), ('CGI', 'O'), ('dinosaurs', 'O'), ('.', 'O')]\n"
     ]
    },
    {
     "ename": "LookupError",
     "evalue": "\n**********************************************************************\n  Resource 'corpora/words' not found.  Please use the NLTK\n  Downloader to obtain the resource:  >>> nltk.download()\n  Searched in:\n    - '/home/mike4537/nltk_data'\n    - '/usr/share/nltk_data'\n    - '/usr/local/share/nltk_data'\n    - '/usr/lib/nltk_data'\n    - '/usr/local/lib/nltk_data'\n**********************************************************************",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mLookupError\u001b[0m                               Traceback (most recent call last)",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/corpus/util.py\u001b[0m in \u001b[0;36m__load\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m     79\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mLookupError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 80\u001b[0;31m                 \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mroot\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnltk\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfind\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'{}/{}'\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msubdir\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mzip_name\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     81\u001b[0m                 \u001b[0;32mexcept\u001b[0m \u001b[0mLookupError\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mraise\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/data.py\u001b[0m in \u001b[0;36mfind\u001b[0;34m(resource_name, paths)\u001b[0m\n\u001b[1;32m    652\u001b[0m     \u001b[0mresource_not_found\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m'\\n%s\\n%s\\n%s'\u001b[0m \u001b[0;34m%\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0msep\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmsg\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msep\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 653\u001b[0;31m     \u001b[0;32mraise\u001b[0m \u001b[0mLookupError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mresource_not_found\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    654\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mLookupError\u001b[0m: \n**********************************************************************\n  Resource 'corpora/words.zip/words/' not found.  Please use the\n  NLTK Downloader to obtain the resource:  >>> nltk.download()\n  Searched in:\n    - '/home/mike4537/nltk_data'\n    - '/usr/share/nltk_data'\n    - '/usr/local/share/nltk_data'\n    - '/usr/lib/nltk_data'\n    - '/usr/local/lib/nltk_data'\n**********************************************************************",
      "\nDuring handling of the above exception, another exception occurred:\n",
      "\u001b[0;31mLookupError\u001b[0m                               Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-11-cea2154022c8>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     19\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     20\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mtweet\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mcb_feat_postText\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 21\u001b[0;31m     \u001b[0mchunk_tree\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mne_chunk\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpos_tag\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mx\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mx\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mpreprocess_tokenize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtweet\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     22\u001b[0m     \u001b[0miob\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtree2conlltags\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mchunk_tree\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     23\u001b[0m     \u001b[0;32mfor\u001b[0m \u001b[0mtag\u001b[0m \u001b[0;32min\u001b[0m \u001b[0miob\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/chunk/__init__.py\u001b[0m in \u001b[0;36mne_chunk\u001b[0;34m(tagged_tokens, binary)\u001b[0m\n\u001b[1;32m    175\u001b[0m         \u001b[0mchunker_pickle\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_MULTICLASS_NE_CHUNKER\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    176\u001b[0m     \u001b[0mchunker\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mload\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mchunker_pickle\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 177\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0mchunker\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mparse\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtagged_tokens\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    178\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    179\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mne_chunk_sents\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtagged_sentences\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbinary\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/chunk/named_entity.py\u001b[0m in \u001b[0;36mparse\u001b[0;34m(self, tokens)\u001b[0m\n\u001b[1;32m    120\u001b[0m         \u001b[0mEach\u001b[0m \u001b[0mtoken\u001b[0m \u001b[0mshould\u001b[0m \u001b[0mbe\u001b[0m \u001b[0ma\u001b[0m \u001b[0mpos\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0mtagged\u001b[0m \u001b[0mword\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    121\u001b[0m         \"\"\"\n\u001b[0;32m--> 122\u001b[0;31m         \u001b[0mtagged\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_tagger\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtag\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtokens\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    123\u001b[0m         \u001b[0mtree\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_tagged_to_parse\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtagged\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    124\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mtree\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/tag/sequential.py\u001b[0m in \u001b[0;36mtag\u001b[0;34m(self, tokens)\u001b[0m\n\u001b[1;32m     61\u001b[0m         \u001b[0mtags\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     62\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtokens\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 63\u001b[0;31m             \u001b[0mtags\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mappend\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtag_one\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtokens\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mi\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtags\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     64\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mlist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mzip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtokens\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtags\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     65\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/tag/sequential.py\u001b[0m in \u001b[0;36mtag_one\u001b[0;34m(self, tokens, index, history)\u001b[0m\n\u001b[1;32m     81\u001b[0m         \u001b[0mtag\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     82\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mtagger\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_taggers\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 83\u001b[0;31m             \u001b[0mtag\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtagger\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mchoose_tag\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtokens\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mindex\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhistory\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     84\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mtag\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     85\u001b[0m                 \u001b[0;32mbreak\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/tag/sequential.py\u001b[0m in \u001b[0;36mchoose_tag\u001b[0;34m(self, tokens, index, history)\u001b[0m\n\u001b[1;32m    630\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mchoose_tag\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtokens\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mindex\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhistory\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    631\u001b[0m         \u001b[0;31m# Use our feature detector to get the featureset.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 632\u001b[0;31m         \u001b[0mfeatureset\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfeature_detector\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtokens\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mindex\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhistory\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    633\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    634\u001b[0m         \u001b[0;31m# Use the classifier to pick a tag.  If a cutoff probability\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/tag/sequential.py\u001b[0m in \u001b[0;36mfeature_detector\u001b[0;34m(self, tokens, index, history)\u001b[0m\n\u001b[1;32m    678\u001b[0m         \u001b[0mSee\u001b[0m\u001b[0;31m \u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;31m`\u001b[0m\u001b[0mclassifier\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    679\u001b[0m         \"\"\"\n\u001b[0;32m--> 680\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_feature_detector\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtokens\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mindex\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhistory\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    681\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    682\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mclassifier\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/chunk/named_entity.py\u001b[0m in \u001b[0;36m_feature_detector\u001b[0;34m(self, tokens, index, history)\u001b[0m\n\u001b[1;32m     96\u001b[0m             \u001b[0;34m'pos'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mpos\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     97\u001b[0m             \u001b[0;34m'word'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mword\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 98\u001b[0;31m             \u001b[0;34m'en-wordlist'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mword\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_english_wordlist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     99\u001b[0m             \u001b[0;34m'prevtag'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mprevtag\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    100\u001b[0m             \u001b[0;34m'prevpos'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mprevpos\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/chunk/named_entity.py\u001b[0m in \u001b[0;36m_english_wordlist\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m     47\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mAttributeError\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     48\u001b[0m             \u001b[0;32mfrom\u001b[0m \u001b[0mnltk\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcorpus\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mwords\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 49\u001b[0;31m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_en_wordlist\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mset\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mwords\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mwords\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'en-basic'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     50\u001b[0m             \u001b[0mwl\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_en_wordlist\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     51\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mwl\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/corpus/util.py\u001b[0m in \u001b[0;36m__getattr__\u001b[0;34m(self, attr)\u001b[0m\n\u001b[1;32m    114\u001b[0m             \u001b[0;32mraise\u001b[0m \u001b[0mAttributeError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"LazyCorpusLoader object has no attribute '__bases__'\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    115\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 116\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__load\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    117\u001b[0m         \u001b[0;31m# This looks circular, but its not, since __load() changes our\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    118\u001b[0m         \u001b[0;31m# __class__ to something new:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/corpus/util.py\u001b[0m in \u001b[0;36m__load\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m     79\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mLookupError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     80\u001b[0m                 \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mroot\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnltk\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfind\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'{}/{}'\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msubdir\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mzip_name\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 81\u001b[0;31m                 \u001b[0;32mexcept\u001b[0m \u001b[0mLookupError\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mraise\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     82\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     83\u001b[0m         \u001b[0;31m# Load the corpus.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/corpus/util.py\u001b[0m in \u001b[0;36m__load\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m     76\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     77\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 78\u001b[0;31m                 \u001b[0mroot\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnltk\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfind\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'{}/{}'\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msubdir\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__name\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     79\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mLookupError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     80\u001b[0m                 \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mroot\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnltk\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfind\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'{}/{}'\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msubdir\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mzip_name\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.5/dist-packages/nltk/data.py\u001b[0m in \u001b[0;36mfind\u001b[0;34m(resource_name, paths)\u001b[0m\n\u001b[1;32m    651\u001b[0m     \u001b[0msep\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m'*'\u001b[0m \u001b[0;34m*\u001b[0m \u001b[0;36m70\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    652\u001b[0m     \u001b[0mresource_not_found\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m'\\n%s\\n%s\\n%s'\u001b[0m \u001b[0;34m%\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0msep\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmsg\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0msep\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 653\u001b[0;31m     \u001b[0;32mraise\u001b[0m \u001b[0mLookupError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mresource_not_found\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    654\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    655\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mLookupError\u001b[0m: \n**********************************************************************\n  Resource 'corpora/words' not found.  Please use the NLTK\n  Downloader to obtain the resource:  >>> nltk.download()\n  Searched in:\n    - '/home/mike4537/nltk_data'\n    - '/usr/share/nltk_data'\n    - '/usr/local/share/nltk_data'\n    - '/usr/lib/nltk_data'\n    - '/usr/local/lib/nltk_data'\n**********************************************************************"
     ]
    }
   ],
   "source": [
    "# TODO NER on tweets is not trivial, dont use this feature jet\n",
    "from nltk.tag import StanfordNERTagger\n",
    "\n",
    "st = StanfordNERTagger('/home/mike4537/opt/stanford-ner/classifiers/english.all.3class.distsim.crf.ser.gz',\n",
    "                       '/home/mike4537/opt/stanford-ner/stanford-ner.jar',\n",
    "                       encoding='utf-8')\n",
    "\n",
    "text = 'While in France, Christine Lagarde discussed short-term stimulus efforts in a recent interview with the Wall Street Journal.'\n",
    "\n",
    "tokenized_text = tokenize(cb_feat_postText[0])\n",
    "classified_text = st.tag(tokenized_text)\n",
    "\n",
    "print(classified_text)\n",
    "\n",
    "from nltk import ne_chunk, pos_tag\n",
    "from nltk.chunk import tree2conlltags\n",
    "\n",
    "# how does this work ????\n",
    "\n",
    "for tweet in cb_feat_postText:\n",
    "    chunk_tree = ne_chunk(pos_tag([x for x in preprocess_tokenize(tweet)]))\n",
    "    iob = tree2conlltags(chunk_tree)\n",
    "    for tag in iob:\n",
    "        if tag[2] != 'O':\n",
    "            print(tag)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(19538, 707010)\n"
     ]
    }
   ],
   "source": [
    "# Word 1-grams matrix\n",
    "'''\n",
    "input=u'content', strip_accents=None, ngram_range=(1, 1) -> all ngrams between (inclusive)\n",
    "analyzer=u'word' -> {‘word’, ‘char’, ‘char_wb’}\n",
    "preprocessor=None, tokenizer=None, stop_words=None, token_pattern=u'(?u)\\b\\w\\w+\\b', \n",
    "'''\n",
    "# there is a difference in feature dimension(100) if i use my tokenizer vs. the build in one??????\n",
    "count_vectorizer = CountVectorizer(preprocessor=preprocess, tokenizer=tokenize, ngram_range=(1, 5))\n",
    "cb_feat_textwordonegram = count_vectorizer.fit_transform(cb_feat_postText)\n",
    "print(cb_feat_textwordonegram.shape) # 2459, 8095"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__tweet length__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[18  7  6 ...,  8 14 14]\n"
     ]
    }
   ],
   "source": [
    "cb_feat_wordlength = []\n",
    "for tweet in cb_feat_postText:\n",
    "    cb_feat_wordlength.append(len([x for x in tokenize(tweet)]))\n",
    "cb_feat_wordlength = numpy.asarray(cb_feat_wordlength)\n",
    "print(cb_feat_wordlength)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "scrolled": false
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(19538, 794933)\n"
     ]
    }
   ],
   "source": [
    "from scipy.sparse import hstack\n",
    "cb_feat_all = hstack((cb_feat_textwordonegram, cb_feat_pos_frequency))\n",
    "cb_feat_all = hstack((cb_feat_all, cb_feat_wordlength[:, numpy.newaxis]))\n",
    "cb_feat_all = hstack((cb_feat_all, cb_feat_pos_ngrams))\n",
    "print(cb_feat_all.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "__Learn from the extracted features from here on__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "# split x,y dataset in train/test:\n",
    "x_train, x_test, y_train, y_test = train_test_split(cb_feat_all, cb_scores.T, random_state=42)\n",
    "y_test_random1 = numpy.full((len(y_test)), 0.5)\n",
    "y_test_random2 = numpy.full((len(y_test)), 0.5)\n",
    "# print(y_test_random1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# model evaluation\n",
    "from sklearn.metrics import explained_variance_score\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from sklearn.metrics import median_absolute_error\n",
    "from sklearn.metrics import r2_score\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn import svm \n",
    "\n",
    "# model = RandomForestRegressor()\n",
    "model = svm.SVR()\n",
    "model.fit(x_train, y_train)\n",
    "\n",
    "y_predicted = model.predict(x_test)\n",
    "\n",
    "# NOTE: mean square is bullshit if values are < 1\n",
    "# cross entropy für differenz zwischen verteilungen \n",
    "\n",
    "# score = cross_val_score(model, x_train, y_train, cv=5)\n",
    "# print(\"cross_val \" + str(score.mean()) + \" +- \" + str(score.std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained variance score: 0.118973022455\n",
      "Mean squared error: 0.0588736219215\n",
      "Mean absolute error: 0.194469281294\n",
      "Median absolute error: 0.16815369654\n",
      "R^2: 0.0921818683213\n",
      "Mean squared error to random1: 0.0943616513166\n",
      "Mean absolute error to random1: 0.270562947805\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXUAAAD5CAYAAADY+KXfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzsnXl4FMXWh3/VQ4AECATClgTEgKAIIgq4XldcuYKKuMD9\nFC+yiIAKqIhBQQiCIIoCglxMRBbZBBRBUFllEQQhCWEJBAhJyEIgC1kmydT5/ujZp2e6ZzKTZJJ6\nnydPpqtrOdVdfbrq1OkqRkQEgUAgENQIpKoWQCAQCATeQyh1gUAgqEEIpS4QCAQ1CKHUBQKBoAYh\nlLpAIBDUIIRSFwgEghpEHS2RUlJSMGvWLPTp0wePP/64zbm4uDisXLkSkiShe/fueO6553wiqEAg\nEAjUUVXqJSUliImJQZcuXRTPx8TE4IMPPkDTpk0xefJk3HnnnYiIiHCZZ3p6ukfChoWFeZzW36mt\ndRf1rn3U1rqr1TssLExTPqrml4CAALz//vsICQlxOJeZmYmGDRsiNDTU3FOPj4/XVLAnFKxfDjqX\n5JW86GoO+MYVIH2JR+n5/h2g+L9dx9n7B+j4P5Yyr2TDMPsD8G0bwH9ZDUpPASUchmHaWNDZk44y\n5irLSEmJ4Dt+cYyfeh588xpQeRn4TyvAN60C378DfONy0LV81TqZ8qXzSeDb1ruOe/Ec+JZ1ICLw\n3zeCkk+5jn8+CXz51+A/rQQZDHIYEfjWH0EpZ52mu7ZlHehUgvmYb9ugqQ3wbethGNoX/OBul/EM\nn7wD/uNS1fzUoJSz4L/K10PxfFqKfM+tzlPCEfB9f8i/k0+B//6TctqCfPkeFl5zLUPqefC1sXLc\nAsv9psICxTZAxw7Jbe/caQV5LzjIW1Ho+D/ge3+Xf+dflWUqLrKcJwL/dR1Kz7puS7ZyOl5X1TQp\nZ2GYOAx0KgHEDeA//wDKdO8lQpczwX9aASorU49bkCfXtcj1/fMWqj11nU4HnU6neC43NxfBwcHm\n48aNGyMjI0O1UK1vHGvKL6Xi0v8+BwC0+cW1MtVC5mdRKD0Zh0bBjdB40HC301/89nOQiiwXY+fa\nxMn4ZDx48mnQKfnFRxuXA8bGyGe865BX5hcfovT4UTRq2AC4/nXzdbs4tC8AoNUzL0EKamgpzxge\nmHsZRTu22OQVeC0fzd6Z6rpOxvSmx6P5o30RENbGZdymnbsiZ9US9WthjA8AIZ06o8GDT6D0XBIy\n18Y6TUuGcqTO+wQwni+/lIpLa75VLQsALq6JkfNYPBthT7+oGKc8Mx2Xkk+Bkk8hbNQEl/mpYapf\ni/sfRd32nZyeD+11D+p362kTFvbc/1nSP/kM6oS2lMON9/vyzK9QvPs3BOmL0fStD1VlAID6eVcQ\nOkG+djmffYii7Zsd2sDFofJvPn28w/U0y3vHvah/Sw9N10ANc30HvIzsRTNQ8vc+NNBJCBk2DgBQ\nmnQCmeu+Q+a67zQ/40rXVWsaPnsimr03HTk/rQD++Bnhq3dorsulyaNQnpaCxmERaNRXuX2ZuDxt\nDor370QQN6DpG67bmSe60R5NNnWtaH1bejK0orTUCqW3x5B2AQBQkHoRhRXIT4sspjiGtIu2J+yu\nl31ehrQUo4wpaKxw/lJaOliDhrCnKOWcY1jaBejdrGdW6kUwKL/QTVxJOW/+rfW+XE1NQV56OijV\ncj2U0lJ5uc15T9uAs7h0yXX5npCddhEssJHT8zlpF8GahzuVLzM1FazUYDMUN6TK7aAw7SJKNMpZ\nnHrBod0Vpac4bQPO6p+TlgoWWnFFY1+WuU6pKSg2lq3WHlyhdF21cMUoBxUWuFWm6dnMS0tFgUo6\nQ6qsawrTUlzev0ozv7giJCQEubm55uMrV66gadOmFcnSBV5eooYx7+ZXFbhTBY8uX1UvC2RfflXL\n4w08aHc1cXkmrz9/VfU8V797U6GeeosWLVBcXIysrCw0a9YMR44cwejRo70lm1fhf2yS73vdekBO\nltUZ924K3/cHkHvFrTRUqgerW0893pXLoA3fgz39f2BNQ7XnX3gN/O3/WAIU7PM4kwi+egmk54fI\naY4dBJ0+DmnAq67zPn0c9NcusEHDwSTHXjvFWYbJxA2g5YvAet0H1kmeWKf8q6C1sbaJMuQeNx3Z\np1im4fOPgDPHIX2x0rlc6SlgYW0ByPZz1A8Cu6c36NsvQClnFOQ8BDoZZ66/w/m0FNDWdaCMNCAz\nHVLUHNDPK8H6DQJr1gKk14OPGgDWbxCkf78A4hy0chHY7feA3XiLUzkdcFOZ8UN7AJPNmwF05gRo\n3x9gT70E+nEpqLgQ0NUBjh1Urld5OWA09+H0cdm2e+YEcDLOLTkAeY6HflwK1A8C2kZCuvcR53L/\ntQvITIPUd6Db5diUefYk6M/fwAa9DuRdBW1cBvbkAPBFnwIBdZXL/mkF0DIc0h33u1UW/34+2B0P\ngHW82bVMhQVu5VvZqCr15ORkLF26FNnZ2dDpdDhw4AB69OiBFi1aoFevXnjttdcwd+5cAMBdd93l\nFZuQMhV7E9MP39gGNDZO/LrZC6KYue6Xvfd3sAf7qMbjKxYCxw6CigqhGxUFrXWmLWsB4urxftsI\n6t0XrGlz8HnT5LDHngYLdpwEN8s0630AALv9bqDzrY4Rjh6w/D59HLT7V9DuX6FbLE/60dpY0H5b\nWyXt3AIMeh20bYNyoYnGyeVzzifM+Nwp0M1cIudntJ+zRo1BB3cpx/9Kth+b6u9wfvIo2+OJw+T4\nVy5DNz4a9MsP8vHG5cC/XwDOnQbt3ALaucVcV19A38yylWvme3L44b1AUaGLhHK7pn/22wZvWuWx\nLHzlYsD6RexCqdP/PpN/VFCp8xnvAgBYl9vAd/0KnDjm0J4cyv5ZvldwU6nT7q2g3VtV76c5/2qK\nqlKPjIzE5MmTnZ7v3LkzoqOjvSmTE7w9zKnE4VpZqbZ4ptnxkmKNGRvr4I4Hj/1LjGu8rkaPFZdY\n2cDNxRVrrYubZSp4glC5uicCuPrLz4Zio+K0r4dCXWXU2pWX2p0rhQ5Y7rPWtqeFYpUyPYCsn2sX\nl4bKytTrXFl46DFXWYgvSmsCbtnWPXw5eqyLKtHm6HbdKqJgK/s6VhE10Z5fUaxNaNXw+njV+6Wy\nMAztC3TqCmnkRLCgBuZwKtWDf/o+cOEMpHemg3XsIvuwbljmmInxvtCebeCSBMrOhPTgE2C33gnD\nlDFA6nlIX/4AFhgk+4svng3WubtDNpRwGPTXLtCBnWDPvgzpCccvamlNDPDoM4Dew14rEXJmRYGH\ntwO74wFzMJ8/DaxVBGj3Vs1Z8feHAnUCLMfvDAYbNAJISXaMO+VNy++5UwAA7L7HIf3fSOW8507W\nLscPixXD6bCynZ3OJdnmry+GYepbYJE3WuIc2Klc1qJPLXF2bwXfshbo2gNQ+c5ATkDgC2eCnNis\nAZh9r81JSorA50VDenIAmILJikr14PMto1v++0bLueNHwL/fj/IJ02GYNdE2YeIxy2/GXCuUC2dg\nGPt/QEGe8zj2ch07CL59E6SRH2iLf3gv+J+/gbUIA23f5Dout4y6DONfAfKuygeH98nPMwBmNb9j\n+PhNSM++bMkgN8dl/nzpPLCkRNB5BZ/7c0ng08eBPT8EKNWrVcsx79i5oL3y9wTssWds8z6yH4Zf\nVkN6fQL41zMAANI3G8GU5k6OHQRxg+LclDfxS6UOADgVD9r3O1jvfuYgOvoXcEGeJOOzJkK3+Cdl\nhW4H7fpVTpP4j2xPSz0vh+/7A+zhp0A/r5LtxaePO6Q1KTsA8iSSglLXjP0zamoYZaUoMikOa/OB\nE5lUsTNT0PKF2kXc/SvgRKm7A/3xs2I4XzhDOXz+NMB+giolGWT9MnKipOnvPy2/t6x1GdeBgjzQ\nRUcXUWuFSrFf2p7avxM4FQ9+Kt7RPsuYXHbiUUv8VUssv1csAgBkvTsUyLpkV6bVvZckdZOYGwod\ngGWe5de1tvI6i79wphwfR9QzNz5TACwK3Q7T3AgA4OI522fr55VA67bO8y/IAzn5YI5Pl/3gafUS\nxfNqmBQ6ANDW9WD3PWY5aXSLNSl0AMDVHMDa0cH6Gl48D1zX3iM5tFKzzC/eHgqZHiINk5CVRjUc\n7vkM67q6awv3hQxawhlTaS9MmweMWn2ZDx9d63kWb7W3imajde6nUqiADa0SzG81S6m7g5Y2Uq0a\nkr9SBXbr6o6W4beaMvXpdxa+uO416F5W83kR/zW/KJGgYRhoIk/Z19zw5cdWR8aGyDV4fphS5GTJ\nfs0njtmEG15/Vktq+d+peLOdEQDIav0ScmKLrm4YhvaFNPxdaH2YretrguKt7qebpgSv4cRUwOdM\nUg6f8S5wi+VzdcPQvpDetmpTDDb2dI/xoVKnzWvMv3nMF7LXiU6nzQPKDr5nG3A1RzafVEyqCqZ3\nkquC6dEwtC+g00H66Cvw7750TKR67QmGj98EWrQGTiUAGtZd8iY1qqdOB7Sv3eAUa1urydc3343J\npr9kP2lu/8GNUxe4mguP+aJC6clqAtGviDtkc8g/t16vRasyVuupV9Kja3Ij9EChAwAtnecFhY7K\n7+gbDPL8kdKHfGr3kABcPAcc3qeg0H3fza9RSt3rVMT8UhOWIagoNWjEXdmorqMkmlfVUc2vfY1Q\n6s6Wv9SyLKZKznIeWj5qsU7FudtpAAAVlreaUUMndakiH8Fo7fGK+RxbDOXuP4elem0fpHkbN8y1\nvsDvbeqUlgI+eRTYU47LX/KR/SuW99b1spuim/DocYBxFUi3uOC4Zkl1hBL/UY9k4uhfvhOkiuBv\nvuR52nmulz82x1Pxy9b+1XENws1nir8xoELFkbP17c+7fk5NS0xUFX7fU6d42X7pk/UYPFm4hwhw\nselDTYD2bdca06dyCARVgsKmItUJP1fqTOiNqkDS2GzEvREIKh3/UerO7LNOXBMFvoO0uhcaap/H\nj0BQ1fiNUqfflN3bnH1uXmU48WuuUbjzPYBAIKhU/Eepq2wgXF2gK9lVLYJAIKiuVIKrs98odb/x\n+/YXOQUCQY3Ef5S6v/g8Z6SqxxEIqhkUf7iqRRB4Cf9R6v5CRlpVSyAQuA3/cop6JIFfIJS6QCAQ\n1CD8W6kL+7VAIBDYoGmZgNjYWCQlJYExhsGDB6NDhw7mc2+88QaaNWsGyfhBypgxY9C0aVPvS6qw\nngJZ7R4jEAgE1R5Pt7R0A1WlnpiYiIyMDERHRyM1NRVff/01oqNt14OeOHEi6tev7zMhASjvBGO3\nxKlAIBBUZ/gP/4Muao5Py1A1v8THx6NnT3nR/4iICBQWFqKoqMinQgkEAkGNpBLWhVLtqefm5iIy\nMtJ8HBwcjNzcXAQFBZnDvvnmG2RnZ+PGG2/EwIEDlXfStiIsLMxtQS+6nUIgEAiqGUQu9Z8nutEe\nt5fetV+8//nnn8ett96Khg0bYtasWfjrr79w5513uswjPT3d3WIFAoGgRuBM/4WFhbnUjVoVvqr5\nJSQkBLm5uebjq1evIiQkxHx8//33o3HjxtDpdOjevTtSUlI0FSwQCAQC76Oq1Lt164YDBw4AAJKT\nkxESEoLAwEAAQFFREaKjo1Fu3H8zMTERbdq08aG4AoFAIHCFqvmlU6dOiIyMRFRUFBhjGDJkCHbu\n3ImgoCD06tUL3bt3xwcffIC6deuiXbt2qqYXgUAgEPgORqo73HofT2zqhqF9fSCJQCAQVC66xcrb\n5FWaTV0gEAgE/oNQ6gKBQFCDEEpdIBAIahBCqQsEAkENwm+UOnvhtaoWQSAQCKo9/qPU73+8qkUQ\nCASCao/fKHW/2c5OIBAIqhD/UeoCgUAgUEUodYFAIKhB+I9SF9YXgUAgUMV/lLrQ6gKBQKCK/yh1\nscm0QCAQqOI3Sp3VrVfVIggEAkG1x2+UukAgEAjUEUpdIBAIahBCqQsEAkENQih1gUAgqCyu6+Dz\nIoRSFwgEgsoitIXPixBKXSAQCCqLSljDSih1gUAgqCRY2/Y+L6OOlkixsbFISkoCYwyDBw9Ghw4W\nu1BcXBxWrlwJSZLQvXt3PPfccz4TVgvs1TdBMXOrVAZBFdP+RuDsyaqWQlBV1AsE9MVVLYUD7K4H\nwR571uflqPbUExMTkZGRgejoaIwYMQIxMTE252NiYjBu3DhMnToVcXFxSE1N9ZmwakiT50G6+2HV\neOyV0a4jdLjJSxIJvIH00ZduxWcP9vGRJB5w211VV3aHzlVXdhUizY51frJJ00qTw4bWbSD9922w\nOpr60RVCVanHx8ejZ8+eAICIiAgUFhaiqKgIAJCZmYmGDRsiNDTU3FOPj4/3rcTegPOqlkDgDsxN\nK2F1WnufV6EskrCuVhsqcZkT1ddGbm4uIiMjzcfBwcHIzc1FUFAQcnNzERwcbD7XuHFjZGRkqBYa\nFhbmkbCX7+2N4j9/R9D9j6Fo11aH86063ghd4yZIC24Mnp9ne1KnAwwG1LulBxp374ms752X0+SB\nx5B75oRHMnqbgBs6oywpsarFqDICIjuieacbkQ6gTusIlF+yjASD7nsURbu3OaRp0b0nMgHUu7UX\n9EcP+kw2Z+3QmpB7HsDVowd8JoMrmtzXG7mnE6qkbHsaPNIXhb/9VCllhbW9Dtm39IA+7m+Hc40e\nfBIF65dVihwAEHBde5RdOIvg+x5FYw16z1PdaI3bYwFy0Qtydc6a9PR0d4sFALQePxXpvfuhJKwt\npJtvByUdBxo3BfvXo0BBHjILi4DCIiD6G7C//zT32NhN3YBGTYCEwyjrcjty6tWDNHkeENIMyLsC\nNGgIJJ8GGjQCGocgP7QlpPDrAQLo2EGwtpFA6zagnVvAHnsaCKgH2rYeaNQYKCsFysvBWkWAsi+B\n3XwbUFYKOpUA1vNe0O8/gd3xANCgAejCWbCQUKB+IMANoO2bgI5dgTp1wJq1AJ1JBOt5H+joAbDm\nrYFGwTC0bgupvBRNc7ORc/ECWPPWoLMnwLrcDjp9HAhqAHbb3UDaBdDxI2D3PgKU6oGcbKBePblO\nQQ2B0wnyiy20Jej8GSAzHaz3U8Dp40DHLkDaeSCsDZCfC74mFlKf50EHdgANGoFdfwPQpJl8E1qG\nAyeOAoENQDlZoMR/wG6/ByzieqBhsHw9W4UDcYdAZWVgwU2AZi2AOnWAvFygYSPQ1h/BHvq33JPU\n1QECAoCyMiAzHQgJBaWnAGV6QK9HyxcG41JWFqRpC8Ebh0AylIP2bQdrG4mSjl3AIm8ESvVgN3YD\nGgUDej0u12sA6eP5KAttCamkBPT3HrDb7wby84CQULmuujqgpOOy3DodKD8X7NY7gPNJoDMnwFqF\ng+L+BntpGGjzarAuPeRrWXQNdGAH2JMDUBISCtaxK1jHm4GzJ0FxfwPNWwKhLcHadQSCgpDXqAmk\nCZ/K9yAzDahbDwABgQ2AwgIg4npZjsZNQX9uA1qEAdyAlk8+i6xiPXApVa7X5Ux5hBkYBJSWgk7F\nA81bAfXqg3W5DbRnG1hYWzks7QLQrRfydXXkdhwcAlxKBV3OALvxFtDmNWDdegHlZaALyWD39pZl\nCG8nt52GwYChXG6neVeBUj3ozAmgIA8IayvLXy8QrEFD0LFDYD3uAW7oDKQky/XiBiC0FWjXZtCW\ndZBeGY3irj0g3X4vKPkUUFwIdusd8nVu3BRUpgdr0Ah04QwQ2hIte92DzJ3bgEZNwEJbyu24IE9u\nK7k58j2sHwi6eE6+Z4UFoPjD8rMV1gaXsrJAI96HdPyI/NxeOAPWrgNQWorCsDaQOnaV72VmKuhS\nKliX2+X8GzQEnT4O1ioCCGoA1Ksvt1GDAZSRJreJYwfBwtqCsjPAwtoABoNc97IyuTdeUgzauQUI\naQZ2UzcYwq+DlH4B18LaolBF74WFhbnUjVoVPiMVTbx69WqEhITgkUceAQCMGjUKs2bNQmBgILKy\nsjB37lxER0cDANasWYNGjRrh8cdd7yfqqVJXq3RNprbWXdS79lFb6+4tpa7aU+/WrRtWr16NRx55\nBMnJyQgJCUFgYCAAoEWLFiguLkZWVhaaNWuGI0eOYPRolUlIN4Tzdlp/p7bWXdS79lFb6+6Neqv2\n1AFg+fLlOHHiBBhjGDJkCM6fP4+goCD06tULiYmJWL58OQDgjjvuQN++fSsslEAgEAg8Q5NSFwgE\nAoF/IHyeBAKBoAYhlLpAIBDUIIRSFwgEghqEUOoCgUBQgxBKXSAQCGoQvl9dxku4WinSn0lJScGs\nWbPQp08fPP7447h8+TLmzZsHzjmaNGmC0aNHIyAgAHv27MHmzZvBGEPv3r3x0EMPoby8HAsWLEB2\ndjYkScLIkSPRsmXLqq6SJpYtW4YTJ06Ac46nn34a7du3r/H11uv1mD9/PvLy8lBWVob+/fvjuuuu\nq/H1tqa0tBTjxo1D//790aVLlxpf9+PHj2POnDlo06YNAKBt27bo27evb+tNfsDx48fpk08+ISKi\nixcv0sSJE6tYIu9QXFxMkydPpoULF9KWLVuIiGj+/Pm0b98+IiJavnw5bd26lYqLi2nMmDFUWFhI\ner2exo4dSwUFBbRjxw5avHgxEREdPXqU5syZU2V1cYf4+HiaPn06ERHl5+fTiBEjakW99+7dSxs2\nbCAioqysLBozZkytqLc1K1asoAkTJtCOHTtqRd0TEhJo9uzZNmG+rrdfmF9crRTpzwQEBOD9999H\nSEiIOez48ePo0aMHAKBHjx6Ii4vDmTNn0L59ewQFBaFu3bro1KkTTp48iYSEBPTq1QsA0LVrV5w6\ndapK6uEunTt3xttvvw0AaNCgAfR6fa2o9913341+/foBAHJyctC0adNaUW8TaWlpSE1NRffu3QHU\njrauhK/r7RdK3X41SNNKkf6OTqdD3bp1bcL0ej0CAgIAWOrprP7W4ZIkgTGG8vLyyquAh0iShPr1\n6wMAtm/fju7du9eKepuIiorC3LlzMXjw4FpV76VLl+KVV14xH9eWuqempmLmzJmYNGkS4uLifF5v\nv7GpW0PiI1hF/O26HDp0CNu3b0dUVBTGjBnjcT7+Vu9p06bh/Pnz+Oqrryokuz/Ve9euXejYsSNa\ntPDOxsv+UvfWrVtjwIABuOuuu5CZmYkpU6bAYDB4nJ+WeotlAgQCgcBPGD58OBYtWuQyTpX01MXS\nu+5TW+su6l37qK1117L07s0336yaj1/Y1P0VKiuFYc4kUBXtfGMN3/ojDEP7gqrJTji+gC6eg2Hm\nBNDlTJfx+O8bwZd8bps2+RQMn04A5V7xpYhVCp1OgGHW+6Br+d7Ls6QYhtkfgBIOey1Pb0LX8uX7\nesqzdk+pxjaV7XpHNyICX/wZ+PZNHpWjlYEDB6rGEUrdCVRWWvFMjh8BThwDnz+94nlVEFobCwDg\nsyZWrSAAeMxcGKa+7f18F88GziSCflzqMh6tWiLv6mSddt40ICkRtHm183RZ6aBi//W64rMmAqeP\ng3ZurlA+VHjNbNulv/8ETsWDz53iDRG9Du3cAiQlgs923e75L6thGPt/oFK9bfj/5gBnEsHXxbou\nyGAAHdwFWvlNBSV2TWhoqGocTUo9NjYWH3zwAaKionDmzBnFOCtWrMDkyZPdEtBdqAITDA55Eclv\n1xWLYJg+3uYcX/89+MjnQFb7YZrTpaWAcq+A8nNhGPkcuJsPSGVNYVBBPgxzJsHw0SgY3h6kGIdv\n+gH8h8WWNHYbctPhvTDMmgjDlDfBVRqr6Xpqkm3fH0DKWafnDbM/AI+dqykv24SO7YNSz4GOHVKW\ngwjEOYgbzJuR047N4DGOZVNJMfgHI8A/GK5YTzp9XO6xFl5zX24NUHkZDBOHgf/i+NIhgwFk9Iig\n1HMwTHgNdC7J87LKymCY8Br4lnWWMCLwhTPBv18A/tZA0NpYua5ebs90JhFZE4Y7HU0QESjhCAzz\no2U5Z7wLvvxrVzlafrnwGqENy+Qt7U7E2ZSFfKOX3eF94If2WF5mJfYvd6ty7F4MlY2qUk9MTERG\nRgaio6MxYsQIxMTEOMRJTU3FiRO+3aiZ9CVI7XuH4gPnLnz3VvBh/cCH9QPt+AU4d9pSTtoF0OY1\n8u/Eo45pJ48Cf2cwKP5veS/S5QtBcYc0DWn56iXgw5+p8E2nE8fAF33qtJFS8inwsf8BThwD0lOA\nawU25w0TXpPjbVwB+uNnWbYVC8GHP23TWPnCmfLepqnn5P1UXcCjx4HPeLci1bJwKh609w/VaFRe\nDiorsw6R/xl3bjdMGQM+5U3weVOVXzjEwScOAx/+jLxfqCl43x+y8si6ZEln6qEX5Mntxi4/Put9\nWe4K9oKdkp0BZGfIyscO/t4Q8NEvyL/XxAI5WeArXUymqenhSylAThbox+8sYfm5oMN7Qbt/lbPY\ntl5W7quWuFkRBXFOxoFvlDfa4bMmQh9/WH6urO4tJSWCsjPAP34LfO5k4OhfoO0/y3vD7tyizaR0\n9bK6LAd3WX6bFL3p+JtZ4B+/Cb5xBfjoF2EY8xIMn39oPGmViY9e7FpRVepaPvxZunQpXnzxRd9I\naMJoJ6V96g+7KygnC/T9fOVzpxPAJ1ttxycx2/M2D7LlHP9qKnjU66Aj+yxxzyQ6mF3ot40AcUDh\nZeEOfM4kedibcBiUlAj+/XxQ6nl5I2rISt8lOVkOQbTDqIwynU/U8F9WgwoLlE9eOAMkK38YQdfy\nQcYy3Rmp0KWLLs/zcS+Dj+wPvnML+IEdll4jY7JCSD3vugDOFa8FANCfv8m98q0/OhGOOwl3v+dK\ner18bfOuOp4rK5UVVn6eQkojeVeAcqMCZMpRDHMnWw7yr8rl6Z11LiyZ0AXlkbkZfbHr83bwPdtA\nif/Yhn0WBdq0SrZbm0ZMf/8JvkB+fshgAP90AvjEYUDqOYtsRpMiAOBMoltyaMHU4bEh9TzINFoq\nLrQ8y9btwck9qCxUvV9yc3MRGRlpPjY5xQcFBQEAdu7cic6dO6N58+aaC/VkH77S0iKYpr+cpefF\nRWD1A8GY8lUtOXoQ2R+MdCpTwV87YP1JU1BuDhoW5+Par+sRcF0HNHyyP0wGGbI3DxQWgH89A62+\nWYeA8OtTq0sIAAAgAElEQVSQMeNdWD/2YWFhMKkoPj8abX7521V1wa8VIO+H/6HR0wNRJ7SlTb1N\n+YQ0boyc6XLvmHZvBQC0+eVv5AcHw4UKcJAn5MJpmPowzZu3QOk/+6Br3gr2/RrasAyBuZfR7J1p\nDvmZ8lK6Nxf7yFscRmzYj9Sn77KRQQlTXs3qBbiOVyT3iMg4/Na1DIcBAP21C61GvINLVnHDWrdG\n8Z7foGsVDpMaD447CGefsNU7cRQlAOocO4hW/x2N8rp1HPJjujooPZeEkn/+Ml/vRo0aoWFgPVwa\n9ixCho1Hg0eeUsy/PDsDuqahYLo6yFu+CPkblqHu2US0mCGbuchQjqCt61Cw9juHtPVWLUbQnfcj\n8K4HbK5XWFgYsusHogRA3YA6aGl13S4mHDH/pp1b5Ov110407PcSGj31gk3+pcUF5metqY4hMCwM\nhsB6UPNH0fJcX1w6DwTYtH+T/C1CmyGDMcuLMeEwwsLCQKV6OBpCbQlKO4+QJ591CM9rFAxTH75l\nixao09p1mwsMDEQzYz1S4WRQw2xPhIWFgZcUI8143KpVa+iaqtu+lfDGHqVuuzRa97SuXbuGHTt2\nYNKkSbhyRbvXgCfuSpSdbf6dlnQafNzLYM/8B9Jj8o2kwgLwt2TbsTQrBqxJM3N8vmkVWIebXPZg\n09PTwQtsh3CFm9eicPNa83H+rXeqypmVfBY8ajSQZVtH+zqrXQP+w2LQHz/j2vGj0L3zCcLCwpB2\n+KBNvlcVrnnahQvgh/aqypmWcsH8+/I0y5xC9uVs8AUznKYrungeeheyu6pX2hcfa44LADnZlxGu\nIZ4JQ7lluJ6ZZqsG0lNTwT/9wCYs9+uZTvMqMZpbykpLkZaaCtj1otPT0sDqBMAw6iWb8IKU8yj4\nZR2oqBBXvpiCvJtvd8ibMlLBJ40EuvWCblQU+EX5XugvnjfXtUlSvKJCB4Ci339G0e8/Q/p8GVhD\ny1eI6enpMBh736WlZarXrfxSKnIXzkL+bfeCf/kxcPUypKHjbeYmrly5ApaeDsp3HEXYk7ZzG1jH\nLqrxTLLak5WZBXuNmbZ/D/iKhar5XduwAoV16pr1gQnrZzozKwtMZWWUomsF5vZNTkdjtofp6ek2\nZsuMzEywEvcdLbS4NGpB1fwSEhJi80n+1atXzWuVJCQkID8/Hx999BFmz56Nc+fOITY2VlPBbmM9\nrE1KAAzloLWx4Evnga+JsTEb0NGDlt9Xc0Abl4N/FoUKj4uyXbvKAQClnndQ6Irx8nNBZaXgy78G\n3/s7+KZVtvZhky3PysWOfzhS9tKw5OKY78pFsi1drXyloSWgONlom1ChTK0TpMYeormooX1hmDbW\nnJ6v+w78943WKVCekQbDwhngf+2C4cM3zO6KdNx2CK9euJtmEW6Mn3IW/MM3QHu2OuSnOGFqtDmb\nj7kBdGSf2WuGOLdM3B47aJ9cjlNWBoOCKcZBxLf/4xhoci08nwQ6vM/xvFJ5PyyW06VdAJ88GpSk\nZMpQf3b4zz/YHm/6AYZxL5vnkOjsSTVJHPP87AOnZj2H1NbmGKXzWvKxvmbOmoydsudb1oH+/N0S\noHCp+IZl4CtcfzTkLVR76t26dcPq1avxyCOPIDk5GSEhIQgMDAQA3HnnnbjzTrn3mpWVhQULFmDw\n4MG+kdTaT9Rguai0ZxsAgN1+tyVs+degLrfJk3stwy3hl1IqJIK1zdxpnEN7lMPt/MP5uJetjozK\nLj0FeLAP2A2drc7JLaTcha3bphzj9VCNd2CnYji38wRywO7B4MsXgv60lEknjoHPmQT237ch3fWg\nuiAXzsgTzgDo13W254hw5ato4OhBs4LiyxZA99YU8C8+cszLyj4uv8StBXXS63KK1ROdmQayU1go\nKgQf/woUsZ7g2/sHaOk8c6+c1sbIcyum83YvUbpwFnza207NQg5S5lhGsPbulnzhDEgjJoAvdD7y\nAuA4CW7tmcQY6OI5Gy8prdDGFfKP+L+B2+9RnEi38SIpUJjsrLALqdX8wP8+A+64X34ZHz8CdOgM\nVj8QdEV9AtUVNhPKxjJJXyKP7pq1ANPpLHb4gcMrVJYWVJV6p06dEBkZiaioKDDGMGTIEOzcuRNB\nQUHm1cMqA2vXQcVGamdH5+8PdYzzj/OPgPjCmapeKWr+z3Ik5de7bQ/bSdJDe0CH9kAa86Eln6x0\nkL4El0Y9r7ksTahNIrqA79kG6V+PyiLYeXvwOZPk8G8/h+HQHvmBVs2QK/Y6ac9vMGTZffRx/B8Y\nhvZVz9Pe0yHJzY9PVCaz6VcnE6gAaK2Vh9gZo1fYiWOyt85vG23i8hHPWA7Ky8Cnuee/zycMsfwe\n4+isoKbQFWG2B/zzD228QJyns1Kg5dZeSUx2G7WWa/MaSE8OAF/8mSXs0wnuy2oHHdkHKi+H1Os+\n5fMGg+zFs3g20CoCuqkLwN/7r22cpES5Y+Xh80V/7QKtlr2CWI97wYZ7yStMI5ps6oMG2fo5t2vX\nziFOixYtfO6n7gq+zJWvqjp0WN0OXSHc6HHwL21tz1xJoQPg331VIZE8hZbOAzVvBXbjLa4jalHo\nAJCRavHesC7n4C54ax0+b38cQ79vVI8EK2+tUj34644TeTZcc+JZVOkw259ueLjwPzYBF5OBRo0t\nYcu/BuxeLrT+e/Ar2UCc8jcEnsK/NpbjRKnjn/0W82SG8vQr/3QC2ID/KrZJLZgUOiB78cBKqVNO\nNlgz7U4lnuCXqzQq4uJjlkrF2eSKLyhxz53Mm/DPoqBb/JN3MtOXeCcfgVegvVb2YTfnoegH40dq\nVhO4znr5tOtXxXBvwPfvAHJzHMvMzgT9+ZtqelrzrfdkWbbAku8/+8F6axhpVgD/Uer+spik27Zb\n/4WvcfwQzbOMas818zcoMx0o1ejJkZJs+e3F9WU8gb79XDnczv5teOdV38viw5eXEmLtF2+jOsNf\nc6Bt672Sj8kOL6h+mHveWnD2YVp1RqE371OsJrZ9hVDqAoFAUElonYupCEKpCwQCQQ3Cf5S6hg9q\nBAKBoLbjP0pdIBAIBKoIpS4QCAQ1CKHUBQKBoAYhlLpAIBDUIPxGqbNnnSyeJBAIBAIz/qPUH+pT\n1SIIBAJBtcdvlLr9KowCgUAgcEQodYFAIKgs6vh+uS3/UeoCgUDg79St5/Mi/FOpV8KFEQgEAn9E\n01ggNjYWSUlJYIxh8ODB6NChg/lcQkICVq5cCUmS0Lp1a4wYMQKS5IN3BbdaeleYYiodadLn4FPd\n25VHIBBUPqraNzExERkZGYiOjsaIESMQE2O7hvY333yDsWPHYurUqSgpKcHRo663AfMYnUVUdktP\nj7NhDzzhDWkqlw6d1eN4GfbiUJsREWvb3nubYtQUOnWF9LZ3d1SyQaercR0Y9tRLvsm4/Y2+ydee\nwKAKJZdG+36ZaVWlHh8fj549ZSUaERGBwsJCFBVZtmabMWMGmjVrBgAIDg7GtWvXfCIoqxOA0Clf\ngg0cAfbKGM8zatte/q+2FVs1QvrP65VfaFBDsMFvuowijXivcmS59U7l8Ka+3RbMGvbQv4F69S0B\nTZpBNz4auO4Gq0jKjxN7ynHvUC1IY6dBem+mR2k148OJO2mh7Xr7bMjbQN263sk7eqFt3td1cBLT\ni9SpA2nyPI+SSjOXQLf4J7BK6KCpKvXc3FwEB1u2pgoODkZurmWv86Ag+c119epVHDt2DN27d/eB\nmDKBPe6G9OCTYPXUberSxM8Uw9k9D4MNGVt5CskrOO+tsReHQvroS9epH34K0hcrLAE3dVMvsVNX\nsPqBriN1uV01n4oSeM/DkAaPdjzRrRekNyb6vHwT7IUhkF61vOTYXQ/IP6xGkNKi9YCScqnvWe+O\ndbwZrP2NkD6z363eBZ26OuZz98Mela+IVtNqSCiYTmcrR6/7gWYtbcNeG+eZHA0b2xyyp/8D6TMN\nG8NrhD32LKQRVhth1w+UnyFTT71OgG0CJy8VdvfDkOatBqvEDojbxm9S2FYuLy8PM2fOxGuvvYZG\njRqp5hEWFubRn3VaNcLvud/muP5tcm8vPKINIp4diPAbOmmprpkWs723ZyEAtPp6tbaIjKGFVX0D\n7+1tczri/4YjvNfd5uNmEz91yCI4LALhN3Q0H4c8/KRqseFdbkHYI/IHX6x+oOJ1b91a/T5UlIZP\nPa94r9pMX4DmzVu4TFu3483m33VaR6iW1XL+D2g6djLqtI1E6KTPELHpEBoPHoVWX69GeEQbhF5v\neXDDR76HsLAwtA6z5BseHo66gY4vwsaNg9Fy3krzcdN3pqnKAljaeviNN6tHNtXh9XccwlqPeh8R\nG/Yrxg9oZxlpNBk+HrrQljayWsPq1UfISFnRNXx6oEs56rWNRFhYGBo8IW+2rWveEuEREQh/6jmb\neBHPvIT6vf5lExZ494Mu8waA1q1b2xyHt++A8Bs7I2L9XjQaMNghvnVbcEWDJ/uj1YJViBgz0UbW\nJq+ORvj1kQhv3wEt5y5D2He/2KRrdMe/7LMCALQcPBLh10e6rd+cndeC6tgrJCTEpmd+9epVhISE\nmI+Lioowffp0vPTSS+jWTb0HCADp6emaBbQmLCzMnJb1uh90cJf8u/8roHW2vZn09HSwnv8CHdoD\nACgd8T4k4jZlsweeAO3cYpOOPfo0aNsGh7JzGoda4vzrUdCebZpklqLmAKWloNMJoA3LzOFZWVma\n0gNAFiw9npLwdmCDXgct/xqA1bWsUwcoL8dVXYBD+vy8PFyzqndurvJGwGZCW5rzNdnRle7ZpUuX\nfD+Byphi2enp6aB6DRyj3/MwaO8fwPUdYXjnE0jHDoLPmwb+yhhgxrsO8U1IizbgsiQBN90GTLoN\nVwHg0iXgnkdxTS4QpLfsLn8pIwMAQGWWsPT0dBj0evng5u5gQQ1Bh/YgP6QlmJWseR21mf7ceU7Y\ngP+CPfRvXFYwp2QWFoHylPcMNQx/D3hnMAAgv1lrsE8W4zIAacYS8AlDbMv4cC7yW7SGtGgDiiUJ\n2LDCMUMj+lK9fI8eew4sOwvU53lLfVqFAxlpQKsIOWzoO8BB+Tll9z6CkBHjUbxvh8v6Xsq2bAsn\nzVtje60efRZYE2sTv7zPC5AGhgAlxeDW7SCyE5B8ynxY3PYGlAQEAqb86tYDSvXIKylFgSksKBgo\nst0s3ZnZOaugEEzjfbTWb87Oa0G1p96tWzccOHAAAJCcnIyQkBAEWvVGli5dij59+uDWW2/VVKC3\nkIaOg7RgLaQ3PwJ75GnbczOWyD/C2pjDGGNgku1wEAEK9j3OAaveizVs0AiwF4dBenmUslB2NlU2\n+E2w6zqA3dAZ7Hq7PK2deZ57FeyJ/hb5F20Abjf1vhkYYwj/YTvYq2+CPfAkJIXJXmn6YkhjpwJN\nmjrKZT8Mbhsp/+jaA9Lny4CIdrZ5DXeu/GwzkidQpajPwXr+C1Aw10hvTJQfYnepFwh0uR31bnKu\nAJmSKaBte0jvz4I09mM5Trdesi1TZSJNMS/7OEq9fZ3dPTe2OdYqAuy/b0P6eD6YdS/R7l4AULw+\nAZEdbY5Z735Aw0aQxkcD3RXmGBgDMyn0egpmMycTrqxJU7AHjSO31hY5WLPmwM1GU+r1HSEtWg/W\nQu4dm64Ve/o/jhk2bGRTHmvQENLwd8Gs2pjpeWX3PuIoz0vDoAtppiirTby69SCN+RDSxwsUzbFK\nJicWfp1szhpnGSmxXve5LEea+BnYI/2U41nZx03n2SD7+S9Hy4avUe2pd+rUCZGRkYiKigJjDEOG\nDMHOnTsRFBSEbt26Yffu3cjIyMD27dsBAPfeey969+6tkqt3YAF1zXZdaco8QF8CFBXKDRJQv55N\nQx2CWO++YM/LPRQ+rJ/NOekBK7NFk2bmTWvZXQ+C9u+A9PE80OF98gMIArOeWLvRbhQTamU6CAwE\nGlnmLZgkgbWMsBFfahQMyYVtlIU0A0KagfKu2ob3ug/sQaMZ5dU3wRoGg13fUZ7wadEaLCDAdgIQ\ncLQXOi3U+OBe1x5s2DswjH8FKCm2jXLrncBG2x4de+wZ+T7d9RD4pxOgBHukH6R+A8F0rpuo9GkM\noJPAx1kWfGOR7pnWKgKTdGB9B4KFXycfvzAUaNse7O6HZSXb2tKxkGZ/B9groEaNwfq8AFoyxzZf\nOy8L6YUhwAtyu9R16grD0L52gljF/WI5YDCA1i8Fu8toynDhRSMNHAF6cZjji80qjUOHCIDU53kY\nrEafbMjboB2bgWunHOLaZPuvR8Fu6gaEtlQ4aSuDtGg9aPFnoL//tAQa2yfr2sN5Gf9+AbTvD0tA\nxy6WczfeAmnyPNCOTWD/elS+VmuUzassvK1ZHzietI53HaRFG8AkCQbjKLqq0DT1PWjQIJvjdu3a\nmX+vWOF8CFaZsLC27qd54EmAc9Aa2U2TPfcqWDMrZdu6DXDponJiU2/l7ochvfomaPCbsjLu87xy\nWXYPjI3Cd/XyqaBHmzR0vOW31UuBhVuul/TqW+Crl4B1vxNIPQ8YFZQ62oSTXhoO/vUnwDXZBCA9\n96rl3Lw1oF9WgbaslXO87zGgZRjY/bZ2f+mtKaDLmWAhzUB6y9CXGXt1JpMUc+YpY88tPSH1fQl8\n2lht8U2EhALlZTZBkpV3CwsMkj1lFGCNQxTDlRoAU7m21qZFOcDK5bdOHaBOHdkt1RzomB977Fmr\n5C5GKgrzaA55vfqWPBGakw1KPgXmYhKdMQY0b+XspO2hpAOGjgc63AT6YbEc9ozCCME+m+atwB57\nBrR1PdjLoyyjGNP58LZg/xkpH7SO8Ep/2nQNpTEfgn8pjxQR1NALObuH7xciqMawOgFgjz4Dg1Gp\n2z9H0nszwN8aZGUKscI0jDb1Vr3+wZWHzUzDA2gPaxkGnSf+s3bXiz3cF/Sj1dyG8cFlHW+G7vNl\njr1LAKxePbBnX4Zh64+y6athY0iPPuMY7+bu5uKU1J30wBOAO98gEDl4UGhBmrHY7TRKmG3/gPKt\nVvFPZ0PHg736FmjnZtDqJarfbjDGwIa9CxbaEny60eOkkVr9tfcopLsfkn880V/uQduZ9DSj9PKR\nJLCHnwL1uBd09C+we7R587BnXgbreR9gMjd6G2cmra49IH2zETAYHF4mlYF/LhOgFa0KztSDsnur\nsgaNIM1bA2m4o/uj9Pr7wE3dwPoNcjjnFUwPXEs37dFV0IhMSE/0hzR3hdmuym6+rcpkUYMpDf21\npJN0iqYIt7EycTHjvAHrO9BiJlDRp4wxsIAASI/0g/TNRjBnPV8rpJ73Os7teBkm6cDaRrrfyTGZ\nUlykY41DIN3/OJhG8yDT6WTToNoHXB50hFTLtp7jqGRqdE+d3fcY6MAOSC8MdRlPevcT0P6dYHc6\nulI584lnba6HbuxUt+SRJn8FPlnB51op//ufkO3OTnol0nszgNwrjukaBoO9OAy4eFY2L/kUhV5V\nUEOwbr3kXqiTCWdFKns+qXFIFRRqgT05AHTuNKSXhoE1aWa2x/I6AaDTCajf4x5o/YxPVWk5RWP9\nfaD07JFGTwKIV6AuFUDBt18L0sARmp/nyqRmK/UmTaGb/o16vBZhYP1c+916BTvvCTbgv6A134J1\nvR04n2TziLGAALC+zj+pdvVlmvSwsk3XK3TrBRw7aBTCyfBz0Ouy94G9jI1DALuJ3EqlfqBlIpeo\nUpSVM1jT5tBN+sJybJqjeewZsJtvRaNe9+Ca0W2yqpAefxY84TAkF5/2swGvAo2aVLgsxhjAvDAC\n8qRse0cBrYS1BSLagXn4UvAVNVqpVz9slaD06NOgR/qBMQZq0lR+oO98oGpE04huVJTFNu6kU8UC\n6tp4G5iQZixxmGSsTKRZsaC4Q6Adm2U3vuIi9USVDJMk2XvGF4vi2aPyUmOdusqmHVeeMwrzH96A\nvTDEwROmusEYg07la+6qQCj1KoaZJ1p1YFZeITURk1eG60g+LL9+oOxPbPQppqJC3xXmD2gYqFSJ\nOQSA1LufeiQvwl4bJ3tg3ey7ZU4qC6HUK5MatuKeTzWwQFCJSHfcD9xxv3pEP6B6j28E1RsvvqRY\nj3vk/+0qYbU9gZGqm1MQ+A7RUxe4jfThXCAny6tDc/bqm2C9+wLXd1SPLPAOnFe1BAIfIJR6JVJV\n9klvw9pcD7S53rt5BtSVF1cSCAQVQphfBLWXKnRpFAh8hVDqglpMLVXqpq9pGwa7jifwS4T5RSCo\nZUhjp4L279C8horAvxBKXVB7qaUddda8lcuvlQX+jTC/CAQCQQ1CKHWBQCCoQQilLqi9CO8XQQ1E\nk009NjYWSUlJYIxh8ODB6NDB8tVfXFwcVq5cCUmS0L17dzz33HMuchIIBAKBL1HtqScmJiIjIwPR\n0dEYMWIEYmJibM7HxMRg3LhxmDp1KuLi4pCamuozYQUC7yJ66oKah6pSj4+PR8+e8lZZERERKCws\nRFGRvGRpZmYmGjZsiNDQUHNPPT4+3rcSCwTeQuh0QQ1E1fySm5uLyEjLHn/BwcHIzc1FUFAQcnNz\nERxs+YChcePGyNCwsH9YWJiH4lYsbXXAtI21J/Xw97p7iq/qXWYoham1VsdrWx1lqixqa929UW+3\n/dTJxeSSq3PWpKenu1ssALnCnqatbrhbj5pUd3fwZb0pK9P8u7pd29p6v4HaW3e1emtV+Krml5CQ\nEOTm5pqPr169ipCQEMVzV65cQdOmTTUVXKuJ8O5iWAKBQGBCVal369YNBw4cAAAkJycjJCQEgYGB\nAIAWLVqguLgYWVlZMBgMOHLkCG655RbfSuznSAvWQZr0eVWLIQDMa6CwB/tUsSACgfdQNb906tQJ\nkZGRiIqKAmMMQ4YMwc6dOxEUFIRevXrhtddew9y5cwEAd911V621hWmFBQRUtQgCIyygruoenAKB\nv6HJpj5o0CCb43bt2pl/d+7cGdHR0V4VSiCoLIRCF9Q0GGmd3RQIBAJBtUcsEyAQCAQ1CKHUBQKB\noAYhlLpAIBDUIIRSFwgEghqEUOoCgUBQgxBKXSAQCGoQQqkLBAJBDcJvNp52tVGHP5OSkoJZs2ah\nT58+ePzxx3H58mXMmzcPnHM0adIEo0ePRkBAAPbs2YPNmzeDMYbevXvjoYceQnl5ORYsWIDs7GxI\nkoSRI0eiZcuWVV0lTSxbtgwnTpwA5xxPP/002rdvX+PrrdfrMX/+fOTl5aGsrAz9+/fHddddV+Pr\nbU1paSnGjRuH/v37o0uXLjW+7sePH8ecOXPQpk0bAEDbtm3Rt29f39ab/IDjx4/TJ598QkREFy9e\npIkTJ1axRN6huLiYJk+eTAsXLqQtW7YQEdH8+fNp3759RES0fPly2rp1KxUXF9OYMWOosLCQ9Ho9\njR07lgoKCmjHjh20ePFiIiI6evQozZkzp8rq4g7x8fE0ffp0IiLKz8+nESNG1Ip67927lzZs2EBE\nRFlZWTRmzJhaUW9rVqxYQRMmTKAdO3bUironJCTQ7NmzbcJ8XW+/ML+42qjDnwkICMD7779vXvUS\nkN/sPXr0AAD06NEDcXFxOHPmDNq3b4+goCDUrVsXnTp1wsmTJ5GQkIBevXoBALp27YpTp05VST3c\npXPnznj77bcBAA0aNIBer68V9b777rvRr18/AEBOTg6aNm1aK+ptIi0tDampqejevTuA2tHWlfB1\nvf1CqdtvxmHaqMPf0el0qFu3rk2YXq9HgHHRL1M9ndXfOlySJDDGUF5eXnkV8BBJklC/fn0AwPbt\n29G9e/daUW8TUVFRmDt3LgYPHlyr6r106VK88sor5uPaUvfU1FTMnDkTkyZNQlxcnM/r7Tc2dWtI\nLFejiL9dl0OHDmH79u2IiorCmDFjPM7H3+o9bdo0nD9/Hl999VWFZPeneu/atQsdO3ZEixYtvJKf\nv9S9devWGDBgAO666y5kZmZiypQpMBgMHuenpd5iQS+BQCDwE4YPH45Fixa5jOOVnrq9B4caYjs7\n96mtdfd1valUDwTUrXZL8NbW+w3U3rpr2c7u5ptvVs2nwjb1kpISxMTEoEuXLhXNSiCoVKikGPyN\nAeBfTa1qUQQCTQwcOFA1ToWVupIHh0DgF1y9LP+P/7tq5RAINBIaGqoap8LmF51OB51O51aaimx5\nV5u3y6utdfdVvcvK9cjwcRkVoTrKVFnU1rp7o95V4v0ibOruU1vr7st6U3aW+Xd1u7a19X4Dtbfu\nWmzqWvALP3WBQCAQaEModYFAIKhBVNj8kpycjKVLlyI7Oxs6nQ4HDhzA+PHj0bBhQ2/IJxD4kOrl\nxlhZUEYqaNdWsGf/DyygrnoCgV9RYaUeGRmJyZMne0EUgaCyqZ3f3fHZHwB5V4HmLcEe+ndViyPw\nMsL8IhDUNvKuyv+L/X9RPIEjQqkLajG10/wiqNkIpS4QVBPEMkwCbyCUuh9CF86A79te1WIIvAhf\nsRB8+DOgsrKqFkXg59QapU5E1aInREQgq6U3KSMNdD7JrTz4tLGgmC9AJbY2USopBqUke0XOyoJO\nJ4AK8qpajCqHdmwGiFuWLhAIPKTWKHU+bSx49DiQwQAqKa46Ob6aCj7iGRCXFTuf9Dp49DjPMjNw\n27w/nQA+9S1QRmpFxVSFEo+Cx84Fca4e2ZQm9Rz48oWgcrk3StkZ4LMmgn80yldiCnwIcQ4qumYb\nduEsDFGvg1LPV41Q1RAq1YOOHjC3e19Ta5Q6Us4CF86AfzgSfPQLlVIkcQ7KvWIbaFo8yhvDbPt5\nvovn5P/ZGQ5RvQ3//EPQ3j+AMye0p4keD9q5GXRgpxyQb9y9yo2eOp04Brqi3pulvKvgG1eYRzPE\nDQ4KyJtQwmHwb79w+ZIjD3fqoeIi8B2/eL8zYjDAEPU6+Ja1HiXnC6aDvzkQZPKmgWxGQmYa+JoY\nj/Kk7AxQWalHad2BL50HvmcbKDcHhqF9wXdu9llZtDYWfP500K/rfFaGNbVHqZvIulRpRdGyBeDv\nDOahojIAABnVSURBVAadOy0fp12wnPv7T1BujuVY4YElItCV7ArLwXf/Cr50nrq8J46Bb1sPyskC\nxR3Sljlx8N83gs6elM0/hS4Up6mnoi/RljeAoj9/B104KxeVmwM+ZxL4e/9VffD5+0NBm34A/bxK\nPp7xnqyAnJStpnCpuAiGedNAZ08qlzd3Cmj/dsB4rx3Or/of+OvP2ihAKizQdH9p1f9AKxaB1n+v\nGlcxvV4Pvm87SK+3felkpQOZaaAfl3qUL44dlP9npLkv08VzMMx8D3Q50xJ2JRt84jBkvTvUM3mc\nlZV6DobPPwRdlZ83ij8M2rMNtHQe6Ohfctjyhdrzy8+F4ZN3wJd/bWNKdRr/rNzxofNnPJDefWqf\nUq9EaM82+X/ScQDyg28+F/sl+PR3zMf8wzeMD98fIL1ejvPTCvD3hoCO7HdSgtxVp4w0GN50vs4y\nfb9AbsRWDzSV6m1eKgDA50wCrYkBn/Aa+FdTtSmcMydAq5aAz3gXfPQL4G8py0HX8lXzckjDDcj5\nZAL4NHmTali9MPjI51wnNin9fKMSNSlbazmsRzoqy+/S7q3AsYPgM99TE1o5+PefjHJYNg7mbw0C\nf2+I6/wA0KWL8n8XZjXKvwr+y2oYRj0PKiq0Pbd+KSjmC/CoEbajODenmCjrEgyfRWkz7yX+A4Pd\ntaITx8D37wAA8MWzgTMnQOu+s0TIlhV86enjljQGAyjHsR0S55pGbADAF34KJB4FbVwmH385RSWF\na+jnH4DkU6CdW0B7f9eQwtjQKmlOz6+UOpXqvTbZSUTmN7cn8DUx4Ns2uJeo2PZhs5kUu3rZ+PDN\nBa2Xe07052/yfzWF88fPgJVpgS+eDTpxTEFoi8LhH40Cf+dVpz1XADZ5Oi17wzLX5w0G2f7+wXCr\nQOV7SCfjYBg/2KzEoGKv57+sdszDfkLcVR7WYhjknjpduQy+fZOjGcV4vrIeTGfQqQTwJXNsRhZU\nVgo+7hX5XuhLgLMnQVnplhGOyb5t9xKng7ucl1NSLN+3Nd+aRxd8xULgZJzjqM+Zu7+daY7PmQT6\n9nNjAWT+TyVF4Ou+AylMEvOFM8AnDAGlpdjKt3SePGI7d9qhDfPtm0Dxhy0BphGiG/M/NmVxbtum\nrG3j7kxsV1Lb8ZuNp6mkCKnP9AVuvQO6Nz6oeH4bl4N+WQ32whCwm7qDhbd1Hre8HHRwN1jrNkBJ\nEVBWCtq2Xj756NMVlsVcTordQ6jycQyfPw3S2GmOvcPiIvA5k+Qcet1nlcAAoI7cQI3DXj7qebCh\n4yFZx7OWqbwMyLsK1kz7hsGUkQbWKlz+/ftG0NpYTel47JdA3hXQ5rVgQ95WyNj2oaANy4A+z1vS\n/7BYfsFZxzm4Gxg6Xl1mkq82nz1R7s02agzW818qaajSt8HjsycCAFj3u0BdbgMd2Q8WantvrHui\nUtTnwKl4zflTSRHo8D7Q1vWAaYSQmQ7dqCiLUsxIs+sQMdC1fNA/B4BLGifpjb19OrwXaBnm3N5s\nMo9cTAYK88G/nApp3DRzD5lPl++t9NFcoEUYWN16oJXfgADoFv+kud4AYHj7P2D/fhHSw7ZLJ/A3\nBgAdu0D3tvG6unvPK7mN+I1SR47xjXj0L1B5GVidAIcoVFYGvuQzSPc/AXZTN5fZ0a4t8v9VSxwa\nAP91HVA/ENIDT8pxdvwCWr3E85VCTLY0DxsDOZsHOH0cfMQzLrOgg7stB8VFck900uu2cRbPhkGh\n1wsA/IvJslJofyOkgSOA0BZgQa4Xa+OTXjdfTzqjZIOWr6SDWclU37/3AEpK3QlUqpcfZjuFbj5v\nPUznHFSQD9YoWPmdaTJPWNm+FfMsLgIf8yLYv1+A1G+QdSU0y+0kZ1BhgW2QQruhowdAX39iTOEc\ns+nKnNB1K6aVi0H7/rANtJ/sL8gDf/dVSxlffgzoPZ/EpX8OKIdbzUEBAF/9LaAvBv95pUNcPuVN\nIDAI0tvOtyakkyovt2v5oB++AeyUOsrLgMR/nOer9eVeSTts+ZX5xQR/vb+i2YCO/gUc3mfupboL\nJRwG5eeC1n1nnjghzkGrl6intR+iWZ87tAd8/ffqa23YJze1k9MJquVrgY9/xUGhm0lPcQjiU960\n9PLOngSf+ha4C9u9VmjVEhjG/p9ltAPZhGDuCZaXg38/X0FbOV5funIZ/I0BMAzt67Q8/t5/Lb8n\nDgMf+x/wP38DjHMXinkXXgNfGwPKV1bu9NdO+f+mVU7LpeIi0OG9mibTzPLFfgn+1iCUXTwvH38/\nHzBNzlq1LzLapt2FVLw8TKNFG9LOu860AgodgHlEYA+fPNrqyOr+OFOwxUXg0xXcg3OMm6FcvQzD\nvGm25zzpqVkpcNqyFnxYPwdTLl//PfjuXx3jax3JVAD/6anbwUc9D2nyV2Dh11kCDVZ2xpNxQIeb\ngAtnAbuJIyUMo15QbpwmtztXsqz8BrR9ExDRDrqPvlSMQ5vXqOZjplQPsnNLzF0yF1xXBwiopz2f\naoGTp8bOjdF+4pN2bwWsTD586TywB55wzMeJt4mqVN99BWrU2BJQXGjzlS5t+kH+fykV7Pa7Qeds\nPxCz9pYwfD3D5hw/sFPu7datBxw7CDZohEtZ+F9Wtu2kRABA6akEoPNt8nUw4QWXRmejGUCe9IaS\nf7nJfl8B33NrLxd+6E+P8wEAuPGSdMDksWOEVmj3enElC3/3VUgL1pqXMjY973RDF1tvrzK9Qxbe\nxm+VOgBQ4lFbpW4F/yzKvcyc9TZUhlXEDbJCB4DU86ALZ4CiQs0z83a5yf/OnQafOMzmTMGPnrmz\nVRWGuZOhe3NyhfKwduGjPdvM3kQ2VMTaYfVioe8XKMdJu6Du3nlkn80hLZlje95mFMRAR/aDG00n\nAED/+8whS15S5Dgh6eELTCuqnj0efPlLSYlgN3S28Vyibz51Ox8UV94Hg1RcJLtptutgY1ahokKw\noAZw1ujo8D6wOx+w6QDwD0faxfK9fd2vlbrD9bnmRqPz1kS0naseHf/HY3/iqvas0Aq3t7kqkXBE\n0TvF67LY9ZK9jsJXgORCudJ5dcVrrdCdcW3TGpDpYzI/hn86AdKiDZo8qawxvDXI5ph+WQU0aeZN\n0ZzCx7wIAGDD3gV63GMJf/MleQLfmV7OyZK/klYyAZnx/TPulzZ1C8aJteIi2ZVtlbrt2x3o9HFV\nRcvHveK9Ap182FLdoJi52uK5Y3Kqrti7ocLicaGEUhukHe5/rWjva17VuJqzUIN/9Ib781z2k8Wu\nPmpzgmHqWxa7ticknwSfMsYmiJZ87jQ6bVjmMMKuCvyop66gXBkDpSSDT30L7JF+7mVn32gU4Mu/\nVpxAdImf9LYrBzK7pPktpb7/ZF2ZGtSOPPji1BEPrkdKsnOzmlbsPHAAgJJPKUTUiu/NL37TUycn\nypWMM+H020bvF+quQhfYwmuQYvISdFTZfc8eg8JXlLUaD9fN8QnV3CzmP0o9VmnIL3auqdZU0qp0\n/gTt1TAfIVDmQuWsneLv+I1SVxx9MaPdWyAQVAjydzOZwIz/KHUlmFRpX2kJBDUZPj+6qkWodpgX\nYfMmlbBkgJ8r9aoWQCAQCKoX/qPUFb1Kqp9WV1u1UCAQCHyJ/yh1pY9ANK7+JxAIBLUF/1HqSlR0\nISGBQCCoROjvCq57owH/VuoCgUDgR1TGV9ZCqQsEAkENQih1gUAgqEEIpS4QCAQ1CKHUBQKBoAYh\nlLpAIBDUILyy9G5sbCySkpLAGMPgwYPRoUMHb2QrEAgEAjepcE89MTERGRkZiI6OxogRIxATE+MN\nuRy5vqNv8hUIBIIaRIWVenx8PHr27AkAiIiIQGFhIYqKiiosmD1S/8Fez1MgEAhqGhU2v+Tm5iIy\nMtJ8HBwcjNzcXAQFBTlNExYW5nY51DwUqbM9ElEgEAiqDa70nye60R6vb2dHGrZz+//2zjw6qirP\n45/frSSkEoipGCArJIZ9MaRBRpiGQbQBRdkCri1LxxaMGicoEiFMg2wiNBgkEOQ4KALaqKi0I8fT\nNqgcOelm4NCsIos0BG3BQNAJWUjqzh8PUimykECFWJX7+afeu+8uv+9bvve9W1X3fffdd55u1mAw\nGLyCmvwvKiqqVm+sq+Ff9/CLw+GgoKCgYv3cuXM4HI7rrdZgMBgM18B1m3piYiK5udZ7F48dO4bD\n4cBut193YAaDwWCoP9c9/NKxY0duueUWMjMzERFSUlI8EZfBYDAYrgGPjKk/8sgjnqjGYDAYDNeJ\n+UepwWAw+BDG1A0Gg8GHMKZuMBgMN4qYuAZvwpi6wWAwXCvNAuuVXU19qYECqdRGg7fwC0Q9N7ex\nQ/BZ1KI3r63c5NkejgTkwcfrl//O+zweg7eh5r0G9uDGDqPhiIz1aHUyaizyyBM1bx89ATV7BWrx\nW6iVHyCBNf/T3lP4jKnL7QPqnrdjd+TuZGTwSGT8M64N3X5Vff6e/15rfWrea3Vuu6FRM5e5J0TG\nIr8Z7p4W0Mx9PTbetVzDPqgrclP9/3imZiwBVcup2LF79W3dNaz6+iY+j5r3GurOe+seRIeuyLCH\n6p4foGWEK5bhD1fd3qUH0ueO+tVZDTLgnvoXanETdOlRv3buuAdpGQFh4e7pV+sc7cGoaYtQL70O\nUsNxrLSvALj1tnrFVl/k8eerpvXuj0qd5tmGmgWiBtxd7SaV8TLym2FIRDTS4iZE2Tzbdg14lakH\ndOwGgDya6r4hIhq5Y2iN5dTclRXLcpdlcGrUONToCUjfga58aX9ArdhY9YRrFenKM3UBKms9KuNl\n1/ZKva+auxKat7givhhsqzbVqq2C4BaoFxaiXv0TRMRU1VoNaupL2FZtwrZqExLdxq0Tsr2Yjbo/\nBfViNjLyUYiNR72yDtuqTaiZy5A7hqJeWOiqa+LUimVJmYw8PNG9Mb8rfgXbNQm6JNUcXLvOtQff\nrjPSJgEQ9/TYeNTCN4j5eAfqscmu+F77CElJRx5IQcb8DrX8vSptSK9fW+ZUDXJ7VZOVlHRsU+Yj\nQc2rHvvKHZ7ddZxl8EjUjFcgoRNqTg5yz/1uM4mqSVNRT/8XMmos9OxbpbOsLo6KsjkfEP3+tor2\nZNRY63xL6FRjGbe6hySjMheDn7+V0DkR9fxLqElTay4zaCQyapzV/pPT3W4EauwcuyZZ593St5H4\nDsjNLVHZG5DhD6MWvI7KXILcnYxa+g5qynxoHuJqr3UU8vgUZPgVP4fu0LV2bYNHop6eYa1ExqJW\nvF8lj5o8G2nfpWr6759DIqJRC/4b9fJqCGtppb+8GpXzAdL7P6o22K4zJPa22u4/2JUeGmalhYRa\nn4+mImOfcrsBkYRON8zIK+PxuV8akvBZWfzr87/Ar/ogt96Gc8oEa4NfAETEACD9BiED78U5Kw0A\ntXgt0iIE9cxM9N+/RMZMcKtTRFBPZKCLixER8PNDYuLRe3a48gx9AL35PWv5soFUvsCaXbrzvSkM\naRWJmrsSvf2v6PffhLIypPOtNWqSCc+gV2eh/rAUCvKhzS1IiHW3a5u9HACd1AfH6VPkL5kJJcVW\nuTuGorf+j1VJVFv3OgePQu/8Ctq65rWXyFgkMhbuGeNKi25TxbQl0G51KCeOIh26oY9+ja5Urwwc\nCsEtcOa8hHRKRA0eacX47TdwKW558HH0O9bTixrxW5yLplvLr6zHuXgGnDhqrS9cDc1vqrpPHnjM\nMubQMEQECb0Z9Z+zrE5OxN0QVYBlGGUX0W8sRecdd6tLZa0HDc4X0+Dsj64ngqBgSOiM6j8I6XG7\nK/+4p3C+tgiVPA6J74A++S3OF5+BVpHY5q6k/PeXng5s/og9CFulzt02bRHO9Tno3M+hx+2IzQah\nN2OblIHzz++g9+2CmDjL7Gw2SElHlxSj9+xAuvVEv5WN3PcgYrOhAu3WE2DBWcQeBAmdKtoqT3sI\nigpdIpvZoaTIWm4djQx/GPHzRz08EadSqPtTXJ1cj3+D3X+rss9VpetCWkYg96fgjIlDLhvfH9+E\nZnacWTPh8AErbdDIqsfOPwC590FrJawl0jbBWrYHYVuytmL/yaCRSGiYdX59tA4iY7G9mA1A6D+/\nIX/Oc+7xvfoOHPkauiZZ1+zKD5Ganu7iO1RcJxVU+oJSLj2J2Ba87p7n/t+BEqTfIGsIKqwlEtwc\n7XRC4c9w5CD6y0+haxJq7FPovTuhW08rvv5DANC9fo1zygTkvgeqj+0GILouM3B5mGud0OvKCW/0\n4QM4l89DPTsHiYlDl5dbFwuXTCas5TUNB+iLpeid25HYeAhqjjhutk5GezC2pW9X5Lt8gqqVH8L/\n/QSBdqTS0IbzLx+hN7yOmjLPMsj80+B0ov/3K/RGa+zZtmoTWmurQ6mDdp33rRXbfQ/hnDjCan/Z\nu0gz9yEVfeQARMdZhlAHnJ9vhoAAVN873evJO17RQdb5aQPcNDk3rkFu6YD0uN3at2tXIINGINGu\nzkhfLMU5Ox0ZOBRVabjhapMc1Tmef+zAuXwuKnOJ9eRl80OufOqoqeyJY9A6CmkWaJn21k9QyzYg\nNXxJVt3x1CXF6D+/jQy4BwlvfdU2a9NdnvEY5J8GQE3KAJsN5+pXUNP/iLSqfdIn55efot+yzFPN\nXIZzbTZq/DNI67pNFqVLS3A+ad0YqOXvI/7+dSpX0f7WT+D096gHXP881/t2QVw75NKdfFRUFCeH\n9nIrV9u5Vz59EjRvYXX23+xDHp9i1bt+JdLpVqRn33rFWBNaaziwGxI6NsjYuKcm9PJqU7+R6J/P\nQ0Cgm3k6//4l5B1HjRpbc7mSkiqGC6CPHwb/Zkh0mzq1X512/a88OPMD0r1nHVXUH601evN7SJce\nSFz7BmunJhrzmDcmtenWp/6J/vhPyG+fQIJbVJunJpzb/4penQXUr5O+kURFRXHqH7vgu5M4l81G\nBtyNquXLyMsWdrUbo186njJ1rxp+aUykRdVhAtW7P/TuX3u5agwdwBMGKRExFcNODYWIIJWGbAyN\nj0S3RSZW/SKwTmVv64f+25eowSM8HJVnkZYR0DICtWLjVZ+ovN3MPY0xdYOhCSH+AdjSZzV2GHWm\nrkNkBhde9esXg8FgMNROo4ypGwwGg6FhMHfqBoPB4EMYUzcYDAYfwpi6wWAw+BDG1A0Gg8GHMKZu\nMBgMPoQxdYPBYPAhjKkbDAaDD+E1f9d64403OHz4MCLC+PHjadeu3dULeQEnTpxg4cKFDB06lCFD\nhvDjjz+ybNkynE4noaGhPP300/j7+7Nt2zY++eQTRIS77rqLgQMHUlZWxvLlyzlz5gxKKVJTU2nd\n+uqTRf0SWLt2LQcPHsTpdDJixAgSEhJ8XndJSQnZ2dmcP3+eixcvkpycTNu2bX1ed2VKS0t59tln\nSU5Oplu3bj6vff/+/SxevJjYWOvlHG3atGHYsGENq1t7Afv379fz58/XWmt98uRJPW3atEaOyDMU\nFRXpmTNn6pycHL1582attdbZ2dl6+/btWmut161bpz/99FNdVFSk09LSdGFhoS4pKdGTJ0/WP//8\ns966datetWqV1lrr3bt368WLFzealvqwd+9ePW/ePK211j/99JOeNGlSk9D91Vdf6Q8//FBrrfXp\n06d1Wlpak9BdmfXr1+uMjAy9devWJqF93759etGiRW5pDa3bK4Zf9u7dy223WS8viImJobCwkAsX\nLjRyVNePv78/L7zwAg6Ha3rg/fv306uXNe1or1692LNnD0eOHCEhIYGgoCACAgLo2LEjX3/9Nfv2\n7aN3b2sC/+7du3Po0KFG0VFfunTpQnp6OgDBwcGUlJQ0Cd19+/Zl+HDr5RP5+fmEhYU1Cd2XOXXq\nFHl5eSQlWS9VaUraK9PQur3C1AsKCggJcb01JSQkhIKCgkaMyDPYbDYCAgLc0kpKSvC/NEf1ZZ01\n6a+crpRCRCgrK7txAq4RpRSBgdZc5Fu2bCEpKalJ6L5MZmYmWVlZjB8/vknpXrNmDePGjatYbyra\n8/LyWLBgATNmzGDPnj0NrttrxtQro810NdXibftlx44dbNmyhczMTNLS0q65Hm/TPWfOHI4fP86r\nr756XbF7k+4vvviCDh060KpVK4/U5y3aIyMjGTNmDH369OGHH35g1qxZlJeXX3N9ddHtFabucDjc\n7szPnTvnNmThSwQGBlJaWkpAQABnz57F4XBU0X/27Fnat2/vll5WVobWGj8vmap09+7dbNy4kenT\npxMUFNQkdB87doyQkBDCw8OJi4ujvLwcu93u87oBdu3axenTp9m1axf5+fn4+/s3iWMeFhZG377W\nm5ciIiIIDQ3l6NGjDarbK4ZfEhMTyc3NBawLw+FwYLfbGzmqhqF79+4VWnNzc+nRowft27fn6NGj\nFBYWUlxczKFDh+jcubPbftm5cyddu9b+0t5fChcuXGDt2rVkZGTQvHlzoGnoPnDgAB9//DFgDSkW\nFxc3Cd0A6enpzJ8/n7lz5zJw4ECSk5ObhPZt27axaZP1hqmCggLOnz/PgAEDGlS310y9u27dOg4e\nPIiIkJKSQlxcXGOHdN0cO3aMNWvWcObMGWw2G2FhYaSlpZGdnc3FixcJDw8nNTUVPz8/cnNz2bRp\nEyLCkCFD6NevH06nk5ycHL7//nv8/f1JTU0lPDy8sWVdlc8++4x3332XyMjIirQnn3ySnJwcn9Zd\nWlrKihUryM/Pp7S0lNGjR1f8lNOXdV/Jhg0baNWqFYmJiT6vvaioiKysLC5cuEBZWRmjR48mPj6+\nQXV7jakbDAaD4ep4xfCLwWAwGOqGMXWDwWDwIYypGwwGgw9hTN1gMBh8CGPqBoPB4EMYUzcYDAYf\nwpi6wWAw+BD/DwuVo++Yj3TmAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f877a875a58>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"Explained variance score: \" + str(explained_variance_score(y_test, y_predicted)))  # lower means: accounts less for variance in data\n",
    "print(\"Mean squared error: \" + str(mean_squared_error(y_test, y_predicted)))  # how close does the model get on average\n",
    "print(\"Mean absolute error: \" + str(mean_absolute_error(y_test, y_predicted)))  # are we biased?\n",
    "print(\"Median absolute error: \" + str(median_absolute_error(y_test, y_predicted)))  # outliers?\n",
    "print(\"R^2: \" + str(r2_score(y_test, y_predicted)))  # future predictions\n",
    "\n",
    "print(\"Mean squared error to random1: \" + str(mean_squared_error(y_test, y_test_random1)))  # how close does the model get on average\n",
    "print(\"Mean absolute error to random1: \" + str(mean_absolute_error(y_test, y_test_random1)))  # are we biased?\n",
    "\n",
    "\n",
    "fig = plt.figure()\n",
    "tmp = fig.add_subplot(3, 1, 1)\n",
    "tmp = plt.plot(y_test)\n",
    "tmp = fig.add_subplot(3, 1, 2)\n",
    "tmp = plt.plot(y_predicted)\n",
    "tmp = fig.add_subplot(3, 1, 3)\n",
    "tmp = plt.plot([b - a for (a, b) in zip(y_predicted, y_test)])\n"
   ]
  }
 ],
 "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": "3.4.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}