Fletcher/Poe Edited.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from textblob import TextBlob\n",
    "import re\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import pandas as pd\n",
    "from textblob import TextBlob\n",
    "import seaborn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import nltk\n",
    "from nltk.tokenize import sent_tokenize\n",
    "from nltk.tokenize import word_tokenize\n",
    "from nltk.corpus import stopwords"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# opens each of 5 poe files, reads to text, decodes, splits into each \"story\"\n",
    "poe_lit=[open('poe'+str(vol)+'.txt').read().decode('utf-8').split(\"\\r\\n\\r\\n\\r\\n\\r\\n\\r\\n\") for vol in range(1,6)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#cut out non-stories by Poe\n",
    "poe_lit_1=poe_lit[0][7:-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "poe_lit_2=poe_lit[1][3:-2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#scrubbed out novel\n",
    "poe_lit_3=poe_lit[2][29:-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "poe_lit_4=poe_lit[3][3:-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#scrubbed out poetry\n",
    "poe_lit_5=poe_lit[4][2:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#all 66 poe short stories in one list\n",
    "poemo=poe_lit_1+poe_lit_2+poe_lit_3+poe_lit_4+poe_lit_5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#yeah it's not a list comprehension, but it's triple nested, so it might look messy\n",
    "#formatting so it's a list containing strings of short stories\n",
    "stories=[]\n",
    "for story in poemo:\n",
    "    story_words=''\n",
    "    for word in word_tokenize(story):\n",
    "            if word.lower() not in stopwords.words('english') and word.isalpha():\n",
    "                story_words+=' '+word.lower()+' '\n",
    "    stories.append(story_words)\n",
    "            \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "vectorizer = CountVectorizer()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "X=vectorizer.fit_transform(stories)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "vocab=vectorizer.get_feature_names()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import lda"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#looking at polarity for each number of topics until 25\n",
    "model_pol=[]\n",
    "for i in range(1,25):\n",
    "    model = lda.LDA(n_topics=i, n_iter=1500, random_state=1)\n",
    "    model.fit(X)\n",
    "    topic_word = model.topic_word_  # model.components_ also works\n",
    "    topic_pol=[]\n",
    "    for j, topic_dist in enumerate(topic_word):\n",
    "        n_top_words = 25\n",
    "        topic_words = np.array(vocab)[np.argsort(topic_dist)][:-(n_top_words+1):-1]\n",
    "        word_pol=0\n",
    "        for word in topic_words:\n",
    "            word_pol+=TextBlob(word).sentiment.polarity\n",
    "        topic_pol.append(word_pol)\n",
    "    model_pol.append(topic_pol)\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "topics_count=[]\n",
    "std=[]\n",
    "for num,pol in enumerate(model_pol):\n",
    "    topics_count.append(num+1)\n",
    "    std.append(np.std(pol))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x119b90790>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfkAAAFwCAYAAAC7CQL0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x119b71190>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(topics_count,std)\n",
    "plt.title('Standard Deviation for Topic Polarity')\n",
    "plt.xlabel(\"Number of Distinct Topics\")\n",
    "plt.ylabel('Standard Deviation')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<lda.lda.LDA instance at 0x1189ca878>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = lda.LDA(n_topics=12, n_iter=1500, random_state=1)\n",
    "model.fit(X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Topic 1: upon could came eyes many length saw night still even fell thought heart head face lay felt among man became within voice words room hand\n",
      "Topic 2: one could however first two little much time found without whole although made great far would three means large manner upon less nearly course doubt\n",
      "Topic 3: one beauty yet art seen valley light scene young hundred character green ellison whose love sense flowers magnificent form became trees eye less sentiment color\n",
      "Topic 4: de said sir jupiter upon legrand let skull goodfellow wid massa te ugh majesty table bug little ye replied parchment tree shuttleworthy angel dat thin\n",
      "Topic 5: would body upon corpse dupin marie one madame murder evidence yet le water girl said rue suspicion committed river prefect house paper question morning us\n",
      "Topic 6: one say upon good thing well old man see way sure like day made take took every make done much would quite never get know\n",
      "Topic 7: us king said like ship sea water let wind thousand perceive rock wyatt river captain come round boat hundred vast monster city legs summit board\n",
      "Topic 8: balloon earth moon car surface great atmosphere feet voyage elevation sun miles day light diameter five actually april sea distance material horizon rotterdam night gas\n",
      "Topic 9: may would upon matter us thus idea nature even general every say let reason said possible case man thought person merely mere present men certain\n",
      "Topic 10: said little gentleman de nose lady replied monsieur madame great acquaintance name pompey really smith oh lalande yes diddler eyes age big john thus sir\n",
      "Topic 11: long death yet spirit life soul within character condition shadow spoke thus days things feeling never lips sense memory fancy died earth god body evil\n",
      "Topic 12: door arm open box back machine automaton within body table person left interior game right baron man hand company never book maelzel drawer main portion\n"
     ]
    }
   ],
   "source": [
    "topic_word = model.topic_word_  \n",
    "n_top_words = 25\n",
    "for i, topic_dist in enumerate(topic_word):\n",
    "    topic_words = np.array(vocab)[np.argsort(topic_dist)][:-(n_top_words+1):-1]\n",
    "    print('Topic {}: {}'.format(i+1, ' '.join(topic_words)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "topic_word = model.topic_word_  # model.components_ also works\n",
    "n_top_words = 25\n",
    "topic_dict={}\n",
    "for i, topic_dist in enumerate(topic_word):\n",
    "    topic_words = np.array(vocab)[np.argsort(topic_dist)][:-(n_top_words+1):-1]\n",
    "    topic_dict['Topic '+str(i+1)]=[(word,TextBlob(word).sentiment.polarity) for word in topic_words]\n",
    "    \n",
    "    \n",
    "        \n",
    "    \n",
    "        \n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Topic 1': [(u'upon', 0.0),\n",
       "  (u'could', 0.0),\n",
       "  (u'came', 0.0),\n",
       "  (u'eyes', 0.0),\n",
       "  (u'many', 0.5),\n",
       "  (u'length', 0.0),\n",
       "  (u'saw', 0.0),\n",
       "  (u'night', 0.0),\n",
       "  (u'still', 0.0),\n",
       "  (u'even', 0.0),\n",
       "  (u'fell', 0.0),\n",
       "  (u'thought', 0.0),\n",
       "  (u'heart', 0.0),\n",
       "  (u'head', 0.0),\n",
       "  (u'face', 0.0),\n",
       "  (u'lay', 0.0),\n",
       "  (u'felt', 0.0),\n",
       "  (u'among', 0.0),\n",
       "  (u'man', 0.0),\n",
       "  (u'became', 0.0),\n",
       "  (u'within', 0.0),\n",
       "  (u'voice', 0.0),\n",
       "  (u'words', 0.0),\n",
       "  (u'room', 0.0),\n",
       "  (u'hand', 0.0)],\n",
       " 'Topic 10': [(u'said', 0.0),\n",
       "  (u'little', -0.1875),\n",
       "  (u'gentleman', 0.0),\n",
       "  (u'de', 0.0),\n",
       "  (u'nose', 0.0),\n",
       "  (u'lady', 0.0),\n",
       "  (u'replied', 0.0),\n",
       "  (u'monsieur', 0.0),\n",
       "  (u'madame', 0.0),\n",
       "  (u'great', 0.8),\n",
       "  (u'acquaintance', 0.0),\n",
       "  (u'name', 0.0),\n",
       "  (u'pompey', 0.0),\n",
       "  (u'really', 0.2),\n",
       "  (u'smith', 0.0),\n",
       "  (u'oh', 0.0),\n",
       "  (u'lalande', 0.0),\n",
       "  (u'yes', 0.0),\n",
       "  (u'diddler', 0.0),\n",
       "  (u'eyes', 0.0),\n",
       "  (u'age', 0.0),\n",
       "  (u'big', 0.0),\n",
       "  (u'john', 0.0),\n",
       "  (u'thus', 0.0),\n",
       "  (u'sir', 0.0)],\n",
       " 'Topic 11': [(u'long', -0.05),\n",
       "  (u'death', 0.0),\n",
       "  (u'yet', 0.0),\n",
       "  (u'spirit', 0.0),\n",
       "  (u'life', 0.0),\n",
       "  (u'soul', 0.0),\n",
       "  (u'within', 0.0),\n",
       "  (u'character', 0.0),\n",
       "  (u'condition', 0.0),\n",
       "  (u'shadow', 0.0),\n",
       "  (u'spoke', 0.0),\n",
       "  (u'thus', 0.0),\n",
       "  (u'days', 0.0),\n",
       "  (u'things', 0.0),\n",
       "  (u'feeling', 0.0),\n",
       "  (u'never', 0.0),\n",
       "  (u'lips', 0.0),\n",
       "  (u'sense', 0.0),\n",
       "  (u'memory', 0.0),\n",
       "  (u'fancy', 0.0),\n",
       "  (u'died', 0.0),\n",
       "  (u'earth', 0.0),\n",
       "  (u'god', 0.0),\n",
       "  (u'body', 0.0),\n",
       "  (u'evil', -1.0)],\n",
       " 'Topic 12': [(u'door', 0.0),\n",
       "  (u'arm', 0.0),\n",
       "  (u'open', 0.0),\n",
       "  (u'box', 0.0),\n",
       "  (u'back', 0.0),\n",
       "  (u'machine', 0.0),\n",
       "  (u'automaton', 0.0),\n",
       "  (u'within', 0.0),\n",
       "  (u'body', 0.0),\n",
       "  (u'table', 0.0),\n",
       "  (u'person', 0.0),\n",
       "  (u'left', 0.0),\n",
       "  (u'interior', 0.0),\n",
       "  (u'game', -0.4),\n",
       "  (u'right', 0.2857142857142857),\n",
       "  (u'baron', 0.0),\n",
       "  (u'man', 0.0),\n",
       "  (u'hand', 0.0),\n",
       "  (u'company', 0.0),\n",
       "  (u'never', 0.0),\n",
       "  (u'book', 0.0),\n",
       "  (u'maelzel', 0.0),\n",
       "  (u'drawer', 0.0),\n",
       "  (u'main', 0.16666666666666666),\n",
       "  (u'portion', 0.0)],\n",
       " 'Topic 2': [(u'one', 0.0),\n",
       "  (u'could', 0.0),\n",
       "  (u'however', 0.0),\n",
       "  (u'first', 0.25),\n",
       "  (u'two', 0.0),\n",
       "  (u'little', -0.1875),\n",
       "  (u'much', 0.2),\n",
       "  (u'time', 0.0),\n",
       "  (u'found', 0.0),\n",
       "  (u'without', 0.0),\n",
       "  (u'whole', 0.2),\n",
       "  (u'although', 0.0),\n",
       "  (u'made', 0.0),\n",
       "  (u'great', 0.8),\n",
       "  (u'far', 0.1),\n",
       "  (u'would', 0.0),\n",
       "  (u'three', 0.0),\n",
       "  (u'means', 0.0),\n",
       "  (u'large', 0.21428571428571427),\n",
       "  (u'manner', 0.0),\n",
       "  (u'upon', 0.0),\n",
       "  (u'less', -0.16666666666666666),\n",
       "  (u'nearly', 0.1),\n",
       "  (u'course', 0.0),\n",
       "  (u'doubt', 0.0)],\n",
       " 'Topic 3': [(u'one', 0.0),\n",
       "  (u'beauty', 0.0),\n",
       "  (u'yet', 0.0),\n",
       "  (u'art', 0.0),\n",
       "  (u'seen', 0.0),\n",
       "  (u'valley', 0.0),\n",
       "  (u'light', 0.4),\n",
       "  (u'scene', 0.0),\n",
       "  (u'young', 0.1),\n",
       "  (u'hundred', 0.0),\n",
       "  (u'character', 0.0),\n",
       "  (u'green', -0.2),\n",
       "  (u'ellison', 0.0),\n",
       "  (u'whose', 0.0),\n",
       "  (u'love', 0.5),\n",
       "  (u'sense', 0.0),\n",
       "  (u'flowers', 0.0),\n",
       "  (u'magnificent', 1.0),\n",
       "  (u'form', 0.0),\n",
       "  (u'became', 0.0),\n",
       "  (u'trees', 0.0),\n",
       "  (u'eye', 0.0),\n",
       "  (u'less', -0.16666666666666666),\n",
       "  (u'sentiment', 0.0),\n",
       "  (u'color', 0.0)],\n",
       " 'Topic 4': [(u'de', 0.0),\n",
       "  (u'said', 0.0),\n",
       "  (u'sir', 0.0),\n",
       "  (u'jupiter', 0.0),\n",
       "  (u'upon', 0.0),\n",
       "  (u'legrand', 0.0),\n",
       "  (u'let', 0.0),\n",
       "  (u'skull', 0.0),\n",
       "  (u'goodfellow', 0.0),\n",
       "  (u'wid', 0.0),\n",
       "  (u'massa', 0.0),\n",
       "  (u'te', 0.0),\n",
       "  (u'ugh', 0.0),\n",
       "  (u'majesty', 0.0),\n",
       "  (u'table', 0.0),\n",
       "  (u'bug', 0.0),\n",
       "  (u'little', -0.1875),\n",
       "  (u'ye', 0.0),\n",
       "  (u'replied', 0.0),\n",
       "  (u'parchment', 0.0),\n",
       "  (u'tree', 0.0),\n",
       "  (u'shuttleworthy', 0.0),\n",
       "  (u'angel', 0.0),\n",
       "  (u'dat', 0.0),\n",
       "  (u'thin', -0.4)],\n",
       " 'Topic 5': [(u'would', 0.0),\n",
       "  (u'body', 0.0),\n",
       "  (u'upon', 0.0),\n",
       "  (u'corpse', 0.0),\n",
       "  (u'dupin', 0.0),\n",
       "  (u'marie', 0.0),\n",
       "  (u'one', 0.0),\n",
       "  (u'madame', 0.0),\n",
       "  (u'murder', 0.0),\n",
       "  (u'evidence', 0.0),\n",
       "  (u'yet', 0.0),\n",
       "  (u'le', 0.0),\n",
       "  (u'water', 0.0),\n",
       "  (u'girl', 0.0),\n",
       "  (u'said', 0.0),\n",
       "  (u'rue', 0.0),\n",
       "  (u'suspicion', 0.0),\n",
       "  (u'committed', 0.0),\n",
       "  (u'river', 0.0),\n",
       "  (u'prefect', 0.0),\n",
       "  (u'house', 0.0),\n",
       "  (u'paper', 0.0),\n",
       "  (u'question', 0.0),\n",
       "  (u'morning', 0.0),\n",
       "  (u'us', 0.0)],\n",
       " 'Topic 6': [(u'one', 0.0),\n",
       "  (u'say', 0.0),\n",
       "  (u'upon', 0.0),\n",
       "  (u'good', 0.7),\n",
       "  (u'thing', 0.0),\n",
       "  (u'well', 0.0),\n",
       "  (u'old', 0.1),\n",
       "  (u'man', 0.0),\n",
       "  (u'see', 0.0),\n",
       "  (u'way', 0.0),\n",
       "  (u'sure', 0.5),\n",
       "  (u'like', 0.0),\n",
       "  (u'day', 0.0),\n",
       "  (u'made', 0.0),\n",
       "  (u'take', 0.0),\n",
       "  (u'took', 0.0),\n",
       "  (u'every', 0.0),\n",
       "  (u'make', 0.0),\n",
       "  (u'done', 0.0),\n",
       "  (u'much', 0.2),\n",
       "  (u'would', 0.0),\n",
       "  (u'quite', 0.0),\n",
       "  (u'never', 0.0),\n",
       "  (u'get', 0.0),\n",
       "  (u'know', 0.0)],\n",
       " 'Topic 7': [(u'us', 0.0),\n",
       "  (u'king', 0.0),\n",
       "  (u'said', 0.0),\n",
       "  (u'like', 0.0),\n",
       "  (u'ship', 0.0),\n",
       "  (u'sea', 0.0),\n",
       "  (u'water', 0.0),\n",
       "  (u'let', 0.0),\n",
       "  (u'wind', 0.0),\n",
       "  (u'thousand', 0.0),\n",
       "  (u'perceive', 0.0),\n",
       "  (u'rock', 0.0),\n",
       "  (u'wyatt', 0.0),\n",
       "  (u'river', 0.0),\n",
       "  (u'captain', 0.0),\n",
       "  (u'come', 0.0),\n",
       "  (u'round', -0.2),\n",
       "  (u'boat', 0.0),\n",
       "  (u'hundred', 0.0),\n",
       "  (u'vast', 0.0),\n",
       "  (u'monster', 0.0),\n",
       "  (u'city', 0.0),\n",
       "  (u'legs', 0.0),\n",
       "  (u'summit', 0.0),\n",
       "  (u'board', 0.0)],\n",
       " 'Topic 8': [(u'balloon', 0.0),\n",
       "  (u'earth', 0.0),\n",
       "  (u'moon', 0.0),\n",
       "  (u'car', 0.0),\n",
       "  (u'surface', 0.0),\n",
       "  (u'great', 0.8),\n",
       "  (u'atmosphere', 0.0),\n",
       "  (u'feet', 0.0),\n",
       "  (u'voyage', 0.0),\n",
       "  (u'elevation', 0.0),\n",
       "  (u'sun', 0.0),\n",
       "  (u'miles', 0.0),\n",
       "  (u'day', 0.0),\n",
       "  (u'light', 0.4),\n",
       "  (u'diameter', 0.0),\n",
       "  (u'five', 0.0),\n",
       "  (u'actually', 0.0),\n",
       "  (u'april', 0.0),\n",
       "  (u'sea', 0.0),\n",
       "  (u'distance', 0.0),\n",
       "  (u'material', 0.0),\n",
       "  (u'horizon', 0.0),\n",
       "  (u'rotterdam', 0.0),\n",
       "  (u'night', 0.0),\n",
       "  (u'gas', 0.0)],\n",
       " 'Topic 9': [(u'may', 0.0),\n",
       "  (u'would', 0.0),\n",
       "  (u'upon', 0.0),\n",
       "  (u'matter', 0.0),\n",
       "  (u'us', 0.0),\n",
       "  (u'thus', 0.0),\n",
       "  (u'idea', 0.0),\n",
       "  (u'nature', 0.0),\n",
       "  (u'even', 0.0),\n",
       "  (u'general', 0.05000000000000002),\n",
       "  (u'every', 0.0),\n",
       "  (u'say', 0.0),\n",
       "  (u'let', 0.0),\n",
       "  (u'reason', 0.0),\n",
       "  (u'said', 0.0),\n",
       "  (u'possible', 0.0),\n",
       "  (u'case', 0.0),\n",
       "  (u'man', 0.0),\n",
       "  (u'thought', 0.0),\n",
       "  (u'person', 0.0),\n",
       "  (u'merely', -0.5),\n",
       "  (u'mere', -0.5),\n",
       "  (u'present', 0.0),\n",
       "  (u'men', 0.0),\n",
       "  (u'certain', 0.21428571428571427)]}"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "topic_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from pymongo import MongoClient\n",
    "client=MongoClient()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "db = client.test_database"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#name poe collection\n",
    "poe = db.poe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "topic_word = model.topic_word_  # model.components_ also works\n",
    "n_top_words = 25\n",
    "for i, topic_dist in enumerate(topic_word):\n",
    "    topic_words = np.array(vocab)[np.argsort(topic_dist)][:-(n_top_words+1):-1]\n",
    "    poe.insert_one({'Topic '+str(i+1):[(word,TextBlob(word).sentiment.polarity) for word in topic_words]})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{u'_id': ObjectId('57b629bf9e65dbc025ea0079'), u'Topic 1': [[u'upon', 0.0], [u'could', 0.0], [u'came', 0.0], [u'eyes', 0.0], [u'many', 0.5], [u'length', 0.0], [u'saw', 0.0], [u'night', 0.0], [u'still', 0.0], [u'even', 0.0], [u'fell', 0.0], [u'thought', 0.0], [u'heart', 0.0], [u'head', 0.0], [u'face', 0.0], [u'lay', 0.0], [u'felt', 0.0], [u'among', 0.0], [u'man', 0.0], [u'became', 0.0], [u'within', 0.0], [u'voice', 0.0], [u'words', 0.0], [u'room', 0.0], [u'hand', 0.0]]}\n",
      "{u'Topic 2': [[u'one', 0.0], [u'could', 0.0], [u'however', 0.0], [u'first', 0.25], [u'two', 0.0], [u'little', -0.1875], [u'much', 0.2], [u'time', 0.0], [u'found', 0.0], [u'without', 0.0], [u'whole', 0.2], [u'although', 0.0], [u'made', 0.0], [u'great', 0.8], [u'far', 0.1], [u'would', 0.0], [u'three', 0.0], [u'means', 0.0], [u'large', 0.21428571428571427], [u'manner', 0.0], [u'upon', 0.0], [u'less', -0.16666666666666666], [u'nearly', 0.1], [u'course', 0.0], [u'doubt', 0.0]], u'_id': ObjectId('57b629bf9e65dbc025ea007a')}\n",
      "{u'Topic 3': [[u'one', 0.0], [u'beauty', 0.0], [u'yet', 0.0], [u'art', 0.0], [u'seen', 0.0], [u'valley', 0.0], [u'light', 0.4], [u'scene', 0.0], [u'young', 0.1], [u'hundred', 0.0], [u'character', 0.0], [u'green', -0.2], [u'ellison', 0.0], [u'whose', 0.0], [u'love', 0.5], [u'sense', 0.0], [u'flowers', 0.0], [u'magnificent', 1.0], [u'form', 0.0], [u'became', 0.0], [u'trees', 0.0], [u'eye', 0.0], [u'less', -0.16666666666666666], [u'sentiment', 0.0], [u'color', 0.0]], u'_id': ObjectId('57b629bf9e65dbc025ea007b')}\n",
      "{u'_id': ObjectId('57b629bf9e65dbc025ea007c'), u'Topic 4': [[u'de', 0.0], [u'said', 0.0], [u'sir', 0.0], [u'jupiter', 0.0], [u'upon', 0.0], [u'legrand', 0.0], [u'let', 0.0], [u'skull', 0.0], [u'goodfellow', 0.0], [u'wid', 0.0], [u'massa', 0.0], [u'te', 0.0], [u'ugh', 0.0], [u'majesty', 0.0], [u'table', 0.0], [u'bug', 0.0], [u'little', -0.1875], [u'ye', 0.0], [u'replied', 0.0], [u'parchment', 0.0], [u'tree', 0.0], [u'shuttleworthy', 0.0], [u'angel', 0.0], [u'dat', 0.0], [u'thin', -0.4]]}\n",
      "{u'_id': ObjectId('57b629bf9e65dbc025ea007d'), u'Topic 5': [[u'would', 0.0], [u'body', 0.0], [u'upon', 0.0], [u'corpse', 0.0], [u'dupin', 0.0], [u'marie', 0.0], [u'one', 0.0], [u'madame', 0.0], [u'murder', 0.0], [u'evidence', 0.0], [u'yet', 0.0], [u'le', 0.0], [u'water', 0.0], [u'girl', 0.0], [u'said', 0.0], [u'rue', 0.0], [u'suspicion', 0.0], [u'committed', 0.0], [u'river', 0.0], [u'prefect', 0.0], [u'house', 0.0], [u'paper', 0.0], [u'question', 0.0], [u'morning', 0.0], [u'us', 0.0]]}\n",
      "{u'_id': ObjectId('57b629bf9e65dbc025ea007e'), u'Topic 6': [[u'one', 0.0], [u'say', 0.0], [u'upon', 0.0], [u'good', 0.7], [u'thing', 0.0], [u'well', 0.0], [u'old', 0.1], [u'man', 0.0], [u'see', 0.0], [u'way', 0.0], [u'sure', 0.5], [u'like', 0.0], [u'day', 0.0], [u'made', 0.0], [u'take', 0.0], [u'took', 0.0], [u'every', 0.0], [u'make', 0.0], [u'done', 0.0], [u'much', 0.2], [u'would', 0.0], [u'quite', 0.0], [u'never', 0.0], [u'get', 0.0], [u'know', 0.0]]}\n",
      "{u'_id': ObjectId('57b629bf9e65dbc025ea007f'), u'Topic 7': [[u'us', 0.0], [u'king', 0.0], [u'said', 0.0], [u'like', 0.0], [u'ship', 0.0], [u'sea', 0.0], [u'water', 0.0], [u'let', 0.0], [u'wind', 0.0], [u'thousand', 0.0], [u'perceive', 0.0], [u'rock', 0.0], [u'wyatt', 0.0], [u'river', 0.0], [u'captain', 0.0], [u'come', 0.0], [u'round', -0.2], [u'boat', 0.0], [u'hundred', 0.0], [u'vast', 0.0], [u'monster', 0.0], [u'city', 0.0], [u'legs', 0.0], [u'summit', 0.0], [u'board', 0.0]]}\n",
      "{u'Topic 8': [[u'balloon', 0.0], [u'earth', 0.0], [u'moon', 0.0], [u'car', 0.0], [u'surface', 0.0], [u'great', 0.8], [u'atmosphere', 0.0], [u'feet', 0.0], [u'voyage', 0.0], [u'elevation', 0.0], [u'sun', 0.0], [u'miles', 0.0], [u'day', 0.0], [u'light', 0.4], [u'diameter', 0.0], [u'five', 0.0], [u'actually', 0.0], [u'april', 0.0], [u'sea', 0.0], [u'distance', 0.0], [u'material', 0.0], [u'horizon', 0.0], [u'rotterdam', 0.0], [u'night', 0.0], [u'gas', 0.0]], u'_id': ObjectId('57b629bf9e65dbc025ea0080')}\n",
      "{u'_id': ObjectId('57b629bf9e65dbc025ea0081'), u'Topic 9': [[u'may', 0.0], [u'would', 0.0], [u'upon', 0.0], [u'matter', 0.0], [u'us', 0.0], [u'thus', 0.0], [u'idea', 0.0], [u'nature', 0.0], [u'even', 0.0], [u'general', 0.05000000000000002], [u'every', 0.0], [u'say', 0.0], [u'let', 0.0], [u'reason', 0.0], [u'said', 0.0], [u'possible', 0.0], [u'case', 0.0], [u'man', 0.0], [u'thought', 0.0], [u'person', 0.0], [u'merely', -0.5], [u'mere', -0.5], [u'present', 0.0], [u'men', 0.0], [u'certain', 0.21428571428571427]]}\n",
      "{u'_id': ObjectId('57b629bf9e65dbc025ea0082'), u'Topic 10': [[u'said', 0.0], [u'little', -0.1875], [u'gentleman', 0.0], [u'de', 0.0], [u'nose', 0.0], [u'lady', 0.0], [u'replied', 0.0], [u'monsieur', 0.0], [u'madame', 0.0], [u'great', 0.8], [u'acquaintance', 0.0], [u'name', 0.0], [u'pompey', 0.0], [u'really', 0.2], [u'smith', 0.0], [u'oh', 0.0], [u'lalande', 0.0], [u'yes', 0.0], [u'diddler', 0.0], [u'eyes', 0.0], [u'age', 0.0], [u'big', 0.0], [u'john', 0.0], [u'thus', 0.0], [u'sir', 0.0]]}\n",
      "{u'_id': ObjectId('57b629bf9e65dbc025ea0083'), u'Topic 11': [[u'long', -0.05], [u'death', 0.0], [u'yet', 0.0], [u'spirit', 0.0], [u'life', 0.0], [u'soul', 0.0], [u'within', 0.0], [u'character', 0.0], [u'condition', 0.0], [u'shadow', 0.0], [u'spoke', 0.0], [u'thus', 0.0], [u'days', 0.0], [u'things', 0.0], [u'feeling', 0.0], [u'never', 0.0], [u'lips', 0.0], [u'sense', 0.0], [u'memory', 0.0], [u'fancy', 0.0], [u'died', 0.0], [u'earth', 0.0], [u'god', 0.0], [u'body', 0.0], [u'evil', -1.0]]}\n",
      "{u'_id': ObjectId('57b629bf9e65dbc025ea0084'), u'Topic 12': [[u'door', 0.0], [u'arm', 0.0], [u'open', 0.0], [u'box', 0.0], [u'back', 0.0], [u'machine', 0.0], [u'automaton', 0.0], [u'within', 0.0], [u'body', 0.0], [u'table', 0.0], [u'person', 0.0], [u'left', 0.0], [u'interior', 0.0], [u'game', -0.4], [u'right', 0.2857142857142857], [u'baron', 0.0], [u'man', 0.0], [u'hand', 0.0], [u'company', 0.0], [u'never', 0.0], [u'book', 0.0], [u'maelzel', 0.0], [u'drawer', 0.0], [u'main', 0.16666666666666666], [u'portion', 0.0]]}\n"
     ]
    }
   ],
   "source": [
    "for doc in poe.find():\n",
    "    print doc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "db.poe.count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}