notes_linking_reactions_to_corpus.ipynb

{
 "metadata": {
  "name": "notes_linking_reactions_to_corpus"
 }, 
 "nbformat": 2, 
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "markdown", 
     "source": [
      "# Linking the reactions to the corpus"
     ]
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "We want to link the individual reactions from the Oct 3 2012 presidential debate to the corpus for the debate."
     ]
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [
      "import pandas as pd", 
      "from pandas.tools.plotting import scatter_matrix", 
      "import random", 
      "import numpy as np", 
      "import reactions"
     ], 
     "language": "python", 
     "outputs": [], 
     "prompt_number": 1
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "## Reactions data"
     ]
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "Read in the Oct3 2012 debate reactions file and get the reactions part."
     ]
    }, 
    {
     "cell_type": "code", 
     "collapsed": false, 
     "input": [
      "%time parts = reactions.split_reactions_file('data/reactions_oct3_4project.csv')"
     ], 
     "language": "python", 
     "outputs": [
      {
       "output_type": "stream", 
       "stream": "stdout", 
       "text": [
        "CPU times: user 4.77 s, sys: 0.44 s, total: 5.21 s", 
        "Wall time: 5.25 s"
       ]
      }
     ], 
     "prompt_number": 2
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [
      "r = parts['reactions'].copy()"
     ], 
     "language": "python", 
     "outputs": [], 
     "prompt_number": 3
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "Add a *start* column to hold the time of the reaction; we will use this to merge with the corpus file later."
     ]
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [
      "r['start'] = r.Time.apply(lambda t: pd.datetime.time(t))"
     ], 
     "language": "python", 
     "outputs": [], 
     "prompt_number": 4
    }, 
    {
     "cell_type": "code", 
     "collapsed": false, 
     "input": [
      "r.head(2)"
     ], 
     "language": "python", 
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">", 
        "<table border=\"1\" class=\"dataframe\">", 
        "  <thead>", 
        "    <tr style=\"text-align: right;\">", 
        "      <th></th>", 
        "      <th>UserID</th>", 
        "      <th>Time</th>", 
        "      <th>Reaction_who</th>", 
        "      <th>Reaction_what</th>", 
        "      <th>start</th>", 
        "    </tr>", 
        "  </thead>", 
        "  <tbody>", 
        "    <tr>", 
        "      <th>0</th>", 
        "      <td> ag1zfnJlYWN0bGFicy00ciwLEgRVc2VyIiJhX2E0Mjc1MD...</td>", 
        "      <td>2012-10-04 01:03:50.179000</td>", 
        "      <td> Moderator</td>", 
        "      <td> Agree</td>", 
        "      <td> 01:03:50.179000</td>", 
        "    </tr>", 
        "    <tr>", 
        "      <th>1</th>", 
        "      <td> ag1zfnJlYWN0bGFicy00ciwLEgRVc2VyIiJhX2E0Mjc1MD...</td>", 
        "      <td>2012-10-04 01:05:17.907000</td>", 
        "      <td>     Obama</td>", 
        "      <td>  Spin</td>", 
        "      <td> 01:05:17.907000</td>", 
        "    </tr>", 
        "  </tbody>", 
        "</table>", 
        "</div>"
       ], 
       "output_type": "pyout", 
       "prompt_number": 5, 
       "text": [
        "                                              UserID                       Time  \\", 
        "0  ag1zfnJlYWN0bGFicy00ciwLEgRVc2VyIiJhX2E0Mjc1MD... 2012-10-04 01:03:50.179000   ", 
        "1  ag1zfnJlYWN0bGFicy00ciwLEgRVc2VyIiJhX2E0Mjc1MD... 2012-10-04 01:05:17.907000   ", 
        "", 
        "  Reaction_who Reaction_what            start  ", 
        "0    Moderator         Agree  01:03:50.179000  ", 
        "1        Obama          Spin  01:05:17.907000  "
       ]
      }
     ], 
     "prompt_number": 5
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "## Transcript corpus"
     ]
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "Load the transcript file."
     ]
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [
      "c = pd.read_csv('corpora/oct3_coded_transcript_sync.csv')"
     ], 
     "language": "python", 
     "outputs": [], 
     "prompt_number": 6
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "Add a *start* column like we did for the reactions data."
     ]
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [
      "c['start'] = pd.to_datetime(c[\"Sync'd start\"]).apply(lambda t: pd.datetime.time(t))"
     ], 
     "language": "python", 
     "outputs": [], 
     "prompt_number": 7
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [
      "c['statement'] = range(len(c))"
     ], 
     "language": "python", 
     "outputs": [], 
     "prompt_number": 12
    }, 
    {
     "cell_type": "code", 
     "collapsed": false, 
     "input": [
      "c.head(2)"
     ], 
     "language": "python", 
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">", 
        "<table border=\"1\" class=\"dataframe\">", 
        "  <thead>", 
        "    <tr style=\"text-align: right;\">", 
        "      <th></th>", 
        "      <th>Sync'd start</th>", 
        "      <th>Sync'd end</th>", 
        "      <th>Speaker</th>", 
        "      <th>Transcript</th>", 
        "      <th>QuestionTopic</th>", 
        "      <th>Topic</th>", 
        "      <th>Frame</th>", 
        "      <th>Tone</th>", 
        "      <th>start</th>", 
        "      <th>statement</th>", 
        "    </tr>", 
        "  </thead>", 
        "  <tbody>", 
        "    <tr>", 
        "      <th>0</th>", 
        "      <td> 1:02:01</td>", 
        "      <td> 1:02:06</td>", 
        "      <td> 0</td>", 
        "      <td> Good evening from the Magness Arena at the Uni...</td>", 
        "      <td> 99</td>", 
        "      <td> 9</td>", 
        "      <td> 9</td>", 
        "      <td> 0</td>", 
        "      <td> 01:02:01</td>", 
        "      <td> 0</td>", 
        "    </tr>", 
        "    <tr>", 
        "      <th>1</th>", 
        "      <td> 1:02:06</td>", 
        "      <td> 1:02:09</td>", 
        "      <td> 0</td>", 
        "      <td>              I'm Jim Lehrer of the PBS NewsHour, </td>", 
        "      <td> 99</td>", 
        "      <td> 9</td>", 
        "      <td> 9</td>", 
        "      <td> 0</td>", 
        "      <td> 01:02:06</td>", 
        "      <td> 1</td>", 
        "    </tr>", 
        "  </tbody>", 
        "</table>", 
        "</div>"
       ], 
       "output_type": "pyout", 
       "prompt_number": 13, 
       "text": [
        "  Sync'd start Sync'd end  Speaker  \\", 
        "0      1:02:01    1:02:06        0   ", 
        "1      1:02:06    1:02:09        0   ", 
        "", 
        "                                          Transcript  QuestionTopic  Topic  Frame  \\", 
        "0  Good evening from the Magness Arena at the Uni...             99      9      9   ", 
        "1               I'm Jim Lehrer of the PBS NewsHour,              99      9      9   ", 
        "", 
        "   Tone     start  statement  ", 
        "0     0  01:02:01          0  ", 
        "1     0  01:02:06          1  "
       ]
      }
     ], 
     "prompt_number": 13
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "## Merge"
     ]
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "The *start* column will be shared, but the times will be out of order, so sort it on *start*."
     ]
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [
      "m = c.append(r)"
     ], 
     "language": "python", 
     "outputs": [], 
     "prompt_number": 15
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [
      "m = m.sort(columns='start')"
     ], 
     "language": "python", 
     "outputs": [], 
     "prompt_number": 16
    }, 
    {
     "cell_type": "code", 
     "collapsed": false, 
     "input": [
      "m.head(2)"
     ], 
     "language": "python", 
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">", 
        "<table border=\"1\" class=\"dataframe\">", 
        "  <thead>", 
        "    <tr style=\"text-align: right;\">", 
        "      <th></th>", 
        "      <th>Frame</th>", 
        "      <th>QuestionTopic</th>", 
        "      <th>Reaction_what</th>", 
        "      <th>Reaction_who</th>", 
        "      <th>Speaker</th>", 
        "      <th>Sync'd end</th>", 
        "      <th>Sync'd start</th>", 
        "      <th>Time</th>", 
        "      <th>Tone</th>", 
        "      <th>Topic</th>", 
        "      <th>Transcript</th>", 
        "      <th>UserID</th>", 
        "      <th>start</th>", 
        "      <th>statement</th>", 
        "    </tr>", 
        "  </thead>", 
        "  <tbody>", 
        "    <tr>", 
        "      <th>56965</th>", 
        "      <td>NaN</td>", 
        "      <td>NaN</td>", 
        "      <td> Disagree</td>", 
        "      <td> Moderator</td>", 
        "      <td>NaN</td>", 
        "      <td> NaN</td>", 
        "      <td> NaN</td>", 
        "      <td> 2012-10-04 00:30:00.429000</td>", 
        "      <td>NaN</td>", 
        "      <td>NaN</td>", 
        "      <td> NaN</td>", 
        "      <td> ag1zfnJlYWN0bGFicy01ciwLEgRVc2VyIiJhX2U4ODllMG...</td>", 
        "      <td> 00:30:00.429000</td>", 
        "      <td>NaN</td>", 
        "    </tr>", 
        "    <tr>", 
        "      <th>92001</th>", 
        "      <td>NaN</td>", 
        "      <td>NaN</td>", 
        "      <td>    Agree</td>", 
        "      <td>     Obama</td>", 
        "      <td>NaN</td>", 
        "      <td> NaN</td>", 
        "      <td> NaN</td>", 
        "      <td> 2012-10-04 00:30:02.337000</td>", 
        "      <td>NaN</td>", 
        "      <td>NaN</td>", 
        "      <td> NaN</td>", 
        "      <td> ag1zfnJlYWN0bGFicy0xciwLEgRVc2VyIiJhX2U5NDA5Zj...</td>", 
        "      <td> 00:30:02.337000</td>", 
        "      <td>NaN</td>", 
        "    </tr>", 
        "  </tbody>", 
        "</table>", 
        "</div>"
       ], 
       "output_type": "pyout", 
       "prompt_number": 17, 
       "text": [
        "       Frame  QuestionTopic Reaction_what Reaction_who  Speaker Sync'd end  \\", 
        "56965    NaN            NaN      Disagree    Moderator      NaN        NaN   ", 
        "92001    NaN            NaN         Agree        Obama      NaN        NaN   ", 
        "", 
        "      Sync'd start                        Time  Tone  Topic Transcript  \\", 
        "56965          NaN  2012-10-04 00:30:00.429000   NaN    NaN        NaN   ", 
        "92001          NaN  2012-10-04 00:30:02.337000   NaN    NaN        NaN   ", 
        "", 
        "                                                  UserID            start  \\", 
        "56965  ag1zfnJlYWN0bGFicy01ciwLEgRVc2VyIiJhX2U4ODllMG...  00:30:00.429000   ", 
        "92001  ag1zfnJlYWN0bGFicy0xciwLEgRVc2VyIiJhX2U5NDA5Zj...  00:30:02.337000   ", 
        "", 
        "       statement  ", 
        "56965        NaN  ", 
        "92001        NaN  "
       ]
      }
     ], 
     "prompt_number": 17
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "Assuming the times are already synchronized and that reactions happen *after* a statement is made, we will assign reactions between two transcript entries to the *prior* entry.  We do this by filling forward the blank columns of the table."
     ]
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [
      "m = m.fillna(method='ffill')"
     ], 
     "language": "python", 
     "outputs": [], 
     "prompt_number": 18
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "It appears that the app accepted clicks a 1/2 hour before the start of the debate (we will see that the same happened for a 1/2 hour at the end as well).  Let's get rid of those reactions at the start, which contain NAs."
     ]
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [
      "m2 = m.dropna()"
     ], 
     "language": "python", 
     "outputs": [], 
     "prompt_number": 19
    }, 
    {
     "cell_type": "code", 
     "collapsed": false, 
     "input": [
      "print len(m), len(m2), len(m)-len(m2)"
     ], 
     "language": "python", 
     "outputs": [
      {
       "output_type": "stream", 
       "stream": "stdout", 
       "text": [
        "194458 193033 1425"
       ]
      }
     ], 
     "prompt_number": 20
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "So that removed ~1400 reactions."
     ]
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "Now notice that the last reaction happened at ~3pm and the last transcript record happened at ~2:30p.  Let's get rid of those really late reactions."
     ]
    }, 
    {
     "cell_type": "code", 
     "collapsed": false, 
     "input": [
      "m2[[\"Sync'd start\",'Time','start','UserID']].tail(1)"
     ], 
     "language": "python", 
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">", 
        "<table border=\"1\" class=\"dataframe\">", 
        "  <thead>", 
        "    <tr style=\"text-align: right;\">", 
        "      <th></th>", 
        "      <th>Sync'd start</th>", 
        "      <th>Time</th>", 
        "      <th>start</th>", 
        "      <th>UserID</th>", 
        "    </tr>", 
        "  </thead>", 
        "  <tbody>", 
        "    <tr>", 
        "      <th>141777</th>", 
        "      <td> 2:32:41</td>", 
        "      <td> 2012-10-04 02:59:59.396000</td>", 
        "      <td> 02:59:59.396000</td>", 
        "      <td> ag1zfnJlYWN0bGFicy0yciwLEgRVc2VyIiJhX2NlOTc1M2...</td>", 
        "    </tr>", 
        "  </tbody>", 
        "</table>", 
        "</div>"
       ], 
       "output_type": "pyout", 
       "prompt_number": 21, 
       "text": [
        "       Sync'd start                        Time            start  \\", 
        "141777      2:32:41  2012-10-04 02:59:59.396000  02:59:59.396000   ", 
        "", 
        "                                                   UserID  ", 
        "141777  ag1zfnJlYWN0bGFicy0yciwLEgRVc2VyIiJhX2NlOTc1M2...  "
       ]
      }
     ], 
     "prompt_number": 21
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [
      "m3 = m2[m2.start < pd.datetime.time(pd.to_datetime('2:33'))]"
     ], 
     "language": "python", 
     "outputs": [], 
     "prompt_number": 22
    }, 
    {
     "cell_type": "code", 
     "collapsed": false, 
     "input": [
      "m3[[\"Sync'd start\",'Time','start','UserID']].tail(1)"
     ], 
     "language": "python", 
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">", 
        "<table border=\"1\" class=\"dataframe\">", 
        "  <thead>", 
        "    <tr style=\"text-align: right;\">", 
        "      <th></th>", 
        "      <th>Sync'd start</th>", 
        "      <th>Time</th>", 
        "      <th>start</th>", 
        "      <th>UserID</th>", 
        "    </tr>", 
        "  </thead>", 
        "  <tbody>", 
        "    <tr>", 
        "      <th>191634</th>", 
        "      <td> 2:32:41</td>", 
        "      <td> 2012-10-04 02:32:59.840000</td>", 
        "      <td> 02:32:59.840000</td>", 
        "      <td> ag1zfnJlYWN0bGFicy0zciwLEgRVc2VyIiJhXzBkNGVlZT...</td>", 
        "    </tr>", 
        "  </tbody>", 
        "</table>", 
        "</div>"
       ], 
       "output_type": "pyout", 
       "prompt_number": 23, 
       "text": [
        "       Sync'd start                        Time            start  \\", 
        "191634      2:32:41  2012-10-04 02:32:59.840000  02:32:59.840000   ", 
        "", 
        "                                                   UserID  ", 
        "191634  ag1zfnJlYWN0bGFicy0zciwLEgRVc2VyIiJhXzBkNGVlZT...  "
       ]
      }
     ], 
     "prompt_number": 23
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "## Turns"
     ]
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "Let's find the turns, i.e. the contiguous records where the same speaker was speaking."
     ]
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [
      "m3['turn'] = (m3.Speaker.shift(1) != m3.Speaker).astype(int).cumsum()"
     ], 
     "language": "python", 
     "outputs": [], 
     "prompt_number": 24
    }, 
    {
     "cell_type": "code", 
     "collapsed": false, 
     "input": [
      "m3[['Speaker','Transcript','start','turn','Reaction_what','Reaction_who','statement']].head(5)"
     ], 
     "language": "python", 
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">", 
        "<table border=\"1\" class=\"dataframe\">", 
        "  <thead>", 
        "    <tr style=\"text-align: right;\">", 
        "      <th></th>", 
        "      <th>Speaker</th>", 
        "      <th>Transcript</th>", 
        "      <th>start</th>", 
        "      <th>turn</th>", 
        "      <th>Reaction_what</th>", 
        "      <th>Reaction_who</th>", 
        "      <th>statement</th>", 
        "    </tr>", 
        "  </thead>", 
        "  <tbody>", 
        "    <tr>", 
        "      <th>0     </th>", 
        "      <td> 0</td>", 
        "      <td> Good evening from the Magness Arena at the Uni...</td>", 
        "      <td>        01:02:01</td>", 
        "      <td> 1</td>", 
        "      <td>    Agree</td>", 
        "      <td> Moderator</td>", 
        "      <td> 0</td>", 
        "    </tr>", 
        "    <tr>", 
        "      <th>56861 </th>", 
        "      <td> 0</td>", 
        "      <td> Good evening from the Magness Arena at the Uni...</td>", 
        "      <td> 01:02:01.401000</td>", 
        "      <td> 1</td>", 
        "      <td> Disagree</td>", 
        "      <td> Moderator</td>", 
        "      <td> 0</td>", 
        "    </tr>", 
        "    <tr>", 
        "      <th>60192 </th>", 
        "      <td> 0</td>", 
        "      <td> Good evening from the Magness Arena at the Uni...</td>", 
        "      <td> 01:02:02.011000</td>", 
        "      <td> 1</td>", 
        "      <td> Disagree</td>", 
        "      <td>    Romney</td>", 
        "      <td> 0</td>", 
        "    </tr>", 
        "    <tr>", 
        "      <th>117568</th>", 
        "      <td> 0</td>", 
        "      <td> Good evening from the Magness Arena at the Uni...</td>", 
        "      <td> 01:02:02.268000</td>", 
        "      <td> 1</td>", 
        "      <td>    Agree</td>", 
        "      <td> Moderator</td>", 
        "      <td> 0</td>", 
        "    </tr>", 
        "    <tr>", 
        "      <th>48989 </th>", 
        "      <td> 0</td>", 
        "      <td> Good evening from the Magness Arena at the Uni...</td>", 
        "      <td> 01:02:02.482000</td>", 
        "      <td> 1</td>", 
        "      <td> Disagree</td>", 
        "      <td> Moderator</td>", 
        "      <td> 0</td>", 
        "    </tr>", 
        "  </tbody>", 
        "</table>", 
        "</div>"
       ], 
       "output_type": "pyout", 
       "prompt_number": 30, 
       "text": [
        "        Speaker                                         Transcript  \\", 
        "0             0  Good evening from the Magness Arena at the Uni...   ", 
        "56861         0  Good evening from the Magness Arena at the Uni...   ", 
        "60192         0  Good evening from the Magness Arena at the Uni...   ", 
        "117568        0  Good evening from the Magness Arena at the Uni...   ", 
        "48989         0  Good evening from the Magness Arena at the Uni...   ", 
        "", 
        "                  start  turn Reaction_what Reaction_who  statement  ", 
        "0              01:02:01     1         Agree    Moderator          0  ", 
        "56861   01:02:01.401000     1      Disagree    Moderator          0  ", 
        "60192   01:02:02.011000     1      Disagree       Romney          0  ", 
        "117568  01:02:02.268000     1         Agree    Moderator          0  ", 
        "48989   01:02:02.482000     1      Disagree    Moderator          0  "
       ]
      }
     ], 
     "prompt_number": 30
    }, 
    {
     "cell_type": "code", 
     "collapsed": false, 
     "input": [
      "m3[['Speaker','Transcript','start','turn','Reaction_what','Reaction_who','statement']].tail(5)"
     ], 
     "language": "python", 
     "outputs": [
      {
       "html": [
        "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">", 
        "<table border=\"1\" class=\"dataframe\">", 
        "  <thead>", 
        "    <tr style=\"text-align: right;\">", 
        "      <th></th>", 
        "      <th>Speaker</th>", 
        "      <th>Transcript</th>", 
        "      <th>start</th>", 
        "      <th>turn</th>", 
        "      <th>Reaction_what</th>", 
        "      <th>Reaction_who</th>", 
        "      <th>statement</th>", 
        "    </tr>", 
        "  </thead>", 
        "  <tbody>", 
        "    <tr>", 
        "      <th>188581</th>", 
        "      <td> 0</td>", 
        "      <td> Thank you, and good night.</td>", 
        "      <td> 02:32:58.844000</td>", 
        "      <td> 190</td>", 
        "      <td> Disagree</td>", 
        "      <td>    Romney</td>", 
        "      <td> 1171</td>", 
        "    </tr>", 
        "    <tr>", 
        "      <th>178766</th>", 
        "      <td> 0</td>", 
        "      <td> Thank you, and good night.</td>", 
        "      <td> 02:32:58.890000</td>", 
        "      <td> 190</td>", 
        "      <td>    Agree</td>", 
        "      <td> Moderator</td>", 
        "      <td> 1171</td>", 
        "    </tr>", 
        "    <tr>", 
        "      <th>167646</th>", 
        "      <td> 0</td>", 
        "      <td> Thank you, and good night.</td>", 
        "      <td> 02:32:59.280000</td>", 
        "      <td> 190</td>", 
        "      <td>    Agree</td>", 
        "      <td>    Romney</td>", 
        "      <td> 1171</td>", 
        "    </tr>", 
        "    <tr>", 
        "      <th>68397 </th>", 
        "      <td> 0</td>", 
        "      <td> Thank you, and good night.</td>", 
        "      <td> 02:32:59.726000</td>", 
        "      <td> 190</td>", 
        "      <td> Disagree</td>", 
        "      <td>    Romney</td>", 
        "      <td> 1171</td>", 
        "    </tr>", 
        "    <tr>", 
        "      <th>191634</th>", 
        "      <td> 0</td>", 
        "      <td> Thank you, and good night.</td>", 
        "      <td> 02:32:59.840000</td>", 
        "      <td> 190</td>", 
        "      <td>    Agree</td>", 
        "      <td>    Romney</td>", 
        "      <td> 1171</td>", 
        "    </tr>", 
        "  </tbody>", 
        "</table>", 
        "</div>"
       ], 
       "output_type": "pyout", 
       "prompt_number": 29, 
       "text": [
        "        Speaker                  Transcript            start  turn Reaction_what  \\", 
        "188581        0  Thank you, and good night.  02:32:58.844000   190      Disagree   ", 
        "178766        0  Thank you, and good night.  02:32:58.890000   190         Agree   ", 
        "167646        0  Thank you, and good night.  02:32:59.280000   190         Agree   ", 
        "68397         0  Thank you, and good night.  02:32:59.726000   190      Disagree   ", 
        "191634        0  Thank you, and good night.  02:32:59.840000   190         Agree   ", 
        "", 
        "       Reaction_who  statement  ", 
        "188581       Romney       1171  ", 
        "178766    Moderator       1171  ", 
        "167646       Romney       1171  ", 
        "68397        Romney       1171  ", 
        "191634       Romney       1171  "
       ]
      }
     ], 
     "prompt_number": 29
    }, 
    {
     "cell_type": "markdown", 
     "source": [
      "It looks like early turns got lot of turns, then there was perhaps some dull spots, then more attention at the end."
     ]
    }, 
    {
     "cell_type": "code", 
     "collapsed": false, 
     "input": [
      "m3.groupby('turn').count().plot(legend=False)"
     ], 
     "language": "python", 
     "outputs": [
      {
       "output_type": "pyout", 
       "prompt_number": 27, 
       "text": [
        "<matplotlib.axes.AxesSubplot at 0x7a44b90>"
       ]
      }, 
      {
       "output_type": "display_data", 
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEKCAYAAAAfGVI8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nqKKKKqo4xlE1BFVUUUUVxzj+H9fMZrjbVOj4AAAAAElFTkSuQmCC\n"
      }
     ], 
     "prompt_number": 27
    }, 
    {
     "cell_type": "code", 
     "collapsed": true, 
     "input": [], 
     "language": "python", 
     "outputs": []
    }
   ]
  }
 ]
}