Marco Bag of words and tf-idf.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "tmp_df = pd.read_csv('/home/spark/Documents/Marco/train_yn.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_train = spark.createDataFrame(tmp_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+--------------------+--------------------+--------+-----------+----------+\n",
      "|            passages|               query|query_id| query_type|answers_yn|\n",
      "+--------------------+--------------------+--------+-----------+----------+\n",
      "|[Row(is_selected=...|was ronald reagan...|   19700|description|       Yes|\n",
      "|[Row(is_selected=...|do game shows pay...|   19749|description|       Yes|\n",
      "|[Row(is_selected=...|does peanut butte...|   19761|description|        No|\n",
      "|[Row(is_selected=...|can you take left...|   19762|description|        No|\n",
      "|[Row(is_selected=...|is armed assault ...|   19774|description|       Yes|\n",
      "|[Row(is_selected=...|is silverbeet the...|   19784|description|       Yes|\n",
      "|[Row(is_selected=...|are springtails h...|   19813|description|        No|\n",
      "|[Row(is_selected=...|are angora goats ...|   19829|description|       Yes|\n",
      "|[Row(is_selected=...|are angora goats ...|   19829|description|        No|\n",
      "|[Row(is_selected=...|does visa signatu...|   19830|description|       yes|\n",
      "+--------------------+--------------------+--------+-----------+----------+\n",
      "only showing top 10 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "df_train.show(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_train_sub = df_train.drop('passages').drop('query_id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "del(tmp_df)\n",
    "del(df_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+--------------------+-----------+----------+\n",
      "|               query| query_type|answers_yn|\n",
      "+--------------------+-----------+----------+\n",
      "|was ronald reagan...|description|       Yes|\n",
      "|do game shows pay...|description|       Yes|\n",
      "|does peanut butte...|description|        No|\n",
      "|can you take left...|description|        No|\n",
      "|is armed assault ...|description|       Yes|\n",
      "|is silverbeet the...|description|       Yes|\n",
      "|are springtails h...|description|        No|\n",
      "|are angora goats ...|description|       Yes|\n",
      "|are angora goats ...|description|        No|\n",
      "|does visa signatu...|description|       yes|\n",
      "+--------------------+-----------+----------+\n",
      "only showing top 10 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "df_train_sub.show(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "+--------------------+-----------+----------+\n",
      "|               query| query_type|answers_yn|\n",
      "+--------------------+-----------+----------+\n",
      "|was ronald reagan...|description|         1|\n",
      "|do game shows pay...|description|         1|\n",
      "|does peanut butte...|description|         0|\n",
      "|can you take left...|description|         0|\n",
      "|is armed assault ...|description|         1|\n",
      "|is silverbeet the...|description|         1|\n",
      "|are springtails h...|description|         0|\n",
      "|are angora goats ...|description|         1|\n",
      "|are angora goats ...|description|         0|\n",
      "|does visa signatu...|description|         1|\n",
      "|can we skip amica...|description|         1|\n",
      "|can i update the ...|description|         1|\n",
      "|is the new tesla ...|description|         1|\n",
      "|does delirium cau...|description|         1|\n",
      "|does claritin hel...|description|         1|\n",
      "|is mount everest ...|description|         1|\n",
      "|do hopi ear candl...|description|         1|\n",
      "|is facility explo...|description|         1|\n",
      "|if a potato gets ...|description|         1|\n",
      "|is psoriasis dang...|description|         1|\n",
      "+--------------------+-----------+----------+\n",
      "only showing top 20 rows\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# convert yes no to 1 o\n",
    "\n",
    "df_train_sub=df_train_sub.na.replace(['yes', 'no'], ['Yes', 'No'], 'answers_yn')\n",
    "df_train_sub=df_train_sub.na.replace(['Yes', 'No'], ['1', '0'], 'answers_yn')\n",
    "\n",
    "df_train_sub.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import nltk"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "sentence = df_train_sub.select('query').limit(1).collect()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "was ronald reagan a democrat\n",
      "['was', 'ronald', 'reagan', 'a', 'democrat']\n"
     ]
    }
   ],
   "source": [
    "print sentence[0][0]\n",
    "token = nltk.word_tokenize(str(sentence[0][0]))\n",
    "print token"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "FreqDist({'a': 1, 'democrat': 1, 'reagan': 1, 'ronald': 1, 'was': 1})"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nltk.FreqDist(token)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_train_sub.createOrReplaceTempView('tmp')\n",
    "# df_1 = spark.sql('select * from tmp limit 1')\n",
    "# df_train_sub.limit(2).show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def term_fre(doc):\n",
    "    token = nltk.word_tokenize(str(doc[0]))\n",
    "    tf = nltk.FreqDist(token)\n",
    "    return dict(tf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "lst_tf = []\n",
    "for i in df_train_sub.select('query').collect():\n",
    "    lst_tf.append(term_fre(i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import h2o"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Checking whether there is an H2O instance running at http://localhost:54321..... not found.\n",
      "Attempting to start a local H2O server...\n",
      "  Java Version: java version \"1.8.0_111\"; Java(TM) SE Runtime Environment (build 1.8.0_111-b14); Java HotSpot(TM) 64-Bit Server VM (build 25.111-b14, mixed mode)\n",
      "  Starting server from /home/spark/anaconda2/lib/python2.7/site-packages/h2o/backend/bin/h2o.jar\n",
      "  Ice root: /tmp/tmpoVYsnO\n",
      "  JVM stdout: /tmp/tmpoVYsnO/h2o_spark_started_from_python.out\n",
      "  JVM stderr: /tmp/tmpoVYsnO/h2o_spark_started_from_python.err\n",
      "  Server is running at http://127.0.0.1:54321\n",
      "Connecting to H2O server at http://127.0.0.1:54321... successful.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div style=\"overflow:auto\"><table style=\"width:50%\"><tr><td>H2O cluster uptime:</td>\n",
       "<td>05 secs</td></tr>\n",
       "<tr><td>H2O cluster version:</td>\n",
       "<td>3.10.0.10</td></tr>\n",
       "<tr><td>H2O cluster version age:</td>\n",
       "<td>1 month and 28 days </td></tr>\n",
       "<tr><td>H2O cluster name:</td>\n",
       "<td>H2O_from_python_spark_ijb36m</td></tr>\n",
       "<tr><td>H2O cluster total nodes:</td>\n",
       "<td>1</td></tr>\n",
       "<tr><td>H2O cluster free memory:</td>\n",
       "<td>875 Mb</td></tr>\n",
       "<tr><td>H2O cluster total cores:</td>\n",
       "<td>4</td></tr>\n",
       "<tr><td>H2O cluster allowed cores:</td>\n",
       "<td>4</td></tr>\n",
       "<tr><td>H2O cluster status:</td>\n",
       "<td>accepting new members, healthy</td></tr>\n",
       "<tr><td>H2O connection url:</td>\n",
       "<td>http://127.0.0.1:54321</td></tr>\n",
       "<tr><td>H2O connection proxy:</td>\n",
       "<td>None</td></tr>\n",
       "<tr><td>Python version:</td>\n",
       "<td>2.7.12 final</td></tr></table></div>"
      ],
      "text/plain": [
       "--------------------------  ------------------------------\n",
       "H2O cluster uptime:         05 secs\n",
       "H2O cluster version:        3.10.0.10\n",
       "H2O cluster version age:    1 month and 28 days\n",
       "H2O cluster name:           H2O_from_python_spark_ijb36m\n",
       "H2O cluster total nodes:    1\n",
       "H2O cluster free memory:    875 Mb\n",
       "H2O cluster total cores:    4\n",
       "H2O cluster allowed cores:  4\n",
       "H2O cluster status:         accepting new members, healthy\n",
       "H2O connection url:         http://127.0.0.1:54321\n",
       "H2O connection proxy:\n",
       "Python version:             2.7.12 final\n",
       "--------------------------  ------------------------------"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "h2o.init()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parse progress: |█████████████████████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "hdf_tf=h2o.H2OFrame(lst_tf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Parse progress: |█████████████████████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "hdf_target=h2o.H2OFrame(df_train_sub.select('answers_yn').collect())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "hdf = h2o.H2OFrame.cbind(hdf_tf,h2o.H2OFrame.asfactor(hdf_target))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead>\n",
       "<tr><th>C1                                                                                                                                                             </th><th style=\"text-align: right;\">  C10</th></tr>\n",
       "</thead>\n",
       "<tbody>\n",
       "<tr><td>{'ronald': 1, 'was': 1, 'reagan': 1, 'a': 1, 'democrat': 1}                                                                                                    </td><td style=\"text-align: right;\">    1</td></tr>\n",
       "<tr><td>{'do': 1, 'game': 1, 'contestants': 1, 'pay': 1, 'their': 1, 'shows': 1}                                                                                       </td><td style=\"text-align: right;\">    1</td></tr>\n",
       "<tr><td>{'butter': 1, 'peanut': 1, 'contain': 1, 'does': 1, 'cholesterol': 1}                                                                                          </td><td style=\"text-align: right;\">    0</td></tr>\n",
       "<tr><td>{'a': 1, 'on': 1, 'lane': 1, 'right': 1, 'to': 1, 'carriageway': 1, 'hand': 1, 'turn': 1, 'can': 1, 'dual': 1, 'you': 1, 'left': 1, 'roundabout': 1, 'take': 1}</td><td style=\"text-align: right;\">    0</td></tr>\n",
       "<tr><td>{'a': 1, 'felony': 1, 'is': 1, 'armed': 1, 'assault': 1}                                                                                                       </td><td style=\"text-align: right;\">    1</td></tr>\n",
       "<tr><td>{'the': 1, 'as': 1, 'is': 1, 'silverbeet': 1, 'same': 1, 'chard': 1}                                                                                           </td><td style=\"text-align: right;\">    1</td></tr>\n",
       "<tr><td>{'harmful': 1, 'springtails': 1, 'are': 1}                                                                                                                     </td><td style=\"text-align: right;\">    0</td></tr>\n",
       "<tr><td>{'good': 1, 'meat': 1, 'for': 1, 'angora': 1, 'goats': 1, 'are': 1}                                                                                            </td><td style=\"text-align: right;\">    1</td></tr>\n",
       "<tr><td>{'good': 1, 'meat': 1, 'for': 1, 'angora': 1, 'goats': 1, 'are': 1}                                                                                            </td><td style=\"text-align: right;\">    0</td></tr>\n",
       "<tr><td>{'does': 1, 'have': 1, 'signature': 1, 'assistance': 1, 'roadside': 1, 'visa': 1}                                                                              </td><td style=\"text-align: right;\">    1</td></tr>\n",
       "</tbody>\n",
       "</table>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": []
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hdf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "hdf_train, hdf_validate=hdf.split_frame(ratios=[0.75], seed =1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from h2o.estimators.naive_bayes import H2ONaiveBayesEstimator\n",
    "from h2o.estimators.gbm import H2OGradientBoostingEstimator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "naivebayes Model Build progress: |████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "nb =H2ONaiveBayesEstimator(seed =1)\n",
    "nb.train(x = hdf.col_names[0], y = hdf.col_names[1], training_frame=hdf_train, validation_frame=hdf_validate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.006909803758099552"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nb.auc()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gbm Model Build progress: |███████████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "gbm_model = H2OGradientBoostingEstimator(seed =1)\n",
    "gbm_model.train(x = hdf.col_names[0], y = hdf.col_names[1], training_frame=hdf_train, validation_frame=hdf_validate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC:  0.751706058627\n",
      "accuracy:  [[0.639517716488152, 0.795816939324912]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAicAAAGHCAYAAABrpPKuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd50d307550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print \"AUC: \", gbm_model.auc()\n",
    "print \"accuracy: \", gbm_model.accuracy()\n",
    "gbm_model.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gbm Model Build progress: |███████████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "gbm_model_2 = H2OGradientBoostingEstimator(ntrees=65,seed=1)\n",
    "gbm_model_2.train(x = hdf.col_names[0], y = hdf.col_names[1], training_frame=hdf_train, validation_frame=hdf_validate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC:  0.764013672553\n",
      "accuracy:  [[0.6404097079885708, 0.7978877614412921]]\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>timestamp</th>\n",
       "      <th>duration</th>\n",
       "      <th>number_of_trees</th>\n",
       "      <th>training_rmse</th>\n",
       "      <th>training_logloss</th>\n",
       "      <th>training_auc</th>\n",
       "      <th>training_lift</th>\n",
       "      <th>training_classification_error</th>\n",
       "      <th>validation_rmse</th>\n",
       "      <th>validation_logloss</th>\n",
       "      <th>validation_auc</th>\n",
       "      <th>validation_lift</th>\n",
       "      <th>validation_classification_error</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:19</td>\n",
       "      <td>0.026 sec</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.430320</td>\n",
       "      <td>0.557216</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.245392</td>\n",
       "      <td>0.424193</td>\n",
       "      <td>0.545668</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.235112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:19</td>\n",
       "      <td>0.066 sec</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.429911</td>\n",
       "      <td>0.556285</td>\n",
       "      <td>0.534468</td>\n",
       "      <td>1.208947</td>\n",
       "      <td>0.245392</td>\n",
       "      <td>0.423927</td>\n",
       "      <td>0.545073</td>\n",
       "      <td>0.523198</td>\n",
       "      <td>1.094551</td>\n",
       "      <td>0.235112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:19</td>\n",
       "      <td>0.104 sec</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.429597</td>\n",
       "      <td>0.555566</td>\n",
       "      <td>0.554162</td>\n",
       "      <td>1.204720</td>\n",
       "      <td>0.245392</td>\n",
       "      <td>0.423883</td>\n",
       "      <td>0.544978</td>\n",
       "      <td>0.528809</td>\n",
       "      <td>1.099388</td>\n",
       "      <td>0.235112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:19</td>\n",
       "      <td>0.146 sec</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.429143</td>\n",
       "      <td>0.554534</td>\n",
       "      <td>0.573588</td>\n",
       "      <td>1.244058</td>\n",
       "      <td>0.243943</td>\n",
       "      <td>0.423810</td>\n",
       "      <td>0.544818</td>\n",
       "      <td>0.529089</td>\n",
       "      <td>1.032142</td>\n",
       "      <td>0.235112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:19</td>\n",
       "      <td>0.196 sec</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.428790</td>\n",
       "      <td>0.553734</td>\n",
       "      <td>0.577178</td>\n",
       "      <td>1.244058</td>\n",
       "      <td>0.243736</td>\n",
       "      <td>0.423780</td>\n",
       "      <td>0.544765</td>\n",
       "      <td>0.517508</td>\n",
       "      <td>1.055961</td>\n",
       "      <td>0.235112</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:19</td>\n",
       "      <td>0.268 sec</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.423587</td>\n",
       "      <td>0.542778</td>\n",
       "      <td>0.647962</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.214951</td>\n",
       "      <td>0.423436</td>\n",
       "      <td>0.544031</td>\n",
       "      <td>0.531633</td>\n",
       "      <td>1.045904</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:19</td>\n",
       "      <td>0.296 sec</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.423280</td>\n",
       "      <td>0.542106</td>\n",
       "      <td>0.649988</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.214951</td>\n",
       "      <td>0.423475</td>\n",
       "      <td>0.544113</td>\n",
       "      <td>0.528778</td>\n",
       "      <td>1.111273</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:19</td>\n",
       "      <td>0.324 sec</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.419160</td>\n",
       "      <td>0.534025</td>\n",
       "      <td>0.663210</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.215158</td>\n",
       "      <td>0.423408</td>\n",
       "      <td>0.543972</td>\n",
       "      <td>0.530064</td>\n",
       "      <td>1.111273</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:19</td>\n",
       "      <td>0.354 sec</td>\n",
       "      <td>8.0</td>\n",
       "      <td>0.419000</td>\n",
       "      <td>0.533667</td>\n",
       "      <td>0.666181</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.215158</td>\n",
       "      <td>0.423361</td>\n",
       "      <td>0.543880</td>\n",
       "      <td>0.530221</td>\n",
       "      <td>1.111273</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:19</td>\n",
       "      <td>0.386 sec</td>\n",
       "      <td>9.0</td>\n",
       "      <td>0.418779</td>\n",
       "      <td>0.533189</td>\n",
       "      <td>0.666333</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.215780</td>\n",
       "      <td>0.423431</td>\n",
       "      <td>0.544034</td>\n",
       "      <td>0.529944</td>\n",
       "      <td>1.111273</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>0.420 sec</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.418550</td>\n",
       "      <td>0.532715</td>\n",
       "      <td>0.667587</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.215780</td>\n",
       "      <td>0.423425</td>\n",
       "      <td>0.544014</td>\n",
       "      <td>0.532875</td>\n",
       "      <td>1.111273</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>0.458 sec</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0.418422</td>\n",
       "      <td>0.532430</td>\n",
       "      <td>0.668379</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.215780</td>\n",
       "      <td>0.423422</td>\n",
       "      <td>0.544019</td>\n",
       "      <td>0.533188</td>\n",
       "      <td>1.111273</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>0.495 sec</td>\n",
       "      <td>12.0</td>\n",
       "      <td>0.418257</td>\n",
       "      <td>0.532082</td>\n",
       "      <td>0.667284</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.215780</td>\n",
       "      <td>0.423504</td>\n",
       "      <td>0.544203</td>\n",
       "      <td>0.532164</td>\n",
       "      <td>1.111273</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>0.536 sec</td>\n",
       "      <td>13.0</td>\n",
       "      <td>0.414964</td>\n",
       "      <td>0.525766</td>\n",
       "      <td>0.673555</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.213916</td>\n",
       "      <td>0.423684</td>\n",
       "      <td>0.544515</td>\n",
       "      <td>0.533410</td>\n",
       "      <td>1.143958</td>\n",
       "      <td>0.233251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>0.579 sec</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.414866</td>\n",
       "      <td>0.525536</td>\n",
       "      <td>0.674057</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.213916</td>\n",
       "      <td>0.423717</td>\n",
       "      <td>0.544591</td>\n",
       "      <td>0.533208</td>\n",
       "      <td>1.143958</td>\n",
       "      <td>0.233251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>0.638 sec</td>\n",
       "      <td>15.0</td>\n",
       "      <td>0.410607</td>\n",
       "      <td>0.516968</td>\n",
       "      <td>0.701361</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.205426</td>\n",
       "      <td>0.423949</td>\n",
       "      <td>0.545018</td>\n",
       "      <td>0.527538</td>\n",
       "      <td>1.143958</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>0.689 sec</td>\n",
       "      <td>16.0</td>\n",
       "      <td>0.410493</td>\n",
       "      <td>0.516769</td>\n",
       "      <td>0.700668</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.205426</td>\n",
       "      <td>0.423957</td>\n",
       "      <td>0.545039</td>\n",
       "      <td>0.528587</td>\n",
       "      <td>1.176642</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>0.759 sec</td>\n",
       "      <td>17.0</td>\n",
       "      <td>0.410408</td>\n",
       "      <td>0.516554</td>\n",
       "      <td>0.701864</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.205426</td>\n",
       "      <td>0.423962</td>\n",
       "      <td>0.545052</td>\n",
       "      <td>0.529457</td>\n",
       "      <td>1.176642</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>0.825 sec</td>\n",
       "      <td>18.0</td>\n",
       "      <td>0.410272</td>\n",
       "      <td>0.516251</td>\n",
       "      <td>0.705731</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.206461</td>\n",
       "      <td>0.423976</td>\n",
       "      <td>0.545089</td>\n",
       "      <td>0.528075</td>\n",
       "      <td>1.176642</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>0.896 sec</td>\n",
       "      <td>19.0</td>\n",
       "      <td>0.409983</td>\n",
       "      <td>0.515659</td>\n",
       "      <td>0.709709</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.206461</td>\n",
       "      <td>0.424072</td>\n",
       "      <td>0.545295</td>\n",
       "      <td>0.527035</td>\n",
       "      <td>1.176642</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>0.954 sec</td>\n",
       "      <td>20.0</td>\n",
       "      <td>0.409880</td>\n",
       "      <td>0.515422</td>\n",
       "      <td>0.708876</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.205840</td>\n",
       "      <td>0.424107</td>\n",
       "      <td>0.545375</td>\n",
       "      <td>0.526547</td>\n",
       "      <td>1.176642</td>\n",
       "      <td>0.233871</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>1.014 sec</td>\n",
       "      <td>21.0</td>\n",
       "      <td>0.409746</td>\n",
       "      <td>0.515140</td>\n",
       "      <td>0.709916</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.206461</td>\n",
       "      <td>0.424127</td>\n",
       "      <td>0.545426</td>\n",
       "      <td>0.525577</td>\n",
       "      <td>1.176642</td>\n",
       "      <td>0.231390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>1.076 sec</td>\n",
       "      <td>22.0</td>\n",
       "      <td>0.409636</td>\n",
       "      <td>0.514917</td>\n",
       "      <td>0.710402</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.206461</td>\n",
       "      <td>0.424129</td>\n",
       "      <td>0.545437</td>\n",
       "      <td>0.526939</td>\n",
       "      <td>1.176642</td>\n",
       "      <td>0.231390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>1.141 sec</td>\n",
       "      <td>23.0</td>\n",
       "      <td>0.409444</td>\n",
       "      <td>0.514533</td>\n",
       "      <td>0.712374</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.206461</td>\n",
       "      <td>0.424153</td>\n",
       "      <td>0.545484</td>\n",
       "      <td>0.526702</td>\n",
       "      <td>1.176642</td>\n",
       "      <td>0.231390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>1.210 sec</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0.409249</td>\n",
       "      <td>0.514114</td>\n",
       "      <td>0.715152</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.206461</td>\n",
       "      <td>0.424326</td>\n",
       "      <td>0.545889</td>\n",
       "      <td>0.525134</td>\n",
       "      <td>1.089484</td>\n",
       "      <td>0.231390</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>1.294 sec</td>\n",
       "      <td>25.0</td>\n",
       "      <td>0.408619</td>\n",
       "      <td>0.512849</td>\n",
       "      <td>0.714484</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.206461</td>\n",
       "      <td>0.424486</td>\n",
       "      <td>0.546146</td>\n",
       "      <td>0.524620</td>\n",
       "      <td>1.089484</td>\n",
       "      <td>0.230769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:20</td>\n",
       "      <td>1.376 sec</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0.408566</td>\n",
       "      <td>0.512702</td>\n",
       "      <td>0.715130</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.205840</td>\n",
       "      <td>0.424499</td>\n",
       "      <td>0.546179</td>\n",
       "      <td>0.526141</td>\n",
       "      <td>1.129101</td>\n",
       "      <td>0.230769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:21</td>\n",
       "      <td>1.441 sec</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0.408480</td>\n",
       "      <td>0.512532</td>\n",
       "      <td>0.714349</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.206461</td>\n",
       "      <td>0.424511</td>\n",
       "      <td>0.546206</td>\n",
       "      <td>0.524675</td>\n",
       "      <td>1.182868</td>\n",
       "      <td>0.230769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:21</td>\n",
       "      <td>1.520 sec</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0.405189</td>\n",
       "      <td>0.506004</td>\n",
       "      <td>0.719823</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.205011</td>\n",
       "      <td>0.424887</td>\n",
       "      <td>0.546866</td>\n",
       "      <td>0.524130</td>\n",
       "      <td>1.167304</td>\n",
       "      <td>0.230769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:21</td>\n",
       "      <td>1.619 sec</td>\n",
       "      <td>29.0</td>\n",
       "      <td>0.405042</td>\n",
       "      <td>0.505694</td>\n",
       "      <td>0.721928</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.205011</td>\n",
       "      <td>0.424973</td>\n",
       "      <td>0.547060</td>\n",
       "      <td>0.524680</td>\n",
       "      <td>1.167304</td>\n",
       "      <td>0.230769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:21</td>\n",
       "      <td>1.718 sec</td>\n",
       "      <td>30.0</td>\n",
       "      <td>0.404936</td>\n",
       "      <td>0.505472</td>\n",
       "      <td>0.721632</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.204183</td>\n",
       "      <td>0.424982</td>\n",
       "      <td>0.547058</td>\n",
       "      <td>0.523161</td>\n",
       "      <td>1.129101</td>\n",
       "      <td>0.230769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:21</td>\n",
       "      <td>1.828 sec</td>\n",
       "      <td>31.0</td>\n",
       "      <td>0.404846</td>\n",
       "      <td>0.505274</td>\n",
       "      <td>0.723259</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.204183</td>\n",
       "      <td>0.425061</td>\n",
       "      <td>0.547247</td>\n",
       "      <td>0.523662</td>\n",
       "      <td>1.129101</td>\n",
       "      <td>0.230769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:21</td>\n",
       "      <td>1.914 sec</td>\n",
       "      <td>32.0</td>\n",
       "      <td>0.404662</td>\n",
       "      <td>0.504913</td>\n",
       "      <td>0.723827</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.204183</td>\n",
       "      <td>0.425141</td>\n",
       "      <td>0.547425</td>\n",
       "      <td>0.524115</td>\n",
       "      <td>1.129101</td>\n",
       "      <td>0.230769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:21</td>\n",
       "      <td>1.998 sec</td>\n",
       "      <td>33.0</td>\n",
       "      <td>0.404584</td>\n",
       "      <td>0.504723</td>\n",
       "      <td>0.724089</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.204183</td>\n",
       "      <td>0.425221</td>\n",
       "      <td>0.547636</td>\n",
       "      <td>0.523595</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.230769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:21</td>\n",
       "      <td>2.084 sec</td>\n",
       "      <td>34.0</td>\n",
       "      <td>0.404163</td>\n",
       "      <td>0.503797</td>\n",
       "      <td>0.724608</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.203562</td>\n",
       "      <td>0.425538</td>\n",
       "      <td>0.548172</td>\n",
       "      <td>0.524881</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:21</td>\n",
       "      <td>2.171 sec</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0.404019</td>\n",
       "      <td>0.503504</td>\n",
       "      <td>0.725672</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.203562</td>\n",
       "      <td>0.425508</td>\n",
       "      <td>0.548109</td>\n",
       "      <td>0.525500</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:21</td>\n",
       "      <td>2.257 sec</td>\n",
       "      <td>36.0</td>\n",
       "      <td>0.403853</td>\n",
       "      <td>0.503119</td>\n",
       "      <td>0.728216</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.203562</td>\n",
       "      <td>0.425526</td>\n",
       "      <td>0.548162</td>\n",
       "      <td>0.528145</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:21</td>\n",
       "      <td>2.358 sec</td>\n",
       "      <td>37.0</td>\n",
       "      <td>0.403707</td>\n",
       "      <td>0.502795</td>\n",
       "      <td>0.729152</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.203562</td>\n",
       "      <td>0.425670</td>\n",
       "      <td>0.548505</td>\n",
       "      <td>0.527023</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:22</td>\n",
       "      <td>2.472 sec</td>\n",
       "      <td>38.0</td>\n",
       "      <td>0.403244</td>\n",
       "      <td>0.501801</td>\n",
       "      <td>0.738898</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.203562</td>\n",
       "      <td>0.425597</td>\n",
       "      <td>0.548332</td>\n",
       "      <td>0.527638</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:22</td>\n",
       "      <td>2.563 sec</td>\n",
       "      <td>39.0</td>\n",
       "      <td>0.403127</td>\n",
       "      <td>0.501543</td>\n",
       "      <td>0.739736</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.203562</td>\n",
       "      <td>0.425687</td>\n",
       "      <td>0.548520</td>\n",
       "      <td>0.525857</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:22</td>\n",
       "      <td>2.665 sec</td>\n",
       "      <td>40.0</td>\n",
       "      <td>0.403063</td>\n",
       "      <td>0.501382</td>\n",
       "      <td>0.740153</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.203562</td>\n",
       "      <td>0.425768</td>\n",
       "      <td>0.548741</td>\n",
       "      <td>0.525409</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:22</td>\n",
       "      <td>2.795 sec</td>\n",
       "      <td>41.0</td>\n",
       "      <td>0.402994</td>\n",
       "      <td>0.501229</td>\n",
       "      <td>0.740341</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.204183</td>\n",
       "      <td>0.425815</td>\n",
       "      <td>0.548866</td>\n",
       "      <td>0.524513</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:22</td>\n",
       "      <td>2.912 sec</td>\n",
       "      <td>42.0</td>\n",
       "      <td>0.400416</td>\n",
       "      <td>0.496039</td>\n",
       "      <td>0.741428</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.203562</td>\n",
       "      <td>0.425138</td>\n",
       "      <td>0.547526</td>\n",
       "      <td>0.526685</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:22</td>\n",
       "      <td>3.011 sec</td>\n",
       "      <td>43.0</td>\n",
       "      <td>0.400263</td>\n",
       "      <td>0.495748</td>\n",
       "      <td>0.741991</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.203562</td>\n",
       "      <td>0.425159</td>\n",
       "      <td>0.547562</td>\n",
       "      <td>0.527785</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:22</td>\n",
       "      <td>3.116 sec</td>\n",
       "      <td>44.0</td>\n",
       "      <td>0.399984</td>\n",
       "      <td>0.495062</td>\n",
       "      <td>0.742246</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.203562</td>\n",
       "      <td>0.425517</td>\n",
       "      <td>0.548172</td>\n",
       "      <td>0.528206</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:22</td>\n",
       "      <td>3.219 sec</td>\n",
       "      <td>45.0</td>\n",
       "      <td>0.399889</td>\n",
       "      <td>0.494851</td>\n",
       "      <td>0.742588</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.203562</td>\n",
       "      <td>0.425610</td>\n",
       "      <td>0.548367</td>\n",
       "      <td>0.526407</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:22</td>\n",
       "      <td>3.327 sec</td>\n",
       "      <td>46.0</td>\n",
       "      <td>0.399811</td>\n",
       "      <td>0.494677</td>\n",
       "      <td>0.742365</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.204183</td>\n",
       "      <td>0.425649</td>\n",
       "      <td>0.548462</td>\n",
       "      <td>0.526417</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:23</td>\n",
       "      <td>3.433 sec</td>\n",
       "      <td>47.0</td>\n",
       "      <td>0.399647</td>\n",
       "      <td>0.494266</td>\n",
       "      <td>0.744502</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.204183</td>\n",
       "      <td>0.425646</td>\n",
       "      <td>0.548472</td>\n",
       "      <td>0.526565</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:23</td>\n",
       "      <td>3.554 sec</td>\n",
       "      <td>48.0</td>\n",
       "      <td>0.399384</td>\n",
       "      <td>0.493685</td>\n",
       "      <td>0.749437</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.204183</td>\n",
       "      <td>0.425620</td>\n",
       "      <td>0.548415</td>\n",
       "      <td>0.526746</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:23</td>\n",
       "      <td>3.750 sec</td>\n",
       "      <td>49.0</td>\n",
       "      <td>0.399013</td>\n",
       "      <td>0.492926</td>\n",
       "      <td>0.751265</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.204183</td>\n",
       "      <td>0.425580</td>\n",
       "      <td>0.548325</td>\n",
       "      <td>0.528832</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:23</td>\n",
       "      <td>3.963 sec</td>\n",
       "      <td>50.0</td>\n",
       "      <td>0.398933</td>\n",
       "      <td>0.492734</td>\n",
       "      <td>0.751706</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.204183</td>\n",
       "      <td>0.425632</td>\n",
       "      <td>0.548453</td>\n",
       "      <td>0.528044</td>\n",
       "      <td>1.153571</td>\n",
       "      <td>0.232630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td></td>\n",
       "      <td>2017-01-05 10:37:23</td>\n",
       "      <td>4.143 sec</td>\n",
       "      <td>65.0</td>\n",
       "      <td>0.396355</td>\n",
       "      <td>0.487219</td>\n",
       "      <td>0.764014</td>\n",
       "      <td>1.325192</td>\n",
       "      <td>0.202112</td>\n",
       "      <td>0.426571</td>\n",
       "      <td>0.550223</td>\n",
       "      <td>0.533586</td>\n",
       "      <td>0.950822</td>\n",
       "      <td>0.233251</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                timestamp    duration  number_of_trees  training_rmse  \\\n",
       "0     2017-01-05 10:37:19   0.026 sec              0.0       0.430320   \n",
       "1     2017-01-05 10:37:19   0.066 sec              1.0       0.429911   \n",
       "2     2017-01-05 10:37:19   0.104 sec              2.0       0.429597   \n",
       "3     2017-01-05 10:37:19   0.146 sec              3.0       0.429143   \n",
       "4     2017-01-05 10:37:19   0.196 sec              4.0       0.428790   \n",
       "5     2017-01-05 10:37:19   0.268 sec              5.0       0.423587   \n",
       "6     2017-01-05 10:37:19   0.296 sec              6.0       0.423280   \n",
       "7     2017-01-05 10:37:19   0.324 sec              7.0       0.419160   \n",
       "8     2017-01-05 10:37:19   0.354 sec              8.0       0.419000   \n",
       "9     2017-01-05 10:37:19   0.386 sec              9.0       0.418779   \n",
       "10    2017-01-05 10:37:20   0.420 sec             10.0       0.418550   \n",
       "11    2017-01-05 10:37:20   0.458 sec             11.0       0.418422   \n",
       "12    2017-01-05 10:37:20   0.495 sec             12.0       0.418257   \n",
       "13    2017-01-05 10:37:20   0.536 sec             13.0       0.414964   \n",
       "14    2017-01-05 10:37:20   0.579 sec             14.0       0.414866   \n",
       "15    2017-01-05 10:37:20   0.638 sec             15.0       0.410607   \n",
       "16    2017-01-05 10:37:20   0.689 sec             16.0       0.410493   \n",
       "17    2017-01-05 10:37:20   0.759 sec             17.0       0.410408   \n",
       "18    2017-01-05 10:37:20   0.825 sec             18.0       0.410272   \n",
       "19    2017-01-05 10:37:20   0.896 sec             19.0       0.409983   \n",
       "20    2017-01-05 10:37:20   0.954 sec             20.0       0.409880   \n",
       "21    2017-01-05 10:37:20   1.014 sec             21.0       0.409746   \n",
       "22    2017-01-05 10:37:20   1.076 sec             22.0       0.409636   \n",
       "23    2017-01-05 10:37:20   1.141 sec             23.0       0.409444   \n",
       "24    2017-01-05 10:37:20   1.210 sec             24.0       0.409249   \n",
       "25    2017-01-05 10:37:20   1.294 sec             25.0       0.408619   \n",
       "26    2017-01-05 10:37:20   1.376 sec             26.0       0.408566   \n",
       "27    2017-01-05 10:37:21   1.441 sec             27.0       0.408480   \n",
       "28    2017-01-05 10:37:21   1.520 sec             28.0       0.405189   \n",
       "29    2017-01-05 10:37:21   1.619 sec             29.0       0.405042   \n",
       "30    2017-01-05 10:37:21   1.718 sec             30.0       0.404936   \n",
       "31    2017-01-05 10:37:21   1.828 sec             31.0       0.404846   \n",
       "32    2017-01-05 10:37:21   1.914 sec             32.0       0.404662   \n",
       "33    2017-01-05 10:37:21   1.998 sec             33.0       0.404584   \n",
       "34    2017-01-05 10:37:21   2.084 sec             34.0       0.404163   \n",
       "35    2017-01-05 10:37:21   2.171 sec             35.0       0.404019   \n",
       "36    2017-01-05 10:37:21   2.257 sec             36.0       0.403853   \n",
       "37    2017-01-05 10:37:21   2.358 sec             37.0       0.403707   \n",
       "38    2017-01-05 10:37:22   2.472 sec             38.0       0.403244   \n",
       "39    2017-01-05 10:37:22   2.563 sec             39.0       0.403127   \n",
       "40    2017-01-05 10:37:22   2.665 sec             40.0       0.403063   \n",
       "41    2017-01-05 10:37:22   2.795 sec             41.0       0.402994   \n",
       "42    2017-01-05 10:37:22   2.912 sec             42.0       0.400416   \n",
       "43    2017-01-05 10:37:22   3.011 sec             43.0       0.400263   \n",
       "44    2017-01-05 10:37:22   3.116 sec             44.0       0.399984   \n",
       "45    2017-01-05 10:37:22   3.219 sec             45.0       0.399889   \n",
       "46    2017-01-05 10:37:22   3.327 sec             46.0       0.399811   \n",
       "47    2017-01-05 10:37:23   3.433 sec             47.0       0.399647   \n",
       "48    2017-01-05 10:37:23   3.554 sec             48.0       0.399384   \n",
       "49    2017-01-05 10:37:23   3.750 sec             49.0       0.399013   \n",
       "50    2017-01-05 10:37:23   3.963 sec             50.0       0.398933   \n",
       "51    2017-01-05 10:37:23   4.143 sec             65.0       0.396355   \n",
       "\n",
       "    training_logloss  training_auc  training_lift  \\\n",
       "0           0.557216      0.500000       1.000000   \n",
       "1           0.556285      0.534468       1.208947   \n",
       "2           0.555566      0.554162       1.204720   \n",
       "3           0.554534      0.573588       1.244058   \n",
       "4           0.553734      0.577178       1.244058   \n",
       "5           0.542778      0.647962       1.325192   \n",
       "6           0.542106      0.649988       1.325192   \n",
       "7           0.534025      0.663210       1.325192   \n",
       "8           0.533667      0.666181       1.325192   \n",
       "9           0.533189      0.666333       1.325192   \n",
       "10          0.532715      0.667587       1.325192   \n",
       "11          0.532430      0.668379       1.325192   \n",
       "12          0.532082      0.667284       1.325192   \n",
       "13          0.525766      0.673555       1.325192   \n",
       "14          0.525536      0.674057       1.325192   \n",
       "15          0.516968      0.701361       1.325192   \n",
       "16          0.516769      0.700668       1.325192   \n",
       "17          0.516554      0.701864       1.325192   \n",
       "18          0.516251      0.705731       1.325192   \n",
       "19          0.515659      0.709709       1.325192   \n",
       "20          0.515422      0.708876       1.325192   \n",
       "21          0.515140      0.709916       1.325192   \n",
       "22          0.514917      0.710402       1.325192   \n",
       "23          0.514533      0.712374       1.325192   \n",
       "24          0.514114      0.715152       1.325192   \n",
       "25          0.512849      0.714484       1.325192   \n",
       "26          0.512702      0.715130       1.325192   \n",
       "27          0.512532      0.714349       1.325192   \n",
       "28          0.506004      0.719823       1.325192   \n",
       "29          0.505694      0.721928       1.325192   \n",
       "30          0.505472      0.721632       1.325192   \n",
       "31          0.505274      0.723259       1.325192   \n",
       "32          0.504913      0.723827       1.325192   \n",
       "33          0.504723      0.724089       1.325192   \n",
       "34          0.503797      0.724608       1.325192   \n",
       "35          0.503504      0.725672       1.325192   \n",
       "36          0.503119      0.728216       1.325192   \n",
       "37          0.502795      0.729152       1.325192   \n",
       "38          0.501801      0.738898       1.325192   \n",
       "39          0.501543      0.739736       1.325192   \n",
       "40          0.501382      0.740153       1.325192   \n",
       "41          0.501229      0.740341       1.325192   \n",
       "42          0.496039      0.741428       1.325192   \n",
       "43          0.495748      0.741991       1.325192   \n",
       "44          0.495062      0.742246       1.325192   \n",
       "45          0.494851      0.742588       1.325192   \n",
       "46          0.494677      0.742365       1.325192   \n",
       "47          0.494266      0.744502       1.325192   \n",
       "48          0.493685      0.749437       1.325192   \n",
       "49          0.492926      0.751265       1.325192   \n",
       "50          0.492734      0.751706       1.325192   \n",
       "51          0.487219      0.764014       1.325192   \n",
       "\n",
       "    training_classification_error  validation_rmse  validation_logloss  \\\n",
       "0                        0.245392         0.424193            0.545668   \n",
       "1                        0.245392         0.423927            0.545073   \n",
       "2                        0.245392         0.423883            0.544978   \n",
       "3                        0.243943         0.423810            0.544818   \n",
       "4                        0.243736         0.423780            0.544765   \n",
       "5                        0.214951         0.423436            0.544031   \n",
       "6                        0.214951         0.423475            0.544113   \n",
       "7                        0.215158         0.423408            0.543972   \n",
       "8                        0.215158         0.423361            0.543880   \n",
       "9                        0.215780         0.423431            0.544034   \n",
       "10                       0.215780         0.423425            0.544014   \n",
       "11                       0.215780         0.423422            0.544019   \n",
       "12                       0.215780         0.423504            0.544203   \n",
       "13                       0.213916         0.423684            0.544515   \n",
       "14                       0.213916         0.423717            0.544591   \n",
       "15                       0.205426         0.423949            0.545018   \n",
       "16                       0.205426         0.423957            0.545039   \n",
       "17                       0.205426         0.423962            0.545052   \n",
       "18                       0.206461         0.423976            0.545089   \n",
       "19                       0.206461         0.424072            0.545295   \n",
       "20                       0.205840         0.424107            0.545375   \n",
       "21                       0.206461         0.424127            0.545426   \n",
       "22                       0.206461         0.424129            0.545437   \n",
       "23                       0.206461         0.424153            0.545484   \n",
       "24                       0.206461         0.424326            0.545889   \n",
       "25                       0.206461         0.424486            0.546146   \n",
       "26                       0.205840         0.424499            0.546179   \n",
       "27                       0.206461         0.424511            0.546206   \n",
       "28                       0.205011         0.424887            0.546866   \n",
       "29                       0.205011         0.424973            0.547060   \n",
       "30                       0.204183         0.424982            0.547058   \n",
       "31                       0.204183         0.425061            0.547247   \n",
       "32                       0.204183         0.425141            0.547425   \n",
       "33                       0.204183         0.425221            0.547636   \n",
       "34                       0.203562         0.425538            0.548172   \n",
       "35                       0.203562         0.425508            0.548109   \n",
       "36                       0.203562         0.425526            0.548162   \n",
       "37                       0.203562         0.425670            0.548505   \n",
       "38                       0.203562         0.425597            0.548332   \n",
       "39                       0.203562         0.425687            0.548520   \n",
       "40                       0.203562         0.425768            0.548741   \n",
       "41                       0.204183         0.425815            0.548866   \n",
       "42                       0.203562         0.425138            0.547526   \n",
       "43                       0.203562         0.425159            0.547562   \n",
       "44                       0.203562         0.425517            0.548172   \n",
       "45                       0.203562         0.425610            0.548367   \n",
       "46                       0.204183         0.425649            0.548462   \n",
       "47                       0.204183         0.425646            0.548472   \n",
       "48                       0.204183         0.425620            0.548415   \n",
       "49                       0.204183         0.425580            0.548325   \n",
       "50                       0.204183         0.425632            0.548453   \n",
       "51                       0.202112         0.426571            0.550223   \n",
       "\n",
       "    validation_auc  validation_lift  validation_classification_error  \n",
       "0         0.500000         1.000000                         0.235112  \n",
       "1         0.523198         1.094551                         0.235112  \n",
       "2         0.528809         1.099388                         0.235112  \n",
       "3         0.529089         1.032142                         0.235112  \n",
       "4         0.517508         1.055961                         0.235112  \n",
       "5         0.531633         1.045904                         0.233871  \n",
       "6         0.528778         1.111273                         0.233871  \n",
       "7         0.530064         1.111273                         0.233871  \n",
       "8         0.530221         1.111273                         0.233871  \n",
       "9         0.529944         1.111273                         0.233871  \n",
       "10        0.532875         1.111273                         0.233871  \n",
       "11        0.533188         1.111273                         0.233871  \n",
       "12        0.532164         1.111273                         0.233871  \n",
       "13        0.533410         1.143958                         0.233251  \n",
       "14        0.533208         1.143958                         0.233251  \n",
       "15        0.527538         1.143958                         0.233871  \n",
       "16        0.528587         1.176642                         0.233871  \n",
       "17        0.529457         1.176642                         0.233871  \n",
       "18        0.528075         1.176642                         0.233871  \n",
       "19        0.527035         1.176642                         0.233871  \n",
       "20        0.526547         1.176642                         0.233871  \n",
       "21        0.525577         1.176642                         0.231390  \n",
       "22        0.526939         1.176642                         0.231390  \n",
       "23        0.526702         1.176642                         0.231390  \n",
       "24        0.525134         1.089484                         0.231390  \n",
       "25        0.524620         1.089484                         0.230769  \n",
       "26        0.526141         1.129101                         0.230769  \n",
       "27        0.524675         1.182868                         0.230769  \n",
       "28        0.524130         1.167304                         0.230769  \n",
       "29        0.524680         1.167304                         0.230769  \n",
       "30        0.523161         1.129101                         0.230769  \n",
       "31        0.523662         1.129101                         0.230769  \n",
       "32        0.524115         1.129101                         0.230769  \n",
       "33        0.523595         1.153571                         0.230769  \n",
       "34        0.524881         1.153571                         0.232630  \n",
       "35        0.525500         1.153571                         0.232630  \n",
       "36        0.528145         1.153571                         0.232630  \n",
       "37        0.527023         1.153571                         0.232630  \n",
       "38        0.527638         1.153571                         0.232630  \n",
       "39        0.525857         1.153571                         0.232630  \n",
       "40        0.525409         1.153571                         0.232630  \n",
       "41        0.524513         1.153571                         0.232630  \n",
       "42        0.526685         1.153571                         0.232630  \n",
       "43        0.527785         1.153571                         0.232630  \n",
       "44        0.528206         1.153571                         0.232630  \n",
       "45        0.526407         1.153571                         0.232630  \n",
       "46        0.526417         1.153571                         0.232630  \n",
       "47        0.526565         1.153571                         0.232630  \n",
       "48        0.526746         1.153571                         0.232630  \n",
       "49        0.528832         1.153571                         0.232630  \n",
       "50        0.528044         1.153571                         0.232630  \n",
       "51        0.533586         0.950822                         0.233251  "
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print \"AUC: \", gbm_model_2.auc()\n",
    "print \"accuracy: \", gbm_model_2.accuracy()\n",
    "# gbm_model_2.plot()\n",
    "\n",
    "H2OGradientBoostingEstimator.scoring_history(gbm_model_2)\n",
    "# plot(h2o.pe)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "gbm prediction progress: |████████████████████████████████████████████████| 100%\n"
     ]
    }
   ],
   "source": [
    "prediction = gbm_model.predict(hdf_validate[0])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "pdf_pred=prediction.as_data_frame()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "prediction unique values: [1 0]\n",
      "0    1\n",
      "1    1\n",
      "2    1\n",
      "3    1\n",
      "4    1\n",
      "5    1\n",
      "6    1\n",
      "7    1\n",
      "8    1\n",
      "9    1\n",
      "Name: predict, dtype: int64\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead>\n",
       "<tr><th>C1                                                                                                                        </th><th style=\"text-align: right;\">  C10</th></tr>\n",
       "</thead>\n",
       "<tbody>\n",
       "<tr><td>{'butter': 1, 'peanut': 1, 'contain': 1, 'does': 1, 'cholesterol': 1}                                                     </td><td style=\"text-align: right;\">    0</td></tr>\n",
       "<tr><td>{'harmful': 1, 'springtails': 1, 'are': 1}                                                                                </td><td style=\"text-align: right;\">    0</td></tr>\n",
       "<tr><td>{'does': 1, 'have': 1, 'signature': 1, 'assistance': 1, 'roadside': 1, 'visa': 1}                                         </td><td style=\"text-align: right;\">    1</td></tr>\n",
       "<tr><td>{'tesla': 1, 'new': 1, 'is': 1, 'hybrid': 1, 'the': 1}                                                                    </td><td style=\"text-align: right;\">    1</td></tr>\n",
       "<tr><td>{'a': 1, 'is': 1, 'plant': 1, 'moss': 1, 'spanish': 1}                                                                    </td><td style=\"text-align: right;\">    1</td></tr>\n",
       "<tr><td>{'did': 1, 'on': 1, 'landed': 1, 'us': 1, 'moon': 1}                                                                      </td><td style=\"text-align: right;\">    1</td></tr>\n",
       "<tr><td>{'farmhouse': 1, 'sinks': 1, 'laminate': 1, 'countertops': 1, 'with': 1, 'can': 1}                                        </td><td style=\"text-align: right;\">    0</td></tr>\n",
       "<tr><td>{'on': 1, 'i': 1, 'saves': 1, 'xbox': 1, 'pc': 1, 'can': 1, 'put': 1, 'my': 1}                                            </td><td style=\"text-align: right;\">    1</td></tr>\n",
       "<tr><td>{'for': 1, 'is': 1, 'there': 1, 'an': 1, 'limit': 1, 'learning': 1, 'speech': 1, 'age': 1}                                </td><td style=\"text-align: right;\">    0</td></tr>\n",
       "<tr><td>{'counsel': 1, 'of': 1, 'an': 1, 'not': 1, 'does': 1, 'are': 1, 'employee': 1, 'partner': 1, 'you': 1, 'or': 1, 'mean': 1}</td><td style=\"text-align: right;\">    0</td></tr>\n",
       "</tbody>\n",
       "</table>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "print 'prediction unique values:', pdf_pred['predict'].unique()\n",
    "print pdf_pred['predict'][0:10]\n",
    "print hdf_validate.head(10)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Playing with data\n",
    "1. clean it\n",
    "2. then word cloud"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# import"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'contestants': 1, 'do': 1, 'game': 1, 'pay': 1, 'shows': 1, 'their': 1}"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lst_tf[1]"
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "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": 1
}