{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# HW4 [Deception] PART 2 -- Check with Myle Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## STEP 1: GET THAT DATA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "def get_data(file, path):\n",
    "    f=open(path+file)\n",
    "    data = f.read()\n",
    "    f.close()\n",
    "    return data\n",
    "    \n",
    "def get_data_from_files(path):\n",
    "    results = [get_data(file, path) for file in os.listdir(path)]\n",
    "    return results\n",
    "\n",
    "# pos = get_data_from_files('../pos_cornell//')\n",
    "# neg = get_data_from_files('../neg_cornell/')\n",
    "\n",
    "# pos = get_data_from_files('../hw4_lie_false/')\n",
    "# neg = get_data_from_files('../hw4_lie_true/')\n",
    "\n",
    "## TRUE IS NEG!!!!\n",
    "neg = get_data_from_files('../myle_pos_deceptive/')\n",
    "pos = get_data_from_files('../myle_pos_truthful/')\n",
    "neg2 = get_data_from_files('../myle_neg_deceptive/')\n",
    "pos2 = get_data_from_files('../myle_neg_truthful/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>PoN</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>I traveled to Chicago with my husband for a ro...</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>I stayed in the Sofitel Chicago Water Tower ho...</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>This hotel was gorgeous! I really enjoyed my s...</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>This is an absolutely exquisite hotel, at a gr...</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>I recently traveled up to Chicago for business...</td>\n",
       "      <td>N</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>310</td>\n",
       "      <td>It's not a bad hotel. It's just so...disappoin...</td>\n",
       "      <td>P</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>311</td>\n",
       "      <td>My wife and I brought our daughter downtown fo...</td>\n",
       "      <td>P</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>312</td>\n",
       "      <td>Excellent Hotel ! Rooms and service were great...</td>\n",
       "      <td>P</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>313</td>\n",
       "      <td>Had a week long stay at the Hilton on south Mi...</td>\n",
       "      <td>P</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>314</td>\n",
       "      <td>We stayed at the James hotel for a 40th birthd...</td>\n",
       "      <td>P</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>315 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                     0 PoN\n",
       "0    I traveled to Chicago with my husband for a ro...   N\n",
       "1    I stayed in the Sofitel Chicago Water Tower ho...   N\n",
       "2    This hotel was gorgeous! I really enjoyed my s...   N\n",
       "3    This is an absolutely exquisite hotel, at a gr...   N\n",
       "4    I recently traveled up to Chicago for business...   N\n",
       "..                                                 ...  ..\n",
       "310  It's not a bad hotel. It's just so...disappoin...   P\n",
       "311  My wife and I brought our daughter downtown fo...   P\n",
       "312  Excellent Hotel ! Rooms and service were great...   P\n",
       "313  Had a week long stay at the Hilton on south Mi...   P\n",
       "314  We stayed at the James hotel for a 40th birthd...   P\n",
       "\n",
       "[315 rows x 2 columns]"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "neg_df = pd.DataFrame(neg)\n",
    "pos_df = pd.DataFrame(pos)\n",
    "pos_df['PoN'] = 'P'\n",
    "neg_df['PoN'] = 'N'\n",
    "neg_df2 = pd.DataFrame(neg2)\n",
    "pos_df2 = pd.DataFrame(pos2)\n",
    "pos_df2['PoN'] = 'P'\n",
    "neg_df2['PoN'] = 'N'\n",
    "all_df = neg_df.append(pos_df)\n",
    "all_df2 = neg_df2.append(pos_df2)\n",
    "all_df = all_df.append(all_df2)\n",
    "all_df.reset_index(drop=True,inplace=True)\n",
    "# all_df2.reset_index(drop=True,inplace=True)\n",
    "all_df[:-5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## STEP 2: TOKENIZE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nltk.tokenize import word_tokenize, sent_tokenize\n",
    "from nltk.sentiment import SentimentAnalyzer\n",
    "from nltk.sentiment.util import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### -- 2a by sentence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_sentence_tokens(review):\n",
    "    return sent_tokenize(review)\n",
    "    \n",
    "all_df['sentences'] = all_df.apply(lambda x: get_sentence_tokens(x[0]), axis=1)\n",
    "all_df['num_sentences'] = all_df.apply(lambda x: len(x['sentences']), axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### -- 2b by word"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_tokens(sentence):\n",
    "    tokens = word_tokenize(sentence)\n",
    "    clean_tokens = [word.lower() for word in tokens if word.isalpha()]\n",
    "    return clean_tokens\n",
    "\n",
    "all_df['tokens'] = all_df.apply(lambda x: get_tokens(x[0]), axis=1)\n",
    "all_df['num_tokens'] = all_df.apply(lambda x: len(x['tokens']), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>PoN</th>\n",
       "      <th>sentences</th>\n",
       "      <th>num_sentences</th>\n",
       "      <th>tokens</th>\n",
       "      <th>num_tokens</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>I traveled to Chicago with my husband for a ro...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I traveled to Chicago with my husband for a r...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, traveled, to, chicago, with, my, husband, ...</td>\n",
       "      <td>68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>I stayed in the Sofitel Chicago Water Tower ho...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I stayed in the Sofitel Chicago Water Tower h...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, stayed, in, the, sofitel, chicago, water, ...</td>\n",
       "      <td>129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>This hotel was gorgeous! I really enjoyed my s...</td>\n",
       "      <td>N</td>\n",
       "      <td>[This hotel was gorgeous!, I really enjoyed my...</td>\n",
       "      <td>7</td>\n",
       "      <td>[this, hotel, was, gorgeous, i, really, enjoye...</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   0 PoN  \\\n",
       "0  I traveled to Chicago with my husband for a ro...   N   \n",
       "1  I stayed in the Sofitel Chicago Water Tower ho...   N   \n",
       "2  This hotel was gorgeous! I really enjoyed my s...   N   \n",
       "\n",
       "                                           sentences  num_sentences  \\\n",
       "0  [I traveled to Chicago with my husband for a r...              6   \n",
       "1  [I stayed in the Sofitel Chicago Water Tower h...              6   \n",
       "2  [This hotel was gorgeous!, I really enjoyed my...              7   \n",
       "\n",
       "                                              tokens  num_tokens  \n",
       "0  [i, traveled, to, chicago, with, my, husband, ...          68  \n",
       "1  [i, stayed, in, the, sofitel, chicago, water, ...         129  \n",
       "2  [this, hotel, was, gorgeous, i, really, enjoye...          69  "
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_df[:3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### -- 2c Remove if tokens < 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>PoN</th>\n",
       "      <th>sentences</th>\n",
       "      <th>num_sentences</th>\n",
       "      <th>tokens</th>\n",
       "      <th>num_tokens</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>I traveled to Chicago with my husband for a ro...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I traveled to Chicago with my husband for a r...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, traveled, to, chicago, with, my, husband, ...</td>\n",
       "      <td>68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>I stayed in the Sofitel Chicago Water Tower ho...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I stayed in the Sofitel Chicago Water Tower h...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, stayed, in, the, sofitel, chicago, water, ...</td>\n",
       "      <td>129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>This hotel was gorgeous! I really enjoyed my s...</td>\n",
       "      <td>N</td>\n",
       "      <td>[This hotel was gorgeous!, I really enjoyed my...</td>\n",
       "      <td>7</td>\n",
       "      <td>[this, hotel, was, gorgeous, i, really, enjoye...</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   0 PoN  \\\n",
       "0  I traveled to Chicago with my husband for a ro...   N   \n",
       "1  I stayed in the Sofitel Chicago Water Tower ho...   N   \n",
       "2  This hotel was gorgeous! I really enjoyed my s...   N   \n",
       "\n",
       "                                           sentences  num_sentences  \\\n",
       "0  [I traveled to Chicago with my husband for a r...              6   \n",
       "1  [I stayed in the Sofitel Chicago Water Tower h...              6   \n",
       "2  [This hotel was gorgeous!, I really enjoyed my...              7   \n",
       "\n",
       "                                              tokens  num_tokens  \n",
       "0  [i, traveled, to, chicago, with, my, husband, ...          68  \n",
       "1  [i, stayed, in, the, sofitel, chicago, water, ...         129  \n",
       "2  [this, hotel, was, gorgeous, i, really, enjoye...          69  "
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_df = all_df.drop(all_df[all_df.num_tokens < 1].index)\n",
    "all_df[:3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## STEP 3: EXPERIMENT\n",
    "#### Experiment with: stopwords, stemming, lemming etc."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### -- 3a remove english stopwords"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nltk.corpus import stopwords\n",
    "stop_words = set(stopwords.words(\"english\"))\n",
    "def remove_stopwords(sentence):\n",
    "    filtered_text = []\n",
    "    for word in sentence:\n",
    "        if word not in stop_words:\n",
    "            filtered_text.append(word)\n",
    "    return filtered_text\n",
    "all_df['no_sw'] = all_df.apply(lambda x: remove_stopwords(x['tokens']),axis=1)\n",
    "all_df['num_no_sw'] = all_df.apply(lambda x: len(x['no_sw']),axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>PoN</th>\n",
       "      <th>sentences</th>\n",
       "      <th>num_sentences</th>\n",
       "      <th>tokens</th>\n",
       "      <th>num_tokens</th>\n",
       "      <th>no_sw</th>\n",
       "      <th>num_no_sw</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>I traveled to Chicago with my husband for a ro...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I traveled to Chicago with my husband for a r...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, traveled, to, chicago, with, my, husband, ...</td>\n",
       "      <td>68</td>\n",
       "      <td>[traveled, chicago, husband, romantic, weekend...</td>\n",
       "      <td>40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>I stayed in the Sofitel Chicago Water Tower ho...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I stayed in the Sofitel Chicago Water Tower h...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, stayed, in, the, sofitel, chicago, water, ...</td>\n",
       "      <td>129</td>\n",
       "      <td>[stayed, sofitel, chicago, water, tower, hotel...</td>\n",
       "      <td>71</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>This hotel was gorgeous! I really enjoyed my s...</td>\n",
       "      <td>N</td>\n",
       "      <td>[This hotel was gorgeous!, I really enjoyed my...</td>\n",
       "      <td>7</td>\n",
       "      <td>[this, hotel, was, gorgeous, i, really, enjoye...</td>\n",
       "      <td>69</td>\n",
       "      <td>[hotel, gorgeous, really, enjoyed, stay, defin...</td>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   0 PoN  \\\n",
       "0  I traveled to Chicago with my husband for a ro...   N   \n",
       "1  I stayed in the Sofitel Chicago Water Tower ho...   N   \n",
       "2  This hotel was gorgeous! I really enjoyed my s...   N   \n",
       "\n",
       "                                           sentences  num_sentences  \\\n",
       "0  [I traveled to Chicago with my husband for a r...              6   \n",
       "1  [I stayed in the Sofitel Chicago Water Tower h...              6   \n",
       "2  [This hotel was gorgeous!, I really enjoyed my...              7   \n",
       "\n",
       "                                              tokens  num_tokens  \\\n",
       "0  [i, traveled, to, chicago, with, my, husband, ...          68   \n",
       "1  [i, stayed, in, the, sofitel, chicago, water, ...         129   \n",
       "2  [this, hotel, was, gorgeous, i, really, enjoye...          69   \n",
       "\n",
       "                                               no_sw  num_no_sw  \n",
       "0  [traveled, chicago, husband, romantic, weekend...         40  \n",
       "1  [stayed, sofitel, chicago, water, tower, hotel...         71  \n",
       "2  [hotel, gorgeous, really, enjoyed, stay, defin...         36  "
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_df[:3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### -- 3b get stems for both tokens and no_sw"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nltk.stem import PorterStemmer\n",
    "def get_stems(sentence):\n",
    "    ps = PorterStemmer()\n",
    "    return [ps.stem(w) for w in sentence]\n",
    "    \n",
    "all_df['stemmed'] = all_df.apply(lambda x: get_stems(x['tokens']),axis=1)\n",
    "all_df['stemmed_no_sw'] = all_df.apply(lambda x: get_stems(x['no_sw']),axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>PoN</th>\n",
       "      <th>sentences</th>\n",
       "      <th>num_sentences</th>\n",
       "      <th>tokens</th>\n",
       "      <th>num_tokens</th>\n",
       "      <th>no_sw</th>\n",
       "      <th>num_no_sw</th>\n",
       "      <th>stemmed</th>\n",
       "      <th>stemmed_no_sw</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>I traveled to Chicago with my husband for a ro...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I traveled to Chicago with my husband for a r...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, traveled, to, chicago, with, my, husband, ...</td>\n",
       "      <td>68</td>\n",
       "      <td>[traveled, chicago, husband, romantic, weekend...</td>\n",
       "      <td>40</td>\n",
       "      <td>[i, travel, to, chicago, with, my, husband, fo...</td>\n",
       "      <td>[travel, chicago, husband, romant, weekend, aw...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>I stayed in the Sofitel Chicago Water Tower ho...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I stayed in the Sofitel Chicago Water Tower h...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, stayed, in, the, sofitel, chicago, water, ...</td>\n",
       "      <td>129</td>\n",
       "      <td>[stayed, sofitel, chicago, water, tower, hotel...</td>\n",
       "      <td>71</td>\n",
       "      <td>[i, stay, in, the, sofitel, chicago, water, to...</td>\n",
       "      <td>[stay, sofitel, chicago, water, tower, hotel, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>This hotel was gorgeous! I really enjoyed my s...</td>\n",
       "      <td>N</td>\n",
       "      <td>[This hotel was gorgeous!, I really enjoyed my...</td>\n",
       "      <td>7</td>\n",
       "      <td>[this, hotel, was, gorgeous, i, really, enjoye...</td>\n",
       "      <td>69</td>\n",
       "      <td>[hotel, gorgeous, really, enjoyed, stay, defin...</td>\n",
       "      <td>36</td>\n",
       "      <td>[thi, hotel, wa, gorgeou, i, realli, enjoy, my...</td>\n",
       "      <td>[hotel, gorgeou, realli, enjoy, stay, definit,...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   0 PoN  \\\n",
       "0  I traveled to Chicago with my husband for a ro...   N   \n",
       "1  I stayed in the Sofitel Chicago Water Tower ho...   N   \n",
       "2  This hotel was gorgeous! I really enjoyed my s...   N   \n",
       "\n",
       "                                           sentences  num_sentences  \\\n",
       "0  [I traveled to Chicago with my husband for a r...              6   \n",
       "1  [I stayed in the Sofitel Chicago Water Tower h...              6   \n",
       "2  [This hotel was gorgeous!, I really enjoyed my...              7   \n",
       "\n",
       "                                              tokens  num_tokens  \\\n",
       "0  [i, traveled, to, chicago, with, my, husband, ...          68   \n",
       "1  [i, stayed, in, the, sofitel, chicago, water, ...         129   \n",
       "2  [this, hotel, was, gorgeous, i, really, enjoye...          69   \n",
       "\n",
       "                                               no_sw  num_no_sw  \\\n",
       "0  [traveled, chicago, husband, romantic, weekend...         40   \n",
       "1  [stayed, sofitel, chicago, water, tower, hotel...         71   \n",
       "2  [hotel, gorgeous, really, enjoyed, stay, defin...         36   \n",
       "\n",
       "                                             stemmed  \\\n",
       "0  [i, travel, to, chicago, with, my, husband, fo...   \n",
       "1  [i, stay, in, the, sofitel, chicago, water, to...   \n",
       "2  [thi, hotel, wa, gorgeou, i, realli, enjoy, my...   \n",
       "\n",
       "                                       stemmed_no_sw  \n",
       "0  [travel, chicago, husband, romant, weekend, aw...  \n",
       "1  [stay, sofitel, chicago, water, tower, hotel, ...  \n",
       "2  [hotel, gorgeou, realli, enjoy, stay, definit,...  "
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_df[:3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### -- 3c get lemmas for both tokens and no_sw"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nltk.stem.wordnet import WordNetLemmatizer\n",
    "def get_lemmas(sentence):\n",
    "    lem = WordNetLemmatizer() \n",
    "    return [lem.lemmatize(w) for w in sentence]\n",
    "    \n",
    "all_df['lemmed'] = all_df.apply(lambda x: get_lemmas(x['tokens']),axis=1)\n",
    "all_df['lemmed_no_sw'] = all_df.apply(lambda x: get_lemmas(x['no_sw']),axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>PoN</th>\n",
       "      <th>sentences</th>\n",
       "      <th>num_sentences</th>\n",
       "      <th>tokens</th>\n",
       "      <th>num_tokens</th>\n",
       "      <th>no_sw</th>\n",
       "      <th>num_no_sw</th>\n",
       "      <th>stemmed</th>\n",
       "      <th>stemmed_no_sw</th>\n",
       "      <th>lemmed</th>\n",
       "      <th>lemmed_no_sw</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>I traveled to Chicago with my husband for a ro...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I traveled to Chicago with my husband for a r...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, traveled, to, chicago, with, my, husband, ...</td>\n",
       "      <td>68</td>\n",
       "      <td>[traveled, chicago, husband, romantic, weekend...</td>\n",
       "      <td>40</td>\n",
       "      <td>[i, travel, to, chicago, with, my, husband, fo...</td>\n",
       "      <td>[travel, chicago, husband, romant, weekend, aw...</td>\n",
       "      <td>[i, traveled, to, chicago, with, my, husband, ...</td>\n",
       "      <td>[traveled, chicago, husband, romantic, weekend...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>I stayed in the Sofitel Chicago Water Tower ho...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I stayed in the Sofitel Chicago Water Tower h...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, stayed, in, the, sofitel, chicago, water, ...</td>\n",
       "      <td>129</td>\n",
       "      <td>[stayed, sofitel, chicago, water, tower, hotel...</td>\n",
       "      <td>71</td>\n",
       "      <td>[i, stay, in, the, sofitel, chicago, water, to...</td>\n",
       "      <td>[stay, sofitel, chicago, water, tower, hotel, ...</td>\n",
       "      <td>[i, stayed, in, the, sofitel, chicago, water, ...</td>\n",
       "      <td>[stayed, sofitel, chicago, water, tower, hotel...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>This hotel was gorgeous! I really enjoyed my s...</td>\n",
       "      <td>N</td>\n",
       "      <td>[This hotel was gorgeous!, I really enjoyed my...</td>\n",
       "      <td>7</td>\n",
       "      <td>[this, hotel, was, gorgeous, i, really, enjoye...</td>\n",
       "      <td>69</td>\n",
       "      <td>[hotel, gorgeous, really, enjoyed, stay, defin...</td>\n",
       "      <td>36</td>\n",
       "      <td>[thi, hotel, wa, gorgeou, i, realli, enjoy, my...</td>\n",
       "      <td>[hotel, gorgeou, realli, enjoy, stay, definit,...</td>\n",
       "      <td>[this, hotel, wa, gorgeous, i, really, enjoyed...</td>\n",
       "      <td>[hotel, gorgeous, really, enjoyed, stay, defin...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   0 PoN  \\\n",
       "0  I traveled to Chicago with my husband for a ro...   N   \n",
       "1  I stayed in the Sofitel Chicago Water Tower ho...   N   \n",
       "2  This hotel was gorgeous! I really enjoyed my s...   N   \n",
       "\n",
       "                                           sentences  num_sentences  \\\n",
       "0  [I traveled to Chicago with my husband for a r...              6   \n",
       "1  [I stayed in the Sofitel Chicago Water Tower h...              6   \n",
       "2  [This hotel was gorgeous!, I really enjoyed my...              7   \n",
       "\n",
       "                                              tokens  num_tokens  \\\n",
       "0  [i, traveled, to, chicago, with, my, husband, ...          68   \n",
       "1  [i, stayed, in, the, sofitel, chicago, water, ...         129   \n",
       "2  [this, hotel, was, gorgeous, i, really, enjoye...          69   \n",
       "\n",
       "                                               no_sw  num_no_sw  \\\n",
       "0  [traveled, chicago, husband, romantic, weekend...         40   \n",
       "1  [stayed, sofitel, chicago, water, tower, hotel...         71   \n",
       "2  [hotel, gorgeous, really, enjoyed, stay, defin...         36   \n",
       "\n",
       "                                             stemmed  \\\n",
       "0  [i, travel, to, chicago, with, my, husband, fo...   \n",
       "1  [i, stay, in, the, sofitel, chicago, water, to...   \n",
       "2  [thi, hotel, wa, gorgeou, i, realli, enjoy, my...   \n",
       "\n",
       "                                       stemmed_no_sw  \\\n",
       "0  [travel, chicago, husband, romant, weekend, aw...   \n",
       "1  [stay, sofitel, chicago, water, tower, hotel, ...   \n",
       "2  [hotel, gorgeou, realli, enjoy, stay, definit,...   \n",
       "\n",
       "                                              lemmed  \\\n",
       "0  [i, traveled, to, chicago, with, my, husband, ...   \n",
       "1  [i, stayed, in, the, sofitel, chicago, water, ...   \n",
       "2  [this, hotel, wa, gorgeous, i, really, enjoyed...   \n",
       "\n",
       "                                        lemmed_no_sw  \n",
       "0  [traveled, chicago, husband, romantic, weekend...  \n",
       "1  [stayed, sofitel, chicago, water, tower, hotel...  \n",
       "2  [hotel, gorgeous, really, enjoyed, stay, defin...  "
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_df[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_df['pos'] = all_df.apply(lambda x: nltk.pos_tag(x['tokens']),axis=1)\n",
    "all_df['pos_no_sw'] = all_df.apply(lambda x: nltk.pos_tag(x['no_sw']),axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>PoN</th>\n",
       "      <th>sentences</th>\n",
       "      <th>num_sentences</th>\n",
       "      <th>tokens</th>\n",
       "      <th>num_tokens</th>\n",
       "      <th>no_sw</th>\n",
       "      <th>num_no_sw</th>\n",
       "      <th>stemmed</th>\n",
       "      <th>stemmed_no_sw</th>\n",
       "      <th>lemmed</th>\n",
       "      <th>lemmed_no_sw</th>\n",
       "      <th>pos</th>\n",
       "      <th>pos_no_sw</th>\n",
       "      <th>pos_dict</th>\n",
       "      <th>pos_dict_no_sw</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>I traveled to Chicago with my husband for a ro...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I traveled to Chicago with my husband for a r...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, traveled, to, chicago, with, my, husband, ...</td>\n",
       "      <td>68</td>\n",
       "      <td>[traveled, chicago, husband, romantic, weekend...</td>\n",
       "      <td>40</td>\n",
       "      <td>[i, travel, to, chicago, with, my, husband, fo...</td>\n",
       "      <td>[travel, chicago, husband, romant, weekend, aw...</td>\n",
       "      <td>[i, traveled, to, chicago, with, my, husband, ...</td>\n",
       "      <td>[traveled, chicago, husband, romantic, weekend...</td>\n",
       "      <td>[(i, NN), (traveled, VBD), (to, TO), (chicago,...</td>\n",
       "      <td>[(traveled, VBN), (chicago, JJ), (husband, NN)...</td>\n",
       "      <td>{'NN': 18, 'VBD': 6, 'TO': 1, 'VB': 3, 'IN': 6...</td>\n",
       "      <td>{'VBN': 1, 'JJ': 6, 'NN': 16, 'RB': 4, 'MD': 2...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>I stayed in the Sofitel Chicago Water Tower ho...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I stayed in the Sofitel Chicago Water Tower h...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, stayed, in, the, sofitel, chicago, water, ...</td>\n",
       "      <td>129</td>\n",
       "      <td>[stayed, sofitel, chicago, water, tower, hotel...</td>\n",
       "      <td>71</td>\n",
       "      <td>[i, stay, in, the, sofitel, chicago, water, to...</td>\n",
       "      <td>[stay, sofitel, chicago, water, tower, hotel, ...</td>\n",
       "      <td>[i, stayed, in, the, sofitel, chicago, water, ...</td>\n",
       "      <td>[stayed, sofitel, chicago, water, tower, hotel...</td>\n",
       "      <td>[(i, JJ), (stayed, VBD), (in, IN), (the, DT), ...</td>\n",
       "      <td>[(stayed, JJ), (sofitel, NN), (chicago, NN), (...</td>\n",
       "      <td>{'JJ': 19, 'VBD': 6, 'IN': 16, 'DT': 14, 'NN':...</td>\n",
       "      <td>{'JJ': 15, 'NN': 29, 'CD': 1, 'NNS': 11, 'RB':...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>This hotel was gorgeous! I really enjoyed my s...</td>\n",
       "      <td>N</td>\n",
       "      <td>[This hotel was gorgeous!, I really enjoyed my...</td>\n",
       "      <td>7</td>\n",
       "      <td>[this, hotel, was, gorgeous, i, really, enjoye...</td>\n",
       "      <td>69</td>\n",
       "      <td>[hotel, gorgeous, really, enjoyed, stay, defin...</td>\n",
       "      <td>36</td>\n",
       "      <td>[thi, hotel, wa, gorgeou, i, realli, enjoy, my...</td>\n",
       "      <td>[hotel, gorgeou, realli, enjoy, stay, definit,...</td>\n",
       "      <td>[this, hotel, wa, gorgeous, i, really, enjoyed...</td>\n",
       "      <td>[hotel, gorgeous, really, enjoyed, stay, defin...</td>\n",
       "      <td>[(this, DT), (hotel, NN), (was, VBD), (gorgeou...</td>\n",
       "      <td>[(hotel, NN), (gorgeous, JJ), (really, RB), (e...</td>\n",
       "      <td>{'DT': 9, 'NN': 15, 'VBD': 6, 'JJ': 10, 'RB': ...</td>\n",
       "      <td>{'NN': 15, 'JJ': 9, 'RB': 6, 'VBN': 1, 'VBG': ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   0 PoN  \\\n",
       "0  I traveled to Chicago with my husband for a ro...   N   \n",
       "1  I stayed in the Sofitel Chicago Water Tower ho...   N   \n",
       "2  This hotel was gorgeous! I really enjoyed my s...   N   \n",
       "\n",
       "                                           sentences  num_sentences  \\\n",
       "0  [I traveled to Chicago with my husband for a r...              6   \n",
       "1  [I stayed in the Sofitel Chicago Water Tower h...              6   \n",
       "2  [This hotel was gorgeous!, I really enjoyed my...              7   \n",
       "\n",
       "                                              tokens  num_tokens  \\\n",
       "0  [i, traveled, to, chicago, with, my, husband, ...          68   \n",
       "1  [i, stayed, in, the, sofitel, chicago, water, ...         129   \n",
       "2  [this, hotel, was, gorgeous, i, really, enjoye...          69   \n",
       "\n",
       "                                               no_sw  num_no_sw  \\\n",
       "0  [traveled, chicago, husband, romantic, weekend...         40   \n",
       "1  [stayed, sofitel, chicago, water, tower, hotel...         71   \n",
       "2  [hotel, gorgeous, really, enjoyed, stay, defin...         36   \n",
       "\n",
       "                                             stemmed  \\\n",
       "0  [i, travel, to, chicago, with, my, husband, fo...   \n",
       "1  [i, stay, in, the, sofitel, chicago, water, to...   \n",
       "2  [thi, hotel, wa, gorgeou, i, realli, enjoy, my...   \n",
       "\n",
       "                                       stemmed_no_sw  \\\n",
       "0  [travel, chicago, husband, romant, weekend, aw...   \n",
       "1  [stay, sofitel, chicago, water, tower, hotel, ...   \n",
       "2  [hotel, gorgeou, realli, enjoy, stay, definit,...   \n",
       "\n",
       "                                              lemmed  \\\n",
       "0  [i, traveled, to, chicago, with, my, husband, ...   \n",
       "1  [i, stayed, in, the, sofitel, chicago, water, ...   \n",
       "2  [this, hotel, wa, gorgeous, i, really, enjoyed...   \n",
       "\n",
       "                                        lemmed_no_sw  \\\n",
       "0  [traveled, chicago, husband, romantic, weekend...   \n",
       "1  [stayed, sofitel, chicago, water, tower, hotel...   \n",
       "2  [hotel, gorgeous, really, enjoyed, stay, defin...   \n",
       "\n",
       "                                                 pos  \\\n",
       "0  [(i, NN), (traveled, VBD), (to, TO), (chicago,...   \n",
       "1  [(i, JJ), (stayed, VBD), (in, IN), (the, DT), ...   \n",
       "2  [(this, DT), (hotel, NN), (was, VBD), (gorgeou...   \n",
       "\n",
       "                                           pos_no_sw  \\\n",
       "0  [(traveled, VBN), (chicago, JJ), (husband, NN)...   \n",
       "1  [(stayed, JJ), (sofitel, NN), (chicago, NN), (...   \n",
       "2  [(hotel, NN), (gorgeous, JJ), (really, RB), (e...   \n",
       "\n",
       "                                            pos_dict  \\\n",
       "0  {'NN': 18, 'VBD': 6, 'TO': 1, 'VB': 3, 'IN': 6...   \n",
       "1  {'JJ': 19, 'VBD': 6, 'IN': 16, 'DT': 14, 'NN':...   \n",
       "2  {'DT': 9, 'NN': 15, 'VBD': 6, 'JJ': 10, 'RB': ...   \n",
       "\n",
       "                                      pos_dict_no_sw  \n",
       "0  {'VBN': 1, 'JJ': 6, 'NN': 16, 'RB': 4, 'MD': 2...  \n",
       "1  {'JJ': 15, 'NN': 29, 'CD': 1, 'NNS': 11, 'RB':...  \n",
       "2  {'NN': 15, 'JJ': 9, 'RB': 6, 'VBN': 1, 'VBG': ...  "
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_pos_dict(pos_tuple):\n",
    "    pos_dict = {}\n",
    "    for t in pos_tuple:\n",
    "        if t[1] in pos_dict.keys():\n",
    "            pos_dict[t[1]] += 1\n",
    "        else:\n",
    "            pos_dict.update({t[1]: 1})\n",
    "    return pos_dict\n",
    "\n",
    "all_df['pos_dict'] = all_df.apply(lambda x: get_pos_dict(x['pos']), axis=1)\n",
    "all_df['pos_dict_no_sw'] = all_df.apply(lambda x: get_pos_dict(x['pos_no_sw']), axis=1)\n",
    "all_df[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>PoN</th>\n",
       "      <th>sentences</th>\n",
       "      <th>num_sentences</th>\n",
       "      <th>tokens</th>\n",
       "      <th>num_tokens</th>\n",
       "      <th>no_sw</th>\n",
       "      <th>num_no_sw</th>\n",
       "      <th>stemmed</th>\n",
       "      <th>stemmed_no_sw</th>\n",
       "      <th>lemmed</th>\n",
       "      <th>lemmed_no_sw</th>\n",
       "      <th>pos</th>\n",
       "      <th>pos_no_sw</th>\n",
       "      <th>pos_dict</th>\n",
       "      <th>pos_dict_no_sw</th>\n",
       "      <th>bow</th>\n",
       "      <th>bow_no_sw</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>I traveled to Chicago with my husband for a ro...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I traveled to Chicago with my husband for a r...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, traveled, to, chicago, with, my, husband, ...</td>\n",
       "      <td>68</td>\n",
       "      <td>[traveled, chicago, husband, romantic, weekend...</td>\n",
       "      <td>40</td>\n",
       "      <td>[i, travel, to, chicago, with, my, husband, fo...</td>\n",
       "      <td>[travel, chicago, husband, romant, weekend, aw...</td>\n",
       "      <td>[i, traveled, to, chicago, with, my, husband, ...</td>\n",
       "      <td>[traveled, chicago, husband, romantic, weekend...</td>\n",
       "      <td>[(i, NN), (traveled, VBD), (to, TO), (chicago,...</td>\n",
       "      <td>[(traveled, VBN), (chicago, JJ), (husband, NN)...</td>\n",
       "      <td>{'NN': 18, 'VBD': 6, 'TO': 1, 'VB': 3, 'IN': 6...</td>\n",
       "      <td>{'VBN': 1, 'JJ': 6, 'NN': 16, 'RB': 4, 'MD': 2...</td>\n",
       "      <td>{'i': 1, 'traveled': 1, 'to': 1, 'chicago': 2,...</td>\n",
       "      <td>{'traveled': 1, 'chicago': 2, 'husband': 1, 'r...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>I stayed in the Sofitel Chicago Water Tower ho...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I stayed in the Sofitel Chicago Water Tower h...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, stayed, in, the, sofitel, chicago, water, ...</td>\n",
       "      <td>129</td>\n",
       "      <td>[stayed, sofitel, chicago, water, tower, hotel...</td>\n",
       "      <td>71</td>\n",
       "      <td>[i, stay, in, the, sofitel, chicago, water, to...</td>\n",
       "      <td>[stay, sofitel, chicago, water, tower, hotel, ...</td>\n",
       "      <td>[i, stayed, in, the, sofitel, chicago, water, ...</td>\n",
       "      <td>[stayed, sofitel, chicago, water, tower, hotel...</td>\n",
       "      <td>[(i, JJ), (stayed, VBD), (in, IN), (the, DT), ...</td>\n",
       "      <td>[(stayed, JJ), (sofitel, NN), (chicago, NN), (...</td>\n",
       "      <td>{'JJ': 19, 'VBD': 6, 'IN': 16, 'DT': 14, 'NN':...</td>\n",
       "      <td>{'JJ': 15, 'NN': 29, 'CD': 1, 'NNS': 11, 'RB':...</td>\n",
       "      <td>{'i': 3, 'stayed': 1, 'in': 1, 'the': 9, 'sofi...</td>\n",
       "      <td>{'stayed': 1, 'sofitel': 1, 'chicago': 1, 'wat...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>This hotel was gorgeous! I really enjoyed my s...</td>\n",
       "      <td>N</td>\n",
       "      <td>[This hotel was gorgeous!, I really enjoyed my...</td>\n",
       "      <td>7</td>\n",
       "      <td>[this, hotel, was, gorgeous, i, really, enjoye...</td>\n",
       "      <td>69</td>\n",
       "      <td>[hotel, gorgeous, really, enjoyed, stay, defin...</td>\n",
       "      <td>36</td>\n",
       "      <td>[thi, hotel, wa, gorgeou, i, realli, enjoy, my...</td>\n",
       "      <td>[hotel, gorgeou, realli, enjoy, stay, definit,...</td>\n",
       "      <td>[this, hotel, wa, gorgeous, i, really, enjoyed...</td>\n",
       "      <td>[hotel, gorgeous, really, enjoyed, stay, defin...</td>\n",
       "      <td>[(this, DT), (hotel, NN), (was, VBD), (gorgeou...</td>\n",
       "      <td>[(hotel, NN), (gorgeous, JJ), (really, RB), (e...</td>\n",
       "      <td>{'DT': 9, 'NN': 15, 'VBD': 6, 'JJ': 10, 'RB': ...</td>\n",
       "      <td>{'NN': 15, 'JJ': 9, 'RB': 6, 'VBN': 1, 'VBG': ...</td>\n",
       "      <td>{'this': 2, 'hotel': 2, 'was': 6, 'gorgeous': ...</td>\n",
       "      <td>{'hotel': 2, 'gorgeous': 1, 'really': 1, 'enjo...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   0 PoN  \\\n",
       "0  I traveled to Chicago with my husband for a ro...   N   \n",
       "1  I stayed in the Sofitel Chicago Water Tower ho...   N   \n",
       "2  This hotel was gorgeous! I really enjoyed my s...   N   \n",
       "\n",
       "                                           sentences  num_sentences  \\\n",
       "0  [I traveled to Chicago with my husband for a r...              6   \n",
       "1  [I stayed in the Sofitel Chicago Water Tower h...              6   \n",
       "2  [This hotel was gorgeous!, I really enjoyed my...              7   \n",
       "\n",
       "                                              tokens  num_tokens  \\\n",
       "0  [i, traveled, to, chicago, with, my, husband, ...          68   \n",
       "1  [i, stayed, in, the, sofitel, chicago, water, ...         129   \n",
       "2  [this, hotel, was, gorgeous, i, really, enjoye...          69   \n",
       "\n",
       "                                               no_sw  num_no_sw  \\\n",
       "0  [traveled, chicago, husband, romantic, weekend...         40   \n",
       "1  [stayed, sofitel, chicago, water, tower, hotel...         71   \n",
       "2  [hotel, gorgeous, really, enjoyed, stay, defin...         36   \n",
       "\n",
       "                                             stemmed  \\\n",
       "0  [i, travel, to, chicago, with, my, husband, fo...   \n",
       "1  [i, stay, in, the, sofitel, chicago, water, to...   \n",
       "2  [thi, hotel, wa, gorgeou, i, realli, enjoy, my...   \n",
       "\n",
       "                                       stemmed_no_sw  \\\n",
       "0  [travel, chicago, husband, romant, weekend, aw...   \n",
       "1  [stay, sofitel, chicago, water, tower, hotel, ...   \n",
       "2  [hotel, gorgeou, realli, enjoy, stay, definit,...   \n",
       "\n",
       "                                              lemmed  \\\n",
       "0  [i, traveled, to, chicago, with, my, husband, ...   \n",
       "1  [i, stayed, in, the, sofitel, chicago, water, ...   \n",
       "2  [this, hotel, wa, gorgeous, i, really, enjoyed...   \n",
       "\n",
       "                                        lemmed_no_sw  \\\n",
       "0  [traveled, chicago, husband, romantic, weekend...   \n",
       "1  [stayed, sofitel, chicago, water, tower, hotel...   \n",
       "2  [hotel, gorgeous, really, enjoyed, stay, defin...   \n",
       "\n",
       "                                                 pos  \\\n",
       "0  [(i, NN), (traveled, VBD), (to, TO), (chicago,...   \n",
       "1  [(i, JJ), (stayed, VBD), (in, IN), (the, DT), ...   \n",
       "2  [(this, DT), (hotel, NN), (was, VBD), (gorgeou...   \n",
       "\n",
       "                                           pos_no_sw  \\\n",
       "0  [(traveled, VBN), (chicago, JJ), (husband, NN)...   \n",
       "1  [(stayed, JJ), (sofitel, NN), (chicago, NN), (...   \n",
       "2  [(hotel, NN), (gorgeous, JJ), (really, RB), (e...   \n",
       "\n",
       "                                            pos_dict  \\\n",
       "0  {'NN': 18, 'VBD': 6, 'TO': 1, 'VB': 3, 'IN': 6...   \n",
       "1  {'JJ': 19, 'VBD': 6, 'IN': 16, 'DT': 14, 'NN':...   \n",
       "2  {'DT': 9, 'NN': 15, 'VBD': 6, 'JJ': 10, 'RB': ...   \n",
       "\n",
       "                                      pos_dict_no_sw  \\\n",
       "0  {'VBN': 1, 'JJ': 6, 'NN': 16, 'RB': 4, 'MD': 2...   \n",
       "1  {'JJ': 15, 'NN': 29, 'CD': 1, 'NNS': 11, 'RB':...   \n",
       "2  {'NN': 15, 'JJ': 9, 'RB': 6, 'VBN': 1, 'VBG': ...   \n",
       "\n",
       "                                                 bow  \\\n",
       "0  {'i': 1, 'traveled': 1, 'to': 1, 'chicago': 2,...   \n",
       "1  {'i': 3, 'stayed': 1, 'in': 1, 'the': 9, 'sofi...   \n",
       "2  {'this': 2, 'hotel': 2, 'was': 6, 'gorgeous': ...   \n",
       "\n",
       "                                           bow_no_sw  \n",
       "0  {'traveled': 1, 'chicago': 2, 'husband': 1, 'r...  \n",
       "1  {'stayed': 1, 'sofitel': 1, 'chicago': 1, 'wat...  \n",
       "2  {'hotel': 2, 'gorgeous': 1, 'really': 1, 'enjo...  "
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# def get_bow_from_tokens(df, column):\n",
    "#     all_column_data = ' '.join(df[column].tolist())\n",
    "#     all_column_fd = Counter(all_column_data.split())\n",
    "#     return all_column_fd\n",
    "\n",
    "# # bow = get_bow_from_column(all_df, 'diy_cleaner')\n",
    "# # bow =\n",
    "from collections import Counter\n",
    "all_df['bow'] = all_df.apply(lambda x: Counter(x['tokens']), axis=1)\n",
    "all_df['bow_no_sw'] = all_df.apply(lambda x: Counter(x['no_sw']), axis=1)\n",
    "all_df[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_df_n = all_df[all_df['PoN'] == 'N']\n",
    "all_df_p = all_df[all_df['PoN'] == 'P']\n",
    "\n",
    "big_bow = [item for review in all_df['bow'].tolist() for item in review]\n",
    "big_bow_n = [item for review in all_df_n['bow'].tolist() for item in review]\n",
    "big_bow_p = [item for review in all_df_p['bow'].tolist() for item in review]\n",
    "\n",
    "df = pd.DataFrame.from_dict(Counter(big_bow), orient='index').reset_index()\n",
    "df = df.rename(columns={'index':'word', 0:'count'})\n",
    "\n",
    "df_n = pd.DataFrame.from_dict(Counter(big_bow_n), orient='index').reset_index()\n",
    "df_n = df_n.rename(columns={'index':'word', 0:'count'})\n",
    "\n",
    "df_p = pd.DataFrame.from_dict(Counter(big_bow_p), orient='index').reset_index()\n",
    "df_p = df_p.rename(columns={'index':'word', 0:'count'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<module 'matplotlib.pyplot' from '/usr/local/lib/python3.7/site-packages/matplotlib/pyplot.py'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt \n",
    "def bar_plot(df, title): \n",
    "    graph = sns.barplot(y = \"count\", x = \"word\", data = df, palette = \"husl\")\n",
    "    plt.title(title)\n",
    "    plt.xlabel(\"Word\")\n",
    "    plt.ylabel(\"Count\")\n",
    "    sns.set_context(\"talk\")\n",
    "    plt.xticks(rotation = 90)\n",
    "    return plt\n",
    "\n",
    "print(bar_plot(df.sort_values(by=[\"count\"], ascending=False)[:20], \"Top 20 Items (ALL) Prior to Cleaning\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<module 'matplotlib.pyplot' from '/usr/local/lib/python3.7/site-packages/matplotlib/pyplot.py'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(bar_plot(df_n.sort_values(by=[\"count\"], ascending=False)[:20], \"Top 20 Items (TRUE) Prior to Cleaning\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<module 'matplotlib.pyplot' from '/usr/local/lib/python3.7/site-packages/matplotlib/pyplot.py'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(bar_plot(df_p.sort_values(by=[\"count\"], ascending=False)[:20], \"Top 20 Items (FALSE) Prior to Cleaning\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_df_n = all_df[all_df['PoN'] == 'N']\n",
    "all_df_p = all_df[all_df['PoN'] == 'P']\n",
    "\n",
    "big_bow = [item for review in all_df['bow_no_sw'].tolist() for item in review]\n",
    "big_bow_n = [item for review in all_df_n['bow_no_sw'].tolist() for item in review]\n",
    "big_bow_p = [item for review in all_df_p['bow_no_sw'].tolist() for item in review]\n",
    "\n",
    "df = pd.DataFrame.from_dict(Counter(big_bow), orient='index').reset_index()\n",
    "df = df.rename(columns={'index':'word', 0:'count'})\n",
    "\n",
    "df_n = pd.DataFrame.from_dict(Counter(big_bow_n), orient='index').reset_index()\n",
    "df_n = df_n.rename(columns={'index':'word', 0:'count'})\n",
    "\n",
    "df_p = pd.DataFrame.from_dict(Counter(big_bow_p), orient='index').reset_index()\n",
    "df_p = df_p.rename(columns={'index':'word', 0:'count'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<module 'matplotlib.pyplot' from '/usr/local/lib/python3.7/site-packages/matplotlib/pyplot.py'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(bar_plot(df.sort_values(by=[\"count\"], ascending=False)[:20], \"Top 20 Items (ALL) Prior to Cleaning\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<module 'matplotlib.pyplot' from '/usr/local/lib/python3.7/site-packages/matplotlib/pyplot.py'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(bar_plot(df_n.sort_values(by=[\"count\"], ascending=False)[:20], \"Top 20 Items (TRUE) Stopwords Removed\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<module 'matplotlib.pyplot' from '/usr/local/lib/python3.7/site-packages/matplotlib/pyplot.py'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(bar_plot(df_p.sort_values(by=[\"count\"], ascending=False)[:20], \"Top 20 Items (FALSE) Stopwords Removed\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_df_n = all_df[all_df['PoN'] == 'N']\n",
    "all_df_p = all_df[all_df['PoN'] == 'P']\n",
    "\n",
    "big_bow = [item for review in all_df['pos_dict'].tolist() for item in review]\n",
    "big_bow_n = [item for review in all_df_n['pos_dict'].tolist() for item in review]\n",
    "big_bow_p = [item for review in all_df_p['pos_dict'].tolist() for item in review]\n",
    "\n",
    "df = pd.DataFrame.from_dict(Counter(big_bow), orient='index').reset_index()\n",
    "df = df.rename(columns={'index':'word', 0:'count'})\n",
    "\n",
    "df_n = pd.DataFrame.from_dict(Counter(big_bow_n), orient='index').reset_index()\n",
    "df_n = df_n.rename(columns={'index':'word', 0:'count'})\n",
    "\n",
    "df_p = pd.DataFrame.from_dict(Counter(big_bow_p), orient='index').reset_index()\n",
    "df_p = df_p.rename(columns={'index':'word', 0:'count'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<module 'matplotlib.pyplot' from '/usr/local/lib/python3.7/site-packages/matplotlib/pyplot.py'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(bar_plot(df.sort_values(by=[\"count\"], ascending=False)[:10], \"Top 10 Items (ALL) Prior to Cleaning\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<module 'matplotlib.pyplot' from '/usr/local/lib/python3.7/site-packages/matplotlib/pyplot.py'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(bar_plot(df_n.sort_values(by=[\"count\"], ascending=False)[:10], \"Top 10 POS (TRUE) Prior to Cleaning\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<module 'matplotlib.pyplot' from '/usr/local/lib/python3.7/site-packages/matplotlib/pyplot.py'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(bar_plot(df_p.sort_values(by=[\"count\"], ascending=False)[:10], \"Top 10 POS (FALSE) Prior to Cleaning\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## STEP 4: TEST EXPERIMENTS!!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import confusion_matrix, classification_report\n",
    "    \n",
    "# def cross_validation():\n",
    "    \n",
    "def get_NB(small_df, labels):\n",
    "    \n",
    "    seeds = [109, 210, 420, 19, 7]\n",
    "    for seed in seeds:\n",
    "        x_train, x_test, y_train, y_test = train_test_split(small_df.values, labels, test_size=0.3, random_state = seed)\n",
    "\n",
    "        gnb = GaussianNB()\n",
    "        gnb.fit(x_train, y_train)\n",
    "        y_pred = gnb.predict(x_test)\n",
    "        print(\"Accuracy:\", metrics.accuracy_score(y_test, y_pred))\n",
    "#         print(\"The accuracy is\", accuracy)\n",
    "        cm = confusion_matrix(y_test, y_pred)\n",
    "        # confusion_matrix_graph(cm, accuracy, \"NB Multinomial Tokenized\")\n",
    "        tn, fp, fn, tp = cm.ravel()\n",
    "        print(tn, fp, fn, tp)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### TEST 1: Parts of speech frequency distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>NN</th>\n",
       "      <th>VBD</th>\n",
       "      <th>TO</th>\n",
       "      <th>VB</th>\n",
       "      <th>IN</th>\n",
       "      <th>PRP$</th>\n",
       "      <th>DT</th>\n",
       "      <th>JJ</th>\n",
       "      <th>RB</th>\n",
       "      <th>MD</th>\n",
       "      <th>...</th>\n",
       "      <th>WDT</th>\n",
       "      <th>PDT</th>\n",
       "      <th>JJR</th>\n",
       "      <th>WP</th>\n",
       "      <th>JJS</th>\n",
       "      <th>EX</th>\n",
       "      <th>RBS</th>\n",
       "      <th>NNP</th>\n",
       "      <th>UH</th>\n",
       "      <th>FW</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PoN</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>18</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>6</td>\n",
       "      <td>3.0</td>\n",
       "      <td>7.0</td>\n",
       "      <td>5</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>28</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>16</td>\n",
       "      <td>2.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>19</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>15</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>10</td>\n",
       "      <td>8.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     NN  VBD   TO   VB  IN  PRP$    DT  JJ   RB   MD  ...  WDT  PDT  JJR  WP  \\\n",
       "PoN                                                   ...                      \n",
       "N    18  6.0  1.0  3.0   6   3.0   7.0   5  4.0  2.0  ...  NaN  NaN  NaN NaN   \n",
       "N    28  6.0  2.0  3.0  16   2.0  14.0  19  4.0  1.0  ...  NaN  NaN  NaN NaN   \n",
       "N    15  6.0  1.0  2.0   3   1.0   9.0  10  8.0  2.0  ...  NaN  NaN  NaN NaN   \n",
       "\n",
       "     JJS  EX  RBS  NNP  UH  FW  \n",
       "PoN                             \n",
       "N    NaN NaN  NaN  NaN NaN NaN  \n",
       "N    NaN NaN  NaN  NaN NaN NaN  \n",
       "N    NaN NaN  NaN  NaN NaN NaN  \n",
       "\n",
       "[3 rows x 31 columns]"
      ]
     },
     "execution_count": 230,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pos_df = pd.DataFrame(all_df['pos_dict'].tolist(), all_df['PoN'])\n",
    "pos_df[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>NN</th>\n",
       "      <th>VBD</th>\n",
       "      <th>TO</th>\n",
       "      <th>VB</th>\n",
       "      <th>IN</th>\n",
       "      <th>PRP$</th>\n",
       "      <th>DT</th>\n",
       "      <th>JJ</th>\n",
       "      <th>RB</th>\n",
       "      <th>MD</th>\n",
       "      <th>...</th>\n",
       "      <th>WDT</th>\n",
       "      <th>PDT</th>\n",
       "      <th>JJR</th>\n",
       "      <th>WP</th>\n",
       "      <th>JJS</th>\n",
       "      <th>EX</th>\n",
       "      <th>RBS</th>\n",
       "      <th>NNP</th>\n",
       "      <th>UH</th>\n",
       "      <th>FW</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PoN</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>18</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>3</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>28</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>16</td>\n",
       "      <td>2</td>\n",
       "      <td>14</td>\n",
       "      <td>19</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>15</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     NN  VBD  TO  VB  IN  PRP$  DT  JJ  RB  MD  ...  WDT  PDT  JJR  WP  JJS  \\\n",
       "PoN                                             ...                           \n",
       "N    18    6   1   3   6     3   7   5   4   2  ...    0    0    0   0    0   \n",
       "N    28    6   2   3  16     2  14  19   4   1  ...    0    0    0   0    0   \n",
       "N    15    6   1   2   3     1   9  10   8   2  ...    0    0    0   0    0   \n",
       "\n",
       "     EX  RBS  NNP  UH  FW  \n",
       "PoN                        \n",
       "N     0    0    0   0   0  \n",
       "N     0    0    0   0   0  \n",
       "N     0    0    0   0   0  \n",
       "\n",
       "[3 rows x 31 columns]"
      ]
     },
     "execution_count": 231,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pos_df = pos_df.fillna(0).astype(int)\n",
    "pos_df[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.5\n",
      "47 3 45 1\n",
      "Accuracy: 0.4895833333333333\n",
      "45 1 48 2\n",
      "Accuracy: 0.4791666666666667\n",
      "42 1 49 4\n",
      "Accuracy: 0.53125\n",
      "47 0 45 4\n",
      "Accuracy: 0.40625\n",
      "35 3 54 4\n"
     ]
    }
   ],
   "source": [
    "get_NB(pos_df, pos_df.index)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### TEST 1b: Normalized parts of speech frequency distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>NN</th>\n",
       "      <th>VBD</th>\n",
       "      <th>TO</th>\n",
       "      <th>VB</th>\n",
       "      <th>IN</th>\n",
       "      <th>PRP$</th>\n",
       "      <th>DT</th>\n",
       "      <th>JJ</th>\n",
       "      <th>RB</th>\n",
       "      <th>MD</th>\n",
       "      <th>...</th>\n",
       "      <th>PDT</th>\n",
       "      <th>JJR</th>\n",
       "      <th>WP</th>\n",
       "      <th>JJS</th>\n",
       "      <th>EX</th>\n",
       "      <th>RBS</th>\n",
       "      <th>NNP</th>\n",
       "      <th>UH</th>\n",
       "      <th>FW</th>\n",
       "      <th>total</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PoN</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>18</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>3</td>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>28</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>16</td>\n",
       "      <td>2</td>\n",
       "      <td>14</td>\n",
       "      <td>19</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>15</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>8</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>69</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 32 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     NN  VBD  TO  VB  IN  PRP$  DT  JJ  RB  MD  ...  PDT  JJR  WP  JJS  EX  \\\n",
       "PoN                                             ...                          \n",
       "N    18    6   1   3   6     3   7   5   4   2  ...    0    0   0    0   0   \n",
       "N    28    6   2   3  16     2  14  19   4   1  ...    0    0   0    0   0   \n",
       "N    15    6   1   2   3     1   9  10   8   2  ...    0    0   0    0   0   \n",
       "\n",
       "     RBS  NNP  UH  FW  total  \n",
       "PoN                           \n",
       "N      0    0   0   0     68  \n",
       "N      0    0   0   0    129  \n",
       "N      0    0   0   0     69  \n",
       "\n",
       "[3 rows x 32 columns]"
      ]
     },
     "execution_count": 146,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pos_df_norm = pos_df.copy()\n",
    "pos_df_norm = pos_df_norm.apply(lambda x: x/x.sum(), axis=1)\n",
    "pos_df_norm[:3]\n",
    "pos_df_norm[1:]\n",
    "test = pos_df.copy()\n",
    "test['total'] = test.sum(axis = 1)\n",
    "test[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>NN</th>\n",
       "      <th>VBD</th>\n",
       "      <th>TO</th>\n",
       "      <th>VB</th>\n",
       "      <th>IN</th>\n",
       "      <th>PRP$</th>\n",
       "      <th>DT</th>\n",
       "      <th>JJ</th>\n",
       "      <th>RB</th>\n",
       "      <th>MD</th>\n",
       "      <th>...</th>\n",
       "      <th>WDT</th>\n",
       "      <th>PDT</th>\n",
       "      <th>JJR</th>\n",
       "      <th>WP</th>\n",
       "      <th>JJS</th>\n",
       "      <th>EX</th>\n",
       "      <th>RBS</th>\n",
       "      <th>NNP</th>\n",
       "      <th>UH</th>\n",
       "      <th>FW</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PoN</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.264706</td>\n",
       "      <td>0.088235</td>\n",
       "      <td>0.014706</td>\n",
       "      <td>0.044118</td>\n",
       "      <td>0.088235</td>\n",
       "      <td>0.044118</td>\n",
       "      <td>0.102941</td>\n",
       "      <td>0.073529</td>\n",
       "      <td>0.058824</td>\n",
       "      <td>0.029412</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.217054</td>\n",
       "      <td>0.046512</td>\n",
       "      <td>0.015504</td>\n",
       "      <td>0.023256</td>\n",
       "      <td>0.124031</td>\n",
       "      <td>0.015504</td>\n",
       "      <td>0.108527</td>\n",
       "      <td>0.147287</td>\n",
       "      <td>0.031008</td>\n",
       "      <td>0.007752</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.217391</td>\n",
       "      <td>0.086957</td>\n",
       "      <td>0.014493</td>\n",
       "      <td>0.028986</td>\n",
       "      <td>0.043478</td>\n",
       "      <td>0.014493</td>\n",
       "      <td>0.130435</td>\n",
       "      <td>0.144928</td>\n",
       "      <td>0.115942</td>\n",
       "      <td>0.028986</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 31 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           NN       VBD        TO        VB        IN      PRP$        DT  \\\n",
       "PoN                                                                         \n",
       "N    0.264706  0.088235  0.014706  0.044118  0.088235  0.044118  0.102941   \n",
       "N    0.217054  0.046512  0.015504  0.023256  0.124031  0.015504  0.108527   \n",
       "N    0.217391  0.086957  0.014493  0.028986  0.043478  0.014493  0.130435   \n",
       "\n",
       "           JJ        RB        MD  ...  WDT  PDT  JJR   WP  JJS   EX  RBS  \\\n",
       "PoN                                ...                                      \n",
       "N    0.073529  0.058824  0.029412  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "N    0.147287  0.031008  0.007752  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "N    0.144928  0.115942  0.028986  ...  0.0  0.0  0.0  0.0  0.0  0.0  0.0   \n",
       "\n",
       "     NNP   UH   FW  \n",
       "PoN                 \n",
       "N    0.0  0.0  0.0  \n",
       "N    0.0  0.0  0.0  \n",
       "N    0.0  0.0  0.0  \n",
       "\n",
       "[3 rows x 31 columns]"
      ]
     },
     "execution_count": 147,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pos_df_norm[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.5104166666666666\n",
      "Accuracy: 0.5\n",
      "Accuracy: 0.4791666666666667\n",
      "Accuracy: 0.5208333333333334\n",
      "Accuracy: 0.3854166666666667\n"
     ]
    }
   ],
   "source": [
    "get_NB(pos_df_norm, pos_df.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.5625\n",
      "Accuracy: 0.65625\n",
      "Accuracy: 0.5625\n",
      "Accuracy: 0.5833333333333334\n",
      "Accuracy: 0.5416666666666666\n"
     ]
    }
   ],
   "source": [
    "# small_df\n",
    "small_df = pos_df_norm.filter(['PRP', 'PRP$','NN'])\n",
    "get_NB(small_df, pos_df.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 192,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1968\n",
      "959\n",
      "1009\n",
      "0.4872967479674797\n",
      "0.5127032520325203\n"
     ]
    }
   ],
   "source": [
    "pos_df_n = pos_df[pos_df.index == 'N']\n",
    "pos_df_p = pos_df[pos_df.index == 'P']\n",
    "print(pos_df['PRP'].sum())\n",
    "print(pos_df_n['PRP'].sum())\n",
    "print(pos_df_p['PRP'].sum())\n",
    "print(pos_df_n['PRP'].sum()/pos_df['PRP'].sum())\n",
    "print(pos_df_p['PRP'].sum()/pos_df['PRP'].sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "888\n",
      "518\n",
      "370\n"
     ]
    }
   ],
   "source": [
    "pos_df_n = pos_df[pos_df.index == 'N']\n",
    "pos_df_p = pos_df[pos_df.index == 'P']\n",
    "print(pos_df['PRP$'].sum())\n",
    "print(pos_df_n['PRP$'].sum())\n",
    "print(pos_df_p['PRP$'].sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13.405406541226437\n",
      "6.357665198600147\n",
      "7.04774134262629\n"
     ]
    }
   ],
   "source": [
    "pos_df_n = pos_df_norm[pos_df_norm.index == 'N']\n",
    "pos_df_p = pos_df_norm[pos_df_norm.index == 'P']\n",
    "print(pos_df_norm['PRP'].sum())\n",
    "print(pos_df_n['PRP'].sum())\n",
    "print(pos_df_p['PRP'].sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.041891895441332615\n",
      "0.039735407491250915\n",
      "0.044048383391414314\n"
     ]
    }
   ],
   "source": [
    "pos_df_n = pos_df_norm[pos_df_norm.index == 'N']\n",
    "pos_df_p = pos_df_norm[pos_df_norm.index == 'P']\n",
    "print(pos_df_norm['PRP'].mean())\n",
    "print(pos_df_n['PRP'].mean())\n",
    "print(pos_df_p['PRP'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 196,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.019007323131444483\n",
      "0.022148272795477524\n",
      "0.015866373467411442\n"
     ]
    }
   ],
   "source": [
    "pos_df_n = pos_df_norm[pos_df_norm.index == 'N']\n",
    "pos_df_p = pos_df_norm[pos_df_norm.index == 'P']\n",
    "print(pos_df_norm['PRP$'].mean())\n",
    "print(pos_df_n['PRP$'].mean())\n",
    "print(pos_df_p['PRP$'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 197,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      [(i, NN), (traveled, VBD), (to, TO), (chicago,...\n",
       "1      [(i, JJ), (stayed, VBD), (in, IN), (the, DT), ...\n",
       "2      [(this, DT), (hotel, NN), (was, VBD), (gorgeou...\n",
       "3      [(this, DT), (is, VBZ), (an, DT), (absolutely,...\n",
       "4      [(i, NN), (recently, RB), (traveled, VBD), (up...\n",
       "                             ...                        \n",
       "315    [(this, DT), (hotel, NN), (was, VBD), (not, RB...\n",
       "316    [(i, JJ), (stayed, VBD), (at, IN), (the, DT), ...\n",
       "317    [(we, PRP), (had, VBD), (a, DT), (reservation,...\n",
       "318    [(i, NN), (am, VBP), (staying, VBG), (here, RB...\n",
       "319    [(we, PRP), (enjoyed, VBD), (the, DT), (hotel,...\n",
       "Name: pos, Length: 320, dtype: object"
      ]
     },
     "execution_count": 197,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_df['pos']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 198,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_df['pos_sent'] = all_df.apply(lambda x: [word[1] for word in x['pos']], axis=1)\n",
    "all_df['pos_sent_str'] = all_df.apply(lambda x: [' '.join(x['pos_sent'])], axis=1)\n",
    "all_df['pos_no_sw_sent'] = all_df.apply(lambda x: [word[1] for word in x['pos_no_sw']], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 199,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "list"
      ]
     },
     "execution_count": 199,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "type(all_df['pos_sent_str'][1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 200,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_df['pos_sent_bi'] = all_df.apply(lambda x: [b for l in x['pos_sent_str'] for b in zip(l.split(\" \")[:-1], l.split(\" \")[1:])], axis=1)\n",
    "# bigrams = [b for l in text for b in zip(l.split(\" \")[:-1], l.split(\" \")[1:])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 201,
   "metadata": {},
   "outputs": [],
   "source": [
    "# all_df['pos_sent_tri'] = all_df.apply(lambda x: [b for l in x['pos_sent_str'] for b in zip(l.split(\" \")[:-1], l.split(\" \")[1:])], axis=1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>PoN</th>\n",
       "      <th>sentences</th>\n",
       "      <th>num_sentences</th>\n",
       "      <th>tokens</th>\n",
       "      <th>num_tokens</th>\n",
       "      <th>no_sw</th>\n",
       "      <th>num_no_sw</th>\n",
       "      <th>stemmed</th>\n",
       "      <th>stemmed_no_sw</th>\n",
       "      <th>...</th>\n",
       "      <th>bow_no_sw</th>\n",
       "      <th>pos_sent</th>\n",
       "      <th>pos_sent_str</th>\n",
       "      <th>pos_no_sw_sent</th>\n",
       "      <th>pos_sent_bi</th>\n",
       "      <th>bow_pos</th>\n",
       "      <th>trigrams</th>\n",
       "      <th>trigrams_pos</th>\n",
       "      <th>trigrams_feats</th>\n",
       "      <th>trigrams_feats_bow</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>I traveled to Chicago with my husband for a ro...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I traveled to Chicago with my husband for a r...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, traveled, to, chicago, with, my, husband, ...</td>\n",
       "      <td>68</td>\n",
       "      <td>[traveled, chicago, husband, romantic, weekend...</td>\n",
       "      <td>40</td>\n",
       "      <td>[i, travel, to, chicago, with, my, husband, fo...</td>\n",
       "      <td>[travel, chicago, husband, romant, weekend, aw...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'traveled': 1, 'chicago': 2, 'husband': 1, 'r...</td>\n",
       "      <td>[NN, VBD, TO, VB, IN, PRP$, NN, IN, DT, JJ, NN...</td>\n",
       "      <td>[NN VBD TO VB IN PRP$ NN IN DT JJ NN RB PRP$ N...</td>\n",
       "      <td>[VBN, JJ, NN, JJ, NN, RB, JJ, JJ, NN, NN, NN, ...</td>\n",
       "      <td>[(NN, VBD), (VBD, TO), (TO, VB), (VB, IN), (IN...</td>\n",
       "      <td>{('NN', 'VBD'): 4, ('VBD', 'TO'): 1, ('TO', 'V...</td>\n",
       "      <td>[(i, traveled, to), (traveled, to, chicago), (...</td>\n",
       "      <td>[(NN, VBD, TO), (VBD, TO, VB), (TO, VB, IN), (...</td>\n",
       "      <td>[NN_VBD_TO, VBD_TO_VB, TO_VB_IN, VB_IN_PRP, IN...</td>\n",
       "      <td>{'NN_VBD_TO': 1, 'VBD_TO_VB': 1, 'TO_VB_IN': 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>I stayed in the Sofitel Chicago Water Tower ho...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I stayed in the Sofitel Chicago Water Tower h...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, stayed, in, the, sofitel, chicago, water, ...</td>\n",
       "      <td>129</td>\n",
       "      <td>[stayed, sofitel, chicago, water, tower, hotel...</td>\n",
       "      <td>71</td>\n",
       "      <td>[i, stay, in, the, sofitel, chicago, water, to...</td>\n",
       "      <td>[stay, sofitel, chicago, water, tower, hotel, ...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'stayed': 1, 'sofitel': 1, 'chicago': 1, 'wat...</td>\n",
       "      <td>[JJ, VBD, IN, DT, NN, NN, NN, NN, NN, IN, PRP$...</td>\n",
       "      <td>[JJ VBD IN DT NN NN NN NN NN IN PRP$ NN CC CD ...</td>\n",
       "      <td>[JJ, NN, NN, NN, NN, NN, NN, CD, NNS, JJ, NN, ...</td>\n",
       "      <td>[(JJ, VBD), (VBD, IN), (IN, DT), (DT, NN), (NN...</td>\n",
       "      <td>{('JJ', 'VBD'): 1, ('VBD', 'IN'): 1, ('IN', 'D...</td>\n",
       "      <td>[(i, stayed, in), (stayed, in, the), (in, the,...</td>\n",
       "      <td>[(JJ, VBD, IN), (VBD, IN, DT), (IN, DT, NN), (...</td>\n",
       "      <td>[JJ_VBD_IN, VBD_IN_DT, IN_DT_NN, DT_NN_NN, NN_...</td>\n",
       "      <td>{'JJ_VBD_IN': 1, 'VBD_IN_DT': 1, 'IN_DT_NN': 2...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>This hotel was gorgeous! I really enjoyed my s...</td>\n",
       "      <td>N</td>\n",
       "      <td>[This hotel was gorgeous!, I really enjoyed my...</td>\n",
       "      <td>7</td>\n",
       "      <td>[this, hotel, was, gorgeous, i, really, enjoye...</td>\n",
       "      <td>69</td>\n",
       "      <td>[hotel, gorgeous, really, enjoyed, stay, defin...</td>\n",
       "      <td>36</td>\n",
       "      <td>[thi, hotel, wa, gorgeou, i, realli, enjoy, my...</td>\n",
       "      <td>[hotel, gorgeou, realli, enjoy, stay, definit,...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'hotel': 2, 'gorgeous': 1, 'really': 1, 'enjo...</td>\n",
       "      <td>[DT, NN, VBD, JJ, JJ, RB, VBN, PRP$, NN, RB, C...</td>\n",
       "      <td>[DT NN VBD JJ JJ RB VBN PRP$ NN RB CC MD RB VB...</td>\n",
       "      <td>[NN, JJ, RB, VBN, NN, RB, VBG, JJ, NN, JJ, NN,...</td>\n",
       "      <td>[(DT, NN), (NN, VBD), (VBD, JJ), (JJ, JJ), (JJ...</td>\n",
       "      <td>{('DT', 'NN'): 8, ('NN', 'VBD'): 6, ('VBD', 'J...</td>\n",
       "      <td>[(this, hotel, was), (hotel, was, gorgeous), (...</td>\n",
       "      <td>[(DT, NN, VBD), (NN, VBD, JJ), (VBD, JJ, JJ), ...</td>\n",
       "      <td>[DT_NN_VBD, NN_VBD_JJ, VBD_JJ_JJ, JJ_JJ_RB, JJ...</td>\n",
       "      <td>{'DT_NN_VBD': 4, 'NN_VBD_JJ': 4, 'VBD_JJ_JJ': ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>This is an absolutely exquisite hotel, at a gr...</td>\n",
       "      <td>N</td>\n",
       "      <td>[This is an absolutely exquisite hotel, at a g...</td>\n",
       "      <td>6</td>\n",
       "      <td>[this, is, an, absolutely, exquisite, hotel, a...</td>\n",
       "      <td>110</td>\n",
       "      <td>[absolutely, exquisite, hotel, great, location...</td>\n",
       "      <td>52</td>\n",
       "      <td>[thi, is, an, absolut, exquisit, hotel, at, a,...</td>\n",
       "      <td>[absolut, exquisit, hotel, great, locat, boast...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'absolutely': 1, 'exquisite': 1, 'hotel': 3, ...</td>\n",
       "      <td>[DT, VBZ, DT, RB, JJ, NN, IN, DT, JJ, NN, CC, ...</td>\n",
       "      <td>[DT VBZ DT RB JJ NN IN DT JJ NN CC NN NN NNS N...</td>\n",
       "      <td>[RB, JJ, NN, JJ, NN, VBG, NN, NNS, JJ, NN, RB,...</td>\n",
       "      <td>[(DT, VBZ), (VBZ, DT), (DT, RB), (RB, JJ), (JJ...</td>\n",
       "      <td>{('DT', 'VBZ'): 1, ('VBZ', 'DT'): 2, ('DT', 'R...</td>\n",
       "      <td>[(this, is, an), (is, an, absolutely), (an, ab...</td>\n",
       "      <td>[(DT, VBZ, DT), (VBZ, DT, RB), (DT, RB, JJ), (...</td>\n",
       "      <td>[DT_VBZ_DT, VBZ_DT_RB, DT_RB_JJ, RB_JJ_NN, JJ_...</td>\n",
       "      <td>{'DT_VBZ_DT': 1, 'VBZ_DT_RB': 1, 'DT_RB_JJ': 1...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 27 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   0 PoN  \\\n",
       "0  I traveled to Chicago with my husband for a ro...   N   \n",
       "1  I stayed in the Sofitel Chicago Water Tower ho...   N   \n",
       "2  This hotel was gorgeous! I really enjoyed my s...   N   \n",
       "3  This is an absolutely exquisite hotel, at a gr...   N   \n",
       "\n",
       "                                           sentences  num_sentences  \\\n",
       "0  [I traveled to Chicago with my husband for a r...              6   \n",
       "1  [I stayed in the Sofitel Chicago Water Tower h...              6   \n",
       "2  [This hotel was gorgeous!, I really enjoyed my...              7   \n",
       "3  [This is an absolutely exquisite hotel, at a g...              6   \n",
       "\n",
       "                                              tokens  num_tokens  \\\n",
       "0  [i, traveled, to, chicago, with, my, husband, ...          68   \n",
       "1  [i, stayed, in, the, sofitel, chicago, water, ...         129   \n",
       "2  [this, hotel, was, gorgeous, i, really, enjoye...          69   \n",
       "3  [this, is, an, absolutely, exquisite, hotel, a...         110   \n",
       "\n",
       "                                               no_sw  num_no_sw  \\\n",
       "0  [traveled, chicago, husband, romantic, weekend...         40   \n",
       "1  [stayed, sofitel, chicago, water, tower, hotel...         71   \n",
       "2  [hotel, gorgeous, really, enjoyed, stay, defin...         36   \n",
       "3  [absolutely, exquisite, hotel, great, location...         52   \n",
       "\n",
       "                                             stemmed  \\\n",
       "0  [i, travel, to, chicago, with, my, husband, fo...   \n",
       "1  [i, stay, in, the, sofitel, chicago, water, to...   \n",
       "2  [thi, hotel, wa, gorgeou, i, realli, enjoy, my...   \n",
       "3  [thi, is, an, absolut, exquisit, hotel, at, a,...   \n",
       "\n",
       "                                       stemmed_no_sw  ...  \\\n",
       "0  [travel, chicago, husband, romant, weekend, aw...  ...   \n",
       "1  [stay, sofitel, chicago, water, tower, hotel, ...  ...   \n",
       "2  [hotel, gorgeou, realli, enjoy, stay, definit,...  ...   \n",
       "3  [absolut, exquisit, hotel, great, locat, boast...  ...   \n",
       "\n",
       "                                           bow_no_sw  \\\n",
       "0  {'traveled': 1, 'chicago': 2, 'husband': 1, 'r...   \n",
       "1  {'stayed': 1, 'sofitel': 1, 'chicago': 1, 'wat...   \n",
       "2  {'hotel': 2, 'gorgeous': 1, 'really': 1, 'enjo...   \n",
       "3  {'absolutely': 1, 'exquisite': 1, 'hotel': 3, ...   \n",
       "\n",
       "                                            pos_sent  \\\n",
       "0  [NN, VBD, TO, VB, IN, PRP$, NN, IN, DT, JJ, NN...   \n",
       "1  [JJ, VBD, IN, DT, NN, NN, NN, NN, NN, IN, PRP$...   \n",
       "2  [DT, NN, VBD, JJ, JJ, RB, VBN, PRP$, NN, RB, C...   \n",
       "3  [DT, VBZ, DT, RB, JJ, NN, IN, DT, JJ, NN, CC, ...   \n",
       "\n",
       "                                        pos_sent_str  \\\n",
       "0  [NN VBD TO VB IN PRP$ NN IN DT JJ NN RB PRP$ N...   \n",
       "1  [JJ VBD IN DT NN NN NN NN NN IN PRP$ NN CC CD ...   \n",
       "2  [DT NN VBD JJ JJ RB VBN PRP$ NN RB CC MD RB VB...   \n",
       "3  [DT VBZ DT RB JJ NN IN DT JJ NN CC NN NN NNS N...   \n",
       "\n",
       "                                      pos_no_sw_sent  \\\n",
       "0  [VBN, JJ, NN, JJ, NN, RB, JJ, JJ, NN, NN, NN, ...   \n",
       "1  [JJ, NN, NN, NN, NN, NN, NN, CD, NNS, JJ, NN, ...   \n",
       "2  [NN, JJ, RB, VBN, NN, RB, VBG, JJ, NN, JJ, NN,...   \n",
       "3  [RB, JJ, NN, JJ, NN, VBG, NN, NNS, JJ, NN, RB,...   \n",
       "\n",
       "                                         pos_sent_bi  \\\n",
       "0  [(NN, VBD), (VBD, TO), (TO, VB), (VB, IN), (IN...   \n",
       "1  [(JJ, VBD), (VBD, IN), (IN, DT), (DT, NN), (NN...   \n",
       "2  [(DT, NN), (NN, VBD), (VBD, JJ), (JJ, JJ), (JJ...   \n",
       "3  [(DT, VBZ), (VBZ, DT), (DT, RB), (RB, JJ), (JJ...   \n",
       "\n",
       "                                             bow_pos  \\\n",
       "0  {('NN', 'VBD'): 4, ('VBD', 'TO'): 1, ('TO', 'V...   \n",
       "1  {('JJ', 'VBD'): 1, ('VBD', 'IN'): 1, ('IN', 'D...   \n",
       "2  {('DT', 'NN'): 8, ('NN', 'VBD'): 6, ('VBD', 'J...   \n",
       "3  {('DT', 'VBZ'): 1, ('VBZ', 'DT'): 2, ('DT', 'R...   \n",
       "\n",
       "                                            trigrams  \\\n",
       "0  [(i, traveled, to), (traveled, to, chicago), (...   \n",
       "1  [(i, stayed, in), (stayed, in, the), (in, the,...   \n",
       "2  [(this, hotel, was), (hotel, was, gorgeous), (...   \n",
       "3  [(this, is, an), (is, an, absolutely), (an, ab...   \n",
       "\n",
       "                                        trigrams_pos  \\\n",
       "0  [(NN, VBD, TO), (VBD, TO, VB), (TO, VB, IN), (...   \n",
       "1  [(JJ, VBD, IN), (VBD, IN, DT), (IN, DT, NN), (...   \n",
       "2  [(DT, NN, VBD), (NN, VBD, JJ), (VBD, JJ, JJ), ...   \n",
       "3  [(DT, VBZ, DT), (VBZ, DT, RB), (DT, RB, JJ), (...   \n",
       "\n",
       "                                      trigrams_feats  \\\n",
       "0  [NN_VBD_TO, VBD_TO_VB, TO_VB_IN, VB_IN_PRP, IN...   \n",
       "1  [JJ_VBD_IN, VBD_IN_DT, IN_DT_NN, DT_NN_NN, NN_...   \n",
       "2  [DT_NN_VBD, NN_VBD_JJ, VBD_JJ_JJ, JJ_JJ_RB, JJ...   \n",
       "3  [DT_VBZ_DT, VBZ_DT_RB, DT_RB_JJ, RB_JJ_NN, JJ_...   \n",
       "\n",
       "                                  trigrams_feats_bow  \n",
       "0  {'NN_VBD_TO': 1, 'VBD_TO_VB': 1, 'TO_VB_IN': 1...  \n",
       "1  {'JJ_VBD_IN': 1, 'VBD_IN_DT': 1, 'IN_DT_NN': 2...  \n",
       "2  {'DT_NN_VBD': 4, 'NN_VBD_JJ': 4, 'VBD_JJ_JJ': ...  \n",
       "3  {'DT_VBZ_DT': 1, 'VBZ_DT_RB': 1, 'DT_RB_JJ': 1...  \n",
       "\n",
       "[4 rows x 27 columns]"
      ]
     },
     "execution_count": 202,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_df[:4]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 203,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['JJ VBD IN DT NN NN NN NN NN IN PRP$ NN CC CD NNS JJ NN CC MD RB VB DT NN RB IN JJ NNS DT JJ NNS IN JJ JJ NN CC JJ NNS VBD DT JJ NN TO DT JJ JJ NNS CC NNS IN NN VBP VBN IN JJ NN NNS DT NNS VBD JJ NN CC IN NNS CC JJ NNS JJ IN JJ NN NN CC VB DT NN VBD JJ RB VBN CC VBN IN DT NNS DT NN NN VBD VBG CC DT NN IN PRP$ NN VBD JJ JJ VBN DT NN IN DT NN NN IN DT JJ NN IN TO DT NN IN VBG NN NN CC VBG RP NNS IN VBG PRP VB RB IN NN']"
      ]
     },
     "execution_count": 203,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test = all_df['pos_sent_str'][1]\n",
    "test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 204,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[('NN', 'NNS'), ('NNS', 'VBP'), ('VBP', 'JJ'), ('JJ', 'JJ'), ('JJ', 'NN'), ('NN', 'NN'), ('NN', 'CC'), ('CC', 'JJ'), ('JJ', 'NN'), ('NN', 'JJ'), ('JJ', 'VBZ'), ('VBZ', 'DT'), ('DT', 'NN'), ('NN', 'VBZ'), ('VBZ', 'RB'), ('RB', 'JJ'), ('JJ', 'RB'), ('RB', 'CC'), ('CC', 'JJ'), ('JJ', 'VB'), ('VB', 'JJ'), ('JJ', 'TO'), ('TO', 'VB'), ('VB', 'DT'), ('DT', 'NN'), ('NN', 'TO'), ('TO', 'VB'), ('VB', 'NN'), ('NN', 'IN'), ('IN', 'RB'), ('RB', 'RB'), ('RB', 'IN'), ('IN', 'PRP'), ('PRP', 'VBP'), ('VBP', 'VBN'), ('VBN', 'IN'), ('IN', 'DT'), ('DT', 'JJ'), ('JJ', 'NNS'), ('NNS', 'DT'), ('DT', 'NN'), ('NN', 'VBP'), ('VBP', 'JJ'), ('JJ', 'IN'), ('IN', 'PRP$'), ('PRP$', 'NNS'), ('NNS', 'CC'), ('CC', 'NN'), ('NN', 'TO'), ('TO', 'VB'), ('VB', 'NN'), ('NN', 'NN')]\n"
     ]
    }
   ],
   "source": [
    "text = [\"this is a sentence\", \"so is this one\"]\n",
    "test2 = [\"NN NNS VBP JJ JJ NN NN CC JJ NN JJ VBZ DT NN VBZ RB JJ RB CC JJ VB JJ TO VB DT NN TO VB NN IN RB RB IN\", \"PRP VBP VBN IN DT JJ NNS DT NN VBP JJ IN PRP$ NNS CC NN TO VB NN NN\"]\n",
    "test1 = ['NN NNS VBP JJ JJ NN NN CC JJ NN JJ VBZ DT NN VBZ RB JJ RB CC JJ VB JJ TO VB DT NN TO VB NN IN RB RB IN PRP VBP VBN IN DT JJ NNS DT NN VBP JJ IN PRP$ NNS CC NN TO VB NN NN']\n",
    "bigrams = [b for l in test1 for b in zip(l.split(\" \")[:-1], l.split(\" \")[1:])]\n",
    "print(bigrams)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('NN', 'VBD'), ('VBD', 'TO'), ('TO', 'VB'), ('VB', 'IN'), ('IN', 'PRP$')]"
      ]
     },
     "execution_count": 205,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# all_bigrams = [bigram for bigram in all_df.pos_sent_bi.tolist()]\n",
    "# flat_list = [item for sublist in l for item in sublist]\n",
    "all_df_n = all_df[all_df['PoN'] == 'N']\n",
    "all_df_p = all_df[all_df['PoN'] == 'P']\n",
    "all_bigrams = [bigram for sublist in all_df.pos_sent_bi.tolist() for bigram in sublist]\n",
    "all_bigrams_n = [bigram for sublist in all_df_n.pos_sent_bi.tolist() for bigram in sublist]\n",
    "all_bigrams_p = [bigram for sublist in all_df_p.pos_sent_bi.tolist() for bigram in sublist]\n",
    "all_bigrams[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [],
   "source": [
    "count = Counter(all_bigrams)\n",
    "count_n = Counter(all_bigrams_n)\n",
    "count_p = Counter(all_bigrams_p)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['TO', 'VB'], dtype='<U3')"
      ]
     },
     "execution_count": 207,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "count.most_common()[:5]\n",
    "\n",
    "import numpy as np\n",
    "\n",
    "# all_df['bow_v3'] = all_df.apply(lambda x: Counter(casual_tokenize(x['pruned'])), axis=1)\n",
    "# new_df = pd.DataFrame(all_df['bow_v3'].tolist(), all_df['PoN'])\n",
    "\n",
    "# most_common_pos = [word[0] for word in big_bow_p.most_common(100)]\n",
    "# print(\"Unique values in array1 that are not in array2:\")\n",
    "most_common_n = [word[0] for word in count_n.most_common(10)]\n",
    "most_common_p = [word[0] for word in count_p.most_common(10)]\n",
    "\n",
    "neg_notpos = np.setdiff1d(most_common_n, most_common_p)\n",
    "neg_notpos\n",
    "\n",
    "# all_bigrams_n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_df_n = all_df[all_df['PoN'] == 'N']\n",
    "all_df_p = all_df[all_df['PoN'] == 'P']\n",
    "\n",
    "big_bow = [item for review in all_df['pos_sent_bi'].tolist() for item in review]\n",
    "big_bow_n = [item for review in all_df_n['pos_sent_bi'].tolist() for item in review]\n",
    "big_bow_p = [item for review in all_df_p['pos_sent_bi'].tolist() for item in review]\n",
    "\n",
    "df = pd.DataFrame.from_dict(Counter(big_bow), orient='index').reset_index()\n",
    "df = df.rename(columns={'index':'word', 0:'count'})\n",
    "\n",
    "df_n = pd.DataFrame.from_dict(Counter(big_bow_n), orient='index').reset_index()\n",
    "df_n = df_n.rename(columns={'index':'word', 0:'count'})\n",
    "\n",
    "df_p = pd.DataFrame.from_dict(Counter(big_bow_p), orient='index').reset_index()\n",
    "df_p = df_p.rename(columns={'index':'word', 0:'count'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<module 'matplotlib.pyplot' from '/usr/local/lib/python3.7/site-packages/matplotlib/pyplot.py'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(bar_plot(df_p.sort_values(by=[\"count\"], ascending=False)[:10], \"Top 10 POS Bigrams (ALL)\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<module 'matplotlib.pyplot' from '/usr/local/lib/python3.7/site-packages/matplotlib/pyplot.py'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(bar_plot(df_n.sort_values(by=[\"count\"], ascending=False)[:10], \"Top 10 POS Bigrams (TRUE)\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<module 'matplotlib.pyplot' from '/usr/local/lib/python3.7/site-packages/matplotlib/pyplot.py'>\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(bar_plot(df_p.sort_values(by=[\"count\"], ascending=False)[:10], \"Top 10 POS Bigrams (FALSE)\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>(NN, VBD)</th>\n",
       "      <th>(VBD, TO)</th>\n",
       "      <th>(TO, VB)</th>\n",
       "      <th>(VB, IN)</th>\n",
       "      <th>(IN, PRP$)</th>\n",
       "      <th>(PRP$, NN)</th>\n",
       "      <th>(NN, IN)</th>\n",
       "      <th>(IN, DT)</th>\n",
       "      <th>(DT, JJ)</th>\n",
       "      <th>(JJ, NN)</th>\n",
       "      <th>...</th>\n",
       "      <th>(RBS, PRP)</th>\n",
       "      <th>(PRP, JJR)</th>\n",
       "      <th>(JJ, RBS)</th>\n",
       "      <th>(RBS, RB)</th>\n",
       "      <th>(PRP$, TO)</th>\n",
       "      <th>(WRB, VBD)</th>\n",
       "      <th>(CC, JJS)</th>\n",
       "      <th>(MD, DT)</th>\n",
       "      <th>(VBN, MD)</th>\n",
       "      <th>(PRP, RBR)</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PoN</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>6</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>4</td>\n",
       "      <td>8</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 553 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     (NN, VBD)  (VBD, TO)  (TO, VB)  (VB, IN)  (IN, PRP$)  (PRP$, NN)  \\\n",
       "PoN                                                                     \n",
       "N            4          1         1         1           1           3   \n",
       "N            3          0         0         0           2           2   \n",
       "N            6          0         0         0           0           1   \n",
       "N            3          1         2         0           2           2   \n",
       "N           12          0         8         4           3           5   \n",
       "\n",
       "     (NN, IN)  (IN, DT)  (DT, JJ)  (JJ, NN)  ...  (RBS, PRP)  (PRP, JJR)  \\\n",
       "PoN                                          ...                           \n",
       "N           3         2         1         1  ...           0           0   \n",
       "N           6         4         4         7  ...           0           0   \n",
       "N           0         2         1         1  ...           0           0   \n",
       "N           5         4         3         6  ...           0           0   \n",
       "N           6        10         4         8  ...           0           0   \n",
       "\n",
       "     (JJ, RBS)  (RBS, RB)  (PRP$, TO)  (WRB, VBD)  (CC, JJS)  (MD, DT)  \\\n",
       "PoN                                                                      \n",
       "N            0          0           0           0          0         0   \n",
       "N            0          0           0           0          0         0   \n",
       "N            0          0           0           0          0         0   \n",
       "N            0          0           0           0          0         0   \n",
       "N            0          0           0           0          0         0   \n",
       "\n",
       "     (VBN, MD)  (PRP, RBR)  \n",
       "PoN                         \n",
       "N            0           0  \n",
       "N            0           0  \n",
       "N            0           0  \n",
       "N            0           0  \n",
       "N            0           0  \n",
       "\n",
       "[5 rows x 553 columns]"
      ]
     },
     "execution_count": 212,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_df['bow_pos'] = all_df.apply(lambda x: Counter(x['pos_sent_bi']), axis=1)\n",
    "\n",
    "new_df = pd.DataFrame(all_df['bow_pos'].tolist(), all_df['PoN'])\n",
    "new_df = new_df.fillna(0).astype(int)\n",
    "new_df[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.5416666666666666\n",
      "Accuracy: 0.5208333333333334\n",
      "Accuracy: 0.4791666666666667\n",
      "Accuracy: 0.5208333333333334\n",
      "Accuracy: 0.5833333333333334\n"
     ]
    }
   ],
   "source": [
    "get_NB(new_df, new_df.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.5625\n",
      "Accuracy: 0.5625\n",
      "Accuracy: 0.46875\n",
      "Accuracy: 0.4895833333333333\n",
      "Accuracy: 0.59375\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>(NN, VBD)</th>\n",
       "      <th>(VBD, TO)</th>\n",
       "      <th>(TO, VB)</th>\n",
       "      <th>(VB, IN)</th>\n",
       "      <th>(IN, PRP$)</th>\n",
       "      <th>(PRP$, NN)</th>\n",
       "      <th>(NN, IN)</th>\n",
       "      <th>(IN, DT)</th>\n",
       "      <th>(DT, JJ)</th>\n",
       "      <th>(JJ, NN)</th>\n",
       "      <th>...</th>\n",
       "      <th>(RBS, PRP)</th>\n",
       "      <th>(PRP, JJR)</th>\n",
       "      <th>(JJ, RBS)</th>\n",
       "      <th>(RBS, RB)</th>\n",
       "      <th>(PRP$, TO)</th>\n",
       "      <th>(WRB, VBD)</th>\n",
       "      <th>(CC, JJS)</th>\n",
       "      <th>(MD, DT)</th>\n",
       "      <th>(VBN, MD)</th>\n",
       "      <th>(PRP, RBR)</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PoN</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.059701</td>\n",
       "      <td>0.014925</td>\n",
       "      <td>0.014925</td>\n",
       "      <td>0.014925</td>\n",
       "      <td>0.014925</td>\n",
       "      <td>0.044776</td>\n",
       "      <td>0.044776</td>\n",
       "      <td>0.029851</td>\n",
       "      <td>0.014925</td>\n",
       "      <td>0.014925</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.023438</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.015625</td>\n",
       "      <td>0.015625</td>\n",
       "      <td>0.046875</td>\n",
       "      <td>0.031250</td>\n",
       "      <td>0.031250</td>\n",
       "      <td>0.054688</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.088235</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.014706</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.029412</td>\n",
       "      <td>0.014706</td>\n",
       "      <td>0.014706</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.027523</td>\n",
       "      <td>0.009174</td>\n",
       "      <td>0.018349</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.018349</td>\n",
       "      <td>0.018349</td>\n",
       "      <td>0.045872</td>\n",
       "      <td>0.036697</td>\n",
       "      <td>0.027523</td>\n",
       "      <td>0.055046</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.046875</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.031250</td>\n",
       "      <td>0.015625</td>\n",
       "      <td>0.011719</td>\n",
       "      <td>0.019531</td>\n",
       "      <td>0.023438</td>\n",
       "      <td>0.039062</td>\n",
       "      <td>0.015625</td>\n",
       "      <td>0.031250</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>P</td>\n",
       "      <td>0.065574</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.032787</td>\n",
       "      <td>0.065574</td>\n",
       "      <td>0.016393</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>P</td>\n",
       "      <td>0.013793</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.006897</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.013793</td>\n",
       "      <td>0.006897</td>\n",
       "      <td>0.075862</td>\n",
       "      <td>0.062069</td>\n",
       "      <td>0.041379</td>\n",
       "      <td>0.048276</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>P</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.007634</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.007634</td>\n",
       "      <td>0.007634</td>\n",
       "      <td>0.061069</td>\n",
       "      <td>0.061069</td>\n",
       "      <td>0.022901</td>\n",
       "      <td>0.053435</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.007634</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>P</td>\n",
       "      <td>0.032258</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.012903</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.006452</td>\n",
       "      <td>0.006452</td>\n",
       "      <td>0.064516</td>\n",
       "      <td>0.051613</td>\n",
       "      <td>0.032258</td>\n",
       "      <td>0.025806</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>P</td>\n",
       "      <td>0.029412</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.029412</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.029412</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.029412</td>\n",
       "      <td>0.058824</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>320 rows × 553 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     (NN, VBD)  (VBD, TO)  (TO, VB)  (VB, IN)  (IN, PRP$)  (PRP$, NN)  \\\n",
       "PoN                                                                     \n",
       "N     0.059701   0.014925  0.014925  0.014925    0.014925    0.044776   \n",
       "N     0.023438   0.000000  0.000000  0.000000    0.015625    0.015625   \n",
       "N     0.088235   0.000000  0.000000  0.000000    0.000000    0.014706   \n",
       "N     0.027523   0.009174  0.018349  0.000000    0.018349    0.018349   \n",
       "N     0.046875   0.000000  0.031250  0.015625    0.011719    0.019531   \n",
       "..         ...        ...       ...       ...         ...         ...   \n",
       "P     0.065574   0.000000  0.000000  0.000000    0.000000    0.000000   \n",
       "P     0.013793   0.000000  0.006897  0.000000    0.013793    0.006897   \n",
       "P     0.000000   0.000000  0.007634  0.000000    0.007634    0.007634   \n",
       "P     0.032258   0.000000  0.012903  0.000000    0.006452    0.006452   \n",
       "P     0.029412   0.000000  0.029412  0.000000    0.000000    0.000000   \n",
       "\n",
       "     (NN, IN)  (IN, DT)  (DT, JJ)  (JJ, NN)  ...  (RBS, PRP)  (PRP, JJR)  \\\n",
       "PoN                                          ...                           \n",
       "N    0.044776  0.029851  0.014925  0.014925  ...         0.0         0.0   \n",
       "N    0.046875  0.031250  0.031250  0.054688  ...         0.0         0.0   \n",
       "N    0.000000  0.029412  0.014706  0.014706  ...         0.0         0.0   \n",
       "N    0.045872  0.036697  0.027523  0.055046  ...         0.0         0.0   \n",
       "N    0.023438  0.039062  0.015625  0.031250  ...         0.0         0.0   \n",
       "..        ...       ...       ...       ...  ...         ...         ...   \n",
       "P    0.032787  0.065574  0.016393  0.000000  ...         0.0         0.0   \n",
       "P    0.075862  0.062069  0.041379  0.048276  ...         0.0         0.0   \n",
       "P    0.061069  0.061069  0.022901  0.053435  ...         0.0         0.0   \n",
       "P    0.064516  0.051613  0.032258  0.025806  ...         0.0         0.0   \n",
       "P    0.029412  0.000000  0.029412  0.058824  ...         0.0         0.0   \n",
       "\n",
       "     (JJ, RBS)  (RBS, RB)  (PRP$, TO)  (WRB, VBD)  (CC, JJS)  (MD, DT)  \\\n",
       "PoN                                                                      \n",
       "N          0.0        0.0         0.0         0.0        0.0       0.0   \n",
       "N          0.0        0.0         0.0         0.0        0.0       0.0   \n",
       "N          0.0        0.0         0.0         0.0        0.0       0.0   \n",
       "N          0.0        0.0         0.0         0.0        0.0       0.0   \n",
       "N          0.0        0.0         0.0         0.0        0.0       0.0   \n",
       "..         ...        ...         ...         ...        ...       ...   \n",
       "P          0.0        0.0         0.0         0.0        0.0       0.0   \n",
       "P          0.0        0.0         0.0         0.0        0.0       0.0   \n",
       "P          0.0        0.0         0.0         0.0        0.0       0.0   \n",
       "P          0.0        0.0         0.0         0.0        0.0       0.0   \n",
       "P          0.0        0.0         0.0         0.0        0.0       0.0   \n",
       "\n",
       "     (VBN, MD)  (PRP, RBR)  \n",
       "PoN                         \n",
       "N          0.0    0.000000  \n",
       "N          0.0    0.000000  \n",
       "N          0.0    0.000000  \n",
       "N          0.0    0.000000  \n",
       "N          0.0    0.000000  \n",
       "..         ...         ...  \n",
       "P          0.0    0.000000  \n",
       "P          0.0    0.000000  \n",
       "P          0.0    0.007634  \n",
       "P          0.0    0.000000  \n",
       "P          0.0    0.000000  \n",
       "\n",
       "[320 rows x 553 columns]"
      ]
     },
     "execution_count": 214,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bi_df_norm = new_df.copy()\n",
    "bi_df_norm = bi_df_norm.apply(lambda x: x/x.sum(), axis=1)\n",
    "# bi_df_norm = bi_df_norm.apply(lambda x: x/x.sum(), axis=1)\n",
    "bi_df_norm\n",
    "\n",
    "get_NB(bi_df_norm, bi_df_norm.index)\n",
    "bi_df_norm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>(DT, NN)</td>\n",
       "      <td>1719</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>14</td>\n",
       "      <td>(IN, DT)</td>\n",
       "      <td>1051</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>(NN, IN)</td>\n",
       "      <td>916</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>(NN, NN)</td>\n",
       "      <td>897</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>(JJ, NN)</td>\n",
       "      <td>859</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>(NN, VBD)</td>\n",
       "      <td>676</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>(DT, JJ)</td>\n",
       "      <td>540</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>30</td>\n",
       "      <td>(PRP, VBD)</td>\n",
       "      <td>458</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>21</td>\n",
       "      <td>(NN, CC)</td>\n",
       "      <td>451</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>(IN, NN)</td>\n",
       "      <td>398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>25</td>\n",
       "      <td>(TO, VB)</td>\n",
       "      <td>374</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>(VBD, DT)</td>\n",
       "      <td>332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>32</td>\n",
       "      <td>(RB, JJ)</td>\n",
       "      <td>308</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>147</td>\n",
       "      <td>(VBD, RB)</td>\n",
       "      <td>281</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>(PRP$, NN)</td>\n",
       "      <td>251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17</td>\n",
       "      <td>(VBD, JJ)</td>\n",
       "      <td>250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>13</td>\n",
       "      <td>(NNS, IN)</td>\n",
       "      <td>233</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>158</td>\n",
       "      <td>(NN, RB)</td>\n",
       "      <td>229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>36</td>\n",
       "      <td>(IN, PRP)</td>\n",
       "      <td>229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>38</td>\n",
       "      <td>(NN, DT)</td>\n",
       "      <td>214</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           word  count\n",
       "1      (DT, NN)   1719\n",
       "14     (IN, DT)   1051\n",
       "3      (NN, IN)    916\n",
       "2      (NN, NN)    897\n",
       "8      (JJ, NN)    859\n",
       "6     (NN, VBD)    676\n",
       "7      (DT, JJ)    540\n",
       "30   (PRP, VBD)    458\n",
       "21     (NN, CC)    451\n",
       "9      (IN, NN)    398\n",
       "25     (TO, VB)    374\n",
       "0     (VBD, DT)    332\n",
       "32     (RB, JJ)    308\n",
       "147   (VBD, RB)    281\n",
       "5    (PRP$, NN)    251\n",
       "17    (VBD, JJ)    250\n",
       "13    (NNS, IN)    233\n",
       "158    (NN, RB)    229\n",
       "36    (IN, PRP)    229\n",
       "38     (NN, DT)    214"
      ]
     },
     "execution_count": 215,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_p.sort_values(by=[\"count\"], ascending=False)[:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>word</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>12</td>\n",
       "      <td>(DT, NN)</td>\n",
       "      <td>1519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>(IN, DT)</td>\n",
       "      <td>1045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>(NN, IN)</td>\n",
       "      <td>877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>13</td>\n",
       "      <td>(NN, NN)</td>\n",
       "      <td>836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>(JJ, NN)</td>\n",
       "      <td>782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>(NN, VBD)</td>\n",
       "      <td>743</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>(DT, JJ)</td>\n",
       "      <td>529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>27</td>\n",
       "      <td>(PRP, VBD)</td>\n",
       "      <td>438</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>(TO, VB)</td>\n",
       "      <td>435</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>49</td>\n",
       "      <td>(NN, CC)</td>\n",
       "      <td>423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>61</td>\n",
       "      <td>(IN, NN)</td>\n",
       "      <td>393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>(PRP$, NN)</td>\n",
       "      <td>377</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>28</td>\n",
       "      <td>(VBD, RB)</td>\n",
       "      <td>337</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>57</td>\n",
       "      <td>(VBD, DT)</td>\n",
       "      <td>292</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>90</td>\n",
       "      <td>(RB, JJ)</td>\n",
       "      <td>284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10</td>\n",
       "      <td>(NN, RB)</td>\n",
       "      <td>241</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>41</td>\n",
       "      <td>(DT, NNS)</td>\n",
       "      <td>240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>(IN, PRP$)</td>\n",
       "      <td>226</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>122</td>\n",
       "      <td>(IN, PRP)</td>\n",
       "      <td>213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>43</td>\n",
       "      <td>(VBD, IN)</td>\n",
       "      <td>209</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           word  count\n",
       "12     (DT, NN)   1519\n",
       "7      (IN, DT)   1045\n",
       "6      (NN, IN)    877\n",
       "13     (NN, NN)    836\n",
       "9      (JJ, NN)    782\n",
       "0     (NN, VBD)    743\n",
       "8      (DT, JJ)    529\n",
       "27   (PRP, VBD)    438\n",
       "2      (TO, VB)    435\n",
       "49     (NN, CC)    423\n",
       "61     (IN, NN)    393\n",
       "5    (PRP$, NN)    377\n",
       "28    (VBD, RB)    337\n",
       "57    (VBD, DT)    292\n",
       "90     (RB, JJ)    284\n",
       "10     (NN, RB)    241\n",
       "41    (DT, NNS)    240\n",
       "4    (IN, PRP$)    226\n",
       "122   (IN, PRP)    213\n",
       "43    (VBD, IN)    209"
      ]
     },
     "execution_count": 216,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_n.sort_values(by=[\"count\"], ascending=False)[:20]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nltk import word_tokenize \n",
    "from nltk.util import ngrams\n",
    "\n",
    "text = ['cant railway station', 'citadel hotel', 'police stn']\n",
    "def get_ngram(line, num):\n",
    "    token = nltk.word_tokenize(line)\n",
    "    grams = list(ngrams(token, num)) \n",
    "    return(grams)\n",
    "\n",
    "# all_df['trigrams'] = all_df.apply(lambda x: get_ngram(x[0],3), axis=1)\n",
    "all_df['trigrams'] = all_df.apply(lambda x: get_ngram(' '.join(x['tokens']),3), axis=1)\n",
    "all_df['trigrams_pos'] = all_df.apply(lambda x: get_ngram(' '.join(x['pos_sent']),3), axis=1)\n",
    "\n",
    "# ' '.join(all_df['tokens'][1])\n",
    "                                  \n",
    "# counter = all_df['trigrams_pos']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_df['trigrams_feats'] = all_df.apply(lambda x: ['_'.join(trigram) for trigram in x['trigrams_pos']], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['IN_PRP_$'], dtype='<U9')"
      ]
     },
     "execution_count": 219,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def flatten_column(df, column):\n",
    "    return [features for row in df[column].tolist() for features in row]\n",
    "#     [bigram for sublist in all_df_n.pos_sent_bi.tolist() for bigram in sublist]\n",
    "\n",
    "flat_trigrams = Counter(flatten_column(all_df, 'trigrams_feats'))\n",
    "flat_trigrams_n = Counter(flatten_column(all_df[all_df['PoN'] == 'N'], 'trigrams_feats'))\n",
    "flat_trigrams_p = Counter(flatten_column(all_df[all_df['PoN'] == 'P'], 'trigrams_feats'))\n",
    "\n",
    "most_common_n = [word[0] for word in flat_trigrams_n.most_common(10)]\n",
    "most_common_p = [word[0] for word in flat_trigrams_p.most_common(10)]\n",
    "\n",
    "neg_notpos = np.setdiff1d(most_common_n, most_common_p)\n",
    "neg_notpos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_df['trigrams_feats_bow'] = all_df.apply(lambda x: Counter(x['trigrams_feats']), axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>NN_VBD_TO</th>\n",
       "      <th>VBD_TO_VB</th>\n",
       "      <th>TO_VB_IN</th>\n",
       "      <th>VB_IN_PRP</th>\n",
       "      <th>IN_PRP_$</th>\n",
       "      <th>PRP_$_NN</th>\n",
       "      <th>$_NN_IN</th>\n",
       "      <th>NN_IN_DT</th>\n",
       "      <th>IN_DT_JJ</th>\n",
       "      <th>DT_JJ_NN</th>\n",
       "      <th>...</th>\n",
       "      <th>VBP_VBN_PDT</th>\n",
       "      <th>WP_VBD_NNS</th>\n",
       "      <th>VBZ_VBD_NN</th>\n",
       "      <th>DT_WRB_PRP</th>\n",
       "      <th>VBD_PRP_RBR</th>\n",
       "      <th>PRP_RBR_IN</th>\n",
       "      <th>JJ_CC_CD</th>\n",
       "      <th>VBN_VBN_RP</th>\n",
       "      <th>NN_PRP_VBN</th>\n",
       "      <th>PRP_VBN_RB</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PoN</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 3696 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     NN_VBD_TO  VBD_TO_VB  TO_VB_IN  VB_IN_PRP  IN_PRP_$  PRP_$_NN  $_NN_IN  \\\n",
       "PoN                                                                           \n",
       "N            1          1         1          1         1         3        2   \n",
       "N            0          0         0          0         2         2        0   \n",
       "N            0          0         0          0         0         1        0   \n",
       "N            0          0         0          0         2         2        1   \n",
       "N            0          0         3          0         3         5        0   \n",
       "\n",
       "     NN_IN_DT  IN_DT_JJ  DT_JJ_NN  ...  VBP_VBN_PDT  WP_VBD_NNS  VBZ_VBD_NN  \\\n",
       "PoN                                ...                                        \n",
       "N           2         1         1  ...            0           0           0   \n",
       "N           2         1         2  ...            0           0           0   \n",
       "N           0         1         1  ...            0           0           0   \n",
       "N           1         2         3  ...            0           0           0   \n",
       "N           2         2         4  ...            0           0           0   \n",
       "\n",
       "     DT_WRB_PRP  VBD_PRP_RBR  PRP_RBR_IN  JJ_CC_CD  VBN_VBN_RP  NN_PRP_VBN  \\\n",
       "PoN                                                                          \n",
       "N             0            0           0         0           0           0   \n",
       "N             0            0           0         0           0           0   \n",
       "N             0            0           0         0           0           0   \n",
       "N             0            0           0         0           0           0   \n",
       "N             0            0           0         0           0           0   \n",
       "\n",
       "     PRP_VBN_RB  \n",
       "PoN              \n",
       "N             0  \n",
       "N             0  \n",
       "N             0  \n",
       "N             0  \n",
       "N             0  \n",
       "\n",
       "[5 rows x 3696 columns]"
      ]
     },
     "execution_count": 221,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_df = pd.DataFrame(all_df['trigrams_feats_bow'].tolist(), all_df['PoN'])\n",
    "new_df = new_df.fillna(0).astype(int)\n",
    "new_df[:5]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.5520833333333334\n",
      "Accuracy: 0.6145833333333334\n",
      "Accuracy: 0.5520833333333334\n",
      "Accuracy: 0.5729166666666666\n",
      "Accuracy: 0.5104166666666666\n"
     ]
    }
   ],
   "source": [
    "get_NB(new_df, new_df.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.53125\n",
      "Accuracy: 0.6041666666666666\n",
      "Accuracy: 0.4791666666666667\n",
      "Accuracy: 0.5833333333333334\n",
      "Accuracy: 0.5208333333333334\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>NN_VBD_TO</th>\n",
       "      <th>VBD_TO_VB</th>\n",
       "      <th>TO_VB_IN</th>\n",
       "      <th>VB_IN_PRP</th>\n",
       "      <th>IN_PRP_$</th>\n",
       "      <th>PRP_$_NN</th>\n",
       "      <th>$_NN_IN</th>\n",
       "      <th>NN_IN_DT</th>\n",
       "      <th>IN_DT_JJ</th>\n",
       "      <th>DT_JJ_NN</th>\n",
       "      <th>...</th>\n",
       "      <th>VBP_VBN_PDT</th>\n",
       "      <th>WP_VBD_NNS</th>\n",
       "      <th>VBZ_VBD_NN</th>\n",
       "      <th>DT_WRB_PRP</th>\n",
       "      <th>VBD_PRP_RBR</th>\n",
       "      <th>PRP_RBR_IN</th>\n",
       "      <th>JJ_CC_CD</th>\n",
       "      <th>VBN_VBN_RP</th>\n",
       "      <th>NN_PRP_VBN</th>\n",
       "      <th>PRP_VBN_RB</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PoN</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.014493</td>\n",
       "      <td>0.014493</td>\n",
       "      <td>0.014493</td>\n",
       "      <td>0.014493</td>\n",
       "      <td>0.014493</td>\n",
       "      <td>0.043478</td>\n",
       "      <td>0.028986</td>\n",
       "      <td>0.028986</td>\n",
       "      <td>0.014493</td>\n",
       "      <td>0.014493</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.015504</td>\n",
       "      <td>0.015504</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.015504</td>\n",
       "      <td>0.007752</td>\n",
       "      <td>0.015504</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.014706</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.014706</td>\n",
       "      <td>0.014706</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.017857</td>\n",
       "      <td>0.017857</td>\n",
       "      <td>0.008929</td>\n",
       "      <td>0.008929</td>\n",
       "      <td>0.017857</td>\n",
       "      <td>0.026786</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>N</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.011494</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.011494</td>\n",
       "      <td>0.019157</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.007663</td>\n",
       "      <td>0.007663</td>\n",
       "      <td>0.015326</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>P</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.016667</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>P</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.013514</td>\n",
       "      <td>0.006757</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.027027</td>\n",
       "      <td>0.027027</td>\n",
       "      <td>0.027027</td>\n",
       "      <td>...</td>\n",
       "      <td>0.006757</td>\n",
       "      <td>0.006757</td>\n",
       "      <td>0.006757</td>\n",
       "      <td>0.006757</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>P</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.007634</td>\n",
       "      <td>0.007634</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.030534</td>\n",
       "      <td>0.007634</td>\n",
       "      <td>0.015267</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.007634</td>\n",
       "      <td>0.007634</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>P</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.006452</td>\n",
       "      <td>0.006452</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.025806</td>\n",
       "      <td>0.019355</td>\n",
       "      <td>0.019355</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.006452</td>\n",
       "      <td>0.006452</td>\n",
       "      <td>0.006452</td>\n",
       "      <td>0.006452</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>P</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.030303</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>320 rows × 3696 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     NN_VBD_TO  VBD_TO_VB  TO_VB_IN  VB_IN_PRP  IN_PRP_$  PRP_$_NN   $_NN_IN  \\\n",
       "PoN                                                                            \n",
       "N     0.014493   0.014493  0.014493   0.014493  0.014493  0.043478  0.028986   \n",
       "N     0.000000   0.000000  0.000000   0.000000  0.015504  0.015504  0.000000   \n",
       "N     0.000000   0.000000  0.000000   0.000000  0.000000  0.014706  0.000000   \n",
       "N     0.000000   0.000000  0.000000   0.000000  0.017857  0.017857  0.008929   \n",
       "N     0.000000   0.000000  0.011494   0.000000  0.011494  0.019157  0.000000   \n",
       "..         ...        ...       ...        ...       ...       ...       ...   \n",
       "P     0.000000   0.000000  0.000000   0.000000  0.000000  0.000000  0.000000   \n",
       "P     0.000000   0.000000  0.000000   0.000000  0.013514  0.006757  0.000000   \n",
       "P     0.000000   0.000000  0.000000   0.000000  0.007634  0.007634  0.000000   \n",
       "P     0.000000   0.000000  0.000000   0.000000  0.006452  0.006452  0.000000   \n",
       "P     0.000000   0.000000  0.000000   0.000000  0.000000  0.000000  0.000000   \n",
       "\n",
       "     NN_IN_DT  IN_DT_JJ  DT_JJ_NN  ...  VBP_VBN_PDT  WP_VBD_NNS  VBZ_VBD_NN  \\\n",
       "PoN                                ...                                        \n",
       "N    0.028986  0.014493  0.014493  ...     0.000000    0.000000    0.000000   \n",
       "N    0.015504  0.007752  0.015504  ...     0.000000    0.000000    0.000000   \n",
       "N    0.000000  0.014706  0.014706  ...     0.000000    0.000000    0.000000   \n",
       "N    0.008929  0.017857  0.026786  ...     0.000000    0.000000    0.000000   \n",
       "N    0.007663  0.007663  0.015326  ...     0.000000    0.000000    0.000000   \n",
       "..        ...       ...       ...  ...          ...         ...         ...   \n",
       "P    0.016667  0.000000  0.000000  ...     0.000000    0.000000    0.000000   \n",
       "P    0.027027  0.027027  0.027027  ...     0.006757    0.006757    0.006757   \n",
       "P    0.030534  0.007634  0.015267  ...     0.000000    0.000000    0.000000   \n",
       "P    0.025806  0.019355  0.019355  ...     0.000000    0.000000    0.000000   \n",
       "P    0.000000  0.000000  0.030303  ...     0.000000    0.000000    0.000000   \n",
       "\n",
       "     DT_WRB_PRP  VBD_PRP_RBR  PRP_RBR_IN  JJ_CC_CD  VBN_VBN_RP  NN_PRP_VBN  \\\n",
       "PoN                                                                          \n",
       "N      0.000000     0.000000    0.000000  0.000000    0.000000    0.000000   \n",
       "N      0.000000     0.000000    0.000000  0.000000    0.000000    0.000000   \n",
       "N      0.000000     0.000000    0.000000  0.000000    0.000000    0.000000   \n",
       "N      0.000000     0.000000    0.000000  0.000000    0.000000    0.000000   \n",
       "N      0.000000     0.000000    0.000000  0.000000    0.000000    0.000000   \n",
       "..          ...          ...         ...       ...         ...         ...   \n",
       "P      0.000000     0.000000    0.000000  0.000000    0.000000    0.000000   \n",
       "P      0.006757     0.000000    0.000000  0.000000    0.000000    0.000000   \n",
       "P      0.000000     0.007634    0.007634  0.000000    0.000000    0.000000   \n",
       "P      0.000000     0.000000    0.000000  0.006452    0.006452    0.006452   \n",
       "P      0.000000     0.000000    0.000000  0.000000    0.000000    0.000000   \n",
       "\n",
       "     PRP_VBN_RB  \n",
       "PoN              \n",
       "N      0.000000  \n",
       "N      0.000000  \n",
       "N      0.000000  \n",
       "N      0.000000  \n",
       "N      0.000000  \n",
       "..          ...  \n",
       "P      0.000000  \n",
       "P      0.000000  \n",
       "P      0.000000  \n",
       "P      0.006452  \n",
       "P      0.000000  \n",
       "\n",
       "[320 rows x 3696 columns]"
      ]
     },
     "execution_count": 223,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tri_df_norm = new_df.copy()\n",
    "tri_df_norm = tri_df_norm.apply(lambda x: x/x.sum(), axis=1)\n",
    "tri_df_norm\n",
    "\n",
    "get_NB(tri_df_norm, tri_df_norm.index)\n",
    "tri_df_norm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 224,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>PoN</th>\n",
       "      <th>sentences</th>\n",
       "      <th>num_sentences</th>\n",
       "      <th>tokens</th>\n",
       "      <th>num_tokens</th>\n",
       "      <th>no_sw</th>\n",
       "      <th>num_no_sw</th>\n",
       "      <th>stemmed</th>\n",
       "      <th>stemmed_no_sw</th>\n",
       "      <th>...</th>\n",
       "      <th>bow_no_sw</th>\n",
       "      <th>pos_sent</th>\n",
       "      <th>pos_sent_str</th>\n",
       "      <th>pos_no_sw_sent</th>\n",
       "      <th>pos_sent_bi</th>\n",
       "      <th>bow_pos</th>\n",
       "      <th>trigrams</th>\n",
       "      <th>trigrams_pos</th>\n",
       "      <th>trigrams_feats</th>\n",
       "      <th>trigrams_feats_bow</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>I traveled to Chicago with my husband for a ro...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I traveled to Chicago with my husband for a r...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, traveled, to, chicago, with, my, husband, ...</td>\n",
       "      <td>68</td>\n",
       "      <td>[traveled, chicago, husband, romantic, weekend...</td>\n",
       "      <td>40</td>\n",
       "      <td>[i, travel, to, chicago, with, my, husband, fo...</td>\n",
       "      <td>[travel, chicago, husband, romant, weekend, aw...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'traveled': 1, 'chicago': 2, 'husband': 1, 'r...</td>\n",
       "      <td>[NN, VBD, TO, VB, IN, PRP$, NN, IN, DT, JJ, NN...</td>\n",
       "      <td>[NN VBD TO VB IN PRP$ NN IN DT JJ NN RB PRP$ N...</td>\n",
       "      <td>[VBN, JJ, NN, JJ, NN, RB, JJ, JJ, NN, NN, NN, ...</td>\n",
       "      <td>[(NN, VBD), (VBD, TO), (TO, VB), (VB, IN), (IN...</td>\n",
       "      <td>{('NN', 'VBD'): 4, ('VBD', 'TO'): 1, ('TO', 'V...</td>\n",
       "      <td>[(i, traveled, to), (traveled, to, chicago), (...</td>\n",
       "      <td>[(NN, VBD, TO), (VBD, TO, VB), (TO, VB, IN), (...</td>\n",
       "      <td>[NN_VBD_TO, VBD_TO_VB, TO_VB_IN, VB_IN_PRP, IN...</td>\n",
       "      <td>{'NN_VBD_TO': 1, 'VBD_TO_VB': 1, 'TO_VB_IN': 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>I stayed in the Sofitel Chicago Water Tower ho...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I stayed in the Sofitel Chicago Water Tower h...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, stayed, in, the, sofitel, chicago, water, ...</td>\n",
       "      <td>129</td>\n",
       "      <td>[stayed, sofitel, chicago, water, tower, hotel...</td>\n",
       "      <td>71</td>\n",
       "      <td>[i, stay, in, the, sofitel, chicago, water, to...</td>\n",
       "      <td>[stay, sofitel, chicago, water, tower, hotel, ...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'stayed': 1, 'sofitel': 1, 'chicago': 1, 'wat...</td>\n",
       "      <td>[JJ, VBD, IN, DT, NN, NN, NN, NN, NN, IN, PRP$...</td>\n",
       "      <td>[JJ VBD IN DT NN NN NN NN NN IN PRP$ NN CC CD ...</td>\n",
       "      <td>[JJ, NN, NN, NN, NN, NN, NN, CD, NNS, JJ, NN, ...</td>\n",
       "      <td>[(JJ, VBD), (VBD, IN), (IN, DT), (DT, NN), (NN...</td>\n",
       "      <td>{('JJ', 'VBD'): 1, ('VBD', 'IN'): 1, ('IN', 'D...</td>\n",
       "      <td>[(i, stayed, in), (stayed, in, the), (in, the,...</td>\n",
       "      <td>[(JJ, VBD, IN), (VBD, IN, DT), (IN, DT, NN), (...</td>\n",
       "      <td>[JJ_VBD_IN, VBD_IN_DT, IN_DT_NN, DT_NN_NN, NN_...</td>\n",
       "      <td>{'JJ_VBD_IN': 1, 'VBD_IN_DT': 1, 'IN_DT_NN': 2...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>This hotel was gorgeous! I really enjoyed my s...</td>\n",
       "      <td>N</td>\n",
       "      <td>[This hotel was gorgeous!, I really enjoyed my...</td>\n",
       "      <td>7</td>\n",
       "      <td>[this, hotel, was, gorgeous, i, really, enjoye...</td>\n",
       "      <td>69</td>\n",
       "      <td>[hotel, gorgeous, really, enjoyed, stay, defin...</td>\n",
       "      <td>36</td>\n",
       "      <td>[thi, hotel, wa, gorgeou, i, realli, enjoy, my...</td>\n",
       "      <td>[hotel, gorgeou, realli, enjoy, stay, definit,...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'hotel': 2, 'gorgeous': 1, 'really': 1, 'enjo...</td>\n",
       "      <td>[DT, NN, VBD, JJ, JJ, RB, VBN, PRP$, NN, RB, C...</td>\n",
       "      <td>[DT NN VBD JJ JJ RB VBN PRP$ NN RB CC MD RB VB...</td>\n",
       "      <td>[NN, JJ, RB, VBN, NN, RB, VBG, JJ, NN, JJ, NN,...</td>\n",
       "      <td>[(DT, NN), (NN, VBD), (VBD, JJ), (JJ, JJ), (JJ...</td>\n",
       "      <td>{('DT', 'NN'): 8, ('NN', 'VBD'): 6, ('VBD', 'J...</td>\n",
       "      <td>[(this, hotel, was), (hotel, was, gorgeous), (...</td>\n",
       "      <td>[(DT, NN, VBD), (NN, VBD, JJ), (VBD, JJ, JJ), ...</td>\n",
       "      <td>[DT_NN_VBD, NN_VBD_JJ, VBD_JJ_JJ, JJ_JJ_RB, JJ...</td>\n",
       "      <td>{'DT_NN_VBD': 4, 'NN_VBD_JJ': 4, 'VBD_JJ_JJ': ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>This is an absolutely exquisite hotel, at a gr...</td>\n",
       "      <td>N</td>\n",
       "      <td>[This is an absolutely exquisite hotel, at a g...</td>\n",
       "      <td>6</td>\n",
       "      <td>[this, is, an, absolutely, exquisite, hotel, a...</td>\n",
       "      <td>110</td>\n",
       "      <td>[absolutely, exquisite, hotel, great, location...</td>\n",
       "      <td>52</td>\n",
       "      <td>[thi, is, an, absolut, exquisit, hotel, at, a,...</td>\n",
       "      <td>[absolut, exquisit, hotel, great, locat, boast...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'absolutely': 1, 'exquisite': 1, 'hotel': 3, ...</td>\n",
       "      <td>[DT, VBZ, DT, RB, JJ, NN, IN, DT, JJ, NN, CC, ...</td>\n",
       "      <td>[DT VBZ DT RB JJ NN IN DT JJ NN CC NN NN NNS N...</td>\n",
       "      <td>[RB, JJ, NN, JJ, NN, VBG, NN, NNS, JJ, NN, RB,...</td>\n",
       "      <td>[(DT, VBZ), (VBZ, DT), (DT, RB), (RB, JJ), (JJ...</td>\n",
       "      <td>{('DT', 'VBZ'): 1, ('VBZ', 'DT'): 2, ('DT', 'R...</td>\n",
       "      <td>[(this, is, an), (is, an, absolutely), (an, ab...</td>\n",
       "      <td>[(DT, VBZ, DT), (VBZ, DT, RB), (DT, RB, JJ), (...</td>\n",
       "      <td>[DT_VBZ_DT, VBZ_DT_RB, DT_RB_JJ, RB_JJ_NN, JJ_...</td>\n",
       "      <td>{'DT_VBZ_DT': 1, 'VBZ_DT_RB': 1, 'DT_RB_JJ': 1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>I recently traveled up to Chicago for business...</td>\n",
       "      <td>N</td>\n",
       "      <td>[I recently traveled up to Chicago for busines...</td>\n",
       "      <td>13</td>\n",
       "      <td>[i, recently, traveled, up, to, chicago, for, ...</td>\n",
       "      <td>257</td>\n",
       "      <td>[recently, traveled, chicago, business, terrif...</td>\n",
       "      <td>116</td>\n",
       "      <td>[i, recent, travel, up, to, chicago, for, busi...</td>\n",
       "      <td>[recent, travel, chicago, busi, terrif, day, n...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'recently': 1, 'traveled': 1, 'chicago': 4, '...</td>\n",
       "      <td>[NN, RB, VBD, RP, TO, VB, IN, NN, CC, VBD, DT,...</td>\n",
       "      <td>[NN RB VBD RP TO VB IN NN CC VBD DT JJ NN NN N...</td>\n",
       "      <td>[RB, VBN, NN, NN, NN, NN, NN, VB, RB, JJ, NN, ...</td>\n",
       "      <td>[(NN, RB), (RB, VBD), (VBD, RP), (RP, TO), (TO...</td>\n",
       "      <td>{('NN', 'RB'): 3, ('RB', 'VBD'): 3, ('VBD', 'R...</td>\n",
       "      <td>[(i, recently, traveled), (recently, traveled,...</td>\n",
       "      <td>[(NN, RB, VBD), (RB, VBD, RP), (VBD, RP, TO), ...</td>\n",
       "      <td>[NN_RB_VBD, RB_VBD_RP, VBD_RP_TO, RP_TO_VB, TO...</td>\n",
       "      <td>{'NN_RB_VBD': 1, 'RB_VBD_RP': 2, 'VBD_RP_TO': ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>315</td>\n",
       "      <td>This hotel was not worth it. From the moment w...</td>\n",
       "      <td>P</td>\n",
       "      <td>[This hotel was not worth it., From the moment...</td>\n",
       "      <td>6</td>\n",
       "      <td>[this, hotel, was, not, worth, it, from, the, ...</td>\n",
       "      <td>62</td>\n",
       "      <td>[hotel, worth, moment, walked, hotel, lobby, c...</td>\n",
       "      <td>27</td>\n",
       "      <td>[thi, hotel, wa, not, worth, it, from, the, mo...</td>\n",
       "      <td>[hotel, worth, moment, walk, hotel, lobbi, che...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'hotel': 2, 'worth': 1, 'moment': 1, 'walked'...</td>\n",
       "      <td>[DT, NN, VBD, RB, JJ, PRP, IN, DT, NN, PRP, VB...</td>\n",
       "      <td>[DT NN VBD RB JJ PRP IN DT NN PRP VBD IN DT NN...</td>\n",
       "      <td>[NN, JJ, NN, VBD, NN, NN, NN, NN, VBP, JJ, NNS...</td>\n",
       "      <td>[(DT, NN), (NN, VBD), (VBD, RB), (RB, JJ), (JJ...</td>\n",
       "      <td>{('DT', 'NN'): 7, ('NN', 'VBD'): 4, ('VBD', 'R...</td>\n",
       "      <td>[(this, hotel, was), (hotel, was, not), (was, ...</td>\n",
       "      <td>[(DT, NN, VBD), (NN, VBD, RB), (VBD, RB, JJ), ...</td>\n",
       "      <td>[DT_NN_VBD, NN_VBD_RB, VBD_RB_JJ, RB_JJ_PRP, J...</td>\n",
       "      <td>{'DT_NN_VBD': 3, 'NN_VBD_RB': 1, 'VBD_RB_JJ': ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>316</td>\n",
       "      <td>I stayed at the hotel during the Dave Matthews...</td>\n",
       "      <td>P</td>\n",
       "      <td>[I stayed at the hotel during the Dave Matthew...</td>\n",
       "      <td>9</td>\n",
       "      <td>[i, stayed, at, the, hotel, during, the, dave,...</td>\n",
       "      <td>146</td>\n",
       "      <td>[stayed, hotel, dave, matthews, caravan, tour,...</td>\n",
       "      <td>76</td>\n",
       "      <td>[i, stay, at, the, hotel, dure, the, dave, mat...</td>\n",
       "      <td>[stay, hotel, dave, matthew, caravan, tour, wo...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'stayed': 1, 'hotel': 3, 'dave': 1, 'matthews...</td>\n",
       "      <td>[JJ, VBD, IN, DT, NN, IN, DT, NN, NNS, VBP, JJ...</td>\n",
       "      <td>[JJ VBD IN DT NN IN DT NN NNS VBP JJ CC MD VB ...</td>\n",
       "      <td>[JJ, NN, VBP, NNS, VB, NNS, MD, VB, NN, NN, RB...</td>\n",
       "      <td>[(JJ, VBD), (VBD, IN), (IN, DT), (DT, NN), (NN...</td>\n",
       "      <td>{('JJ', 'VBD'): 1, ('VBD', 'IN'): 1, ('IN', 'D...</td>\n",
       "      <td>[(i, stayed, at), (stayed, at, the), (at, the,...</td>\n",
       "      <td>[(JJ, VBD, IN), (VBD, IN, DT), (IN, DT, NN), (...</td>\n",
       "      <td>[JJ_VBD_IN, VBD_IN_DT, IN_DT_NN, DT_NN_IN, NN_...</td>\n",
       "      <td>{'JJ_VBD_IN': 1, 'VBD_IN_DT': 1, 'IN_DT_NN': 3...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>317</td>\n",
       "      <td>We had a reservation for 3 rooms with 5 adults...</td>\n",
       "      <td>P</td>\n",
       "      <td>[We had a reservation for 3 rooms with 5 adult...</td>\n",
       "      <td>9</td>\n",
       "      <td>[we, had, a, reservation, for, rooms, with, ad...</td>\n",
       "      <td>132</td>\n",
       "      <td>[reservation, rooms, adults, kids, got, rooms,...</td>\n",
       "      <td>58</td>\n",
       "      <td>[we, had, a, reserv, for, room, with, adult, a...</td>\n",
       "      <td>[reserv, room, adult, kid, got, room, arriv, c...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'reservation': 1, 'rooms': 3, 'adults': 1, 'k...</td>\n",
       "      <td>[PRP, VBD, DT, NN, IN, NNS, IN, NNS, CC, NNS, ...</td>\n",
       "      <td>[PRP VBD DT NN IN NNS IN NNS CC NNS VBD RB NNS...</td>\n",
       "      <td>[NN, NNS, NNS, NNS, VBD, NNS, JJ, NN, NNS, JJ,...</td>\n",
       "      <td>[(PRP, VBD), (VBD, DT), (DT, NN), (NN, IN), (I...</td>\n",
       "      <td>{('PRP', 'VBD'): 8, ('VBD', 'DT'): 1, ('DT', '...</td>\n",
       "      <td>[(we, had, a), (had, a, reservation), (a, rese...</td>\n",
       "      <td>[(PRP, VBD, DT), (VBD, DT, NN), (DT, NN, IN), ...</td>\n",
       "      <td>[PRP_VBD_DT, VBD_DT_NN, DT_NN_IN, NN_IN_NNS, I...</td>\n",
       "      <td>{'PRP_VBD_DT': 1, 'VBD_DT_NN': 1, 'DT_NN_IN': ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>318</td>\n",
       "      <td>I am staying here now and actually am compelle...</td>\n",
       "      <td>P</td>\n",
       "      <td>[I am staying here now and actually am compell...</td>\n",
       "      <td>6</td>\n",
       "      <td>[i, am, staying, here, now, and, actually, am,...</td>\n",
       "      <td>156</td>\n",
       "      <td>[staying, actually, compelled, write, review, ...</td>\n",
       "      <td>72</td>\n",
       "      <td>[i, am, stay, here, now, and, actual, am, comp...</td>\n",
       "      <td>[stay, actual, compel, write, review, fall, as...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'staying': 1, 'actually': 1, 'compelled': 1, ...</td>\n",
       "      <td>[NN, VBP, VBG, RB, RB, CC, RB, VBP, VBN, TO, V...</td>\n",
       "      <td>[NN VBP VBG RB RB CC RB VBP VBN TO VB DT NN IN...</td>\n",
       "      <td>[VBG, RB, VBN, JJ, NN, NN, JJ, NN, JJ, NN, NN,...</td>\n",
       "      <td>[(NN, VBP), (VBP, VBG), (VBG, RB), (RB, RB), (...</td>\n",
       "      <td>{('NN', 'VBP'): 1, ('VBP', 'VBG'): 1, ('VBG', ...</td>\n",
       "      <td>[(i, am, staying), (am, staying, here), (stayi...</td>\n",
       "      <td>[(NN, VBP, VBG), (VBP, VBG, RB), (VBG, RB, RB)...</td>\n",
       "      <td>[NN_VBP_VBG, VBP_VBG_RB, VBG_RB_RB, RB_RB_CC, ...</td>\n",
       "      <td>{'NN_VBP_VBG': 1, 'VBP_VBG_RB': 1, 'VBG_RB_RB'...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>319</td>\n",
       "      <td>We enjoyed the Hotel Monaco. Great location fo...</td>\n",
       "      <td>P</td>\n",
       "      <td>[We enjoyed the Hotel Monaco., Great location ...</td>\n",
       "      <td>4</td>\n",
       "      <td>[we, enjoyed, the, hotel, monaco, great, locat...</td>\n",
       "      <td>35</td>\n",
       "      <td>[enjoyed, hotel, monaco, great, location, walk...</td>\n",
       "      <td>19</td>\n",
       "      <td>[we, enjoy, the, hotel, monaco, great, locat, ...</td>\n",
       "      <td>[enjoy, hotel, monaco, great, locat, walk, bea...</td>\n",
       "      <td>...</td>\n",
       "      <td>{'enjoyed': 2, 'hotel': 1, 'monaco': 1, 'great...</td>\n",
       "      <td>[PRP, VBD, DT, NN, VBZ, JJ, NN, IN, NN, CC, NN...</td>\n",
       "      <td>[PRP VBD DT NN VBZ JJ NN IN NN CC NN NNS DT NN...</td>\n",
       "      <td>[JJ, NN, NN, JJ, NN, VBG, JJ, NNS, NN, RB, RB,...</td>\n",
       "      <td>[(PRP, VBD), (VBD, DT), (DT, NN), (NN, VBZ), (...</td>\n",
       "      <td>{('PRP', 'VBD'): 1, ('VBD', 'DT'): 2, ('DT', '...</td>\n",
       "      <td>[(we, enjoyed, the), (enjoyed, the, hotel), (t...</td>\n",
       "      <td>[(PRP, VBD, DT), (VBD, DT, NN), (DT, NN, VBZ),...</td>\n",
       "      <td>[PRP_VBD_DT, VBD_DT_NN, DT_NN_VBZ, NN_VBZ_JJ, ...</td>\n",
       "      <td>{'PRP_VBD_DT': 1, 'VBD_DT_NN': 1, 'DT_NN_VBZ':...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>320 rows × 27 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                     0 PoN  \\\n",
       "0    I traveled to Chicago with my husband for a ro...   N   \n",
       "1    I stayed in the Sofitel Chicago Water Tower ho...   N   \n",
       "2    This hotel was gorgeous! I really enjoyed my s...   N   \n",
       "3    This is an absolutely exquisite hotel, at a gr...   N   \n",
       "4    I recently traveled up to Chicago for business...   N   \n",
       "..                                                 ...  ..   \n",
       "315  This hotel was not worth it. From the moment w...   P   \n",
       "316  I stayed at the hotel during the Dave Matthews...   P   \n",
       "317  We had a reservation for 3 rooms with 5 adults...   P   \n",
       "318  I am staying here now and actually am compelle...   P   \n",
       "319  We enjoyed the Hotel Monaco. Great location fo...   P   \n",
       "\n",
       "                                             sentences  num_sentences  \\\n",
       "0    [I traveled to Chicago with my husband for a r...              6   \n",
       "1    [I stayed in the Sofitel Chicago Water Tower h...              6   \n",
       "2    [This hotel was gorgeous!, I really enjoyed my...              7   \n",
       "3    [This is an absolutely exquisite hotel, at a g...              6   \n",
       "4    [I recently traveled up to Chicago for busines...             13   \n",
       "..                                                 ...            ...   \n",
       "315  [This hotel was not worth it., From the moment...              6   \n",
       "316  [I stayed at the hotel during the Dave Matthew...              9   \n",
       "317  [We had a reservation for 3 rooms with 5 adult...              9   \n",
       "318  [I am staying here now and actually am compell...              6   \n",
       "319  [We enjoyed the Hotel Monaco., Great location ...              4   \n",
       "\n",
       "                                                tokens  num_tokens  \\\n",
       "0    [i, traveled, to, chicago, with, my, husband, ...          68   \n",
       "1    [i, stayed, in, the, sofitel, chicago, water, ...         129   \n",
       "2    [this, hotel, was, gorgeous, i, really, enjoye...          69   \n",
       "3    [this, is, an, absolutely, exquisite, hotel, a...         110   \n",
       "4    [i, recently, traveled, up, to, chicago, for, ...         257   \n",
       "..                                                 ...         ...   \n",
       "315  [this, hotel, was, not, worth, it, from, the, ...          62   \n",
       "316  [i, stayed, at, the, hotel, during, the, dave,...         146   \n",
       "317  [we, had, a, reservation, for, rooms, with, ad...         132   \n",
       "318  [i, am, staying, here, now, and, actually, am,...         156   \n",
       "319  [we, enjoyed, the, hotel, monaco, great, locat...          35   \n",
       "\n",
       "                                                 no_sw  num_no_sw  \\\n",
       "0    [traveled, chicago, husband, romantic, weekend...         40   \n",
       "1    [stayed, sofitel, chicago, water, tower, hotel...         71   \n",
       "2    [hotel, gorgeous, really, enjoyed, stay, defin...         36   \n",
       "3    [absolutely, exquisite, hotel, great, location...         52   \n",
       "4    [recently, traveled, chicago, business, terrif...        116   \n",
       "..                                                 ...        ...   \n",
       "315  [hotel, worth, moment, walked, hotel, lobby, c...         27   \n",
       "316  [stayed, hotel, dave, matthews, caravan, tour,...         76   \n",
       "317  [reservation, rooms, adults, kids, got, rooms,...         58   \n",
       "318  [staying, actually, compelled, write, review, ...         72   \n",
       "319  [enjoyed, hotel, monaco, great, location, walk...         19   \n",
       "\n",
       "                                               stemmed  \\\n",
       "0    [i, travel, to, chicago, with, my, husband, fo...   \n",
       "1    [i, stay, in, the, sofitel, chicago, water, to...   \n",
       "2    [thi, hotel, wa, gorgeou, i, realli, enjoy, my...   \n",
       "3    [thi, is, an, absolut, exquisit, hotel, at, a,...   \n",
       "4    [i, recent, travel, up, to, chicago, for, busi...   \n",
       "..                                                 ...   \n",
       "315  [thi, hotel, wa, not, worth, it, from, the, mo...   \n",
       "316  [i, stay, at, the, hotel, dure, the, dave, mat...   \n",
       "317  [we, had, a, reserv, for, room, with, adult, a...   \n",
       "318  [i, am, stay, here, now, and, actual, am, comp...   \n",
       "319  [we, enjoy, the, hotel, monaco, great, locat, ...   \n",
       "\n",
       "                                         stemmed_no_sw  ...  \\\n",
       "0    [travel, chicago, husband, romant, weekend, aw...  ...   \n",
       "1    [stay, sofitel, chicago, water, tower, hotel, ...  ...   \n",
       "2    [hotel, gorgeou, realli, enjoy, stay, definit,...  ...   \n",
       "3    [absolut, exquisit, hotel, great, locat, boast...  ...   \n",
       "4    [recent, travel, chicago, busi, terrif, day, n...  ...   \n",
       "..                                                 ...  ...   \n",
       "315  [hotel, worth, moment, walk, hotel, lobbi, che...  ...   \n",
       "316  [stay, hotel, dave, matthew, caravan, tour, wo...  ...   \n",
       "317  [reserv, room, adult, kid, got, room, arriv, c...  ...   \n",
       "318  [stay, actual, compel, write, review, fall, as...  ...   \n",
       "319  [enjoy, hotel, monaco, great, locat, walk, bea...  ...   \n",
       "\n",
       "                                             bow_no_sw  \\\n",
       "0    {'traveled': 1, 'chicago': 2, 'husband': 1, 'r...   \n",
       "1    {'stayed': 1, 'sofitel': 1, 'chicago': 1, 'wat...   \n",
       "2    {'hotel': 2, 'gorgeous': 1, 'really': 1, 'enjo...   \n",
       "3    {'absolutely': 1, 'exquisite': 1, 'hotel': 3, ...   \n",
       "4    {'recently': 1, 'traveled': 1, 'chicago': 4, '...   \n",
       "..                                                 ...   \n",
       "315  {'hotel': 2, 'worth': 1, 'moment': 1, 'walked'...   \n",
       "316  {'stayed': 1, 'hotel': 3, 'dave': 1, 'matthews...   \n",
       "317  {'reservation': 1, 'rooms': 3, 'adults': 1, 'k...   \n",
       "318  {'staying': 1, 'actually': 1, 'compelled': 1, ...   \n",
       "319  {'enjoyed': 2, 'hotel': 1, 'monaco': 1, 'great...   \n",
       "\n",
       "                                              pos_sent  \\\n",
       "0    [NN, VBD, TO, VB, IN, PRP$, NN, IN, DT, JJ, NN...   \n",
       "1    [JJ, VBD, IN, DT, NN, NN, NN, NN, NN, IN, PRP$...   \n",
       "2    [DT, NN, VBD, JJ, JJ, RB, VBN, PRP$, NN, RB, C...   \n",
       "3    [DT, VBZ, DT, RB, JJ, NN, IN, DT, JJ, NN, CC, ...   \n",
       "4    [NN, RB, VBD, RP, TO, VB, IN, NN, CC, VBD, DT,...   \n",
       "..                                                 ...   \n",
       "315  [DT, NN, VBD, RB, JJ, PRP, IN, DT, NN, PRP, VB...   \n",
       "316  [JJ, VBD, IN, DT, NN, IN, DT, NN, NNS, VBP, JJ...   \n",
       "317  [PRP, VBD, DT, NN, IN, NNS, IN, NNS, CC, NNS, ...   \n",
       "318  [NN, VBP, VBG, RB, RB, CC, RB, VBP, VBN, TO, V...   \n",
       "319  [PRP, VBD, DT, NN, VBZ, JJ, NN, IN, NN, CC, NN...   \n",
       "\n",
       "                                          pos_sent_str  \\\n",
       "0    [NN VBD TO VB IN PRP$ NN IN DT JJ NN RB PRP$ N...   \n",
       "1    [JJ VBD IN DT NN NN NN NN NN IN PRP$ NN CC CD ...   \n",
       "2    [DT NN VBD JJ JJ RB VBN PRP$ NN RB CC MD RB VB...   \n",
       "3    [DT VBZ DT RB JJ NN IN DT JJ NN CC NN NN NNS N...   \n",
       "4    [NN RB VBD RP TO VB IN NN CC VBD DT JJ NN NN N...   \n",
       "..                                                 ...   \n",
       "315  [DT NN VBD RB JJ PRP IN DT NN PRP VBD IN DT NN...   \n",
       "316  [JJ VBD IN DT NN IN DT NN NNS VBP JJ CC MD VB ...   \n",
       "317  [PRP VBD DT NN IN NNS IN NNS CC NNS VBD RB NNS...   \n",
       "318  [NN VBP VBG RB RB CC RB VBP VBN TO VB DT NN IN...   \n",
       "319  [PRP VBD DT NN VBZ JJ NN IN NN CC NN NNS DT NN...   \n",
       "\n",
       "                                        pos_no_sw_sent  \\\n",
       "0    [VBN, JJ, NN, JJ, NN, RB, JJ, JJ, NN, NN, NN, ...   \n",
       "1    [JJ, NN, NN, NN, NN, NN, NN, CD, NNS, JJ, NN, ...   \n",
       "2    [NN, JJ, RB, VBN, NN, RB, VBG, JJ, NN, JJ, NN,...   \n",
       "3    [RB, JJ, NN, JJ, NN, VBG, NN, NNS, JJ, NN, RB,...   \n",
       "4    [RB, VBN, NN, NN, NN, NN, NN, VB, RB, JJ, NN, ...   \n",
       "..                                                 ...   \n",
       "315  [NN, JJ, NN, VBD, NN, NN, NN, NN, VBP, JJ, NNS...   \n",
       "316  [JJ, NN, VBP, NNS, VB, NNS, MD, VB, NN, NN, RB...   \n",
       "317  [NN, NNS, NNS, NNS, VBD, NNS, JJ, NN, NNS, JJ,...   \n",
       "318  [VBG, RB, VBN, JJ, NN, NN, JJ, NN, JJ, NN, NN,...   \n",
       "319  [JJ, NN, NN, JJ, NN, VBG, JJ, NNS, NN, RB, RB,...   \n",
       "\n",
       "                                           pos_sent_bi  \\\n",
       "0    [(NN, VBD), (VBD, TO), (TO, VB), (VB, IN), (IN...   \n",
       "1    [(JJ, VBD), (VBD, IN), (IN, DT), (DT, NN), (NN...   \n",
       "2    [(DT, NN), (NN, VBD), (VBD, JJ), (JJ, JJ), (JJ...   \n",
       "3    [(DT, VBZ), (VBZ, DT), (DT, RB), (RB, JJ), (JJ...   \n",
       "4    [(NN, RB), (RB, VBD), (VBD, RP), (RP, TO), (TO...   \n",
       "..                                                 ...   \n",
       "315  [(DT, NN), (NN, VBD), (VBD, RB), (RB, JJ), (JJ...   \n",
       "316  [(JJ, VBD), (VBD, IN), (IN, DT), (DT, NN), (NN...   \n",
       "317  [(PRP, VBD), (VBD, DT), (DT, NN), (NN, IN), (I...   \n",
       "318  [(NN, VBP), (VBP, VBG), (VBG, RB), (RB, RB), (...   \n",
       "319  [(PRP, VBD), (VBD, DT), (DT, NN), (NN, VBZ), (...   \n",
       "\n",
       "                                               bow_pos  \\\n",
       "0    {('NN', 'VBD'): 4, ('VBD', 'TO'): 1, ('TO', 'V...   \n",
       "1    {('JJ', 'VBD'): 1, ('VBD', 'IN'): 1, ('IN', 'D...   \n",
       "2    {('DT', 'NN'): 8, ('NN', 'VBD'): 6, ('VBD', 'J...   \n",
       "3    {('DT', 'VBZ'): 1, ('VBZ', 'DT'): 2, ('DT', 'R...   \n",
       "4    {('NN', 'RB'): 3, ('RB', 'VBD'): 3, ('VBD', 'R...   \n",
       "..                                                 ...   \n",
       "315  {('DT', 'NN'): 7, ('NN', 'VBD'): 4, ('VBD', 'R...   \n",
       "316  {('JJ', 'VBD'): 1, ('VBD', 'IN'): 1, ('IN', 'D...   \n",
       "317  {('PRP', 'VBD'): 8, ('VBD', 'DT'): 1, ('DT', '...   \n",
       "318  {('NN', 'VBP'): 1, ('VBP', 'VBG'): 1, ('VBG', ...   \n",
       "319  {('PRP', 'VBD'): 1, ('VBD', 'DT'): 2, ('DT', '...   \n",
       "\n",
       "                                              trigrams  \\\n",
       "0    [(i, traveled, to), (traveled, to, chicago), (...   \n",
       "1    [(i, stayed, in), (stayed, in, the), (in, the,...   \n",
       "2    [(this, hotel, was), (hotel, was, gorgeous), (...   \n",
       "3    [(this, is, an), (is, an, absolutely), (an, ab...   \n",
       "4    [(i, recently, traveled), (recently, traveled,...   \n",
       "..                                                 ...   \n",
       "315  [(this, hotel, was), (hotel, was, not), (was, ...   \n",
       "316  [(i, stayed, at), (stayed, at, the), (at, the,...   \n",
       "317  [(we, had, a), (had, a, reservation), (a, rese...   \n",
       "318  [(i, am, staying), (am, staying, here), (stayi...   \n",
       "319  [(we, enjoyed, the), (enjoyed, the, hotel), (t...   \n",
       "\n",
       "                                          trigrams_pos  \\\n",
       "0    [(NN, VBD, TO), (VBD, TO, VB), (TO, VB, IN), (...   \n",
       "1    [(JJ, VBD, IN), (VBD, IN, DT), (IN, DT, NN), (...   \n",
       "2    [(DT, NN, VBD), (NN, VBD, JJ), (VBD, JJ, JJ), ...   \n",
       "3    [(DT, VBZ, DT), (VBZ, DT, RB), (DT, RB, JJ), (...   \n",
       "4    [(NN, RB, VBD), (RB, VBD, RP), (VBD, RP, TO), ...   \n",
       "..                                                 ...   \n",
       "315  [(DT, NN, VBD), (NN, VBD, RB), (VBD, RB, JJ), ...   \n",
       "316  [(JJ, VBD, IN), (VBD, IN, DT), (IN, DT, NN), (...   \n",
       "317  [(PRP, VBD, DT), (VBD, DT, NN), (DT, NN, IN), ...   \n",
       "318  [(NN, VBP, VBG), (VBP, VBG, RB), (VBG, RB, RB)...   \n",
       "319  [(PRP, VBD, DT), (VBD, DT, NN), (DT, NN, VBZ),...   \n",
       "\n",
       "                                        trigrams_feats  \\\n",
       "0    [NN_VBD_TO, VBD_TO_VB, TO_VB_IN, VB_IN_PRP, IN...   \n",
       "1    [JJ_VBD_IN, VBD_IN_DT, IN_DT_NN, DT_NN_NN, NN_...   \n",
       "2    [DT_NN_VBD, NN_VBD_JJ, VBD_JJ_JJ, JJ_JJ_RB, JJ...   \n",
       "3    [DT_VBZ_DT, VBZ_DT_RB, DT_RB_JJ, RB_JJ_NN, JJ_...   \n",
       "4    [NN_RB_VBD, RB_VBD_RP, VBD_RP_TO, RP_TO_VB, TO...   \n",
       "..                                                 ...   \n",
       "315  [DT_NN_VBD, NN_VBD_RB, VBD_RB_JJ, RB_JJ_PRP, J...   \n",
       "316  [JJ_VBD_IN, VBD_IN_DT, IN_DT_NN, DT_NN_IN, NN_...   \n",
       "317  [PRP_VBD_DT, VBD_DT_NN, DT_NN_IN, NN_IN_NNS, I...   \n",
       "318  [NN_VBP_VBG, VBP_VBG_RB, VBG_RB_RB, RB_RB_CC, ...   \n",
       "319  [PRP_VBD_DT, VBD_DT_NN, DT_NN_VBZ, NN_VBZ_JJ, ...   \n",
       "\n",
       "                                    trigrams_feats_bow  \n",
       "0    {'NN_VBD_TO': 1, 'VBD_TO_VB': 1, 'TO_VB_IN': 1...  \n",
       "1    {'JJ_VBD_IN': 1, 'VBD_IN_DT': 1, 'IN_DT_NN': 2...  \n",
       "2    {'DT_NN_VBD': 4, 'NN_VBD_JJ': 4, 'VBD_JJ_JJ': ...  \n",
       "3    {'DT_VBZ_DT': 1, 'VBZ_DT_RB': 1, 'DT_RB_JJ': 1...  \n",
       "4    {'NN_RB_VBD': 1, 'RB_VBD_RP': 2, 'VBD_RP_TO': ...  \n",
       "..                                                 ...  \n",
       "315  {'DT_NN_VBD': 3, 'NN_VBD_RB': 1, 'VBD_RB_JJ': ...  \n",
       "316  {'JJ_VBD_IN': 1, 'VBD_IN_DT': 1, 'IN_DT_NN': 3...  \n",
       "317  {'PRP_VBD_DT': 1, 'VBD_DT_NN': 1, 'DT_NN_IN': ...  \n",
       "318  {'NN_VBP_VBG': 1, 'VBP_VBG_RB': 1, 'VBG_RB_RB'...  \n",
       "319  {'PRP_VBD_DT': 1, 'VBD_DT_NN': 1, 'DT_NN_VBZ':...  \n",
       "\n",
       "[320 rows x 27 columns]"
      ]
     },
     "execution_count": 224,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}