train.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## LDA : Latent Dirichlet Allocation is an example of topic model and is used to classify text in a document to a particular topic. It builds a topic per document model and words per topic model, modeled as Dirichlet distributions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/onion8/miniconda3/lib/python2.7/site-packages/IPython/core/magics/pylab.py:161: UserWarning: pylab import has clobbered these variables: ['stem']\n",
      "`%matplotlib` prevents importing * from pylab and numpy\n",
      "  \"\\n`%matplotlib` prevents importing * from pylab and numpy\"\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Populating the interactive namespace from numpy and matplotlib\n"
     ]
    }
   ],
   "source": [
    "from __future__ import unicode_literals\n",
    "from nltk.corpus import stopwords\n",
    "from nltk import stem,pos_tag\n",
    "import re\n",
    "from collections import Counter\n",
    "import numpy as np\n",
    "from sklearn.cluster import KMeans\n",
    "%pylab inline\n",
    "from textblob import TextBlob\n",
    "from sklearn.metrics import silhouette_score,confusion_matrix,accuracy_score,roc_curve\n",
    "import pandas as pd\n",
    "from nltk.stem.wordnet import WordNetLemmatizer\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.ensemble import RandomForestClassifier,GradientBoostingClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.preprocessing import label_binarize,Normalizer\n",
    "from sklearn.multiclass import OneVsRestClassifier\n",
    "from scipy import interp\n",
    "import ftfy\n",
    "import pyLDAvis.sklearn\n",
    "pyLDAvis.enable_notebook()\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "import statsmodels.api as sm\n",
    "from seaborn import lmplot\n",
    "import os  # for os.path.basename\n",
    "import matplotlib as mpl\n",
    "from sklearn.manifold import MDS,TSNE\n",
    "%pylab inline\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "import sklearn\n",
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.naive_bayes import BernoulliNB, MultinomialNB\n",
    "from sklearn.utils.extmath import density\n",
    "from sklearn import metrics\n",
    "from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer, TfidfTransformer\n",
    "from sklearn.decomposition import TruncatedSVD,PCA,NMF\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.decomposition.online_lda import LatentDirichletAllocation\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df=pd.read_csv('dataset/train.csv',encoding='utf-8')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "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>Company Name</th>\n",
       "      <th>Exchange:Ticker</th>\n",
       "      <th>Company Type</th>\n",
       "      <th>Company Status</th>\n",
       "      <th>Geographic Locations</th>\n",
       "      <th>Business Description</th>\n",
       "      <th>Industry Classifications</th>\n",
       "      <th>Security Tickers</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>024 Pharma, Inc. (OTCPK:EEIG)</td>\n",
       "      <td>OTCPK:EEIG</td>\n",
       "      <td>Public Company</td>\n",
       "      <td>Operating</td>\n",
       "      <td>United States of America (Primary)</td>\n",
       "      <td>024 Pharma, Inc. provides healthcare products ...</td>\n",
       "      <td>Beauty Care Products (Primary); Consumer Stapl...</td>\n",
       "      <td>OTCPK:EEIG</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1-800-FLOWERS.COM, Inc. (NasdaqGS:FLWS)</td>\n",
       "      <td>NasdaqGS:FLWS</td>\n",
       "      <td>Public Company</td>\n",
       "      <td>Operating</td>\n",
       "      <td>United States of America (Primary)</td>\n",
       "      <td>1-800-Flowers.com, Inc., together with its sub...</td>\n",
       "      <td>Catalog Flowers, Gifts and Novelties (Primary)...</td>\n",
       "      <td>NasdaqGS:FLWS; DB:FWC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12 Retech Corporation (OTCPK:RETC)</td>\n",
       "      <td>OTCPK:RETC</td>\n",
       "      <td>Public Company</td>\n",
       "      <td>Operating</td>\n",
       "      <td>United States of America (Primary)</td>\n",
       "      <td>12 Retech Corporation, through its subsidiarie...</td>\n",
       "      <td>Information Technology (Primary); Internet Sof...</td>\n",
       "      <td>OTCPK:RETC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1347 Property Insurance Holdings, Inc. (Nasdaq...</td>\n",
       "      <td>NasdaqGM:PIH</td>\n",
       "      <td>Public Company</td>\n",
       "      <td>Operating</td>\n",
       "      <td>United States of America (Primary)</td>\n",
       "      <td>1347 Property Insurance Holdings, Inc., throug...</td>\n",
       "      <td>Casualty (Primary); Financials (Primary); Fire...</td>\n",
       "      <td>NasdaqGM:PIH</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1847 Holdings LLC (OTCPK:EFSH)</td>\n",
       "      <td>OTCPK:EFSH</td>\n",
       "      <td>Public Company</td>\n",
       "      <td>Operating</td>\n",
       "      <td>United States of America (Primary)</td>\n",
       "      <td>1847 Holdings LLC, through its subsidiaries, p...</td>\n",
       "      <td>Commercial and Professional Services (Primary)...</td>\n",
       "      <td>OTCPK:EFSH</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1867 Western Financial Corporation (OTCPK:WFCL)</td>\n",
       "      <td>OTCPK:WFCL</td>\n",
       "      <td>Public Company</td>\n",
       "      <td>Operating</td>\n",
       "      <td>United States of America (Primary)</td>\n",
       "      <td>1867 Western Financial Corporation, through it...</td>\n",
       "      <td>Banks (Primary); Banks (Primary); Diversified ...</td>\n",
       "      <td>OTCPK:WFCL</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1mage Software, Inc. (OTCPK:ISOL)</td>\n",
       "      <td>OTCPK:ISOL</td>\n",
       "      <td>Public Company</td>\n",
       "      <td>Operating</td>\n",
       "      <td>United States of America (Primary)</td>\n",
       "      <td>1mage Software Inc operates in the technology ...</td>\n",
       "      <td>Application Software (Primary); Information Te...</td>\n",
       "      <td>OTCPK:ISOL</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1PM Industries, Inc. (OTCPK:OPMZ)</td>\n",
       "      <td>OTCPK:OPMZ</td>\n",
       "      <td>Public Company</td>\n",
       "      <td>Operating</td>\n",
       "      <td>United States of America (Primary)</td>\n",
       "      <td>1PM Industries, Inc. provides consulting servi...</td>\n",
       "      <td>Commercial and Professional Services (Primary)...</td>\n",
       "      <td>OTCPK:OPMZ</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1st Capital Bank (OTCPK:FISB)</td>\n",
       "      <td>OTCPK:FISB</td>\n",
       "      <td>Public Company</td>\n",
       "      <td>Operating</td>\n",
       "      <td>United States of America (Primary)</td>\n",
       "      <td>1st Capital Bank provides various banking prod...</td>\n",
       "      <td>Banks (Primary); Banks (Primary); Financials (...</td>\n",
       "      <td>OTCPK:FISB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1st Colonial Bancorp, Inc. (OTCPK:FCOB)</td>\n",
       "      <td>OTCPK:FCOB</td>\n",
       "      <td>Public Company</td>\n",
       "      <td>Operating</td>\n",
       "      <td>United States of America (Primary)</td>\n",
       "      <td>1st Colonial Bancorp, Inc. operates as the ban...</td>\n",
       "      <td>Banks (Primary); Banks (Primary); Financials (...</td>\n",
       "      <td>OTCPK:FCOB</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        Company Name Exchange:Ticker  \\\n",
       "0                      024 Pharma, Inc. (OTCPK:EEIG)      OTCPK:EEIG   \n",
       "1            1-800-FLOWERS.COM, Inc. (NasdaqGS:FLWS)   NasdaqGS:FLWS   \n",
       "2                 12 Retech Corporation (OTCPK:RETC)      OTCPK:RETC   \n",
       "3  1347 Property Insurance Holdings, Inc. (Nasdaq...    NasdaqGM:PIH   \n",
       "4                     1847 Holdings LLC (OTCPK:EFSH)      OTCPK:EFSH   \n",
       "5    1867 Western Financial Corporation (OTCPK:WFCL)      OTCPK:WFCL   \n",
       "6                  1mage Software, Inc. (OTCPK:ISOL)      OTCPK:ISOL   \n",
       "7                  1PM Industries, Inc. (OTCPK:OPMZ)      OTCPK:OPMZ   \n",
       "8                      1st Capital Bank (OTCPK:FISB)      OTCPK:FISB   \n",
       "9            1st Colonial Bancorp, Inc. (OTCPK:FCOB)      OTCPK:FCOB   \n",
       "\n",
       "     Company Type Company Status                Geographic Locations  \\\n",
       "0  Public Company      Operating  United States of America (Primary)   \n",
       "1  Public Company      Operating  United States of America (Primary)   \n",
       "2  Public Company      Operating  United States of America (Primary)   \n",
       "3  Public Company      Operating  United States of America (Primary)   \n",
       "4  Public Company      Operating  United States of America (Primary)   \n",
       "5  Public Company      Operating  United States of America (Primary)   \n",
       "6  Public Company      Operating  United States of America (Primary)   \n",
       "7  Public Company      Operating  United States of America (Primary)   \n",
       "8  Public Company      Operating  United States of America (Primary)   \n",
       "9  Public Company      Operating  United States of America (Primary)   \n",
       "\n",
       "                                Business Description  \\\n",
       "0  024 Pharma, Inc. provides healthcare products ...   \n",
       "1  1-800-Flowers.com, Inc., together with its sub...   \n",
       "2  12 Retech Corporation, through its subsidiarie...   \n",
       "3  1347 Property Insurance Holdings, Inc., throug...   \n",
       "4  1847 Holdings LLC, through its subsidiaries, p...   \n",
       "5  1867 Western Financial Corporation, through it...   \n",
       "6  1mage Software Inc operates in the technology ...   \n",
       "7  1PM Industries, Inc. provides consulting servi...   \n",
       "8  1st Capital Bank provides various banking prod...   \n",
       "9  1st Colonial Bancorp, Inc. operates as the ban...   \n",
       "\n",
       "                            Industry Classifications       Security Tickers  \n",
       "0  Beauty Care Products (Primary); Consumer Stapl...             OTCPK:EEIG  \n",
       "1  Catalog Flowers, Gifts and Novelties (Primary)...  NasdaqGS:FLWS; DB:FWC  \n",
       "2  Information Technology (Primary); Internet Sof...             OTCPK:RETC  \n",
       "3  Casualty (Primary); Financials (Primary); Fire...           NasdaqGM:PIH  \n",
       "4  Commercial and Professional Services (Primary)...             OTCPK:EFSH  \n",
       "5  Banks (Primary); Banks (Primary); Diversified ...             OTCPK:WFCL  \n",
       "6  Application Software (Primary); Information Te...             OTCPK:ISOL  \n",
       "7  Commercial and Professional Services (Primary)...             OTCPK:OPMZ  \n",
       "8  Banks (Primary); Banks (Primary); Financials (...             OTCPK:FISB  \n",
       "9  Banks (Primary); Banks (Primary); Financials (...             OTCPK:FCOB  "
      ]
     },
     "execution_count": 160,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Only considering Business Description and Industry Classifications as there is no variation in other columns of DF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = df[['Business Description','Industry Classifications']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/onion8/miniconda3/lib/python2.7/site-packages/ipykernel_launcher.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  \"\"\"Entry point for launching an IPython kernel.\n"
     ]
    }
   ],
   "source": [
    "data['label']=data['Industry Classifications'].str.split('(',expand=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "rawdata=data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Grouping data by the labels to visualize the frequency of each label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "vis_data=rawdata.groupby('label').size()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "vis_data=pd.DataFrame(vis_data,columns=['Frequency'])\n",
    "vis_data['label']=vis_data.index\n",
    "vis_data['idx']=(arange(0,len(vis_data)))\n",
    "vis_data=vis_data.set_index('idx')\n",
    "vis_data=vis_data.iloc[1:]\n",
    "vis_data=vis_data.sort_values(by=['Frequency'], ascending=False)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>Frequency</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>idx</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>241</td>\n",
       "      <td>Banks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>185</th>\n",
       "      <td>153</td>\n",
       "      <td>Healthcare</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>145</td>\n",
       "      <td>Biotechnology</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>133</td>\n",
       "      <td>Energy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>117</td>\n",
       "      <td>Consumer Discretionary</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Frequency                    label\n",
       "idx                                    \n",
       "59         241                   Banks \n",
       "185        153              Healthcare \n",
       "69         145           Biotechnology \n",
       "175        133                  Energy \n",
       "138        117  Consumer Discretionary "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vis_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualizing top 5 labels frequency"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.text.Text at 0x1c1fcea5d0>,\n",
       " <matplotlib.text.Text at 0x1c1fcea350>,\n",
       " <matplotlib.text.Text at 0x1c1fb01150>,\n",
       " <matplotlib.text.Text at 0x1c1fb01250>,\n",
       " <matplotlib.text.Text at 0x1c1fb018d0>]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c1fdb6a50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set(style=\"whitegrid\")\n",
    "ax = sns.barplot(x=\"label\", y=\"Frequency\", data=vis_data.head(5))\n",
    "ax.set_xticklabels(ax.get_xticklabels(),rotation=40)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>Frequency</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>idx</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>1</td>\n",
       "      <td>Apparel and Textile Distribution</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>142</th>\n",
       "      <td>1</td>\n",
       "      <td>Cosmetic Dental Implants and Supplies</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>1</td>\n",
       "      <td>Casinos and Gaming</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>1</td>\n",
       "      <td>Casualty Reinsurance Carriers</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>1</td>\n",
       "      <td>Catering Services</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Frequency                                   label\n",
       "idx                                                   \n",
       "38           1       Apparel and Textile Distribution \n",
       "142          1  Cosmetic Dental Implants and Supplies \n",
       "96           1                     Casinos and Gaming \n",
       "98           1          Casualty Reinsurance Carriers \n",
       "102          1                      Catering Services "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vis_data.tail()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualizing last 5 labels frequency"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.text.Text at 0x1c1d448190>,\n",
       " <matplotlib.text.Text at 0x1c1c2602d0>,\n",
       " <matplotlib.text.Text at 0x1c1d563410>,\n",
       " <matplotlib.text.Text at 0x1c1d563950>,\n",
       " <matplotlib.text.Text at 0x1c1d563e90>]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x106e7f0d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "sns.set(style=\"whitegrid\")\n",
    "ax = sns.barplot(x=\"label\", y=\"Frequency\", data=vis_data.tail(5))\n",
    "ax.set_xticklabels(ax.get_xticklabels(),rotation=40)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data is highly imbalanced as companies of some labels are very less"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": true
   },
   "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>Frequency</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>idx</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>241</td>\n",
       "      <td>Banks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>185</th>\n",
       "      <td>153</td>\n",
       "      <td>Healthcare</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>145</td>\n",
       "      <td>Biotechnology</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>133</td>\n",
       "      <td>Energy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>117</td>\n",
       "      <td>Consumer Discretionary</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>190</th>\n",
       "      <td>106</td>\n",
       "      <td>Information Technology</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>88</td>\n",
       "      <td>Capital Goods</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>125</th>\n",
       "      <td>85</td>\n",
       "      <td>Commercial and Professional Services</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>50</td>\n",
       "      <td>Application Software</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>128</th>\n",
       "      <td>37</td>\n",
       "      <td>Communications Equipment</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>32</td>\n",
       "      <td>Asset Management and Custody Banks</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>31</td>\n",
       "      <td>Consumer Staples</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>28</td>\n",
       "      <td>Chemicals</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>27</td>\n",
       "      <td>Application Hosting Services</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>23</td>\n",
       "      <td>Aerospace and Defense</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>172</th>\n",
       "      <td>22</td>\n",
       "      <td>Electronic Equipment and Instruments</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>18</td>\n",
       "      <td>Advertising</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>17</td>\n",
       "      <td>Auto Components</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>147</th>\n",
       "      <td>17</td>\n",
       "      <td>Data Processing and Outsourced Services</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>184</th>\n",
       "      <td>17</td>\n",
       "      <td>Health Care Technology</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>182</th>\n",
       "      <td>15</td>\n",
       "      <td>Gold</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>157</th>\n",
       "      <td>15</td>\n",
       "      <td>Diversified Financials</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>178</th>\n",
       "      <td>15</td>\n",
       "      <td>Financials</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>14</td>\n",
       "      <td>Building Products</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>13</td>\n",
       "      <td>Automotive Retail</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>195</th>\n",
       "      <td>13</td>\n",
       "      <td>Materials</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>12</td>\n",
       "      <td>Anti-infective Drugs</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>11</td>\n",
       "      <td>Apparel Retail</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>11</td>\n",
       "      <td>Beverages</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>10</td>\n",
       "      <td>Capital Markets</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>198</th>\n",
       "      <td>10</td>\n",
       "      <td>Nonresidential Building Operators and Lessors</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>202</th>\n",
       "      <td>10</td>\n",
       "      <td>Utilities</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>199</th>\n",
       "      <td>10</td>\n",
       "      <td>Real Estate</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>10</td>\n",
       "      <td>Beauty Care Products</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>10</td>\n",
       "      <td>Alternative Carriers</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>116</th>\n",
       "      <td>9</td>\n",
       "      <td>Coal and Consumable Fuels</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>9</td>\n",
       "      <td>Consumer Finance</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>159</th>\n",
       "      <td>9</td>\n",
       "      <td>Diversified Metals and Mining</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>156</th>\n",
       "      <td>9</td>\n",
       "      <td>Diversified Financial Services</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>173</th>\n",
       "      <td>9</td>\n",
       "      <td>Electronic Equipment, Instruments and Components</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>161</th>\n",
       "      <td>8</td>\n",
       "      <td>Diversified Telecommunication Services</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>8</td>\n",
       "      <td>Agricultural Products</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>8</td>\n",
       "      <td>Cardiovascular Equipment</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>170</th>\n",
       "      <td>8</td>\n",
       "      <td>Electronic Components</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>8</td>\n",
       "      <td>Diversified Metal Ores</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>8</td>\n",
       "      <td>Casinos</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>188</th>\n",
       "      <td>7</td>\n",
       "      <td>Industrials</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>6</td>\n",
       "      <td>Multi-Utilities</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>6</td>\n",
       "      <td>Broadcasting</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>6</td>\n",
       "      <td>Biological Products</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>141</th>\n",
       "      <td>6</td>\n",
       "      <td>Containers and Packaging</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>6</td>\n",
       "      <td>Accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>6</td>\n",
       "      <td>Analog and Mixed Signal</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Frequency                                              label\n",
       "idx                                                              \n",
       "59         241                                             Banks \n",
       "185        153                                        Healthcare \n",
       "69         145                                     Biotechnology \n",
       "175        133                                            Energy \n",
       "138        117                            Consumer Discretionary \n",
       "190        106                            Information Technology \n",
       "89          88                                     Capital Goods \n",
       "125         85              Commercial and Professional Services \n",
       "42          50                              Application Software \n",
       "128         37                          Communications Equipment \n",
       "44          32                Asset Management and Custody Banks \n",
       "140         31                                  Consumer Staples \n",
       "107         28                                         Chemicals \n",
       "41          27                      Application Hosting Services \n",
       "10          23                             Aerospace and Defense \n",
       "172         22              Electronic Equipment and Instruments \n",
       "7           18                                       Advertising \n",
       "48          17                                   Auto Components \n",
       "147         17           Data Processing and Outsourced Services \n",
       "184         17                            Health Care Technology \n",
       "182         15                                              Gold \n",
       "157         15                            Diversified Financials \n",
       "178         15                                        Financials \n",
       "80          14                                 Building Products \n",
       "54          13                                 Automotive Retail \n",
       "195         13                                         Materials \n",
       "34          12                              Anti-infective Drugs \n",
       "37          11                                    Apparel Retail \n",
       "66          11                                         Beverages \n",
       "91          10                                   Capital Markets \n",
       "198         10     Nonresidential Building Operators and Lessors \n",
       "202         10                                         Utilities \n",
       "199         10                                       Real Estate \n",
       "62          10                              Beauty Care Products \n",
       "22          10                              Alternative Carriers \n",
       "116          9                         Coal and Consumable Fuels \n",
       "139          9                                  Consumer Finance \n",
       "159          9                     Diversified Metals and Mining \n",
       "156          9                    Diversified Financial Services \n",
       "173          9  Electronic Equipment, Instruments and Components \n",
       "161          8            Diversified Telecommunication Services \n",
       "11           8                             Agricultural Products \n",
       "94           8                          Cardiovascular Equipment \n",
       "170          8                             Electronic Components \n",
       "158          8                            Diversified Metal Ores \n",
       "95           8                                           Casinos \n",
       "188          7                                       Industrials \n",
       "197          6                                   Multi-Utilities \n",
       "76           6                                      Broadcasting \n",
       "67           6                               Biological Products \n",
       "141          6                          Containers and Packaging \n",
       "1            6                                       Accessories \n",
       "28           6                           Analog and Mixed Signal "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(vis_data[vis_data['Frequency']>5])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Dependencies for LDA and visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/onion8/miniconda3/lib/python2.7/site-packages/scipy/sparse/sparsetools.py:21: DeprecationWarning: `scipy.sparse.sparsetools` is deprecated!\n",
      "scipy.sparse.sparsetools is a private module for scipy.sparse, and should not be used.\n",
      "  _deprecated()\n"
     ]
    }
   ],
   "source": [
    "import re\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from pprint import pprint\n",
    "\n",
    "# Gensim\n",
    "import gensim\n",
    "import gensim.corpora as corpora\n",
    "from gensim.utils import simple_preprocess\n",
    "from gensim.models import CoherenceModel\n",
    "\n",
    "# spacy for lemmatization\n",
    "import spacy\n",
    "\n",
    "# Plotting tools\n",
    "import pyLDAvis\n",
    "import pyLDAvis.gensim  # don't skip this\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       024 Pharma, Inc. provides healthcare products ...\n",
       "1       1-800-Flowers.com, Inc., together with its sub...\n",
       "2       12 Retech Corporation, through its subsidiarie...\n",
       "3       1347 Property Insurance Holdings, Inc., throug...\n",
       "4       1847 Holdings LLC, through its subsidiaries, p...\n",
       "5       1867 Western Financial Corporation, through it...\n",
       "6       1mage Software Inc operates in the technology ...\n",
       "7       1PM Industries, Inc. provides consulting servi...\n",
       "8       1st Capital Bank provides various banking prod...\n",
       "9       1st Colonial Bancorp, Inc. operates as the ban...\n",
       "10      1st Constitution Bancorp operates as the bank ...\n",
       "11      1st NRG Corp. explores, develops, produces, an...\n",
       "12      1st Prestige Wealth Management provides wealth...\n",
       "13      1st Source Corporation operates as the bank ho...\n",
       "14      1ST SUMMIT BANCORP of Johnstown, Inc. operates...\n",
       "15      RM Investors, Inc. supplies fruits and vegetab...\n",
       "16      2050 Motors, Inc., a development stage company...\n",
       "17      22nd Century Group, Inc., a plant biotechnolog...\n",
       "18      24/7 Kid Doc, Inc. provides pediatric services...\n",
       "19      24Holdings Inc. does not have significant oper...\n",
       "20      2U, Inc. operates as an education technology c...\n",
       "21      30DC, Inc. provides digital media solutions. T...\n",
       "22      3D Entertainment Holdings, Inc. operates as a ...\n",
       "23      3D MakerJet, Inc. develops, imports, and sells...\n",
       "24      3D Pioneer Systems Inc. engages in 3D printing...\n",
       "25      3D Systems Corporation, through its subsidiari...\n",
       "26      3DShopping.com doing business as O2 Essential ...\n",
       "27      3DX Industries, Inc. manufactures and sells co...\n",
       "28      3M Company operates as a diversified technolog...\n",
       "29      3Pea International, Inc. provides prepaid card...\n",
       "                              ...                        \n",
       "1972    Dawson Geophysical Company provides onshore se...\n",
       "1973    Day Tradexchange, Inc. operates as a consultin...\n",
       "1974    Daybreak Oil and Gas, Inc. operates as an inde...\n",
       "1975    DCP Midstream, LP, together with its subsidiar...\n",
       "1976    Deal A Day Group Corp., a development stage co...\n",
       "1977    Dean Foods Company, a food and beverage compan...\n",
       "1978    Debt Resolve, Inc. provides software solutions...\n",
       "1979    Deciphera Pharmaceuticals, Inc., a clinical-st...\n",
       "1980    Decision Diagnostics Corp. develops smart phon...\n",
       "1981    DecisionPoint Systems, Inc. operates as an ent...\n",
       "1982    Decker Manufacturing Corporation manufactures ...\n",
       "1983    Deckers Outdoor Corporation, together with its...\n",
       "1984    Deep Blue Marine, Inc. engages in underwater e...\n",
       "1985    Deep Down, Inc., an oilfield services company,...\n",
       "1986    Deep Earth Resources, Inc., a development stag...\n",
       "1987    Deep Field Technologies, Inc. engages in paint...\n",
       "1988    Deep Green Waste & Recycling, Inc. provides fu...\n",
       "1989    Deep Rock Oil & Gas, Inc. owns oil and gas pro...\n",
       "1990    Deere & Company, together with its subsidiarie...\n",
       "1991    Defense Technology Systems Inc. offers data an...\n",
       "1992    Defentect Group, Inc. develops and provides IP...\n",
       "1993    Del Frisco's Restaurant Group, Inc. develops, ...\n",
       "1994    Del Taco Restaurants, Inc. develops, franchise...\n",
       "1995    Delcath Systems, Inc., an interventional oncol...\n",
       "1996    Delek US Holdings, Inc. engages in the integra...\n",
       "1997    Delhi Bank Corp. operates as a bank holding co...\n",
       "1998    dELiA*s, Inc. operates as a multi-channel reta...\n",
       "1999    Delivery Technology Solutions, Inc. offers res...\n",
       "2000    Delmar Bancorp operates as the bank holding co...\n",
       "2001    Delmarva Bancshares, Inc. operates as the bank...\n",
       "Name: Business Description, Length: 2002, dtype: object"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rawdata['Business Description']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package stopwords to\n",
      "[nltk_data]     /Users/onion8/nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import nltk\n",
    "nltk.download('stopwords')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Stopwords are extended with other common words in the description"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "from nltk.corpus import stopwords\n",
    "stop_words = stopwords.words('english')\n",
    "stop_words.extend(['Inc.', 'Company', 'Corporation','inc','inc.'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[u'Beauty Care Products ' u'Catalog Flowers, Gifts and Novelties '\n",
      " u'Information Technology ' u'Casualty '\n",
      " u'Commercial and Professional Services ' u'Banks '\n",
      " u'Application Software ' u'Energy '\n",
      " u'Asset Management and Custody Banks ' u'Consumer Staples '\n",
      " u'Consumer Discretionary ' u'Cigarettes ' u'Healthcare '\n",
      " u'Computer and Computer Software Stores '\n",
      " u'Application Hosting Services ' u'Industry Specific Electronics '\n",
      " u'B2B Commerce ' u'Capital Goods '\n",
      " u'Data Processing and Outsourced Services '\n",
      " u'Nonresidential Building Operators and Lessors '\n",
      " u'Communications Equipment ' u'Alternative Carriers '\n",
      " u'Independent Power and Renewable Electricity Producers '\n",
      " u'Business Credit Agencies ' u'Building Products '\n",
      " u'Adhesives And Sealants ' u'Agricultural Products '\n",
      " u'Banks; Equity Real Estate Investment Trusts '\n",
      " u'Air and Gas Compressors ' u'Aerospace and Defense ' u'Biotechnology '\n",
      " u'Automotive Retail '\n",
      " u'Electronic Equipment, Instruments and Components ' u'Accessory Stores '\n",
      " u'Cardiovascular Equipment ' u'Advertising ' u\"Children's Hospitals \"\n",
      " u'Health Care Technology ' u'Calendars, Planners and Organizers '\n",
      " u'Chemicals ' u'Containers and Packaging '\n",
      " u'Electronic Equipment and Instruments ' u'Blood and Organ Banks '\n",
      " u'Computer Games ' u'Anti-infective Drugs ' u'Cardiovascular Drugs '\n",
      " u'Financials '\n",
      " u'Banking related Technology Hardware, Storage and Peripherals '\n",
      " u'Commercial Electronics Distribution ' u'Digital '\n",
      " u'Biological Products ' u'Auto Components '\n",
      " u'Colleges, Universities, Professional Schools, and Junior Colleges '\n",
      " u'Architectural Services ' u'Automatic Testing Equipment '\n",
      " u'Central Nervous System Drugs ' u'Coffee And Tea ' u'Gold '\n",
      " u'Broadcasting; Consumer Discretionary; Media; Media; Television; Television Content and Programming; Television Networks'\n",
      " u'Diversified Financials ' u'Casinos and Gaming '\n",
      " u'Agricultural and Farm Machinery ' u'Diversified Metal Ores '\n",
      " u'Air Freight and Logistics ' u'Antennas ' u'Materials '\n",
      " u'Business-to-Business ' u'Electronic Components ' u'Utilities '\n",
      " u'Airlines ' u'Diversified Telecommunication Services '\n",
      " u'Electric Power Generation By Fossil Fuels ' u'Aluminum '\n",
      " u'Real Estate ' u'Diversified Metals and Mining ' u'Beverages '\n",
      " u'Animal Feed ' u'Iron Ores ' u'Casualty Reinsurance Carriers '\n",
      " u'Iron And Steel Foundries ' u'Coal and Consumable Fuels '\n",
      " u'Electric Power Distribution '\n",
      " u'Anesthesia Equipment, Instruments and Accessories ' u'Car Loans '\n",
      " u'Analog and Mixed Signal ' u'Cable And Other Pay Television '\n",
      " u'Chewing, Pipe Tobacco And Snuff ' u'Back-End Equipment '\n",
      " u'Catering Services ' u'Consumer Finance ' u'Broadcasting '\n",
      " u'Alcoholic Beverage and Tobacco Distribution ' u'Industrials '\n",
      " u'Multi-Utilities ' u'Agriculture and Economic Consulting '\n",
      " u'Apparel Retail ' u'Dental Care Equipment ' u'Dental Services '\n",
      " u'Colleges, Universities, and Professional Schools '\n",
      " u'Apartment Building Operators and Lessors ' u'Gas Utilities '\n",
      " u'Accessories ' u'Iron And Steel Forgings ' u'Generic Pharmaceuticals '\n",
      " u'Amplifier ' u'Accident Insurance ' u'Data Management '\n",
      " u'Crude Petroleum Pipelines ' u'Capital Markets '\n",
      " u'Communications Equipment Distribution '\n",
      " u'Beer, Ale and Malt Beverages ' u'Apparel ' u'Computer Components '\n",
      " u'All Terrain Vehicles ' u'Beverage and Tobacco Bars '\n",
      " u'Bituminous Coal And Lignite Mining ' u'Casinos ' u'Automobiles '\n",
      " u'Co-location/Data Centers ' u'Diversified Financial Services '\n",
      " u'Claim Adjusters ' u'Automobile Manufacturers '\n",
      " u'Diversified Real Estate Activities ' u'Alternative Energy Resources '\n",
      " u'Entertainment Software ' u'Concrete '\n",
      " u'Data Processing and Entry Services ' u'Data Storage '\n",
      " u'Capital Goods; Industrial Machinery; Industrials; Lifting Equipment; Machinery; Material Handling Machinery and Equipment; Overhead Traveling Cranes, Hoists, and Monorail Systems'\n",
      " u'Batteries ' u'Development Consulting ' u'Bread and Bakery Products '\n",
      " u'Aerosol Packaging ' u'Apparel, Accessories and Luxury Goods '\n",
      " u'Book Stores ' u'Materials; Materials; Metals and Mining' u'-'\n",
      " u'Land Sub Dividers And Developers ' u'Building Product Distribution '\n",
      " u'Coffee ' u'Electronic Design and Engineering Services '\n",
      " u'Appliance Stores ' u'Accounting Services ' u'Cleaning Products '\n",
      " u'Biometrics Software ' u'Electrotherapeutic Apparatus '\n",
      " u'Cosmetic Dental Implants and Supplies ' u'Sewage Treatment Systems '\n",
      " u'Apparel and Textile Distribution ' u'Accounting Software '\n",
      " u'Alcohol and Narcotic Testing Devices ' u'Buses '\n",
      " u'Beverage Distribution ' u'Catalog Consumer Electronics Retail '\n",
      " u'Electrocardiographs ' u'Telecommunication Services '\n",
      " u'Forest Products ' u'Cellulosic Manmade Fibers '\n",
      " u'Animal Care Product Distribution ' u'Bowling '\n",
      " u'Industrial Service Paper ' u'Child Day Care Services '\n",
      " u'Agricultural Services ' u'Assisted Living Facilities and Services '\n",
      " u'Alcohol Cocktails and Drink Mixes ' u'Aerobic Training Equipment '\n",
      " u'Automation Products and Services ' u'Crude Petroleum and Natural Gas '\n",
      " u'Biscuits ' u'Cemeteries ' u'Amusement And Recreation Services '\n",
      " u'Commercial and Household Goods Moving Services '\n",
      " u'Animal Care Products ' u'Electric Power By Solar Energy '\n",
      " u'Gynecologic and Obstetric Services '\n",
      " u'Cold-Rolled Steel Sheet, Strip, And Bars ' u'Audio Conferencing '\n",
      " u'Building Care and Improvement Services ' u'Amplifiers '\n",
      " u'Cable Network Equipment ' u'Dental Insurance Plans '\n",
      " u'Camping, Hiking and Outdoor Equipment ' u'E-Commerce Software '\n",
      " u'Construction Materials '\n",
      " u'Commercial and Professional Services; Consulting Services; Industrials; Management Consulting Services; Professional Services; Research and Consulting Services'\n",
      " u'Electric Power Transmission and Control '\n",
      " u'Backup and Recovery Software '\n",
      " u'Chemical And Fertilizer Mineral Mining '\n",
      " u'Commercial and Professional Services; Human Resource and Employment Services; Industrials; Professional Services; Temporary Help Supply'\n",
      " u'Competitive Local Exchange Carriers ' u'Consumer Digital Lending '\n",
      " u'Electronic Component Receivers ' u'Disaster Recovery '\n",
      " u'Computer and Electronics Retail '\n",
      " u'Equity Real Estate Investment Trusts '\n",
      " u'Calling Card Provider Services ' u'Automatic Vending Machines '\n",
      " u'Domain Services ' u'Collectibles ' u'Chicken ' u'Development Tools '\n",
      " u'Athletic Footwear ' u'Catalog Apparel and Accessory Retail ']\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>Business Description</th>\n",
       "      <th>Industry Classifications</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>024 Pharma, Inc. provides healthcare products ...</td>\n",
       "      <td>Beauty Care Products (Primary); Consumer Stapl...</td>\n",
       "      <td>Beauty Care Products</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1-800-Flowers.com, Inc., together with its sub...</td>\n",
       "      <td>Catalog Flowers, Gifts and Novelties (Primary)...</td>\n",
       "      <td>Catalog Flowers, Gifts and Novelties</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12 Retech Corporation, through its subsidiarie...</td>\n",
       "      <td>Information Technology (Primary); Internet Sof...</td>\n",
       "      <td>Information Technology</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1347 Property Insurance Holdings, Inc., throug...</td>\n",
       "      <td>Casualty (Primary); Financials (Primary); Fire...</td>\n",
       "      <td>Casualty</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1847 Holdings LLC, through its subsidiaries, p...</td>\n",
       "      <td>Commercial and Professional Services (Primary)...</td>\n",
       "      <td>Commercial and Professional Services</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                Business Description  \\\n",
       "0  024 Pharma, Inc. provides healthcare products ...   \n",
       "1  1-800-Flowers.com, Inc., together with its sub...   \n",
       "2  12 Retech Corporation, through its subsidiarie...   \n",
       "3  1347 Property Insurance Holdings, Inc., throug...   \n",
       "4  1847 Holdings LLC, through its subsidiaries, p...   \n",
       "\n",
       "                            Industry Classifications  \\\n",
       "0  Beauty Care Products (Primary); Consumer Stapl...   \n",
       "1  Catalog Flowers, Gifts and Novelties (Primary)...   \n",
       "2  Information Technology (Primary); Internet Sof...   \n",
       "3  Casualty (Primary); Financials (Primary); Fire...   \n",
       "4  Commercial and Professional Services (Primary)...   \n",
       "\n",
       "                                   label  \n",
       "0                  Beauty Care Products   \n",
       "1  Catalog Flowers, Gifts and Novelties   \n",
       "2                Information Technology   \n",
       "3                              Casualty   \n",
       "4  Commercial and Professional Services   "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(rawdata['label'].unique())\n",
    "rawdata.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Remove special characters and newline characters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[u'024 Pharma, Inc. provides healthcare products worldwide. Its products include vitamin and mineral supplements; stress release, joint, heart health, and weight-loss products; and skin care, hair, and anti-aging products. The company was formerly known as B Green Innovations, Inc. and changed its name to 024 Pharma, Inc. in October 2016. 024 Pharma, Inc. was incorporated in 2004 and is based in West Palm Beach, Florida.']\n"
     ]
    }
   ],
   "source": [
    "# Convert to list\n",
    "data = rawdata['Business Description'].values.tolist()\n",
    "\n",
    "# Remove Emails\n",
    "data = [re.sub('\\S*@\\S*\\s?', '', sent) for sent in data]\n",
    "\n",
    "# Remove new line characters\n",
    "data = [re.sub('\\s+', ' ', sent) for sent in data]\n",
    "\n",
    "# Remove distracting single quotes\n",
    "data = [re.sub(\"\\'\", \"\", sent) for sent in data]\n",
    "\n",
    "\n",
    "\n",
    "pprint(data[:1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tokenize words and Clean-up text(removing punctuations and unnecessary characters)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "def sent_to_words(sentences):\n",
    "    for sentence in sentences:\n",
    "#         sentence = list(sentence)\n",
    "#         sentence.append([x.encode('utf-8') for x in sentence])\n",
    "        sentence = sentence.encode('utf-8')\n",
    "        #print(sentence.encode('utf-8'))\n",
    "        yield((gensim.utils.simple_preprocess(str(sentence), deacc=True)))\n",
    " # deacc=True removes punctuations\n",
    "\n",
    "data_words=list(sent_to_words(data))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[u'pharma',\n",
       "  u'inc',\n",
       "  u'provides',\n",
       "  u'healthcare',\n",
       "  u'products',\n",
       "  u'worldwide',\n",
       "  u'its',\n",
       "  u'products',\n",
       "  u'include',\n",
       "  u'vitamin',\n",
       "  u'and',\n",
       "  u'mineral',\n",
       "  u'supplements',\n",
       "  u'stress',\n",
       "  u'release',\n",
       "  u'joint',\n",
       "  u'heart',\n",
       "  u'health',\n",
       "  u'and',\n",
       "  u'weight',\n",
       "  u'loss',\n",
       "  u'products',\n",
       "  u'and',\n",
       "  u'skin',\n",
       "  u'care',\n",
       "  u'hair',\n",
       "  u'and',\n",
       "  u'anti',\n",
       "  u'aging',\n",
       "  u'products',\n",
       "  u'the',\n",
       "  u'company',\n",
       "  u'was',\n",
       "  u'formerly',\n",
       "  u'known',\n",
       "  u'as',\n",
       "  u'green',\n",
       "  u'innovations',\n",
       "  u'inc',\n",
       "  u'and',\n",
       "  u'changed',\n",
       "  u'its',\n",
       "  u'name',\n",
       "  u'to',\n",
       "  u'pharma',\n",
       "  u'inc',\n",
       "  u'in',\n",
       "  u'october',\n",
       "  u'pharma',\n",
       "  u'inc',\n",
       "  u'was',\n",
       "  u'incorporated',\n",
       "  u'in',\n",
       "  u'and',\n",
       "  u'is',\n",
       "  u'based',\n",
       "  u'in',\n",
       "  u'west',\n",
       "  u'palm',\n",
       "  u'beach',\n",
       "  u'florida']]"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_words[:1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Creating Bigram and Trigram Models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[u'pharma', u'inc', u'provides', u'healthcare', u'products', u'worldwide', u'its', u'products', u'include', u'vitamin', u'and', u'mineral', u'supplements', u'stress', u'release', u'joint', u'heart', u'health', u'and', u'weight_loss', u'products', u'and', u'skin', u'care', u'hair', u'and', u'anti', u'aging', u'products', u'the', u'company', u'was', u'formerly_known', u'as', u'green', u'innovations', u'inc', u'and', u'changed', u'its', u'name', u'to', u'pharma', u'inc', u'in', u'october', u'pharma', u'inc', u'was', u'incorporated', u'in', u'and', u'is', u'based', u'in', u'west_palm_beach_florida']\n"
     ]
    }
   ],
   "source": [
    "# Build the bigram and trigram models\n",
    "bigram = gensim.models.Phrases(data_words, min_count=5, threshold=100) # higher threshold fewer phrases.\n",
    "trigram = gensim.models.Phrases(bigram[data_words], threshold=100)  \n",
    "\n",
    "# Faster way to get a sentence clubbed as a trigram/bigram\n",
    "bigram_mod = gensim.models.phrases.Phraser(bigram)\n",
    "trigram_mod = gensim.models.phrases.Phraser(trigram)\n",
    "\n",
    "# See trigram example\n",
    "print(trigram_mod[bigram_mod[data_words[0]]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[u'flowers', u'com', u'inc', u'together_with', u'its', u'subsidiaries', u'provides', u'gourmet', u'food', u'and', u'floral', u'gifts', u'for', u'various', u'occasions', u'in', u'the', u'united_states', u'it', u'operates', u'in', u'three_segments', u'consumer', u'floral', u'gourmet', u'food', u'and', u'gift', u'baskets', u'and', u'bloomnet', u'wire', u'service', u'the', u'company', u'offers', u'range', u'of', u'products', u'including', u'fresh', u'cut', u'flowers', u'floral', u'and', u'fruit', u'arrangements', u'and', u'plants', u'gifts', u'popcorn', u'gourmet', u'foods', u'and', u'gift', u'baskets', u'cookies', u'chocolates', u'candies', u'wine', u'and', u'gift', u'quality', u'fruits', u'as', u'well', u'as', u'balloons', u'candles', u'keepsake', u'gifts', u'and', u'plush', u'stuffed', u'animals', u'it', u'also', u'provides', u'floral', u'wire', u'service', u'through', u'mybloomnet', u'net', u'gourmet', u'gifts', u'such', u'as', u'fruits', u'and', u'other', u'gourmet', u'items', u'through', u'or', u'harryanddavid', u'com', u'popcorn', u'and', u'specialty', u'treats', u'through', u'or', u'com', u'cookies', u'and', u'baked', u'gifts', u'from', u'or', u'cheryls', u'com', u'gift', u'baskets', u'and', u'towers', u'from', u'baskets', u'com', u'english', u'muffins', u'and', u'other', u'breakfast', u'treats', u'from', u'or', u'wolfermans', u'com', u'carved', u'fresh', u'fruit', u'arrangements', u'from', u'fruitbouquets', u'com', u'and', u'steaks', u'and', u'chops', u'from', u'stockyards', u'com', u'the', u'company', u'offers', u'its', u'products', u'to', u'consumers', u'and', u'wholesalers', u'flowers', u'com', u'inc', u'was', u'founded', u'in', u'and', u'is', u'headquartered', u'in', u'carle', u'place', u'new_york']\n"
     ]
    }
   ],
   "source": [
    "print(trigram_mod[bigram_mod[data_words[1]]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## functions for Remove Stopwords, Make Bigrams and Lemmatize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define functions for stopwords, bigrams, trigrams and lemmatization\n",
    "def remove_stopwords(texts):\n",
    "    return [[word for word in simple_preprocess(str(doc)) if word not in stop_words] for doc in texts]\n",
    "\n",
    "def make_bigrams(texts):\n",
    "    return [bigram_mod[doc] for doc in texts]\n",
    "\n",
    "def make_trigrams(texts):\n",
    "    return [trigram_mod[bigram_mod[doc]] for doc in texts]\n",
    "\n",
    "def lemmatization(texts, allowed_postags=['NOUN', 'ADJ', 'VERB', 'ADV']):\n",
    "    \"\"\"https://spacy.io/api/annotation\"\"\"\n",
    "    texts_out = []\n",
    "    for sent in texts:\n",
    "        doc = nlp(\" \".join(sent)) \n",
    "        texts_out.append([token.lemma_ for token in doc if token.pos_ in allowed_postags])\n",
    "    return texts_out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[u'pharma', u'provide', u'healthcare', u'product', u'worldwide', u'product', u'include', u'vitamin', u'mineral', u'supplement', u'stress', u'release', u'joint', u'heart', u'health', u'weight_loss', u'product', u'skin', u'care', u'hair', u'anti', u'age', u'product', u'company', u'green', u'innovation', u'change', u'name', u'pharma', u'october', u'pharma', u'incorporate', u'base', u'west_palm', u'beach_florida']]\n"
     ]
    }
   ],
   "source": [
    "# Remove Stop Words\n",
    "data_words_nostops = remove_stopwords(data_words)\n",
    "\n",
    "# Form Bigrams\n",
    "data_words_bigrams = make_bigrams(data_words_nostops)\n",
    "\n",
    "\n",
    "# Initialize spacy 'en' model, keeping only tagger component (for efficiency)\n",
    "# python3 -m spacy download en\n",
    "nlp = spacy.load('en', disable=['parser', 'ner'])\n",
    "\n",
    "\n",
    "# Do lemmatization keeping only noun, adj, vb, adv\n",
    "data_lemmatized = lemmatization(data_words_bigrams, allowed_postags=['NOUN', 'ADJ', 'VERB', 'ADV'])\n",
    "\n",
    "print(data_lemmatized[:1])\n",
    "\n",
    "# for dg in data_words_bigrams:\n",
    "#     dg=[x.decode('utf-8') for x in dg]\n",
    "#     dw_bigram.append(dg)\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create the Dictionary and Corpus needed for Topic Modeling()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[(0, 1), (1, 1), (2, 1), (3, 1), (4, 1), (5, 1), (6, 1), (7, 1), (8, 1), (9, 1), (10, 1), (11, 1), (12, 1), (13, 1), (14, 1), (15, 1), (16, 1), (17, 1), (18, 1), (19, 3), (20, 4), (21, 1), (22, 1), (23, 1), (24, 1), (25, 1), (26, 1), (27, 1), (28, 1), (29, 1)]]\n"
     ]
    }
   ],
   "source": [
    "# Create Dictionary\n",
    "id2word = corpora.Dictionary(data_lemmatized)\n",
    "\n",
    "# Create Corpus\n",
    "texts = data_lemmatized\n",
    "\n",
    "# Term Document Frequency\n",
    "corpus = [id2word.doc2bow(text) for text in texts]\n",
    "\n",
    "# View\n",
    "print(corpus[:1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[(u'age', 1),\n",
       "  (u'anti', 1),\n",
       "  (u'base', 1),\n",
       "  (u'beach_florida', 1),\n",
       "  (u'care', 1),\n",
       "  (u'change', 1),\n",
       "  (u'company', 1),\n",
       "  (u'green', 1),\n",
       "  (u'hair', 1),\n",
       "  (u'health', 1),\n",
       "  (u'healthcare', 1),\n",
       "  (u'heart', 1),\n",
       "  (u'include', 1),\n",
       "  (u'incorporate', 1),\n",
       "  (u'innovation', 1),\n",
       "  (u'joint', 1),\n",
       "  (u'mineral', 1),\n",
       "  (u'name', 1),\n",
       "  (u'october', 1),\n",
       "  (u'pharma', 3),\n",
       "  (u'product', 4),\n",
       "  (u'provide', 1),\n",
       "  (u'release', 1),\n",
       "  (u'skin', 1),\n",
       "  (u'stress', 1),\n",
       "  (u'supplement', 1),\n",
       "  (u'vitamin', 1),\n",
       "  (u'weight_loss', 1),\n",
       "  (u'west_palm', 1),\n",
       "  (u'worldwide', 1)]]"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Human readable format of corpus (term-frequency)\n",
    "[[(id2word[id], freq) for id, freq in cp] for cp in corpus[:1]]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Building the Topic Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "lda_model = gensim.models.ldamodel.LdaModel(corpus=corpus,\n",
    "                                           id2word=id2word,\n",
    "                                           num_topics=8, \n",
    "                                           random_state=100,\n",
    "                                           update_every=1,\n",
    "                                           chunksize=100,\n",
    "                                           passes=10,\n",
    "                                           alpha='auto',\n",
    "                                           per_word_topics=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## View the topics in LDA model and also the weightage of keywords in each topic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(0,\n",
      "  u'0.050*\"product\" + 0.029*\"system\" + 0.028*\"use\" + 0.015*\"manufacture\" + 0.013*\"technology\" + 0.012*\"market\" + 0.010*\"application\" + 0.010*\"sell\" + 0.009*\"chemical\" + 0.008*\"device\"'),\n",
      " (1,\n",
      "  u'0.047*\"product\" + 0.026*\"store\" + 0.025*\"brand\" + 0.020*\"food\" + 0.019*\"company\" + 0.018*\"retail\" + 0.015*\"sell\" + 0.014*\"operate\" + 0.010*\"beverage\" + 0.009*\"specialty\"'),\n",
      " (2,\n",
      "  u'0.020*\"care\" + 0.018*\"treatment\" + 0.016*\"patient\" + 0.016*\"health\" + 0.014*\"medical\" + 0.014*\"product\" + 0.012*\"company\" + 0.010*\"pharmaceutical\" + 0.007*\"waste\" + 0.007*\"include\"'),\n",
      " (3,\n",
      "  u'0.037*\"segment\" + 0.025*\"service\" + 0.020*\"product\" + 0.015*\"system\" + 0.013*\"company\" + 0.013*\"well\" + 0.012*\"provide\" + 0.012*\"marine\" + 0.012*\"equipment\" + 0.012*\"industrial\"'),\n",
      " (4,\n",
      "  u'0.059*\"company\" + 0.027*\"develop\" + 0.024*\"development\" + 0.018*\"base\" + 0.016*\"corp\" + 0.016*\"focus\" + 0.016*\"found\" + 0.015*\"name\" + 0.015*\"change\" + 0.014*\"treatment\"'),\n",
      " (5,\n",
      "  u'0.032*\"service\" + 0.030*\"company\" + 0.025*\"solution\" + 0.020*\"provide\" + 0.016*\"offer\" + 0.014*\"base\" + 0.013*\"technology\" + 0.012*\"system\" + 0.012*\"management\" + 0.012*\"software\"'),\n",
      " (6,\n",
      "  u'0.042*\"service\" + 0.041*\"loan\" + 0.037*\"company\" + 0.025*\"account\" + 0.022*\"provide\" + 0.019*\"commercial\" + 0.018*\"operate\" + 0.018*\"check\" + 0.018*\"deposit\" + 0.018*\"banking\"'),\n",
      " (7,\n",
      "  u'0.044*\"company\" + 0.036*\"maryland\" + 0.025*\"own\" + 0.022*\"delmarva\" + 0.021*\"operate\" + 0.017*\"oil\" + 0.017*\"energy\" + 0.013*\"gas\" + 0.013*\"united_state\" + 0.013*\"found\"')]\n"
     ]
    }
   ],
   "source": [
    "pprint(lda_model.print_topics())\n",
    "doc_lda = lda_model[corpus]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Perplexity and Coherence Score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(u'\\nPerplexity: ', -7.712608244622393)\n",
      "(u'\\nCoherence Score: ', 0.47945523115286265)\n"
     ]
    }
   ],
   "source": [
    "# Compute Perplexity\n",
    "print('\\nPerplexity: ', lda_model.log_perplexity(corpus))  # a measure of how good the model is. lower the better.\n",
    "\n",
    "# Compute Coherence Score\n",
    "coherence_model_lda = CoherenceModel(model=lda_model, texts=data_lemmatized, dictionary=id2word, coherence='c_v')\n",
    "coherence_lda = coherence_model_lda.get_coherence()\n",
    "print('\\nCoherence Score: ', coherence_lda)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualize the topics-keywords(it is interactive)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "<link rel=\"stylesheet\" type=\"text/css\" href=\"https://cdn.rawgit.com/bmabey/pyLDAvis/files/ldavis.v1.0.0.css\">\n",
       "\n",
       "\n",
       "<div id=\"ldavis_el14441209944222246259283215\"></div>\n",
       "<script type=\"text/javascript\">\n",
       "\n",
       "var ldavis_el14441209944222246259283215_data = {\"plot.opts\": {\"xlab\": \"PC1\", \"ylab\": \"PC2\"}, \"topic.order\": [6, 7, 4, 5, 8, 3, 1, 2], \"token.table\": {\"Topic\": [1, 2, 1, 3, 8, 2, 5, 6, 5, 6, 1, 2, 3, 5, 6, 7, 3, 3, 3, 3, 1, 2, 3, 4, 5, 6, 7, 8, 4, 4, 7, 1, 5, 6, 1, 3, 7, 1, 3, 5, 8, 7, 3, 2, 2, 1, 3, 2, 2, 2, 2, 1, 2, 3, 4, 5, 5, 8, 8, 8, 7, 4, 3, 6, 2, 1, 3, 8, 1, 2, 1, 1, 2, 4, 5, 8, 1, 2, 4, 6, 4, 6, 5, 2, 6, 2, 6, 8, 2, 8, 7, 1, 2, 1, 4, 5, 8, 1, 2, 3, 5, 7, 6, 6, 4, 6, 4, 1, 5, 3, 7, 8, 1, 8, 4, 1, 2, 3, 4, 1, 1, 2, 3, 4, 5, 6, 7, 8, 3, 7, 1, 1, 7, 2, 3, 1, 1, 2, 7, 8, 1, 1, 3, 5, 7, 8, 7, 2, 3, 4, 2, 1, 2, 3, 4, 5, 7, 8, 8, 7, 2, 5, 1, 2, 3, 5, 3, 3, 6, 5, 8, 1, 1, 8, 1, 3, 1, 2, 4, 5, 6, 6, 3, 2, 5, 2, 1, 3, 4, 7, 1, 4, 5, 6, 7, 1, 2, 4, 5, 1, 6, 7, 6, 1, 4, 6, 4, 6, 1, 1, 2, 3, 6, 8, 1, 3, 5, 8, 1, 3, 7, 8, 6, 3, 8, 4, 6, 4, 8, 3, 5, 1, 2, 3, 7, 4, 7, 1, 3, 5, 7, 1, 4, 5, 3, 1, 2, 1, 3, 2, 3, 7, 5, 5, 5, 2, 4, 7, 8, 8, 1, 4, 5, 1, 2, 3, 4, 5, 6, 7, 8, 8, 3, 5, 7, 3, 3, 7, 1, 4, 8, 5, 8, 3, 1, 1, 2, 3, 4, 5, 6, 8, 2, 6, 8, 1, 2, 6, 5, 3, 1, 2, 3, 4, 8, 1, 2, 5, 6, 8, 2, 6, 1, 5, 8, 2, 4, 6, 4, 1, 2, 3, 4, 5, 6, 7, 8, 1, 2, 4, 5, 7, 7, 2, 6, 2, 2, 3, 5, 7, 1, 3, 5, 7, 1, 7, 8, 4, 7, 3, 3, 7, 7, 2, 7, 1, 4, 2, 5, 2, 1, 8, 5, 7, 7, 2, 5, 8, 2, 3, 5, 8, 8, 6, 2, 3, 6, 7, 6, 2, 2, 5, 3, 4, 7, 5, 4, 1, 3, 1, 2, 3, 6, 1, 3, 6, 7, 8, 1, 3, 7, 3, 1, 2, 3, 6, 7, 8, 1, 3, 4, 5, 3, 7, 8, 4, 6, 7, 1, 3, 4, 3, 1, 5, 5, 1, 3, 5, 5, 3, 5, 1, 2, 7, 2, 6, 2, 3, 1, 3, 4, 5, 8, 4, 5, 5, 1, 2, 1, 2, 3, 4, 8, 3, 2, 3, 5, 6, 7, 8, 5, 4, 6, 8, 8, 1, 2, 3, 6, 7, 8, 3, 5, 4, 6, 1, 2, 1, 2, 3, 5, 6, 8, 7, 4, 6, 3, 6, 2, 5, 7, 6, 3, 4, 6, 2, 8, 8, 4, 6, 7, 6, 4, 4, 6, 1, 4, 7, 6, 3, 7, 7, 6, 8, 1, 2, 3, 5, 7, 3, 4, 5, 7, 5, 1, 2, 3, 6, 7, 8, 1, 3, 5, 7, 1, 2, 1, 2, 3, 4, 6, 2, 5, 1, 2, 3, 6, 7, 7, 3, 2, 2, 4, 6, 5, 3, 6, 7, 1, 4, 6, 7, 4, 2, 3, 3, 4, 5, 1, 2, 3, 8, 1, 3, 8, 8, 2, 1, 7, 4, 2, 4, 7, 8, 2, 1, 2, 2, 3, 6, 1, 3, 5, 7, 8, 7, 1, 7, 7, 1, 2, 3, 6, 1, 2, 3, 6, 8, 4, 6, 4, 8, 1, 1, 3, 3, 6, 7, 8, 5, 4, 4, 5, 6, 2, 3, 5, 1, 3, 1, 8, 4, 6, 1, 2, 3, 4, 5, 8, 1, 3, 7, 8, 1, 3, 6, 6, 6, 1, 3, 6, 7, 4, 3, 1, 4, 7, 7, 6, 7, 5, 4, 6, 4, 6, 6, 3, 3, 3, 5, 4, 6, 4, 6, 4, 6, 3, 4, 1, 2, 3, 4, 5, 6, 8, 1, 3, 4, 6, 7, 1, 3, 5, 1, 2, 3, 4, 5, 6, 7, 8, 6, 2, 3, 7, 1, 1, 6, 5, 3, 5, 1, 1, 1, 2, 3, 4, 5, 6, 7, 8, 5, 1, 2, 5, 8, 7, 8, 1, 8], \"Freq\": [0.9409817457199097, 0.05683782324194908, 0.12359403073787689, 0.2913287878036499, 0.5826575756072998, 0.9994418025016785, 0.8495730757713318, 0.14343442022800446, 0.9681644439697266, 0.9685014486312866, 0.2291191667318344, 0.5373586416244507, 0.08241696655750275, 0.026373429223895073, 0.08571364730596542, 0.036263465881347656, 0.9894142150878906, 0.9847718477249146, 0.986503005027771, 0.9860095977783203, 0.2457411289215088, 0.3055807054042816, 0.1268598735332489, 0.12127484381198883, 0.08537110686302185, 0.06303100287914276, 0.03111657127737999, 0.019946521148085594, 0.9722467064857483, 0.9776299595832825, 0.9692777991294861, 0.991094172000885, 0.9617965817451477, 0.9799805283546448, 0.6367132067680359, 0.1325758993625641, 0.23026339709758759, 0.1301184743642807, 0.12511391937732697, 0.6606014966964722, 0.07506835460662842, 0.9789031147956848, 0.9739632606506348, 0.9924812912940979, 0.9977512955665588, 0.07331545650959015, 0.920516312122345, 0.9974794387817383, 0.9966132044792175, 0.997868001461029, 0.9986664056777954, 0.4332185983657837, 0.14147545397281647, 0.0007992963073775172, 0.278155118227005, 0.1462712287902832, 0.9862880110740662, 0.9598460793495178, 0.9504011273384094, 0.9924931526184082, 0.9661670327186584, 0.9767543077468872, 0.9780567288398743, 0.9916162490844727, 0.9925720691680908, 0.27771326899528503, 0.11593854427337646, 0.6039589643478394, 0.23861366510391235, 0.7606567144393921, 0.9891415238380432, 0.20925407111644745, 0.11593806743621826, 0.5542405247688293, 0.06786618381738663, 0.05089963972568512, 0.9906434416770935, 0.9975343942642212, 0.8988451957702637, 0.09630484133958817, 0.9532111883163452, 0.03530411794781685, 0.97447669506073, 0.9950616955757141, 0.9732392430305481, 0.012167650274932384, 0.9399510025978088, 0.04562868922948837, 0.9929875731468201, 0.9388952851295471, 0.9563531875610352, 0.9935574531555176, 0.9971073865890503, 0.3040395677089691, 0.531130850315094, 0.1576501429080963, 0.0056303623132407665, 0.9869459867477417, 0.9981496930122375, 0.11009809374809265, 0.007339872885495424, 0.8807847499847412, 0.9873673915863037, 0.9944086670875549, 0.43440040946006775, 0.5577486753463745, 0.9929544925689697, 0.9940721392631531, 0.994715690612793, 0.1549776941537857, 0.8191678524017334, 0.9754146337509155, 0.5508025884628296, 0.4479861259460449, 0.9754477739334106, 0.9894251227378845, 0.7927478551864624, 0.2062433362007141, 0.9911609292030334, 0.9958954453468323, 0.22328343987464905, 0.2769407033920288, 0.060388460755348206, 0.21866776049137115, 0.13981659710407257, 0.03519454970955849, 0.014039355330169201, 0.03173279017210007, 0.7391006350517273, 0.25522899627685547, 0.9951336979866028, 0.9884271621704102, 0.9584342241287231, 0.6045034527778625, 0.39219003915786743, 0.9879786968231201, 0.2745821475982666, 0.4311196506023407, 0.10264753550291061, 0.18989793956279755, 0.9956788420677185, 0.238957941532135, 0.4408942461013794, 0.02692483924329281, 0.29280760884284973, 0.9339767694473267, 0.9781670570373535, 0.03325099125504494, 0.02418253943324089, 0.9400961995124817, 0.9919032454490662, 0.19336599111557007, 0.13208803534507751, 0.12936457991600037, 0.2859638035297394, 0.17974866926670074, 0.0789804756641388, 0.9486414790153503, 0.9279404878616333, 0.9727163910865784, 0.999101459980011, 0.9843888878822327, 0.44356590509414673, 0.2819693982601166, 0.2308521419763565, 0.04287254065275192, 0.9878122806549072, 0.9744019508361816, 0.9812377691268921, 0.08946220576763153, 0.8946220278739929, 0.9869551062583923, 0.9985798001289368, 0.9659379720687866, 0.07439662516117096, 0.9175583720207214, 0.08707977831363678, 0.09433642774820328, 0.07982312887907028, 0.7111514806747437, 0.029026592150330544, 0.9869949817657471, 0.9924576282501221, 0.9973897933959961, 0.9962467551231384, 0.9984046816825867, 0.7461739182472229, 0.15104734897613525, 0.07552367448806763, 0.02718852274119854, 0.11862033605575562, 0.7258434891700745, 0.04377655312418938, 0.050837285816669464, 0.06213445961475372, 0.10948973149061203, 0.17406059801578522, 0.6414975523948669, 0.07299315184354782, 0.486734539270401, 0.1400195211172104, 0.37005162239074707, 0.98694908618927, 0.9974390268325806, 0.9288030862808228, 0.06517916172742844, 0.5422486662864685, 0.4509791135787964, 0.9901349544525146, 0.29792162775993347, 0.022626958787441254, 0.3431755304336548, 0.04148275777697563, 0.2941504716873169, 0.284917414188385, 0.5873681902885437, 0.09643357992172241, 0.03068341314792633, 0.14761461317539215, 0.1528865545988083, 0.42702797055244446, 0.26886945962905884, 0.9907012581825256, 0.9821208119392395, 0.962475597858429, 0.3922814428806305, 0.5996301770210266, 0.987244725227356, 0.9592812061309814, 0.8929849863052368, 0.09345191717147827, 0.4931242763996124, 0.08651302754878998, 0.10814128816127777, 0.3114469051361084, 0.9767711758613586, 0.975462019443512, 0.9931617975234985, 0.14414149522781372, 0.7744072675704956, 0.08196280896663666, 0.11244384944438934, 0.45289885997772217, 0.43103477358818054, 0.9881632924079895, 0.9815311431884766, 0.011028439737856388, 0.9864058494567871, 0.9796004295349121, 0.3198370635509491, 0.576921284198761, 0.1032385379076004, 0.9914515614509583, 0.9721507430076599, 0.9810214042663574, 0.9971900582313538, 0.995583713054657, 0.11328036338090897, 0.8816167712211609, 0.9786490201950073, 0.34121397137641907, 0.5118209719657898, 0.14469200372695923, 0.19299574196338654, 0.257561594247818, 0.09684877842664719, 0.21194441616535187, 0.14808037877082825, 0.026668503880500793, 0.025966700166463852, 0.04000275582075119, 0.9690640568733215, 0.2088601291179657, 0.7634197473526001, 0.02520725689828396, 0.9830505847930908, 0.23043444752693176, 0.7622062563896179, 0.9872260689735413, 0.987602949142456, 0.9637001752853394, 0.9819833040237427, 0.9554606080055237, 0.9745659232139587, 0.9901731014251709, 0.28045690059661865, 0.0598728209733963, 0.141804039478302, 0.31827130913734436, 0.12919923663139343, 0.05672161653637886, 0.01575600542128086, 0.1274891346693039, 0.7984846234321594, 0.07045452296733856, 0.5332831740379333, 0.006273919716477394, 0.4579961597919464, 0.9812317490577698, 0.9694377779960632, 0.3383093476295471, 0.016583791002631187, 0.08623571693897247, 0.3980109989643097, 0.1625211536884308, 0.189956396818161, 0.5635372996330261, 0.053820978850126266, 0.053820978850126266, 0.1393013596534729, 0.9968266487121582, 0.9860906600952148, 0.02058490924537182, 0.9674907326698303, 0.9445582628250122, 0.03585117310285568, 0.9500560760498047, 0.9921597838401794, 0.9816386103630066, 0.29689592123031616, 0.23546917736530304, 0.17010481655597687, 0.10316542536020279, 0.02913832478225231, 0.08426488935947418, 0.04173868149518967, 0.040163636207580566, 0.30543839931488037, 0.050346989184617996, 0.486687570810318, 0.12754571437835693, 0.030208194628357887, 0.9679194092750549, 0.9446723461151123, 0.05176286771893501, 0.9911564588546753, 0.07886708527803421, 0.736092746257782, 0.02628902718424797, 0.15773417055606842, 0.4573352634906769, 0.4168420433998108, 0.03096540831029415, 0.0928962230682373, 0.9959855675697327, 0.3490085005760193, 0.6345609426498413, 0.994293749332428, 0.9811787605285645, 0.974084734916687, 0.9873762726783752, 0.9788818359375, 0.9846463203430176, 0.9960198402404785, 0.9465639591217041, 0.9891683459281921, 0.9902631640434265, 0.17385409772396088, 0.8149411082267761, 0.9933329820632935, 0.2231074869632721, 0.7585654258728027, 0.9696834087371826, 0.956788957118988, 0.9886677265167236, 0.5355821847915649, 0.46106642484664917, 0.978417694568634, 0.23817208409309387, 0.04253073036670685, 0.7145162224769592, 0.9817502498626709, 0.9738883376121521, 0.9786251187324524, 0.9122813940048218, 0.07286592572927475, 0.013115867041051388, 0.9846869707107544, 0.9792883396148682, 0.9991695284843445, 0.17194615304470062, 0.8253415822982788, 0.993693470954895, 0.96821129322052, 0.9870320558547974, 0.9785639047622681, 0.9892556667327881, 0.20302386581897736, 0.7951768040657043, 0.517925500869751, 0.3655944764614105, 0.01128377951681614, 0.10381077229976654, 0.12166659533977509, 0.2706460952758789, 0.012414958328008652, 0.4990813434123993, 0.09435368329286575, 0.21103061735630035, 0.331619530916214, 0.4522084593772888, 0.9954950213432312, 0.36335840821266174, 0.006606516428291798, 0.29200804233551025, 0.027747370302677155, 0.21405114233493805, 0.09645514190196991, 0.7616025805473328, 0.12167614698410034, 0.11266310513019562, 0.998485803604126, 0.47425076365470886, 0.5194174647331238, 0.9233437776565552, 0.07608374208211899, 0.7244495153427124, 0.1984793245792389, 0.9470197558403015, 0.050569016486406326, 0.9792042374610901, 0.9938598871231079, 0.1184065118432045, 0.858447253704071, 0.9838272333145142, 0.06862194836139679, 0.9195341467857361, 0.9675776362419128, 0.9790413975715637, 0.21846328675746918, 0.7791857123374939, 0.3042108416557312, 0.6937658786773682, 0.980560839176178, 0.9956943988800049, 0.970517098903656, 0.9964309930801392, 0.9802285432815552, 0.3072957992553711, 0.011487693525850773, 0.4135569632053375, 0.16657155752182007, 0.10051731765270233, 0.979533851146698, 0.995297908782959, 0.968298614025116, 0.9587752223014832, 0.03844712674617767, 0.17618857324123383, 0.2259809970855713, 0.09958484768867493, 0.3983393907546997, 0.09958484768867493, 0.9846868515014648, 0.18969950079917908, 0.12449029088020325, 0.10077785700559616, 0.45053631067276, 0.09484975039958954, 0.04149676486849785, 0.9777209162712097, 0.9919101595878601, 0.9677646160125732, 0.9642057418823242, 0.9648054242134094, 0.3475702702999115, 0.3392152190208435, 0.15261898934841156, 0.05737137421965599, 0.06127040088176727, 0.042332276701927185, 0.03477473184466362, 0.9632600545883179, 0.9884747266769409, 0.987846314907074, 0.2370658963918686, 0.7611063122749329, 0.12291395664215088, 0.4584358334541321, 0.09633796662092209, 0.23187552392482758, 0.009301596321165562, 0.08039237558841705, 0.9823896288871765, 0.07164864242076874, 0.9211968183517456, 0.9759945869445801, 0.9678964614868164, 0.04985639080405235, 0.947271466255188, 0.9769141674041748, 0.9861658811569214, 0.9949762225151062, 0.9810474514961243, 0.9961847066879272, 0.9435974955558777, 0.05242208391427994, 0.9688278436660767, 0.391909658908844, 0.564056932926178, 0.040289778262376785, 0.9858074188232422, 0.9821224808692932, 0.9833452105522156, 0.9866095781326294, 0.9073581695556641, 0.09132500737905502, 0.9683734178543091, 0.9701372981071472, 0.7332072854042053, 0.2637757956981659, 0.9793180227279663, 0.9876946806907654, 0.9429315328598022, 0.514144241809845, 0.03126552700996399, 0.1424318552017212, 0.052109211683273315, 0.2570721209049225, 0.1892206370830536, 0.022261250764131546, 0.6734028458595276, 0.11130625754594803, 0.9779017567634583, 0.14084170758724213, 0.19687551259994507, 0.179080992937088, 0.07837159931659698, 0.2468516081571579, 0.15750041604042053, 0.18046478927135468, 0.07567878812551498, 0.6403589248657227, 0.09896456450223923, 0.8181142210960388, 0.17490717768669128, 0.29699891805648804, 0.12419953942298889, 0.021599920466542244, 0.13499949872493744, 0.42119845747947693, 0.4806815981864929, 0.516802191734314, 0.3900657594203949, 0.4076406955718994, 0.14304038882255554, 0.044425513595342636, 0.014645772986114025, 0.9752503037452698, 0.9894840717315674, 0.9972946047782898, 0.08586826175451279, 0.08586826175451279, 0.8157484531402588, 0.9844534993171692, 0.9064070582389832, 0.08031454682350159, 0.9628270864486694, 0.13740034401416779, 0.5496013760566711, 0.12595030665397644, 0.1774754375219345, 0.9861301183700562, 0.8936662077903748, 0.10331401228904724, 0.07910594344139099, 0.9176289439201355, 0.9904943704605103, 0.20284324884414673, 0.2057410180568695, 0.13039922714233398, 0.4607439637184143, 0.283077597618103, 0.07279138267040253, 0.6308586597442627, 0.9663035273551941, 0.995746910572052, 0.2999352216720581, 0.6921581625938416, 0.9664446711540222, 0.9975911378860474, 0.981417179107666, 0.9642071723937988, 0.9421371221542358, 0.9932039380073547, 0.7707926034927368, 0.22568251192569733, 0.001077378517948091, 0.938396692276001, 0.06033319607377052, 0.19381214678287506, 0.34110939502716064, 0.06046938896179199, 0.2000141441822052, 0.20621612668037415, 0.9889515042304993, 0.1282692402601242, 0.8693804144859314, 0.9818258285522461, 0.37570521235466003, 0.10867506265640259, 0.4471202492713928, 0.06520503759384155, 0.357395738363266, 0.459264874458313, 0.16816948354244232, 0.015081188641488552, 0.9600353240966797, 0.9840940237045288, 0.987865149974823, 0.9900927543640137, 0.971077561378479, 0.9969842433929443, 0.9724768400192261, 0.025853166356682777, 0.2618280053138733, 0.1787080019712448, 0.22442400455474854, 0.3324800133705139, 0.9799758791923523, 0.9924890995025635, 0.09432745724916458, 0.6854462027549744, 0.2200973927974701, 0.9958546161651611, 0.9678230881690979, 0.02580861561000347, 0.259392112493515, 0.7403483390808105, 0.17686541378498077, 0.8206555247306824, 0.03894790634512901, 0.9542236924171448, 0.4554871916770935, 0.024547213688492775, 0.10091632604598999, 0.04909442737698555, 0.2672919034957886, 0.10091632604598999, 0.9899352788925171, 0.6863009333610535, 0.23897980153560638, 0.07353224605321884, 0.8272485136985779, 0.13136564195156097, 0.03550422936677933, 0.9783470630645752, 0.9778414964675903, 0.37360671162605286, 0.2889863848686218, 0.037520330399274826, 0.2993643581867218, 0.9921713471412659, 0.9669055342674255, 0.5888828635215759, 0.22221994400024414, 0.18888695538043976, 0.9741181135177612, 0.22193977236747742, 0.7667010426521301, 0.9909475445747375, 0.9112009406089783, 0.08283644914627075, 0.40292271971702576, 0.5888870358467102, 0.9872984290122986, 0.9871188402175903, 0.9818594455718994, 0.8982905149459839, 0.0979953333735466, 0.36162906885147095, 0.6361992955207825, 0.49851125478744507, 0.5003306269645691, 0.8554024696350098, 0.1333094835281372, 0.9859664440155029, 0.984923243522644, 0.2592213749885559, 0.15394806861877441, 0.09961345791816711, 0.11772499233484268, 0.2399778664112091, 0.05886249616742134, 0.070182204246521, 0.1364157348871231, 0.22501564025878906, 0.04500312730669975, 0.08578720688819885, 0.5090978741645813, 0.9924625158309937, 0.9914950132369995, 0.9607396125793457, 0.24008989334106445, 0.42682647705078125, 0.11946502327919006, 0.08350952714681625, 0.009278836660087109, 0.03595549240708351, 0.05219345539808273, 0.03131607174873352, 0.9893463253974915, 0.09608981758356094, 0.9014139771461487, 0.971303403377533, 0.9971553683280945, 0.9855443835258484, 0.9929481148719788, 0.9701619148254395, 0.10101664811372757, 0.8931998014450073, 0.9913234114646912, 0.98948073387146, 0.24430027604103088, 0.2672281265258789, 0.23560349643230438, 0.02292785793542862, 0.08854896575212479, 0.082224041223526, 0.02292785793542862, 0.03557771071791649, 0.9855016469955444, 0.3373727798461914, 0.04089367017149925, 0.21469175815582275, 0.38848984241485596, 0.9669685959815979, 0.9897770285606384, 0.9949004054069519, 0.9629669785499573], \"Term\": [\"access\", \"access\", \"accessory\", \"accessory\", \"accessory\", \"account\", \"acquire\", \"acquire\", \"acre\", \"acute\", \"addition\", \"addition\", \"addition\", \"addition\", \"addition\", \"addition\", \"advertising\", \"aerospace\", \"air\", \"aircraft\", \"also\", \"also\", \"also\", \"also\", \"also\", \"also\", \"also\", \"also\", \"alternative\", \"america\", \"analog\", \"analytic\", \"animal\", \"anti\", \"application\", \"application\", \"application\", \"approximately\", \"approximately\", \"approximately\", \"approximately\", \"assembly\", \"assessment\", \"auto\", \"automobile\", \"automotive\", \"automotive\", \"bancorp\", \"bank\", \"bank_hold\", \"banking\", \"base\", \"base\", \"base\", \"base\", \"base\", \"basin\", \"beauty\", \"beer\", \"beverage\", \"biologic\", \"bioscience\", \"black\", \"blood\", \"branch\", \"brand\", \"brand\", \"brand\", \"business\", \"business\", \"cable\", \"california\", \"california\", \"california\", \"california\", \"california\", \"call\", \"cambridge\", \"cancer\", \"cancer\", \"candidate\", \"candidate\", \"capacity\", \"card\", \"cardiac\", \"care\", \"care\", \"care\", \"cash\", \"casual\", \"catalyst\", \"cell\", \"certificate\", \"change\", \"change\", \"change\", \"change\", \"channel\", \"check\", \"chemical\", \"chemical\", \"chemical\", \"chronic\", \"claim\", \"clinical\", \"clinical\", \"clinical_trial\", \"cloud\", \"coal\", \"coating\", \"coating\", \"coffee\", \"com\", \"com\", \"combination\", \"commerce\", \"commercial\", \"commercial\", \"commercialize\", \"communication\", \"company\", \"company\", \"company\", \"company\", \"company\", \"company\", \"company\", \"company\", \"component\", \"component\", \"computer\", \"connectivity\", \"connector\", \"construction\", \"construction\", \"consult\", \"consumer\", \"consumer\", \"consumer\", \"consumer\", \"content\", \"control\", \"control\", \"control\", \"control\", \"convenience\", \"converter\", \"corp\", \"corp\", \"corp\", \"corporate\", \"corporation\", \"corporation\", \"corporation\", \"corporation\", \"corporation\", \"corporation\", \"craft\", \"cream\", \"creation\", \"credit\", \"crude_oil\", \"customer\", \"customer\", \"customer\", \"customer\", \"customize\", \"cyber\", \"daily\", \"dairy\", \"dairy\", \"data\", \"datum\", \"davi\", \"dealer\", \"dealer\", \"december\", \"december\", \"december\", \"december\", \"december\", \"deep\", \"defense\", \"delmar\", \"delmarva\", \"deposit\", \"design\", \"design\", \"design\", \"design\", \"develop\", \"develop\", \"develop\", \"develop\", \"develop\", \"development\", \"development\", \"development\", \"development\", \"device\", \"device\", \"device\", \"dialysis\", \"digital\", \"discovery\", \"discovery\", \"disease\", \"disease\", \"display\", \"distribute\", \"distribute\", \"distribute\", \"distribute\", \"distribute\", \"distribution\", \"distribution\", \"distribution\", \"distribution\", \"distributor\", \"distributor\", \"distributor\", \"distributor\", \"district\", \"division\", \"drink\", \"drug\", \"drug\", \"drug_candidate\", \"eastern\", \"electric\", \"electric\", \"electronic\", \"electronic\", \"electronic\", \"electronic\", \"email\", \"emission\", \"enable\", \"energy\", \"energy\", \"energy\", \"engage\", \"engage\", \"engage\", \"engine\", \"enterprise\", \"enterprise\", \"entertainment\", \"environmental\", \"equipment\", \"equipment\", \"equipment\", \"exploration\", \"explore\", \"feed\", \"financial\", \"focus\", \"food\", \"food\", \"footwear\", \"formerly_known\", \"formerly_known\", \"formerly_known\", \"found\", \"found\", \"found\", \"found\", \"found\", \"found\", \"found\", \"found\", \"garden\", \"gas\", \"gas\", \"gas\", \"generator\", \"glass\", \"glass\", \"global\", \"gold\", \"golf\", \"green\", \"grocery\", \"handle\", \"hardware\", \"headquarter\", \"headquarter\", \"headquarter\", \"headquarter\", \"headquarter\", \"headquarter\", \"headquarter\", \"health\", \"health\", \"health\", \"healthcare\", \"healthcare\", \"healthcare\", \"heat\", \"highway\", \"holding\", \"holding\", \"holding\", \"holding\", \"holding\", \"home\", \"home\", \"home\", \"home\", \"home\", \"home_equity\", \"hospital\", \"hotel\", \"hotel\", \"household\", \"houston_texas\", \"houston_texas\", \"human\", \"immune\", \"include\", \"include\", \"include\", \"include\", \"include\", \"include\", \"include\", \"include\", \"incorporate\", \"incorporate\", \"incorporate\", \"incorporate\", \"incorporate\", \"indication\", \"individual\", \"individual\", \"individual_retirement\", \"industrial\", \"industrial\", \"industrial\", \"industrial\", \"industry\", \"industry\", \"industry\", \"industry\", \"information\", \"ingredient\", \"ingredient\", \"inhibitor\", \"injection\", \"inspection\", \"installation\", \"instrument\", \"instrumentation\", \"insurance\", \"integrated_circuit\", \"integration\", \"intend\", \"interest\", \"interest\", \"investment\", \"item\", \"item\", \"joint\", \"kansa\", \"kit\", \"land\", \"land\", \"las_vegas\", \"lease\", \"lease\", \"lease\", \"liability\", \"limit\", \"limited\", \"line\", \"line\", \"line\", \"liquid\", \"llc\", \"loan\", \"locate\", \"locate\", \"logistic\", \"los_angele\", \"louisiana\", \"lp\", \"ltd\", \"maintenance\", \"maintenance\", \"management\", \"management\", \"management\", \"management\", \"manufacture\", \"manufacture\", \"manufacture\", \"manufacture\", \"manufacture\", \"manufacturer\", \"manufacturer\", \"manufacturer\", \"marine\", \"market\", \"market\", \"market\", \"market\", \"market\", \"market\", \"marketing\", \"marketing\", \"marketing\", \"maryland\", \"material\", \"material\", \"mattress\", \"medical\", \"medical\", \"medical\", \"medium\", \"medium\", \"memory\", \"metal\", \"mexico\", \"mexico\", \"mile\", \"military\", \"military\", \"mine\", \"mineral\", \"mining\", \"mining\", \"mobile\", \"mobile\", \"module\", \"money_market\", \"monitoring\", \"mortgage\", \"motor\", \"name\", \"name\", \"name\", \"name\", \"name\", \"natural\", \"natural_gas\", \"net\", \"network\", \"network\", \"new_york\", \"new_york\", \"new_york\", \"new_york\", \"new_york\", \"news\", \"non\", \"non\", \"non\", \"non\", \"non\", \"non\", \"northern\", \"novel\", \"nurse\", \"nut\", \"nutrition\", \"offer\", \"offer\", \"offer\", \"offer\", \"offer\", \"offer\", \"oil\", \"oil\", \"oklahoma\", \"oncology\", \"online\", \"online\", \"operate\", \"operate\", \"operate\", \"operate\", \"operate\", \"operate\", \"optical\", \"oral\", \"oral\", \"outdoor\", \"outpatient\", \"own\", \"own\", \"packaging\", \"pain\", \"part\", \"passenger\", \"patient\", \"personal\", \"personal\", \"pet\", \"pharmaceutical\", \"pharmaceutical\", \"pharmaceutical\", \"pharmacy\", \"phase\", \"phase_ii\", \"physician\", \"platform\", \"platform\", \"policy\", \"post\", \"power\", \"power\", \"precision\", \"prevention\", \"previously\", \"process\", \"process\", \"process\", \"process\", \"process\", \"produce\", \"produce\", \"produce\", \"produce\", \"producer\", \"product\", \"product\", \"product\", \"product\", \"product\", \"product\", \"production\", \"production\", \"production\", \"production\", \"professional\", \"professional\", \"program\", \"program\", \"program\", \"program\", \"program\", \"property\", \"property\", \"provide\", \"provide\", \"provide\", \"provide\", \"provide\", \"pump\", \"rail\", \"real_estate\", \"record\", \"record\", \"record\", \"recovery\", \"repair\", \"repair\", \"representative\", \"research\", \"research\", \"research\", \"research\", \"reserve\", \"residential\", \"residential\", \"resource\", \"resource\", \"restaurant\", \"retail\", \"retail\", \"retail\", \"retail\", \"retailer\", \"retailer\", \"retailer\", \"rhode_island\", \"safe_deposit\", \"safety\", \"safety\", \"san_francisco\", \"saving\", \"science\", \"seal\", \"seasonal\", \"secure\", \"security\", \"security\", \"segment\", \"segment\", \"segment\", \"sell\", \"sell\", \"sell\", \"sell\", \"sell\", \"semiconductor\", \"sensor\", \"sensor\", \"series\", \"serve\", \"serve\", \"serve\", \"serve\", \"service\", \"service\", \"service\", \"service\", \"significant_operation\", \"silver\", \"skin\", \"small_molecule\", \"snack\", \"software\", \"solution\", \"solution\", \"specialty\", \"specialty\", \"specialty\", \"specialty\", \"spring\", \"stage\", \"state\", \"state\", \"state\", \"statement\", \"steel\", \"steel\", \"storage\", \"storage\", \"store\", \"store\", \"student\", \"student\", \"subsidiary\", \"subsidiary\", \"subsidiary\", \"subsidiary\", \"subsidiary\", \"subsidiary\", \"suite\", \"supply\", \"supply\", \"supply\", \"support\", \"support\", \"support\", \"surgery\", \"surgical\", \"system\", \"system\", \"system\", \"system\", \"target\", \"technical\", \"technology\", \"technology\", \"technology\", \"temperature\", \"test\", \"test\", \"texas\", \"therapeutic\", \"therapeutic\", \"therapy\", \"therapy\", \"tissue\", \"track\", \"transmission\", \"transportation\", \"transportation\", \"treat\", \"treat\", \"treatment\", \"treatment\", \"trial\", \"trial\", \"truck\", \"tumor\", \"united_state\", \"united_state\", \"united_state\", \"united_state\", \"united_state\", \"united_state\", \"united_state\", \"use\", \"use\", \"use\", \"use\", \"use\", \"user\", \"utility\", \"valley\", \"various\", \"various\", \"various\", \"various\", \"various\", \"various\", \"various\", \"various\", \"vascular\", \"vehicle\", \"vehicle\", \"vessel\", \"video\", \"voice\", \"waste\", \"wastewater\", \"water\", \"water\", \"web\", \"website\", \"well\", \"well\", \"well\", \"well\", \"well\", \"well\", \"well\", \"well\", \"west\", \"wholesale\", \"wholesale\", \"wholesale\", \"wholesale\", \"wind\", \"wine\", \"wireless\", \"world\"]}, \"mdsDat\": {\"y\": [-0.046505147183783686, 0.03823345219389828, -0.1603530621328391, 0.21625366759261186, 0.21667785517596502, -0.015366285812733325, -0.18890146479965322, -0.06003901503346609], \"cluster\": [1, 1, 1, 1, 1, 1, 1, 1], \"Freq\": [22.563846588134766, 22.073501586914062, 13.500492095947266, 10.986976623535156, 9.492788314819336, 8.726937294006348, 7.547307014465332, 5.108153820037842], \"topics\": [1, 2, 3, 4, 5, 6, 7, 8], \"x\": [-0.10330494697164824, -0.3243442255814425, -0.051973563832479364, 0.12337322716850496, 0.021018751171086765, 0.11171499274085138, 0.1306465331087156, 0.09286923219641081]}, \"R\": 30, \"lambda.step\": 0.01, \"tinfo\": {\"Category\": [\"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Default\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic1\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic2\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic3\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic4\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic5\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic6\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic7\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\", \"Topic8\"], \"Term\": [\"loan\", \"product\", \"service\", \"segment\", \"account\", \"maryland\", \"solution\", \"system\", \"company\", \"check\", \"deposit\", \"banking\", \"own\", \"commercial\", \"operate\", \"credit\", \"use\", \"develop\", \"treatment\", \"development\", \"technology\", \"delmarva\", \"brand\", \"base\", \"business\", \"store\", \"line\", \"care\", \"sell\", \"financial\", \"software\", \"datum\", \"communication\", \"information\", \"digital\", \"video\", \"enable\", \"cloud\", \"cell\", \"web\", \"wireless\", \"analytic\", \"display\", \"computer\", \"content\", \"channel\", \"website\", \"entertainment\", \"voice\", \"hardware\", \"global\", \"call\", \"consult\", \"data\", \"user\", \"suite\", \"cable\", \"connectivity\", \"commerce\", \"integration\", \"solution\", \"enterprise\", \"network\", \"medium\", \"platform\", \"access\", \"support\", \"security\", \"design\", \"technology\", \"application\", \"service\", \"management\", \"provide\", \"base\", \"offer\", \"company\", \"marketing\", \"system\", \"professional\", \"customer\", \"include\", \"market\", \"also\", \"well\", \"product\", \"found\", \"united_state\", \"mobile\", \"name\", \"loan\", \"account\", \"check\", \"deposit\", \"banking\", \"credit\", \"financial\", \"real_estate\", \"bank\", \"saving\", \"mortgage\", \"certificate\", \"money_market\", \"home_equity\", \"bancorp\", \"bank_hold\", \"card\", \"automobile\", \"statement\", \"cambridge\", \"delmar\", \"branch\", \"cash\", \"corporate\", \"insurance\", \"secure\", \"individual_retirement\", \"auto\", \"investment\", \"safe_deposit\", \"line\", \"personal\", \"residential\", \"commercial\", \"individual\", \"business\", \"mobile\", \"service\", \"online\", \"provide\", \"operate\", \"company\", \"offer\", \"product\", \"various\", \"addition\", \"also\", \"found\", \"well\", \"management\", \"marine\", \"defense\", \"engine\", \"part\", \"utility\", \"installation\", \"advertising\", \"metal\", \"air\", \"truck\", \"logistic\", \"transmission\", \"environmental\", \"news\", \"division\", \"motor\", \"outdoor\", \"cyber\", \"track\", \"aerospace\", \"handle\", \"rail\", \"aircraft\", \"inspection\", \"customize\", \"technical\", \"black\", \"generator\", \"assessment\", \"highway\", \"segment\", \"steel\", \"vehicle\", \"automotive\", \"transportation\", \"dealer\", \"repair\", \"electric\", \"industrial\", \"power\", \"component\", \"storage\", \"equipment\", \"military\", \"maintenance\", \"supply\", \"system\", \"distribution\", \"service\", \"product\", \"sell\", \"industry\", \"well\", \"market\", \"serve\", \"provide\", \"offer\", \"control\", \"include\", \"commercial\", \"company\", \"use\", \"construction\", \"also\", \"customer\", \"operate\", \"found\", \"focus\", \"clinical_trial\", \"stage\", \"target\", \"gold\", \"inhibitor\", \"commercialize\", \"ltd\", \"phase_ii\", \"intend\", \"natural\", \"novel\", \"drug_candidate\", \"phase\", \"oklahoma\", \"small_molecule\", \"bioscience\", \"combination\", \"america\", \"reserve\", \"tumor\", \"silver\", \"science\", \"alternative\", \"passenger\", \"immune\", \"email\", \"san_francisco\", \"los_angele\", \"memory\", \"corp\", \"resource\", \"therapeutic\", \"cancer\", \"candidate\", \"develop\", \"development\", \"discovery\", \"trial\", \"houston_texas\", \"change\", \"treatment\", \"formerly_known\", \"california\", \"company\", \"name\", \"base\", \"incorporate\", \"engage\", \"found\", \"corporation\", \"technology\", \"disease\", \"holding\", \"research\", \"pharmaceutical\", \"also\", \"new_york\", \"include\", \"headquarter\", \"united_state\", \"maryland\", \"delmarva\", \"texas\", \"natural_gas\", \"restaurant\", \"exploration\", \"coal\", \"crude_oil\", \"mineral\", \"capacity\", \"recovery\", \"mile\", \"northern\", \"green\", \"net\", \"basin\", \"west\", \"lp\", \"explore\", \"spring\", \"producer\", \"valley\", \"acre\", \"feed\", \"animal\", \"wastewater\", \"heat\", \"hotel\", \"joint\", \"mine\", \"oil\", \"own\", \"water\", \"energy\", \"locate\", \"gas\", \"acquire\", \"mining\", \"interest\", \"mexico\", \"approximately\", \"produce\", \"december\", \"state\", \"property\", \"production\", \"lease\", \"company\", \"operate\", \"engage\", \"united_state\", \"land\", \"found\", \"base\", \"corporation\", \"name\", \"subsidiary\", \"well\", \"also\", \"patient\", \"waste\", \"hospital\", \"claim\", \"physician\", \"pharmacy\", \"blood\", \"human\", \"oncology\", \"vascular\", \"anti\", \"deep\", \"tissue\", \"surgery\", \"skin\", \"district\", \"prevention\", \"llc\", \"pain\", \"post\", \"daily\", \"limited\", \"dialysis\", \"acute\", \"monitoring\", \"nurse\", \"cardiac\", \"chronic\", \"outpatient\", \"surgical\", \"care\", \"oral\", \"health\", \"medical\", \"student\", \"treatment\", \"pharmaceutical\", \"drug\", \"treat\", \"clinical\", \"record\", \"therapy\", \"disease\", \"non\", \"product\", \"program\", \"healthcare\", \"company\", \"include\", \"well\", \"offer\", \"management\", \"provide\", \"also\", \"liquid\", \"semiconductor\", \"module\", \"instrument\", \"assembly\", \"packaging\", \"optical\", \"injection\", \"temperature\", \"series\", \"analog\", \"kit\", \"louisiana\", \"precision\", \"vessel\", \"pump\", \"indication\", \"policy\", \"kansa\", \"instrumentation\", \"seal\", \"catalyst\", \"biologic\", \"wind\", \"representative\", \"emission\", \"creation\", \"converter\", \"integrated_circuit\", \"connector\", \"chemical\", \"sensor\", \"test\", \"use\", \"manufacture\", \"safety\", \"product\", \"system\", \"material\", \"coating\", \"glass\", \"manufacturer\", \"device\", \"distributor\", \"market\", \"technology\", \"application\", \"sell\", \"control\", \"electronic\", \"process\", \"offer\", \"power\", \"medical\", \"industrial\", \"company\", \"beverage\", \"las_vegas\", \"wine\", \"footwear\", \"pet\", \"liability\", \"nutrition\", \"golf\", \"grocery\", \"beauty\", \"garden\", \"coffee\", \"limit\", \"eastern\", \"world\", \"craft\", \"beer\", \"significant_operation\", \"davi\", \"household\", \"drink\", \"convenience\", \"previously\", \"seasonal\", \"snack\", \"casual\", \"mattress\", \"cream\", \"rhode_island\", \"nut\", \"food\", \"store\", \"dairy\", \"brand\", \"retailer\", \"retail\", \"accessory\", \"item\", \"product\", \"ingredient\", \"com\", \"specialty\", \"sell\", \"distribute\", \"consumer\", \"operate\", \"company\", \"distributor\", \"market\", \"name\", \"offer\", \"united_state\", \"holding\", \"wholesale\", \"found\", \"include\", \"home\", \"well\"], \"loglift\": [30.0, 29.0, 28.0, 27.0, 26.0, 25.0, 24.0, 23.0, 22.0, 21.0, 20.0, 19.0, 18.0, 17.0, 16.0, 15.0, 14.0, 13.0, 12.0, 11.0, 10.0, 9.0, 8.0, 7.0, 6.0, 5.0, 4.0, 3.0, 2.0, 1.0, 1.486799955368042, 1.4866000413894653, 1.4853999614715576, 1.4852999448776245, 1.485200047492981, 1.4837000370025635, 1.4836000204086304, 1.4829000234603882, 1.482800006866455, 1.4812999963760376, 1.4809000492095947, 1.4808000326156616, 1.4807000160217285, 1.4805999994277954, 1.4805999994277954, 1.4802000522613525, 1.4802000522613525, 1.4794000387191772, 1.479200005531311, 1.479099988937378, 1.4789999723434448, 1.4788000583648682, 1.4782999753952026, 1.4780999422073364, 1.4779000282287598, 1.4778000116348267, 1.4774999618530273, 1.4763000011444092, 1.476199984550476, 1.476099967956543, 1.461300015449524, 1.4702999591827393, 1.44760000705719, 1.4349000453948975, 1.3912999629974365, 1.4258999824523926, 1.3003000020980835, 1.228600025177002, 1.1962000131607056, 0.9584000110626221, 1.0377999544143677, 0.460099995136261, 0.8313999772071838, 0.5468000173568726, 0.6517999768257141, 0.43160000443458557, -0.010400000028312206, 1.2148000001907349, 0.5040000081062317, 1.2896000146865845, 0.6744999885559082, 0.2741999924182892, 0.4772999882698059, 0.0869000032544136, 0.07810000330209732, -0.4702000021934509, -0.15549999475479126, 0.13689999282360077, 0.299699991941452, 0.30809998512268066, 1.510200023651123, 1.5096999406814575, 1.5092999935150146, 1.5092999935150146, 1.5092999935150146, 1.5091999769210815, 1.5085999965667725, 1.5085999965667725, 1.5084999799728394, 1.5082000494003296, 1.5080000162124634, 1.507599949836731, 1.5075000524520874, 1.5075000524520874, 1.507200002670288, 1.507200002670288, 1.5069999694824219, 1.5069999694824219, 1.5063999891281128, 1.506100058555603, 1.506100058555603, 1.5051000118255615, 1.5051000118255615, 1.5051000118255615, 1.504699945449829, 1.504699945449829, 1.5045000314712524, 1.5041999816894531, 1.5041999816894531, 1.50409996509552, 1.419700026512146, 1.4519000053405762, 1.3990000486373901, 1.2781000137329102, 1.4522000551223755, 1.2375999689102173, 1.1460000276565552, 0.7329000234603882, 1.236799955368042, 0.613099992275238, 0.7307999730110168, 0.22669999301433563, 0.4302000105381012, -0.11410000175237656, 0.6586999893188477, 0.8910999894142151, 0.325300008058548, 0.15389999747276306, 0.1923999935388565, 0.5041000247001648, 1.9990999698638916, 1.9946000576019287, 1.9940999746322632, 1.993899941444397, 1.993499994277954, 1.9930000305175781, 1.9925999641418457, 1.9924999475479126, 1.9910999536514282, 1.990399956703186, 1.9900000095367432, 1.989400029182434, 1.9881999492645264, 1.9881000518798828, 1.9879000186920166, 1.9869999885559082, 1.9853999614715576, 1.9852999448776245, 1.9851000308990479, 1.9848999977111816, 1.9844000339508057, 1.983299970626831, 1.9807000160217285, 1.979699969291687, 1.9789999723434448, 1.9789999723434448, 1.978600025177002, 1.978600025177002, 1.9780999422073364, 1.9768999814987183, 1.9386999607086182, 1.9642000198364258, 1.8986999988555908, 1.9220000505447388, 1.8954999446868896, 1.9182000160217285, 1.9057999849319458, 1.894700050354004, 1.6949000358581543, 1.6919000148773193, 1.6996999979019165, 1.6990000009536743, 1.451799988746643, 1.916200041770935, 1.7692999839782715, 1.6223000288009644, 0.7616000175476074, 1.4722000360488892, 0.21879999339580536, 0.28209999203681946, 0.9257000088691711, 1.1262999773025513, 0.5579000115394592, 0.7713000178337097, 1.1983000040054321, 0.05900000035762787, 0.1216999962925911, 1.1855000257492065, 0.22910000383853912, 0.4259999990463257, -0.8054999709129333, 0.508899986743927, 1.0664000511169434, -0.061400000005960464, 0.536899983882904, -0.33660000562667847, -0.3312999904155731, 2.205399990081787, 2.2030999660491943, 2.202500104904175, 2.2002999782562256, 2.1982998847961426, 2.1979000568389893, 2.1977999210357666, 2.197700023651123, 2.196500062942505, 2.195499897003174, 2.1951000690460205, 2.1949000358581543, 2.1947999000549316, 2.194700002670288, 2.1935999393463135, 2.1926000118255615, 2.1923000812530518, 2.191699981689453, 2.1905999183654785, 2.18969988822937, 2.188499927520752, 2.1881000995635986, 2.187700033187866, 2.1874001026153564, 2.186500072479248, 2.1861000061035156, 2.185499906539917, 2.18530011177063, 2.18530011177063, 2.184799909591675, 2.14739990234375, 2.1233999729156494, 2.1166000366210938, 2.100600004196167, 2.165299892425537, 1.8873000144958496, 1.7639000415802002, 2.134200096130371, 2.053999900817871, 2.1480000019073486, 1.5760999917984009, 1.5124000310897827, 1.5400999784469604, 1.6205999851226807, 0.6881999969482422, 1.3250000476837158, 0.9283000230789185, 1.4859999418258667, 1.4148999452590942, 0.6564000248908997, 0.9571999907493591, 0.7057999968528748, 1.5992000102996826, 1.2860000133514404, 1.6121000051498413, 1.273300051689148, 0.10159999877214432, 1.2858999967575073, -0.06620000302791595, 1.0592999458312988, 0.07010000199079514, 2.3529999256134033, 2.351900100708008, 2.348599910736084, 2.3485000133514404, 2.3466999530792236, 2.3452000617980957, 2.3443000316619873, 2.3422000408172607, 2.339200019836426, 2.3373000621795654, 2.336699962615967, 2.3352999687194824, 2.333199977874756, 2.3326001167297363, 2.331199884414673, 2.330199956893921, 2.330199956893921, 2.3297998905181885, 2.329200029373169, 2.3289999961853027, 2.3282999992370605, 2.3280999660491943, 2.327899932861328, 2.327399969100952, 2.327199935913086, 2.3269999027252197, 2.32669997215271, 2.3257999420166016, 2.3250999450683594, 2.324399948120117, 2.3173000812530518, 2.3008999824523926, 2.2421998977661133, 2.097899913787842, 2.160900115966797, 2.0869998931884766, 2.189300060272217, 2.104099988937378, 2.154099941253662, 2.2070999145507812, 1.942199945449829, 1.9586999416351318, 2.015199899673462, 1.9744000434875488, 1.6928000450134277, 1.9069000482559204, 2.01419997215271, 0.387800008058548, 0.8938999772071838, 1.5164999961853027, 0.927299976348877, 1.577299952507019, 0.4456999897956848, 0.4307999908924103, 0.6410999894142151, 0.5656999945640564, 1.0389000177383423, -0.06710000336170197, -0.10239999741315842, 2.434999942779541, 2.430500030517578, 2.4291999340057373, 2.428999900817871, 2.428499937057495, 2.4282000064849854, 2.427999973297119, 2.427799940109253, 2.426300048828125, 2.4256999492645264, 2.4244000911712646, 2.4235000610351562, 2.42330002784729, 2.421799898147583, 2.4209001064300537, 2.4205000400543213, 2.4202001094818115, 2.4191999435424805, 2.419100046157837, 2.4189000129699707, 2.418299913406372, 2.417799949645996, 2.4177000522613525, 2.41759991645813, 2.4175000190734863, 2.417099952697754, 2.4165000915527344, 2.415600061416626, 2.414799928665161, 2.414599895477295, 2.376800060272217, 2.3545000553131104, 2.215399980545044, 2.115600109100342, 2.3852999210357666, 1.7457000017166138, 1.868299961090088, 1.930400013923645, 1.9864000082015991, 1.8595999479293823, 2.2307000160217285, 1.9076000452041626, 1.6448999643325806, 1.6414999961853027, -0.10849999636411667, 1.573699951171875, 1.6642999649047852, -0.9089999794960022, -0.03180000185966492, -0.05810000002384186, -0.42289999127388, 0.17499999701976776, -0.6700999736785889, -0.32510000467300415, 2.5739998817443848, 2.5731000900268555, 2.5708999633789062, 2.5694000720977783, 2.568700075149536, 2.56850004196167, 2.567199945449829, 2.565200090408325, 2.5648999214172363, 2.5643999576568604, 2.563699960708618, 2.5634000301361084, 2.5632998943328857, 2.5631000995635986, 2.560499906539917, 2.560499906539917, 2.5571000576019287, 2.556299924850464, 2.554800033569336, 2.552999973297119, 2.5513999462127686, 2.5504000186920166, 2.5501999855041504, 2.5501999855041504, 2.5488998889923096, 2.5485000610351562, 2.546999931335449, 2.546999931335449, 2.5467000007629395, 2.5462000370025635, 2.4535999298095703, 2.4358999729156494, 2.3231000900268555, 1.9088000059127808, 1.8866000175476074, 2.2139999866485596, 1.1857999563217163, 1.3782999515533447, 1.929800033569336, 2.3889999389648438, 2.3046998977661133, 1.7919000387191772, 1.5923000574111938, 1.7288000583648682, 1.0394999980926514, 0.914900004863739, 1.1152000427246094, 0.9728000164031982, 1.3515000343322754, 1.4146000146865845, 1.2263000011444092, -0.2094999998807907, 1.25, 0.9689000248908997, 0.7333999872207642, -1.6848000288009644, 2.9639999866485596, 2.9542999267578125, 2.9540998935699463, 2.952399969100952, 2.9516000747680664, 2.950000047683716, 2.9498000144958496, 2.949199914932251, 2.944499969482422, 2.9444000720977783, 2.9440999031066895, 2.9439001083374023, 2.9435999393463135, 2.942199945449829, 2.93969988822937, 2.9375, 2.9358999729156494, 2.935499906539917, 2.9310998916625977, 2.930500030517578, 2.9296998977661133, 2.928800106048584, 2.9258999824523926, 2.925800085067749, 2.9244000911712646, 2.9233999252319336, 2.921099901199341, 2.920799970626831, 2.918600082397461, 2.9184999465942383, 2.848099946975708, 2.7767999172210693, 2.860100030899048, 2.4707999229431152, 2.517199993133545, 2.201200008392334, 2.4268999099731445, 2.6865999698638916, 1.1267999410629272, 2.528599977493286, 2.166599988937378, 1.8672000169754028, 1.392899990081787, 1.7452000379562378, 1.316499948501587, 0.45559999346733093, -0.475600004196167, 1.6648000478744507, 0.6359000205993652, 0.6729999780654907, -0.1915999948978424, 0.32030001282691956, 1.1548999547958374, 2.0404000282287598, -0.24060000479221344, -0.2433999925851822, 1.0085999965667725, -0.35659998655319214], \"Freq\": [1582.0, 2641.0, 3514.0, 928.0, 951.0, 600.0, 1005.0, 1252.0, 5199.0, 685.0, 685.0, 677.0, 441.0, 930.0, 1505.0, 633.0, 711.0, 708.0, 549.0, 712.0, 900.0, 373.0, 370.0, 1251.0, 825.0, 282.0, 686.0, 328.0, 644.0, 459.0, 456.44708251953125, 422.6692199707031, 270.18048095703125, 263.1275329589844, 255.72486877441406, 180.58419799804688, 178.2930450439453, 158.0100555419922, 156.0753173828125, 124.15323638916016, 117.6725082397461, 115.10094451904297, 114.20276641845703, 112.62008666992188, 112.55819702148438, 107.48247528076172, 107.20464324951172, 98.41825103759766, 96.47592163085938, 95.00994110107422, 94.28317260742188, 92.94609069824219, 88.13812255859375, 86.20417785644531, 84.71289825439453, 83.92172241210938, 81.96765899658203, 73.93376922607422, 75.84556579589844, 72.86703491210938, 978.3711547851562, 178.027587890625, 399.33514404296875, 206.09947204589844, 307.9021301269531, 148.69046020507812, 233.27349853515625, 222.03431701660156, 247.05093383789062, 529.531005859375, 365.1376647949219, 1256.2086181640625, 459.24072265625, 798.55810546875, 541.730712890625, 623.7011108398438, 1161.0771484375, 168.70785522460938, 467.88592529296875, 145.22152709960938, 268.6248474121094, 376.9248046875, 275.22772216796875, 308.4705810546875, 308.5801696777344, 372.4261169433594, 275.2220764160156, 228.58140563964844, 221.1990966796875, 213.82455444335938, 1581.3057861328125, 950.5226440429688, 684.2594604492188, 684.0845336914062, 676.895751953125, 632.5609130859375, 458.28167724609375, 455.2254943847656, 441.4902648925781, 387.9285888671875, 365.2989501953125, 316.91119384765625, 310.3322448730469, 303.9595031738281, 283.7093811035156, 282.5963439941406, 264.3009033203125, 262.5835876464844, 229.94888305664062, 215.52444458007812, 212.54803466796875, 177.3146209716797, 176.23455810546875, 176.42715454101562, 164.65054321289062, 164.1129608154297, 159.41041564941406, 152.1415252685547, 152.01100158691406, 150.63673400878906, 626.4192504882812, 269.7787780761719, 346.22314453125, 737.6402587890625, 218.62425231933594, 628.2315673828125, 504.4097900390625, 1614.40087890625, 304.68988037109375, 834.7457885742188, 689.9486694335938, 1439.8604736328125, 609.3335571289062, 520.1362915039062, 367.75433349609375, 326.4460144042969, 383.0073547363281, 366.8552551269531, 338.4144287109375, 323.86553955078125, 285.340087890625, 118.96763610839844, 112.40473937988281, 108.61347961425781, 103.95513153076172, 98.3161849975586, 95.07909393310547, 93.6440658569336, 82.18551635742188, 77.15977478027344, 74.53916931152344, 71.37539672851562, 65.41665649414062, 65.07293701171875, 64.226806640625, 60.27347946166992, 54.39042663574219, 54.47587966918945, 53.76478958129883, 52.883201599121094, 51.39409255981445, 48.583221435546875, 42.67408752441406, 41.14375686645508, 39.55536651611328, 39.398345947265625, 38.937198638916016, 38.73602294921875, 38.077003479003906, 36.19832229614258, 870.8729858398438, 74.58715057373047, 197.00807189941406, 113.26888275146484, 165.0495147705078, 74.13077545166016, 79.13074493408203, 86.46913146972656, 279.66961669921875, 163.9696502685547, 138.93763732910156, 136.61138916015625, 284.83245849609375, 66.84561157226562, 93.63043975830078, 111.58501434326172, 362.19891357421875, 134.24798583984375, 590.5133666992188, 472.7789306640625, 219.74208068847656, 174.80247497558594, 298.327392578125, 220.96267700195312, 144.11627197265625, 293.3326110839844, 273.7461242675781, 131.26661682128906, 215.56637573242188, 192.4339141845703, 313.7057800292969, 159.68743896484375, 132.99185180664062, 159.13153076171875, 140.06600952148438, 145.11953735351562, 138.1195831298828, 302.4176025390625, 172.29884338378906, 153.23602294921875, 111.96175384521484, 90.20709228515625, 86.57781982421875, 85.84425354003906, 85.00077056884766, 76.36560821533203, 70.77633666992188, 68.49613952636719, 67.63301849365234, 66.94385528564453, 66.27985382080078, 61.799354553222656, 57.658912658691406, 56.41118621826172, 54.43655014038086, 51.24308776855469, 48.76583480834961, 45.7819938659668, 44.80518341064453, 43.91083908081055, 43.304100036621094, 41.88007736206055, 40.845035552978516, 40.022186279296875, 39.43079376220703, 39.3573112487793, 38.8966178894043, 311.2357482910156, 174.1632080078125, 176.20291137695312, 167.8024139404297, 54.25968551635742, 513.631103515625, 456.7099609375, 56.975711822509766, 77.13182830810547, 52.51104736328125, 283.1136779785156, 274.01348876953125, 237.33154296875, 196.4511260986328, 1136.9134521484375, 287.86749267578125, 347.7868957519531, 144.6696014404297, 144.78114318847656, 301.7982177734375, 210.12460327148438, 200.28172302246094, 101.2824935913086, 119.85797882080078, 96.20818328857422, 107.16609191894531, 152.4297332763672, 103.78504943847656, 130.57394409179688, 100.5684814453125, 104.10369110107422, 599.9425048828125, 372.3772888183594, 157.47938537597656, 154.778076171875, 120.1961898803711, 100.91649627685547, 91.53434753417969, 76.25114440917969, 61.351478576660156, 54.459815979003906, 52.87966537475586, 48.851009368896484, 44.04594421386719, 42.832672119140625, 40.354244232177734, 38.58819580078125, 38.61812210083008, 37.87812042236328, 37.10505676269531, 36.80001449584961, 35.8582649230957, 35.47469711303711, 35.196861267089844, 34.72002029418945, 34.392425537109375, 34.09080123901367, 33.694576263427734, 47.196834564208984, 32.04069137573242, 31.084537506103516, 277.0146789550781, 418.17718505859375, 168.08901977539062, 273.6971435546875, 143.7442169189453, 212.48123168945312, 76.81979370117188, 106.88477325439453, 75.31197357177734, 58.29575729370117, 132.29132080078125, 120.93330383300781, 98.14289093017578, 108.72023010253906, 185.6692657470703, 109.77528381347656, 83.63662719726562, 727.4105834960938, 349.2926330566406, 138.48036193847656, 211.97335815429688, 98.68769836425781, 211.22225952148438, 182.71543884277344, 132.3478240966797, 116.39335632324219, 98.36388397216797, 112.28120422363281, 107.39253234863281, 250.02894592285156, 112.8718490600586, 96.42558288574219, 94.60336303710938, 90.29316711425781, 87.32403564453125, 85.79853820800781, 84.73868560791016, 73.98165130615234, 70.83614349365234, 64.37786102294922, 60.86716842651367, 59.842674255371094, 54.26689529418945, 51.7082633972168, 50.545475006103516, 49.69384765625, 47.064491271972656, 46.73114776611328, 46.485137939453125, 44.930030822753906, 44.02992630004883, 43.653541564941406, 43.470367431640625, 43.376346588134766, 42.47049331665039, 41.19907760620117, 39.58335876464844, 38.33219909667969, 37.932804107666016, 308.994873046875, 89.80464935302734, 238.39801025390625, 218.8181610107422, 48.67694854736328, 274.8531494140625, 154.32855224609375, 107.334228515625, 94.98731231689453, 104.48432922363281, 56.750396728515625, 75.8787841796875, 84.20931243896484, 76.00791931152344, 206.8007354736328, 77.96574401855469, 73.46817779541016, 182.84228515625, 107.3504638671875, 104.15181732177734, 102.6454086303711, 92.12925720214844, 91.46278381347656, 79.02147674560547, 92.49955749511719, 85.0216064453125, 70.45894622802734, 63.43061065673828, 60.36415100097656, 59.465118408203125, 55.05679702758789, 49.01042175292969, 48.344520568847656, 46.94021987915039, 45.49531555175781, 44.587135314941406, 44.658634185791016, 43.99930191040039, 39.22203063964844, 39.06112289428711, 34.19450759887695, 33.146175384521484, 31.4682674407959, 29.539281845092773, 28.109960556030273, 27.301162719726562, 27.016374588012695, 26.993574142456055, 26.074005126953125, 25.725568771362305, 24.7685604095459, 24.62936782836914, 24.42678451538086, 24.112293243408203, 119.5880126953125, 60.506954193115234, 76.36306762695312, 361.9794616699219, 200.51638793945312, 59.875938415527344, 652.4910888671875, 375.14727783203125, 92.08223724365234, 37.166046142578125, 42.6696662902832, 90.1404800415039, 111.27474212646484, 80.65431213378906, 161.52630615234375, 169.5735321044922, 131.97299194335938, 128.7685089111328, 86.63436126708984, 71.79528045654297, 74.05657196044922, 109.8899154663086, 58.920753479003906, 60.117366790771484, 59.77805709838867, 72.7863540649414, 91.74246978759766, 47.082950592041016, 46.53315353393555, 42.98588562011719, 41.36677551269531, 38.77008056640625, 38.430625915527344, 37.441097259521484, 31.492719650268555, 31.344507217407227, 31.0371150970459, 30.829124450683594, 30.86873435974121, 29.27400779724121, 27.084508895874023, 25.400606155395508, 24.299842834472656, 24.046524047851562, 21.812210083007812, 21.27958106994629, 20.866432189941406, 20.461084365844727, 19.19696617126465, 19.21242904663086, 18.61360740661621, 18.219070434570312, 17.457429885864258, 17.364736557006836, 16.638507843017578, 16.673946380615234, 178.9491424560547, 232.03663635253906, 29.914085388183594, 224.16165161132812, 78.2753677368164, 159.28158569335938, 65.52207946777344, 33.61476516723633, 416.3449401855469, 40.36647415161133, 60.71138381958008, 79.52865600585938, 132.6573486328125, 77.5784912109375, 74.25503540039062, 121.24981689453125, 165.0733642578125, 51.205081939697266, 73.01874542236328, 69.72908020019531, 75.71961212158203, 62.16162872314453, 48.8781623840332, 38.441795349121094, 57.219905853271484, 50.850563049316406, 44.23583221435547, 45.22863006591797], \"Total\": [1582.0, 2641.0, 3514.0, 928.0, 951.0, 600.0, 1005.0, 1252.0, 5199.0, 685.0, 685.0, 677.0, 441.0, 930.0, 1505.0, 633.0, 711.0, 708.0, 549.0, 712.0, 900.0, 373.0, 370.0, 1251.0, 825.0, 282.0, 686.0, 328.0, 644.0, 459.0, 457.37933349609375, 423.6015930175781, 271.11279296875, 264.06005859375, 256.65728759765625, 181.51634216308594, 179.2255859375, 158.9421844482422, 157.0115509033203, 125.08531188964844, 118.6048355102539, 116.03337097167969, 115.13581848144531, 113.55258178710938, 113.49040985107422, 108.41525268554688, 108.13752746582031, 99.3505859375, 97.40809631347656, 95.94281768798828, 95.2162857055664, 93.87837982177734, 89.07074737548828, 87.13668823242188, 85.64555358886719, 84.85403442382812, 82.9001693725586, 74.86641693115234, 76.8122787475586, 73.79936981201172, 1005.6795043945312, 181.3493194580078, 416.15594482421875, 217.52450561523438, 339.4469909667969, 158.3452606201172, 281.6565856933594, 288.0152282714844, 331.02203369140625, 900.0092163085938, 573.2565307617188, 3514.311767578125, 886.2278442382812, 2048.37255859375, 1251.1004638671875, 1795.3204345703125, 5199.6689453125, 221.9005126953125, 1252.6541748046875, 177.2368621826172, 606.4487915039062, 1269.8052978515625, 756.8284912109375, 1253.3514404296875, 1264.8369140625, 2641.262939453125, 1424.90185546875, 883.414794921875, 726.4698486328125, 696.3973999023438, 1582.3140869140625, 951.5311279296875, 685.2679443359375, 685.0929565429688, 677.904052734375, 633.5692749023438, 459.29058837890625, 456.2342834472656, 442.4986572265625, 388.9368896484375, 366.3073425292969, 317.91961669921875, 311.34051513671875, 304.9677734375, 284.7176513671875, 283.6046447753906, 265.3101806640625, 263.5927429199219, 230.95741271972656, 216.5338897705078, 213.55743408203125, 178.3245849609375, 177.24290466308594, 177.43666076660156, 165.6593475341797, 165.12217712402344, 160.41867065429688, 153.15150451660156, 153.0201873779297, 151.64495849609375, 686.1917724609375, 286.1389465332031, 387.1691589355469, 930.9391479492188, 231.826416015625, 825.6023559570312, 726.4698486328125, 3514.311767578125, 400.7324523925781, 2048.37255859375, 1505.117919921875, 5199.6689453125, 1795.3204345703125, 2641.262939453125, 862.1770629882812, 606.6712036132812, 1253.3514404296875, 1424.90185546875, 1264.8369140625, 886.2278442382812, 286.28973388671875, 119.90436553955078, 113.34159088134766, 109.55035400390625, 104.89210510253906, 99.25294494628906, 96.01641082763672, 94.58073425292969, 83.12189483642578, 78.09596252441406, 75.47599029541016, 72.31177520751953, 66.35358428955078, 66.01083374023438, 65.16510009765625, 61.2102165222168, 55.32817459106445, 55.418609619140625, 54.70465850830078, 53.81957244873047, 52.330989837646484, 49.520755767822266, 43.61012268066406, 42.09079360961914, 40.49352264404297, 40.334861755371094, 39.8749885559082, 39.67242431640625, 39.015846252441406, 37.134925842285156, 928.1788940429688, 77.49349975585938, 218.5455322265625, 122.75719451904297, 183.68222045898438, 80.64881896972656, 87.15731048583984, 96.30621337890625, 380.3868408203125, 223.6748046875, 188.06640625, 185.04803466796875, 494.0015563964844, 72.86298370361328, 118.21269989013672, 163.19371032714844, 1252.6541748046875, 228.13629150390625, 3514.311767578125, 2641.262939453125, 644.9544067382812, 419.82330322265625, 1264.8369140625, 756.8284912109375, 322.0610046386719, 2048.37255859375, 1795.3204345703125, 297.1234130859375, 1269.8052978515625, 930.9391479492188, 5199.6689453125, 711.0617065429688, 339.1213073730469, 1253.3514404296875, 606.4487915039062, 1505.117919921875, 1424.90185546875, 303.3396301269531, 173.22042846679688, 154.15786743164062, 112.88372802734375, 91.1297378540039, 87.4992904663086, 86.76693725585938, 85.92318725585938, 77.28720092773438, 71.69811248779297, 69.4207763671875, 68.55459594726562, 67.86564636230469, 67.2013931274414, 62.722900390625, 58.58037185668945, 57.332740783691406, 55.35919189453125, 52.16697692871094, 49.68918228149414, 46.70414733886719, 45.72733688354492, 44.833126068115234, 44.22745895385742, 42.81138610839844, 41.76689910888672, 40.95124816894531, 40.354095458984375, 40.28046417236328, 39.8282585144043, 330.81719970703125, 189.61912536621094, 193.1516876220703, 186.906494140625, 56.65061569213867, 708.1416625976562, 712.3955688476562, 61.36930465698242, 90.01610565185547, 55.78618240356445, 532.8253784179688, 549.6365356445312, 463.05255126953125, 353.6370849609375, 5199.6689453125, 696.3973999023438, 1251.1004638671875, 297.9324035644531, 320.1597900390625, 1424.90185546875, 734.3587036132812, 900.0092163085938, 186.2613983154297, 301.49920654296875, 174.67205810546875, 273.0220947265625, 1253.3514404296875, 261.0838928222656, 1269.8052978515625, 317.3393249511719, 883.414794921875, 600.909912109375, 373.4014587402344, 158.43421936035156, 155.73226928710938, 121.15161895751953, 101.87083435058594, 92.48873901367188, 77.20526123046875, 62.305843353271484, 55.41435623168945, 53.83697509765625, 49.80549240112305, 45.00261688232422, 43.78893280029297, 41.3095703125, 39.54220199584961, 39.573753356933594, 38.83241653442383, 38.059940338134766, 37.75603103637695, 36.813514709472656, 36.430267333984375, 36.150882720947266, 35.677101135253906, 35.35050964355469, 35.04569625854492, 34.650325775146484, 48.57927703857422, 33.00046157836914, 32.03877258300781, 287.56512451171875, 441.26739501953125, 188.0878143310547, 353.8190002441406, 174.4732208251953, 277.69781494140625, 90.63375854492188, 137.32284545898438, 92.03118896484375, 67.56385040283203, 199.81788635253906, 179.68441772460938, 137.80467224121094, 159.0205078125, 359.9056091308594, 171.77865600585938, 117.56205749511719, 5199.6689453125, 1505.117919921875, 320.1597900390625, 883.414794921875, 214.7196044921875, 1424.90185546875, 1251.1004638671875, 734.3587036132812, 696.3973999023438, 366.640380859375, 1264.8369140625, 1253.3514404296875, 250.9574737548828, 113.80252075195312, 97.35413360595703, 95.53416442871094, 91.22149658203125, 88.25253295898438, 86.72709655761719, 85.67168426513672, 74.91043853759766, 71.76455688476562, 65.30741882324219, 61.80376052856445, 60.77190017700195, 55.19513702392578, 52.638763427734375, 51.47868728637695, 50.62293243408203, 47.99403762817383, 47.659324645996094, 47.41596984863281, 45.860443115234375, 44.961036682128906, 44.58183288574219, 44.39848709106445, 44.306278228759766, 43.398983001708984, 42.12736129760742, 40.51176834106445, 39.260398864746094, 38.86110305786133, 328.74053955078125, 97.69898986816406, 298.0646057128906, 302.2984924316406, 51.350643157958984, 549.6365356445312, 273.0220947265625, 178.4433135986328, 149.32427978515625, 186.4639129638672, 69.87448120117188, 129.0570068359375, 186.2613983154297, 168.68785095214844, 2641.262939453125, 185.1858673095703, 159.38998413085938, 5199.6689453125, 1269.8052978515625, 1264.8369140625, 1795.3204345703125, 886.2278442382812, 2048.37255859375, 1253.3514404296875, 93.43070983886719, 85.94961547851562, 71.3877182006836, 64.35914611816406, 61.293094635009766, 60.39425277709961, 55.985931396484375, 49.9399299621582, 49.27533721923828, 47.869998931884766, 46.42631912231445, 45.51579666137695, 45.591224670410156, 44.92922592163086, 40.1522331237793, 39.9897346496582, 35.12689208984375, 34.077762603759766, 32.40003967285156, 30.467792510986328, 29.03940200805664, 28.232248306274414, 27.945478439331055, 27.922313690185547, 27.00381088256836, 26.654035568237305, 25.701221466064453, 25.558006286621094, 25.3548641204834, 25.040842056274414, 136.2421417236328, 70.16490936279297, 99.12599182128906, 711.0617065429688, 402.7399597167969, 86.68538665771484, 2641.262939453125, 1252.6541748046875, 177.12149047851562, 45.167789459228516, 56.41517639160156, 199.02325439453125, 299.9581604003906, 189.68312072753906, 756.8284912109375, 900.0092163085938, 573.2565307617188, 644.9544067382812, 297.1234130859375, 231.17904663085938, 287.8569641113281, 1795.3204345703125, 223.6748046875, 302.2984924316406, 380.3868408203125, 5199.6689453125, 92.69585418701172, 48.0367431640625, 47.485443115234375, 43.9381217956543, 42.31917953491211, 39.724971771240234, 39.38618087768555, 38.39368438720703, 32.44508361816406, 32.296844482421875, 31.9896297454834, 31.781356811523438, 31.831165313720703, 30.230968475341797, 28.038345336914062, 26.35347557067871, 25.25249481201172, 24.999080657958984, 22.775789260864258, 22.232614517211914, 21.818735122680664, 21.413808822631836, 20.149925231933594, 20.166915893554688, 19.565893173217773, 19.171466827392578, 18.411344528198242, 18.320140838623047, 17.592815399169922, 17.631092071533203, 203.03607177734375, 282.7008361816406, 33.533714294433594, 370.8861389160156, 123.64100646972656, 345.09405517578125, 113.27407836914062, 44.82144546508789, 2641.262939453125, 63.03571319580078, 136.16493225097656, 240.615966796875, 644.9544067382812, 265.17041015625, 389.6830139160156, 1505.117919921875, 5199.6689453125, 189.68312072753906, 756.8284912109375, 696.3973999023438, 1795.3204345703125, 883.414794921875, 301.49920654296875, 97.81465148925781, 1424.90185546875, 1269.8052978515625, 315.8619689941406, 1264.8369140625], \"logprob\": [30.0, 29.0, 28.0, 27.0, 26.0, 25.0, 24.0, 23.0, 22.0, 21.0, 20.0, 19.0, 18.0, 17.0, 16.0, 15.0, 14.0, 13.0, 12.0, 11.0, 10.0, 9.0, 8.0, 7.0, 6.0, 5.0, 4.0, 3.0, 2.0, 1.0, -4.456500053405762, -4.533400058746338, -4.980899810791016, -5.007299900054932, -5.035900115966797, -5.383800029754639, -5.396500110626221, -5.517300128936768, -5.529600143432617, -5.758399963378906, -5.812099933624268, -5.834199905395508, -5.8420000076293945, -5.855899810791016, -5.856500148773193, -5.902599811553955, -5.905200004577637, -5.990699768066406, -6.010700225830078, -6.026000022888184, -6.033699989318848, -6.047900199890137, -6.101099967956543, -6.123199939727783, -6.140699863433838, -6.150100231170654, -6.173600196838379, -6.276800155639648, -6.251299858093262, -6.291299819946289, -3.6940999031066895, -5.3979997634887695, -4.590199947357178, -5.2515997886657715, -4.850200176239014, -5.578100204467773, -5.127799987792969, -5.17710018157959, -5.070400238037109, -4.308000087738037, -4.679699897766113, -3.4440999031066895, -4.450399875640869, -3.897200107574463, -4.285200119018555, -4.1442999839782715, -3.522900104522705, -5.4517998695373535, -4.431700229644775, -5.6016998291015625, -4.986599922180176, -4.647900104522705, -4.962399959564209, -4.848299980163574, -4.8480000495910645, -4.659900188446045, -4.962399959564209, -5.148099899291992, -5.1809000968933105, -5.214799880981445, -3.191999912261963, -3.7009999752044678, -4.029699802398682, -4.029900074005127, -4.040500164031982, -4.1082000732421875, -4.430500030517578, -4.43720006942749, -4.467800140380859, -4.5971999168396, -4.657299995422363, -4.7993998527526855, -4.820400238037109, -4.841100215911865, -4.909999847412109, -4.914000034332275, -4.980899810791016, -4.987400054931641, -5.120100021362305, -5.184899806976318, -5.198800086975098, -5.380099773406982, -5.386199951171875, -5.3850998878479, -5.45419979095459, -5.457399845123291, -5.486499786376953, -5.533199787139893, -5.533999919891357, -5.543099880218506, -4.118000030517578, -4.960400104522705, -4.710899829864502, -3.9544999599456787, -5.170599937438965, -4.115099906921387, -4.33459997177124, -3.171299934387207, -4.838699817657471, -3.830899953842163, -4.021399974822998, -3.2857000827789307, -4.145599842071533, -4.303899765014648, -4.650599956512451, -4.769700050354004, -4.609899997711182, -4.6529998779296875, -4.733699798583984, -4.777699947357178, -4.412700176239014, -5.287499904632568, -5.344200134277344, -5.378499984741211, -5.422399997711182, -5.4781999588012695, -5.511600017547607, -5.526800155639648, -5.657400131225586, -5.7204999923706055, -5.755000114440918, -5.798399925231934, -5.8856000900268555, -5.8907999992370605, -5.903900146484375, -5.967400074005127, -6.070199966430664, -6.068600177764893, -6.081699848175049, -6.098299980163574, -6.126800060272217, -6.18310022354126, -6.31279993057251, -6.349299907684326, -6.388599872589111, -6.392600059509277, -6.404399871826172, -6.409599781036377, -6.426700115203857, -6.47730016708374, -3.296799898147583, -5.75439977645874, -4.783100128173828, -5.336599826812744, -4.960100173950195, -5.760499954223633, -5.695199966430664, -5.606599807739258, -4.432700157165527, -4.966700077056885, -5.132299900054932, -5.149199962615967, -4.414400100708008, -5.863999843597412, -5.5269999504089355, -5.351600170135498, -4.174099922180176, -5.1666998863220215, -3.68530011177063, -3.9077000617980957, -4.673900127410889, -4.902699947357178, -4.368199825286865, -4.668300151824951, -5.095699787139893, -4.385000228881836, -4.454100131988525, -5.1890997886657715, -4.6930999755859375, -4.806600093841553, -4.31790018081665, -4.993100166320801, -5.17609977722168, -4.996600151062012, -5.124199867248535, -5.088799953460693, -5.138199806213379, -4.148499965667725, -4.711100101470947, -4.8282999992370605, -5.142199993133545, -5.3582000732421875, -5.3993000984191895, -5.407800197601318, -5.417699813842773, -5.524799823760986, -5.600800037384033, -5.633500099182129, -5.646200180053711, -5.656499862670898, -5.666399955749512, -5.736400127410889, -5.805799961090088, -5.827700138092041, -5.86329984664917, -5.923699855804443, -5.973299980163574, -6.036399841308594, -6.058000087738037, -6.078199863433838, -6.092100143432617, -6.125500202178955, -6.1504998207092285, -6.170899868011475, -6.185800075531006, -6.187600135803223, -6.199399948120117, -4.119800090789795, -4.700300216674805, -4.688700199127197, -4.737500190734863, -5.866499900817871, -3.618799924850464, -3.736299991607666, -5.817699909210205, -5.514800071716309, -5.8993000984191895, -4.2144999504089355, -4.247099876403809, -4.390900135040283, -4.579899787902832, -2.8243000507354736, -4.197800159454346, -4.008699893951416, -4.885900020599365, -4.8850998878479, -4.150599956512451, -4.512599945068359, -4.5605998039245605, -5.242400169372559, -5.073999881744385, -5.293799877166748, -5.1859002113342285, -4.833600044250488, -5.2179999351501465, -4.988399982452393, -5.249499797821045, -5.214900016784668, -3.3173000812530518, -3.7941999435424805, -4.654900074005127, -4.6722002029418945, -4.925000190734863, -5.099899768829346, -5.197400093078613, -5.380099773406982, -5.597499847412109, -5.716700077056885, -5.746099948883057, -5.825399875640869, -5.928899765014648, -5.9567999839782715, -6.016499996185303, -6.061200141906738, -6.060400009155273, -6.079800128936768, -6.100399971008301, -6.108699798583984, -6.1346001625061035, -6.145299911499023, -6.153200149536133, -6.166800022125244, -6.176300048828125, -6.185100078582764, -6.196800231933594, -5.859799861907959, -6.247099876403809, -6.277400016784668, -4.090099811553955, -3.6782000064849854, -4.589700222015381, -4.102099895477295, -4.746099948883057, -4.355299949645996, -5.372700214385986, -5.042399883270264, -5.392499923706055, -5.648600101470947, -4.829100131988525, -4.918900012969971, -5.127699851989746, -5.025400161743164, -4.490200042724609, -5.015699863433838, -5.287700176239014, -3.1247000694274902, -3.8582000732421875, -4.783400058746338, -4.357699871063232, -5.122200012207031, -4.361199855804443, -4.506199836730957, -4.828700065612793, -4.957200050354004, -5.125500202178955, -4.993100166320801, -5.037700176239014, -4.108500003814697, -4.903800010681152, -5.061299800872803, -5.0802998542785645, -5.126999855041504, -5.160399913787842, -5.177999973297119, -5.190499782562256, -5.326200008392334, -5.369699954986572, -5.4653000831604, -5.521299839019775, -5.538300037384033, -5.636099815368652, -5.6844000816345215, -5.707200050354004, -5.724100112915039, -5.778500080108643, -5.785600185394287, -5.790900230407715, -5.824900150299072, -5.845200061798096, -5.853700160980225, -5.857999801635742, -5.860099792480469, -5.881199836730957, -5.911600112915039, -5.951600074768066, -5.983699798583984, -5.994200229644775, -3.896699905395508, -5.132400035858154, -4.156099796295166, -4.241799831390381, -5.744800090789795, -4.013800144195557, -4.59089994430542, -4.954100131988525, -5.076300144195557, -4.980999946594238, -5.591400146484375, -5.300899982452393, -5.196700096130371, -5.299200057983398, -4.298299789428711, -5.273799896240234, -5.333199977874756, -4.42140007019043, -4.95389986038208, -4.9842000007629395, -4.998700141906738, -5.106800079345703, -5.114099979400635, -5.260300159454346, -4.957600116729736, -5.041900157928467, -5.229800224304199, -5.33489990234375, -5.384399890899658, -5.399400234222412, -5.476399898529053, -5.592800140380859, -5.606500148773193, -5.635900020599365, -5.667200088500977, -5.687399864196777, -5.685800075531006, -5.7006001472473145, -5.8155999183654785, -5.819699764251709, -5.952700138092041, -5.98390007019043, -6.035799980163574, -6.099100112915039, -6.14870023727417, -6.1778998374938965, -6.188399791717529, -6.189199924468994, -6.223899841308594, -6.237299919128418, -6.275199890136719, -6.280900001525879, -6.289100170135498, -6.30210018157959, -4.700799942016602, -5.381999969482422, -5.1493000984191895, -3.5931999683380127, -4.183899879455566, -5.392499923706055, -3.003999948501587, -3.557499885559082, -4.962100028991699, -5.8694000244140625, -5.731299877166748, -4.983399868011475, -4.772799968719482, -5.094600200653076, -4.400100231170654, -4.351500034332275, -4.602200031280518, -4.626800060272217, -5.023099899291992, -5.210999965667725, -5.179999828338623, -4.785299777984619, -5.408599853515625, -5.388500213623047, -5.394199848175049, -5.197299957275391, -4.575500011444092, -5.242499828338623, -5.254300117492676, -5.333600044250488, -5.372000217437744, -5.436800003051758, -5.4456000328063965, -5.471700191497803, -5.644700050354004, -5.649400234222412, -5.659299850463867, -5.665999889373779, -5.664700031280518, -5.717800140380859, -5.795499801635742, -5.8597002029418945, -5.9039998054504395, -5.9145002365112305, -6.01200008392334, -6.0366997718811035, -6.056300163269043, -6.075900077819824, -6.139699935913086, -6.138899803161621, -6.170599937438965, -6.191999912261963, -6.2347002029418945, -6.239999771118164, -6.282700061798096, -6.280600070953369, -3.907399892807007, -3.647599935531616, -5.696100234985352, -3.6821000576019287, -4.7342000007629395, -4.023799896240234, -4.912099838256836, -5.579500198364258, -3.0629000663757324, -5.396500110626221, -4.98829984664917, -4.718299865722656, -4.206699848175049, -4.743199825286865, -4.786900043487549, -4.296599864959717, -3.988100051879883, -5.158599853515625, -4.803699970245361, -4.849800109863281, -4.767399787902832, -4.964700222015381, -5.205100059509277, -5.445300102233887, -5.047599792480469, -5.165599822998047, -5.304900169372559, -5.282700061798096]}};\n",
       "\n",
       "function LDAvis_load_lib(url, callback){\n",
       "  var s = document.createElement('script');\n",
       "  s.src = url;\n",
       "  s.async = true;\n",
       "  s.onreadystatechange = s.onload = callback;\n",
       "  s.onerror = function(){console.warn(\"failed to load library \" + url);};\n",
       "  document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "}\n",
       "\n",
       "if(typeof(LDAvis) !== \"undefined\"){\n",
       "   // already loaded: just create the visualization\n",
       "   !function(LDAvis){\n",
       "       new LDAvis(\"#\" + \"ldavis_el14441209944222246259283215\", ldavis_el14441209944222246259283215_data);\n",
       "   }(LDAvis);\n",
       "}else if(typeof define === \"function\" && define.amd){\n",
       "   // require.js is available: use it to load d3/LDAvis\n",
       "   require.config({paths: {d3: \"https://cdnjs.cloudflare.com/ajax/libs/d3/3.5.5/d3.min\"}});\n",
       "   require([\"d3\"], function(d3){\n",
       "      window.d3 = d3;\n",
       "      LDAvis_load_lib(\"https://cdn.rawgit.com/bmabey/pyLDAvis/files/ldavis.v1.0.0.js\", function(){\n",
       "        new LDAvis(\"#\" + \"ldavis_el14441209944222246259283215\", ldavis_el14441209944222246259283215_data);\n",
       "      });\n",
       "    });\n",
       "}else{\n",
       "    // require.js not available: dynamically load d3 & LDAvis\n",
       "    LDAvis_load_lib(\"https://cdnjs.cloudflare.com/ajax/libs/d3/3.5.5/d3.min.js\", function(){\n",
       "         LDAvis_load_lib(\"https://cdn.rawgit.com/bmabey/pyLDAvis/files/ldavis.v1.0.0.js\", function(){\n",
       "                 new LDAvis(\"#\" + \"ldavis_el14441209944222246259283215\", ldavis_el14441209944222246259283215_data);\n",
       "            })\n",
       "         });\n",
       "}\n",
       "</script>"
      ],
      "text/plain": [
       "PreparedData(topic_coordinates=            Freq  cluster  topics         x         y\n",
       "topic                                                \n",
       "5      22.563847        1       1 -0.103305 -0.046505\n",
       "6      22.073502        1       2 -0.324344  0.038233\n",
       "3      13.500492        1       3 -0.051974 -0.160353\n",
       "4      10.986977        1       4  0.123373  0.216254\n",
       "7       9.492788        1       5  0.021019  0.216678\n",
       "2       8.726937        1       6  0.111715 -0.015366\n",
       "0       7.547307        1       7  0.130647 -0.188901\n",
       "1       5.108154        1       8  0.092869 -0.060039, topic_info=      Category         Freq          Term        Total  loglift  logprob\n",
       "term                                                                    \n",
       "2447   Default  1582.000000          loan  1582.000000  30.0000  30.0000\n",
       "4860   Default  2641.000000       product  2641.000000  29.0000  29.0000\n",
       "357    Default  3514.000000       service  3514.000000  28.0000  28.0000\n",
       "3281   Default   928.000000       segment   928.000000  27.0000  27.0000\n",
       "10349  Default   951.000000       account   951.000000  26.0000  26.0000\n",
       "10199  Default   600.000000      maryland   600.000000  25.0000  25.0000\n",
       "7441   Default  1005.000000      solution  1005.000000  24.0000  24.0000\n",
       "3207   Default  1252.000000        system  1252.000000  23.0000  23.0000\n",
       "11988  Default  5199.000000       company  5199.000000  22.0000  22.0000\n",
       "11974  Default   685.000000         check   685.000000  21.0000  21.0000\n",
       "8535   Default   685.000000       deposit   685.000000  20.0000  20.0000\n",
       "2371   Default   677.000000       banking   677.000000  19.0000  19.0000\n",
       "14642  Default   441.000000           own   441.000000  18.0000  18.0000\n",
       "6566   Default   930.000000    commercial   930.000000  17.0000  17.0000\n",
       "7612   Default  1505.000000       operate  1505.000000  16.0000  16.0000\n",
       "7507   Default   633.000000        credit   633.000000  15.0000  15.0000\n",
       "13969  Default   711.000000           use   711.000000  14.0000  14.0000\n",
       "5303   Default   708.000000       develop   708.000000  13.0000  13.0000\n",
       "5366   Default   549.000000     treatment   549.000000  12.0000  12.0000\n",
       "9901   Default   712.000000   development   712.000000  11.0000  11.0000\n",
       "1557   Default   900.000000    technology   900.000000  10.0000  10.0000\n",
       "12525  Default   373.000000      delmarva   373.000000   9.0000   9.0000\n",
       "11003  Default   370.000000         brand   370.000000   8.0000   8.0000\n",
       "5888   Default  1251.000000          base  1251.000000   7.0000   7.0000\n",
       "5441   Default   825.000000      business   825.000000   6.0000   6.0000\n",
       "13306  Default   282.000000         store   282.000000   5.0000   5.0000\n",
       "2728   Default   686.000000          line   686.000000   4.0000   4.0000\n",
       "4994   Default   328.000000          care   328.000000   3.0000   3.0000\n",
       "2272   Default   644.000000          sell   644.000000   2.0000   2.0000\n",
       "274    Default   459.000000     financial   459.000000   1.0000   1.0000\n",
       "...        ...          ...           ...          ...      ...      ...\n",
       "6714    Topic8    16.638508  rhode_island    17.592815   2.9186  -6.2827\n",
       "7308    Topic8    16.673946           nut    17.631092   2.9185  -6.2806\n",
       "5818    Topic8   178.949142          food   203.036072   2.8481  -3.9074\n",
       "13306   Topic8   232.036636         store   282.700836   2.7768  -3.6476\n",
       "12965   Topic8    29.914085         dairy    33.533714   2.8601  -5.6961\n",
       "11003   Topic8   224.161652         brand   370.886139   2.4708  -3.6821\n",
       "10069   Topic8    78.275368      retailer   123.641006   2.5172  -4.7342\n",
       "10541   Topic8   159.281586        retail   345.094055   2.2012  -4.0238\n",
       "8514    Topic8    65.522079     accessory   113.274078   2.4269  -4.9121\n",
       "8335    Topic8    33.614765          item    44.821445   2.6866  -5.5795\n",
       "4860    Topic8   416.344940       product  2641.262939   1.1268  -3.0629\n",
       "10202   Topic8    40.366474    ingredient    63.035713   2.5286  -5.3965\n",
       "4308    Topic8    60.711384           com   136.164932   2.1666  -4.9883\n",
       "6401    Topic8    79.528656     specialty   240.615967   1.8672  -4.7183\n",
       "2272    Topic8   132.657349          sell   644.954407   1.3929  -4.2067\n",
       "692     Topic8    77.578491    distribute   265.170410   1.7452  -4.7432\n",
       "7828    Topic8    74.255035      consumer   389.683014   1.3165  -4.7869\n",
       "7612    Topic8   121.249817       operate  1505.117920   0.4556  -4.2966\n",
       "11988   Topic8   165.073364       company  5199.668945  -0.4756  -3.9881\n",
       "12312   Topic8    51.205082   distributor   189.683121   1.6648  -5.1586\n",
       "4779    Topic8    73.018745        market   756.828491   0.6359  -4.8037\n",
       "12295   Topic8    69.729080          name   696.397400   0.6730  -4.8498\n",
       "835     Topic8    75.719612         offer  1795.320435  -0.1916  -4.7674\n",
       "9809    Topic8    62.161629  united_state   883.414795   0.3203  -4.9647\n",
       "4620    Topic8    48.878162       holding   301.499207   1.1549  -5.2051\n",
       "12874   Topic8    38.441795     wholesale    97.814651   2.0404  -5.4453\n",
       "2142    Topic8    57.219906         found  1424.901855  -0.2406  -5.0476\n",
       "13867   Topic8    50.850563       include  1269.805298  -0.2434  -5.1656\n",
       "7220    Topic8    44.235832          home   315.861969   1.0086  -5.3049\n",
       "7140    Topic8    45.228630          well  1264.836914  -0.3566  -5.2827\n",
       "\n",
       "[495 rows x 6 columns], token_table=       Topic      Freq         Term\n",
       "term                               \n",
       "10253      1  0.940982       access\n",
       "10253      2  0.056838       access\n",
       "8514       1  0.123594    accessory\n",
       "8514       3  0.291329    accessory\n",
       "8514       8  0.582658    accessory\n",
       "10349      2  0.999442      account\n",
       "4181       5  0.849573      acquire\n",
       "4181       6  0.143434      acquire\n",
       "8909       5  0.968164         acre\n",
       "12763      6  0.968501        acute\n",
       "7055       1  0.229119     addition\n",
       "7055       2  0.537359     addition\n",
       "7055       3  0.082417     addition\n",
       "7055       5  0.026373     addition\n",
       "7055       6  0.085714     addition\n",
       "7055       7  0.036263     addition\n",
       "5681       3  0.989414  advertising\n",
       "2607       3  0.984772    aerospace\n",
       "4365       3  0.986503          air\n",
       "9794       3  0.986010     aircraft\n",
       "2280       1  0.245741         also\n",
       "2280       2  0.305581         also\n",
       "2280       3  0.126860         also\n",
       "2280       4  0.121275         also\n",
       "2280       5  0.085371         also\n",
       "2280       6  0.063031         also\n",
       "2280       7  0.031117         also\n",
       "2280       8  0.019947         also\n",
       "9947       4  0.972247  alternative\n",
       "5112       4  0.977630      america\n",
       "...      ...       ...          ...\n",
       "13481      8  0.031316      various\n",
       "9781       6  0.989346     vascular\n",
       "6611       2  0.096090      vehicle\n",
       "6611       3  0.901414      vehicle\n",
       "6146       7  0.971303       vessel\n",
       "2887       1  0.997155        video\n",
       "4372       1  0.985544        voice\n",
       "3458       6  0.992948        waste\n",
       "11078      5  0.970162   wastewater\n",
       "12110      3  0.101017        water\n",
       "12110      5  0.893200        water\n",
       "3569       1  0.991323          web\n",
       "7972       1  0.989481      website\n",
       "7140       1  0.244300         well\n",
       "7140       2  0.267228         well\n",
       "7140       3  0.235603         well\n",
       "7140       4  0.022928         well\n",
       "7140       5  0.088549         well\n",
       "7140       6  0.082224         well\n",
       "7140       7  0.022928         well\n",
       "7140       8  0.035578         well\n",
       "1505       5  0.985502         west\n",
       "12874      1  0.337373    wholesale\n",
       "12874      2  0.040894    wholesale\n",
       "12874      5  0.214692    wholesale\n",
       "12874      8  0.388490    wholesale\n",
       "215        7  0.966969         wind\n",
       "216        8  0.989777         wine\n",
       "11936      1  0.994900     wireless\n",
       "1634       8  0.962967        world\n",
       "\n",
       "[730 rows x 3 columns], R=30, lambda_step=0.01, plot_opts={'xlab': 'PC1', 'ylab': 'PC2'}, topic_order=[6, 7, 4, 5, 8, 3, 1, 2])"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Visualize the topics\n",
    "pyLDAvis.enable_notebook()\n",
    "vis = pyLDAvis.gensim.prepare(lda_model, corpus, id2word)\n",
    "vis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Building LDA Mallet Mode(normally gives better quality of topics)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "mallet_path = '/Users/onion8/Downloads/utilities/mallet-2.0.8/bin/mallet' # update this path\n",
    "ldamallet = gensim.models.wrappers.LdaMallet(mallet_path, corpus=corpus, num_topics=8, id2word=id2word)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(0,\n",
      "  [(u'product', 0.07619203670326069),\n",
      "   (u'company', 0.03774100169315637),\n",
      "   (u'brand', 0.023868043038942598),\n",
      "   (u'sell', 0.020973291823693264),\n",
      "   (u'operate', 0.017805450871156262),\n",
      "   (u'store', 0.014637609918619258),\n",
      "   (u'retail', 0.013053689442350757),\n",
      "   (u'food', 0.012999071494893221),\n",
      "   (u'offer', 0.012398274072860342),\n",
      "   (u'include', 0.012179802283030203)]),\n",
      " (1,\n",
      "  [(u'loan', 0.05333805699889781),\n",
      "   (u'service', 0.03593922216973705),\n",
      "   (u'commercial', 0.03424657534246575),\n",
      "   (u'business', 0.03212092583845064),\n",
      "   (u'banking', 0.025350338529365454),\n",
      "   (u'account', 0.02440560541646985),\n",
      "   (u'credit', 0.02039048968666352),\n",
      "   (u'company', 0.020114942528735632),\n",
      "   (u'deposit', 0.018934026137616123),\n",
      "   (u'real_estate', 0.0182648401826484)]),\n",
      " (2,\n",
      "  [(u'company', 0.03403281561906617),\n",
      "   (u'develop', 0.02916402940746872),\n",
      "   (u'treatment', 0.028677150786308975),\n",
      "   (u'development', 0.01864745119041823),\n",
      "   (u'pharmaceutical', 0.015336676566531964),\n",
      "   (u'patient', 0.013194410633429085),\n",
      "   (u'disease', 0.011246896148790106),\n",
      "   (u'therapeutic', 0.011003456838210234),\n",
      "   (u'clinical', 0.009737572423194897),\n",
      "   (u'focus', 0.009494133112615025)]),\n",
      " (3,\n",
      "  [(u'service', 0.13409747933255567),\n",
      "   (u'provide', 0.04853679565856875),\n",
      "   (u'management', 0.041588476948825885),\n",
      "   (u'offer', 0.030836334127909926),\n",
      "   (u'company', 0.020236344271440887),\n",
      "   (u'segment', 0.018207638078815237),\n",
      "   (u'include', 0.017852614495105747),\n",
      "   (u'health', 0.017497590911396257),\n",
      "   (u'care', 0.016128214231373943),\n",
      "   (u'customer', 0.012222954810569559)]),\n",
      " (4,\n",
      "  [(u'company', 0.08282278193782619),\n",
      "   (u'base', 0.03074653959609712),\n",
      "   (u'corporation', 0.03000907646925346),\n",
      "   (u'found', 0.024903562514181982),\n",
      "   (u'group', 0.021670070342636714),\n",
      "   (u'insurance', 0.021386430678466076),\n",
      "   (u'new_york', 0.018436578171091445),\n",
      "   (u'holding', 0.017699115044247787),\n",
      "   (u'sale', 0.014749262536873156),\n",
      "   (u'product', 0.014068527342863625)]),\n",
      " (5,\n",
      "  [(u'product', 0.036437768240343345),\n",
      "   (u'system', 0.03489270386266095),\n",
      "   (u'segment', 0.019699570815450644),\n",
      "   (u'equipment', 0.017467811158798283),\n",
      "   (u'industrial', 0.01613733905579399),\n",
      "   (u'manufacture', 0.013261802575107296),\n",
      "   (u'market', 0.01296137339055794),\n",
      "   (u'sell', 0.010987124463519314),\n",
      "   (u'power', 0.010772532188841201),\n",
      "   (u'control', 0.010085836909871245)]),\n",
      " (6,\n",
      "  [(u'company', 0.05475195542697953),\n",
      "   (u'oil', 0.019982856530590377),\n",
      "   (u'base', 0.019982856530590377),\n",
      "   (u'energy', 0.019554269795349834),\n",
      "   (u'formerly_known', 0.01800064288010286),\n",
      "   (u'corporation', 0.017250616093431908),\n",
      "   (u'corp', 0.016768456016286296),\n",
      "   (u'found', 0.016018429229615345),\n",
      "   (u'operate', 0.015696989178184934),\n",
      "   (u'engage', 0.015696989178184934)]),\n",
      " (7,\n",
      "  [(u'solution', 0.03494316981450966),\n",
      "   (u'technology', 0.0232391093083196),\n",
      "   (u'software', 0.019056069632822074),\n",
      "   (u'network', 0.018887057928761566),\n",
      "   (u'system', 0.017323699666201883),\n",
      "   (u'application', 0.01491528288333967),\n",
      "   (u'provide', 0.013647695102885875),\n",
      "   (u'datum', 0.013098407064689229),\n",
      "   (u'digital', 0.012168842692356446),\n",
      "   (u'platform', 0.012084336840326193)])]\n",
      "(u'\\nCoherence Score: ', 0.5168848885376561)\n"
     ]
    }
   ],
   "source": [
    "pprint(ldamallet.show_topics(formatted=False))\n",
    "\n",
    "# Compute Coherence Score\n",
    "coherence_model_ldamallet = CoherenceModel(model=ldamallet, texts=data_lemmatized, dictionary=id2word, coherence='c_v')\n",
    "coherence_ldamallet = coherence_model_ldamallet.get_coherence()\n",
    "print('\\nCoherence Score: ', coherence_ldamallet)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Finding optimal number of topics for LDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_coherence_values(dictionary, corpus, texts, limit, start=2, step=3):\n",
    "    \"\"\"\n",
    "    Compute c_v coherence for various number of topics\n",
    "\n",
    "    Parameters:\n",
    "    ----------\n",
    "    dictionary : Gensim dictionary\n",
    "    corpus : Gensim corpus\n",
    "    texts : List of input texts\n",
    "    limit : Max num of topics\n",
    "\n",
    "    Returns:\n",
    "    -------\n",
    "    model_list : List of LDA topic models\n",
    "    coherence_values : Coherence values corresponding to the LDA model with respective number of topics\n",
    "    \"\"\"\n",
    "    coherence_values = []\n",
    "    model_list = []\n",
    "    for num_topics in range(start, limit, step):\n",
    "        model = gensim.models.wrappers.LdaMallet(mallet_path, corpus=corpus, num_topics=num_topics, id2word=id2word)\n",
    "        model_list.append(model)\n",
    "        coherencemodel = CoherenceModel(model=model, texts=texts, dictionary=dictionary, coherence='c_v')\n",
    "        coherence_values.append(coherencemodel.get_coherence())\n",
    "\n",
    "    return model_list, coherence_values\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Can take a long time to run.\n",
    "model_list, coherence_values = compute_coherence_values(dictionary=id2word, corpus=corpus, texts=data_lemmatized, start=2, limit=40, step=6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1c2bc7fd90>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "limit=40; start=2; step=6;\n",
    "x = range(start, limit, step)\n",
    "plt.plot(x, coherence_values)\n",
    "plt.xlabel(\"Num Topics\")\n",
    "plt.ylabel(\"Coherence score\")\n",
    "plt.legend((\"coherence_values\"), loc='best')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(u'Num Topics =', 2, u' has Coherence Value of', 0.4163)\n",
      "(u'Num Topics =', 8, u' has Coherence Value of', 0.5477)\n",
      "(u'Num Topics =', 14, u' has Coherence Value of', 0.5267)\n",
      "(u'Num Topics =', 20, u' has Coherence Value of', 0.5131)\n",
      "(u'Num Topics =', 26, u' has Coherence Value of', 0.5203)\n",
      "(u'Num Topics =', 32, u' has Coherence Value of', 0.5325)\n",
      "(u'Num Topics =', 38, u' has Coherence Value of', 0.5262)\n"
     ]
    }
   ],
   "source": [
    "for m, cv in zip(x, coherence_values):\n",
    "    print(\"Num Topics =\", m, \" has Coherence Value of\", round(cv, 4))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Selecting the optimal number of topics by their coherence score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[(0,\n",
      "  u'0.038*\"product\" + 0.036*\"system\" + 0.020*\"segment\" + 0.017*\"equipment\" + 0.014*\"industrial\" + 0.013*\"offer\" + 0.012*\"manufacture\" + 0.012*\"market\" + 0.011*\"control\" + 0.011*\"power\"'),\n",
      " (1,\n",
      "  u'0.030*\"company\" + 0.030*\"treatment\" + 0.027*\"develop\" + 0.016*\"pharmaceutical\" + 0.015*\"development\" + 0.014*\"patient\" + 0.012*\"disease\" + 0.011*\"therapeutic\" + 0.010*\"research\" + 0.010*\"clinical\"'),\n",
      " (2,\n",
      "  u'0.086*\"product\" + 0.030*\"company\" + 0.023*\"brand\" + 0.021*\"sell\" + 0.020*\"offer\" + 0.020*\"operate\" + 0.019*\"market\" + 0.014*\"store\" + 0.014*\"include\" + 0.014*\"retail\"'),\n",
      " (3,\n",
      "  u'0.112*\"service\" + 0.044*\"provide\" + 0.036*\"management\" + 0.028*\"offer\" + 0.021*\"segment\" + 0.016*\"insurance\" + 0.014*\"include\" + 0.014*\"health\" + 0.013*\"care\" + 0.010*\"company\"'),\n",
      " (4,\n",
      "  u'0.120*\"company\" + 0.071*\"base\" + 0.034*\"change\" + 0.032*\"corporation\" + 0.031*\"found\" + 0.030*\"technology\" + 0.029*\"formerly_known\" + 0.024*\"development\" + 0.020*\"holding\" + 0.020*\"california\"'),\n",
      " (5,\n",
      "  u'0.047*\"company\" + 0.025*\"operate\" + 0.023*\"energy\" + 0.021*\"oil\" + 0.018*\"found\" + 0.016*\"gas\" + 0.016*\"water\" + 0.014*\"united_state\" + 0.013*\"property\" + 0.012*\"customer\"'),\n",
      " (6,\n",
      "  u'0.053*\"loan\" + 0.034*\"commercial\" + 0.033*\"service\" + 0.032*\"business\" + 0.025*\"banking\" + 0.024*\"account\" + 0.024*\"company\" + 0.020*\"credit\" + 0.019*\"deposit\" + 0.018*\"real_estate\"'),\n",
      " (7,\n",
      "  u'0.038*\"solution\" + 0.021*\"provide\" + 0.019*\"software\" + 0.018*\"network\" + 0.016*\"service\" + 0.016*\"system\" + 0.014*\"application\" + 0.013*\"datum\" + 0.012*\"platform\" + 0.012*\"offer\"')]\n"
     ]
    }
   ],
   "source": [
    "optimal_model = model_list[1]\n",
    "model_topics = optimal_model.show_topics(formatted=False)\n",
    "pprint(optimal_model.print_topics(num_words=10))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Finding the dominant topic in each sentence (depends on the distribution of keyword specific to that topic)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "scrolled": false
   },
   "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>Document_No</th>\n",
       "      <th>Dominant_Topic</th>\n",
       "      <th>Topic_Perc_Contrib</th>\n",
       "      <th>Keywords</th>\n",
       "      <th>Text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.2027</td>\n",
       "      <td>company, base, change, corporation, found, tec...</td>\n",
       "      <td>024 Pharma, Inc. provides healthcare products ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.5647</td>\n",
       "      <td>product, company, brand, sell, offer, operate,...</td>\n",
       "      <td>1-800-Flowers.com, Inc., together with its sub...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.3254</td>\n",
       "      <td>solution, provide, software, network, service,...</td>\n",
       "      <td>12 Retech Corporation, through its subsidiarie...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.2461</td>\n",
       "      <td>service, provide, management, offer, segment, ...</td>\n",
       "      <td>1347 Property Insurance Holdings, Inc., throug...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.1903</td>\n",
       "      <td>company, operate, energy, oil, found, gas, wat...</td>\n",
       "      <td>1847 Holdings LLC, through its subsidiaries, p...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.1793</td>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>1867 Western Financial Corporation, through it...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.6013</td>\n",
       "      <td>solution, provide, software, network, service,...</td>\n",
       "      <td>1mage Software Inc operates in the technology ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.1806</td>\n",
       "      <td>company, base, change, corporation, found, tec...</td>\n",
       "      <td>1PM Industries, Inc. provides consulting servi...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>8</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.5858</td>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>1st Capital Bank provides various banking prod...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>9</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.5887</td>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>1st Colonial Bancorp, Inc. operates as the ban...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>10</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.6415</td>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>1st Constitution Bancorp operates as the bank ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>11</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.4064</td>\n",
       "      <td>company, operate, energy, oil, found, gas, wat...</td>\n",
       "      <td>1st NRG Corp. explores, develops, produces, an...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>12</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.1976</td>\n",
       "      <td>service, provide, management, offer, segment, ...</td>\n",
       "      <td>1st Prestige Wealth Management provides wealth...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>13</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.4877</td>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>1st Source Corporation operates as the bank ho...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>14</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.5513</td>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>1ST SUMMIT BANCORP of Johnstown, Inc. operates...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Document_No  Dominant_Topic  Topic_Perc_Contrib  \\\n",
       "0             0             4.0              0.2027   \n",
       "1             1             2.0              0.5647   \n",
       "2             2             7.0              0.3254   \n",
       "3             3             3.0              0.2461   \n",
       "4             4             5.0              0.1903   \n",
       "5             5             6.0              0.1793   \n",
       "6             6             7.0              0.6013   \n",
       "7             7             4.0              0.1806   \n",
       "8             8             6.0              0.5858   \n",
       "9             9             6.0              0.5887   \n",
       "10           10             6.0              0.6415   \n",
       "11           11             5.0              0.4064   \n",
       "12           12             3.0              0.1976   \n",
       "13           13             6.0              0.4877   \n",
       "14           14             6.0              0.5513   \n",
       "\n",
       "                                             Keywords  \\\n",
       "0   company, base, change, corporation, found, tec...   \n",
       "1   product, company, brand, sell, offer, operate,...   \n",
       "2   solution, provide, software, network, service,...   \n",
       "3   service, provide, management, offer, segment, ...   \n",
       "4   company, operate, energy, oil, found, gas, wat...   \n",
       "5   loan, commercial, service, business, banking, ...   \n",
       "6   solution, provide, software, network, service,...   \n",
       "7   company, base, change, corporation, found, tec...   \n",
       "8   loan, commercial, service, business, banking, ...   \n",
       "9   loan, commercial, service, business, banking, ...   \n",
       "10  loan, commercial, service, business, banking, ...   \n",
       "11  company, operate, energy, oil, found, gas, wat...   \n",
       "12  service, provide, management, offer, segment, ...   \n",
       "13  loan, commercial, service, business, banking, ...   \n",
       "14  loan, commercial, service, business, banking, ...   \n",
       "\n",
       "                                                 Text  \n",
       "0   024 Pharma, Inc. provides healthcare products ...  \n",
       "1   1-800-Flowers.com, Inc., together with its sub...  \n",
       "2   12 Retech Corporation, through its subsidiarie...  \n",
       "3   1347 Property Insurance Holdings, Inc., throug...  \n",
       "4   1847 Holdings LLC, through its subsidiaries, p...  \n",
       "5   1867 Western Financial Corporation, through it...  \n",
       "6   1mage Software Inc operates in the technology ...  \n",
       "7   1PM Industries, Inc. provides consulting servi...  \n",
       "8   1st Capital Bank provides various banking prod...  \n",
       "9   1st Colonial Bancorp, Inc. operates as the ban...  \n",
       "10  1st Constitution Bancorp operates as the bank ...  \n",
       "11  1st NRG Corp. explores, develops, produces, an...  \n",
       "12  1st Prestige Wealth Management provides wealth...  \n",
       "13  1st Source Corporation operates as the bank ho...  \n",
       "14  1ST SUMMIT BANCORP of Johnstown, Inc. operates...  "
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def format_topics_sentences(ldamodel=lda_model, corpus=corpus, texts=data):\n",
    "    # Init output\n",
    "    sent_topics_df = pd.DataFrame()\n",
    "\n",
    "    # Get main topic in each document\n",
    "    for i, row in enumerate(ldamodel[corpus]):\n",
    "        row = sorted(row, key=lambda x: (x[1]), reverse=True)\n",
    "        # Get the Dominant topic, Perc Contribution and Keywords for each document\n",
    "        for j, (topic_num, prop_topic) in enumerate(row):\n",
    "            if j == 0:  # => dominant topic\n",
    "                wp = ldamodel.show_topic(topic_num)\n",
    "                topic_keywords = \", \".join([word for word, prop in wp])\n",
    "                sent_topics_df = sent_topics_df.append(pd.Series([int(topic_num), round(prop_topic,4), topic_keywords]), ignore_index=True)\n",
    "            else:\n",
    "                break\n",
    "    sent_topics_df.columns = ['Dominant_Topic', 'Perc_Contribution', 'Topic_Keywords']\n",
    "\n",
    "    # Add original text to the end of the output\n",
    "    contents = pd.Series(texts)\n",
    "    sent_topics_df = pd.concat([sent_topics_df, contents], axis=1)\n",
    "    return(sent_topics_df)\n",
    "\n",
    "\n",
    "df_topic_sents_keywords = format_topics_sentences(ldamodel=optimal_model, corpus=corpus, texts=data)\n",
    "\n",
    "# Format\n",
    "df_dominant_topic = df_topic_sents_keywords.reset_index()\n",
    "df_dominant_topic.columns = ['Document_No', 'Dominant_Topic', 'Topic_Perc_Contrib', 'Keywords', 'Text']\n",
    "\n",
    "# Show\n",
    "df_dominant_topic.head(15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_dominant_topic['act_label'] = rawdata['label']\n",
    "df_dominant_topic['desc'] = rawdata['Business Description']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "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>Document_No</th>\n",
       "      <th>Dominant_Topic</th>\n",
       "      <th>Topic_Perc_Contrib</th>\n",
       "      <th>Keywords</th>\n",
       "      <th>Text</th>\n",
       "      <th>act_label</th>\n",
       "      <th>desc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1992</th>\n",
       "      <td>1992</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.3976</td>\n",
       "      <td>solution, provide, software, network, service,...</td>\n",
       "      <td>Defentect Group, Inc. develops and provides IP...</td>\n",
       "      <td>Information Technology</td>\n",
       "      <td>Defentect Group, Inc. develops and provides IP...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1993</th>\n",
       "      <td>1993</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.3238</td>\n",
       "      <td>product, company, brand, sell, offer, operate,...</td>\n",
       "      <td>Del Friscos Restaurant Group, Inc. develops, o...</td>\n",
       "      <td>Consumer Discretionary</td>\n",
       "      <td>Del Frisco's Restaurant Group, Inc. develops, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1994</th>\n",
       "      <td>1994</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.3102</td>\n",
       "      <td>product, company, brand, sell, offer, operate,...</td>\n",
       "      <td>Del Taco Restaurants, Inc. develops, franchise...</td>\n",
       "      <td>Consumer Discretionary</td>\n",
       "      <td>Del Taco Restaurants, Inc. develops, franchise...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1995</th>\n",
       "      <td>1995</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.3596</td>\n",
       "      <td>company, treatment, develop, pharmaceutical, d...</td>\n",
       "      <td>Delcath Systems, Inc., an interventional oncol...</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Delcath Systems, Inc., an interventional oncol...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1996</th>\n",
       "      <td>1996</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.4592</td>\n",
       "      <td>company, operate, energy, oil, found, gas, wat...</td>\n",
       "      <td>Delek US Holdings, Inc. engages in the integra...</td>\n",
       "      <td>Energy</td>\n",
       "      <td>Delek US Holdings, Inc. engages in the integra...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1997</th>\n",
       "      <td>1997</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.5784</td>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>Delhi Bank Corp. operates as a bank holding co...</td>\n",
       "      <td>Banks</td>\n",
       "      <td>Delhi Bank Corp. operates as a bank holding co...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1998</th>\n",
       "      <td>1998</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.4014</td>\n",
       "      <td>product, company, brand, sell, offer, operate,...</td>\n",
       "      <td>dELiA*s, Inc. operates as a multi-channel reta...</td>\n",
       "      <td>Catalog Apparel and Accessory Retail</td>\n",
       "      <td>dELiA*s, Inc. operates as a multi-channel reta...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999</th>\n",
       "      <td>1999</td>\n",
       "      <td>7.0</td>\n",
       "      <td>0.4168</td>\n",
       "      <td>solution, provide, software, network, service,...</td>\n",
       "      <td>Delivery Technology Solutions, Inc. offers res...</td>\n",
       "      <td>Commercial and Professional Services</td>\n",
       "      <td>Delivery Technology Solutions, Inc. offers res...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000</th>\n",
       "      <td>2000</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.6051</td>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>Delmar Bancorp operates as the bank holding co...</td>\n",
       "      <td>Banks</td>\n",
       "      <td>Delmar Bancorp operates as the bank holding co...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2001</th>\n",
       "      <td>2001</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.5140</td>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>Delmarva Bancshares, Inc. operates as the bank...</td>\n",
       "      <td>Banks</td>\n",
       "      <td>Delmarva Bancshares, Inc. operates as the bank...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Document_No  Dominant_Topic  Topic_Perc_Contrib  \\\n",
       "1992         1992             7.0              0.3976   \n",
       "1993         1993             2.0              0.3238   \n",
       "1994         1994             2.0              0.3102   \n",
       "1995         1995             1.0              0.3596   \n",
       "1996         1996             5.0              0.4592   \n",
       "1997         1997             6.0              0.5784   \n",
       "1998         1998             2.0              0.4014   \n",
       "1999         1999             7.0              0.4168   \n",
       "2000         2000             6.0              0.6051   \n",
       "2001         2001             6.0              0.5140   \n",
       "\n",
       "                                               Keywords  \\\n",
       "1992  solution, provide, software, network, service,...   \n",
       "1993  product, company, brand, sell, offer, operate,...   \n",
       "1994  product, company, brand, sell, offer, operate,...   \n",
       "1995  company, treatment, develop, pharmaceutical, d...   \n",
       "1996  company, operate, energy, oil, found, gas, wat...   \n",
       "1997  loan, commercial, service, business, banking, ...   \n",
       "1998  product, company, brand, sell, offer, operate,...   \n",
       "1999  solution, provide, software, network, service,...   \n",
       "2000  loan, commercial, service, business, banking, ...   \n",
       "2001  loan, commercial, service, business, banking, ...   \n",
       "\n",
       "                                                   Text  \\\n",
       "1992  Defentect Group, Inc. develops and provides IP...   \n",
       "1993  Del Friscos Restaurant Group, Inc. develops, o...   \n",
       "1994  Del Taco Restaurants, Inc. develops, franchise...   \n",
       "1995  Delcath Systems, Inc., an interventional oncol...   \n",
       "1996  Delek US Holdings, Inc. engages in the integra...   \n",
       "1997  Delhi Bank Corp. operates as a bank holding co...   \n",
       "1998  dELiA*s, Inc. operates as a multi-channel reta...   \n",
       "1999  Delivery Technology Solutions, Inc. offers res...   \n",
       "2000  Delmar Bancorp operates as the bank holding co...   \n",
       "2001  Delmarva Bancshares, Inc. operates as the bank...   \n",
       "\n",
       "                                  act_label  \\\n",
       "1992                Information Technology    \n",
       "1993                Consumer Discretionary    \n",
       "1994                Consumer Discretionary    \n",
       "1995                            Healthcare    \n",
       "1996                                Energy    \n",
       "1997                                 Banks    \n",
       "1998  Catalog Apparel and Accessory Retail    \n",
       "1999  Commercial and Professional Services    \n",
       "2000                                 Banks    \n",
       "2001                                 Banks    \n",
       "\n",
       "                                                   desc  \n",
       "1992  Defentect Group, Inc. develops and provides IP...  \n",
       "1993  Del Frisco's Restaurant Group, Inc. develops, ...  \n",
       "1994  Del Taco Restaurants, Inc. develops, franchise...  \n",
       "1995  Delcath Systems, Inc., an interventional oncol...  \n",
       "1996  Delek US Holdings, Inc. engages in the integra...  \n",
       "1997  Delhi Bank Corp. operates as a bank holding co...  \n",
       "1998  dELiA*s, Inc. operates as a multi-channel reta...  \n",
       "1999  Delivery Technology Solutions, Inc. offers res...  \n",
       "2000  Delmar Bancorp operates as the bank holding co...  \n",
       "2001  Delmarva Bancshares, Inc. operates as the bank...  "
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_dominant_topic.tail(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {
    "scrolled": false
   },
   "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>Keywords</th>\n",
       "      <th>act_label</th>\n",
       "      <th>desc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>company, base, change, corporation, found, tec...</td>\n",
       "      <td>Beauty Care Products</td>\n",
       "      <td>024 Pharma, Inc. provides healthcare products ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>product, company, brand, sell, offer, operate,...</td>\n",
       "      <td>Catalog Flowers, Gifts and Novelties</td>\n",
       "      <td>1-800-Flowers.com, Inc., together with its sub...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>solution, provide, software, network, service,...</td>\n",
       "      <td>Information Technology</td>\n",
       "      <td>12 Retech Corporation, through its subsidiarie...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>service, provide, management, offer, segment, ...</td>\n",
       "      <td>Casualty</td>\n",
       "      <td>1347 Property Insurance Holdings, Inc., throug...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>company, operate, energy, oil, found, gas, wat...</td>\n",
       "      <td>Commercial and Professional Services</td>\n",
       "      <td>1847 Holdings LLC, through its subsidiaries, p...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>Banks</td>\n",
       "      <td>1867 Western Financial Corporation, through it...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>solution, provide, software, network, service,...</td>\n",
       "      <td>Application Software</td>\n",
       "      <td>1mage Software Inc operates in the technology ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>company, base, change, corporation, found, tec...</td>\n",
       "      <td>Commercial and Professional Services</td>\n",
       "      <td>1PM Industries, Inc. provides consulting servi...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>Banks</td>\n",
       "      <td>1st Capital Bank provides various banking prod...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>Banks</td>\n",
       "      <td>1st Colonial Bancorp, Inc. operates as the ban...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>Banks</td>\n",
       "      <td>1st Constitution Bancorp operates as the bank ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>company, operate, energy, oil, found, gas, wat...</td>\n",
       "      <td>Energy</td>\n",
       "      <td>1st NRG Corp. explores, develops, produces, an...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>service, provide, management, offer, segment, ...</td>\n",
       "      <td>Asset Management and Custody Banks</td>\n",
       "      <td>1st Prestige Wealth Management provides wealth...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>Banks</td>\n",
       "      <td>1st Source Corporation operates as the bank ho...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>Banks</td>\n",
       "      <td>1ST SUMMIT BANCORP of Johnstown, Inc. operates...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                             Keywords  \\\n",
       "0   company, base, change, corporation, found, tec...   \n",
       "1   product, company, brand, sell, offer, operate,...   \n",
       "2   solution, provide, software, network, service,...   \n",
       "3   service, provide, management, offer, segment, ...   \n",
       "4   company, operate, energy, oil, found, gas, wat...   \n",
       "5   loan, commercial, service, business, banking, ...   \n",
       "6   solution, provide, software, network, service,...   \n",
       "7   company, base, change, corporation, found, tec...   \n",
       "8   loan, commercial, service, business, banking, ...   \n",
       "9   loan, commercial, service, business, banking, ...   \n",
       "10  loan, commercial, service, business, banking, ...   \n",
       "11  company, operate, energy, oil, found, gas, wat...   \n",
       "12  service, provide, management, offer, segment, ...   \n",
       "13  loan, commercial, service, business, banking, ...   \n",
       "14  loan, commercial, service, business, banking, ...   \n",
       "\n",
       "                                act_label  \\\n",
       "0                   Beauty Care Products    \n",
       "1   Catalog Flowers, Gifts and Novelties    \n",
       "2                 Information Technology    \n",
       "3                               Casualty    \n",
       "4   Commercial and Professional Services    \n",
       "5                                  Banks    \n",
       "6                   Application Software    \n",
       "7   Commercial and Professional Services    \n",
       "8                                  Banks    \n",
       "9                                  Banks    \n",
       "10                                 Banks    \n",
       "11                                Energy    \n",
       "12    Asset Management and Custody Banks    \n",
       "13                                 Banks    \n",
       "14                                 Banks    \n",
       "\n",
       "                                                 desc  \n",
       "0   024 Pharma, Inc. provides healthcare products ...  \n",
       "1   1-800-Flowers.com, Inc., together with its sub...  \n",
       "2   12 Retech Corporation, through its subsidiarie...  \n",
       "3   1347 Property Insurance Holdings, Inc., throug...  \n",
       "4   1847 Holdings LLC, through its subsidiaries, p...  \n",
       "5   1867 Western Financial Corporation, through it...  \n",
       "6   1mage Software Inc operates in the technology ...  \n",
       "7   1PM Industries, Inc. provides consulting servi...  \n",
       "8   1st Capital Bank provides various banking prod...  \n",
       "9   1st Colonial Bancorp, Inc. operates as the ban...  \n",
       "10  1st Constitution Bancorp operates as the bank ...  \n",
       "11  1st NRG Corp. explores, develops, produces, an...  \n",
       "12  1st Prestige Wealth Management provides wealth...  \n",
       "13  1st Source Corporation operates as the bank ho...  \n",
       "14  1ST SUMMIT BANCORP of Johnstown, Inc. operates...  "
      ]
     },
     "execution_count": 158,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_dominant_topic[['Keywords','act_label','desc']].head(15)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Finding the most representative document for each topic (which sentences represents a topic best way)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "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>Topic_Num</th>\n",
       "      <th>Topic_Perc_Contrib</th>\n",
       "      <th>Keywords</th>\n",
       "      <th>Text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.6481</td>\n",
       "      <td>product, system, segment, equipment, industria...</td>\n",
       "      <td>Dana Incorporated provides drive and motion pr...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.7103</td>\n",
       "      <td>company, treatment, develop, pharmaceutical, d...</td>\n",
       "      <td>Agios Pharmaceuticals, Inc., a biopharmaceutic...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.0</td>\n",
       "      <td>0.6421</td>\n",
       "      <td>product, company, brand, sell, offer, operate,...</td>\n",
       "      <td>Carters, Inc., together with its subsidiaries,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3.0</td>\n",
       "      <td>0.6299</td>\n",
       "      <td>service, provide, management, offer, segment, ...</td>\n",
       "      <td>Cross Country Healthcare, Inc. provides health...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.0</td>\n",
       "      <td>0.3892</td>\n",
       "      <td>company, base, change, corporation, found, tec...</td>\n",
       "      <td>Buscar Company focuses on buying, breeding, ra...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>5.0</td>\n",
       "      <td>0.6386</td>\n",
       "      <td>company, operate, energy, oil, found, gas, wat...</td>\n",
       "      <td>Andeavor Logistics LP operates as a diversifie...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>6.0</td>\n",
       "      <td>0.6892</td>\n",
       "      <td>loan, commercial, service, business, banking, ...</td>\n",
       "      <td>Citizens First Corporation operates as the ban...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>7.0</td>\n",
       "      <td>0.6761</td>\n",
       "      <td>solution, provide, software, network, service,...</td>\n",
       "      <td>Akamai Technologies, Inc. provides cloud servi...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Topic_Num  Topic_Perc_Contrib  \\\n",
       "0        0.0              0.6481   \n",
       "1        1.0              0.7103   \n",
       "2        2.0              0.6421   \n",
       "3        3.0              0.6299   \n",
       "4        4.0              0.3892   \n",
       "5        5.0              0.6386   \n",
       "6        6.0              0.6892   \n",
       "7        7.0              0.6761   \n",
       "\n",
       "                                            Keywords  \\\n",
       "0  product, system, segment, equipment, industria...   \n",
       "1  company, treatment, develop, pharmaceutical, d...   \n",
       "2  product, company, brand, sell, offer, operate,...   \n",
       "3  service, provide, management, offer, segment, ...   \n",
       "4  company, base, change, corporation, found, tec...   \n",
       "5  company, operate, energy, oil, found, gas, wat...   \n",
       "6  loan, commercial, service, business, banking, ...   \n",
       "7  solution, provide, software, network, service,...   \n",
       "\n",
       "                                                Text  \n",
       "0  Dana Incorporated provides drive and motion pr...  \n",
       "1  Agios Pharmaceuticals, Inc., a biopharmaceutic...  \n",
       "2  Carters, Inc., together with its subsidiaries,...  \n",
       "3  Cross Country Healthcare, Inc. provides health...  \n",
       "4  Buscar Company focuses on buying, breeding, ra...  \n",
       "5  Andeavor Logistics LP operates as a diversifie...  \n",
       "6  Citizens First Corporation operates as the ban...  \n",
       "7  Akamai Technologies, Inc. provides cloud servi...  "
      ]
     },
     "execution_count": 117,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "sent_topics_sorteddf_mallet = pd.DataFrame()\n",
    "\n",
    "sent_topics_outdf_grpd = df_topic_sents_keywords.groupby('Dominant_Topic')\n",
    "\n",
    "for i, grp in sent_topics_outdf_grpd:\n",
    "    sent_topics_sorteddf_mallet = pd.concat([sent_topics_sorteddf_mallet, \n",
    "                                             grp.sort_values(['Perc_Contribution'], ascending=[0]).head(1)], \n",
    "                                            axis=0)\n",
    "\n",
    "# Reset Index    \n",
    "sent_topics_sorteddf_mallet.reset_index(drop=True, inplace=True)\n",
    "\n",
    "# Format\n",
    "sent_topics_sorteddf_mallet.columns = ['Topic_Num', \"Topic_Perc_Contrib\", \"Keywords\", \"Text\"]\n",
    "\n",
    "# Show\n",
    "sent_topics_sorteddf_mallet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_labels=['Components manufacturer','Biotechnology and healthcare','Facilities','Management firm','Consumer Discretionary','Petroleum','Petroleum','Banking and loans','SW solutions']\n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "keywrds=[]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(len(sent_topics_sorteddf_mallet)):\n",
    "    keywrds.append(str(sent_topics_sorteddf_mallet['Keywords'][i]))\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['product, system, segment, equipment, industrial, offer, manufacture, market, control, power',\n",
       " 'company, treatment, develop, pharmaceutical, development, patient, disease, therapeutic, research, clinical',\n",
       " 'product, company, brand, sell, offer, operate, market, store, include, retail',\n",
       " 'service, provide, management, offer, segment, insurance, include, health, care, company',\n",
       " 'company, base, change, corporation, found, technology, formerly_known, development, holding, california',\n",
       " 'company, operate, energy, oil, found, gas, water, united_state, property, customer',\n",
       " 'loan, commercial, service, business, banking, account, company, credit, deposit, real_estate',\n",
       " 'solution, provide, software, network, service, system, application, datum, platform, offer']"
      ]
     },
     "execution_count": 128,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "keywrds"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}