ibm-recommender/LDA.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import re, nltk, gensim"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.decomposition import LatentDirichletAllocation, TruncatedSVD\n",
    "from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from pprint import pprint"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pyLDAvis\n",
    "import pyLDAvis.sklearn\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['Tip to deal with Fear :We’re all scared of something.  Do you get heart palpitations at the idea of speaking in front of big groups at work? Are you worried you’ll never actually learn to swim? Do you stare at the ceiling, thinking you’ll never write that book you’ve been dreaming about for years?  Me, too. I’ve deeply felt all three of those exact fears, along with many others. But those fears are just a few I’ve started overcoming using a little happiness hack I can share with you.  Are you ready?  O.K., to start with, here is the thought process most of us follow when it comes to facing our fears:  CAN DO IT —> WANT TO DO IT —> DO IT  Before you do anything, you have to feel like you can do it first—and then you have to actually want to do it second.']\n"
     ]
    }
   ],
   "source": [
    "\n",
    "df=pd.read_excel(\"C:\\\\Users\\\\Amitabh Tiwari\\\\ibm-hack\\\\articles.xlsx\")\n",
    "data=[]\n",
    "\n",
    "for v in df['content']:\n",
    "    data.append(v)\n",
    "print(data[3:4])\n",
    "#len(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "#word_tokenizing\n",
    "\n",
    "def sent_to_words(sentences):\n",
    "    for sentence in sentences:\n",
    "        yield(gensim.utils.simple_preprocess(str(sentence), deacc=True))\n",
    "        \n",
    "data_words = list(sent_to_words(data))\n",
    "\n",
    "#print(data_words[:1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "#stop words removal : removing stop words of language=english\n",
    "#stemming : getting the root words\n",
    "\n",
    "from nltk.stem.porter import PorterStemmer\n",
    "from stop_words import get_stop_words\n",
    "\n",
    "p_stemmer = PorterStemmer()\n",
    "en_stop = get_stop_words('en')\n",
    "\n",
    "data_lemmatized = []\n",
    "\n",
    "for i in data_words:\n",
    "    tokens = i\n",
    "    stopped_tokens = [i for i in tokens if not i in en_stop]\n",
    "    stemmed_tokens = [p_stemmer.stem(i) for i in stopped_tokens]\n",
    "    \n",
    "    data_lemmatized.append(' '.join(stemmed_tokens))\n",
    "    \n",
    "#print(data_lemmatized[:1])\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "vectorizer = CountVectorizer(analyzer='word',       \n",
    "                             #min_df=10,                        # minimum reqd occurences of a word \n",
    "                             stop_words='english',             # remove stop words\n",
    "                             lowercase=True,                   # convert all words to lowercase\n",
    "                             token_pattern='[a-zA-Z0-9]{3,}',  # num chars > 3\n",
    "                             # max_features=30000,             # max number of uniq words\n",
    "                            )\n",
    "\n",
    "data_vectorized = vectorizer.fit_transform(data_lemmatized)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sparsicity:  3.9750354080092136 %\n"
     ]
    }
   ],
   "source": [
    "#analyzing sparcity : percentage of non-zero data-points in document-word matrix\n",
    "data_dense = data_vectorized.todense()\n",
    "print(\"Sparsicity: \", ((data_dense > 0).sum()/data_dense.size)*100, \"%\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LatentDirichletAllocation(batch_size=2, doc_topic_prior=None,\n",
      "             evaluate_every=-1, learning_decay=0.7,\n",
      "             learning_method='online', learning_offset=10.0,\n",
      "             max_doc_update_iter=100, max_iter=20, mean_change_tol=0.001,\n",
      "             n_components=5, n_jobs=-1, n_topics=None, perp_tol=0.1,\n",
      "             random_state=100, topic_word_prior=None,\n",
      "             total_samples=1000000.0, verbose=0)\n"
     ]
    }
   ],
   "source": [
    "#Building LDA model\n",
    "lda_model = LatentDirichletAllocation(n_components=5,               # Number of topics\n",
    "                                      max_iter=20,               # Max learning iterations\n",
    "                                      learning_method='online',   \n",
    "                                      random_state=100,          # Random state\n",
    "                                      batch_size=2,            # n docs in each learning iter\n",
    "                                      evaluate_every = -1,       # compute perplexity every n iters, default: Don't\n",
    "                                      n_jobs = -1,               # Use all available CPUs\n",
    "                                     )\n",
    "lda_output = lda_model.fit_transform(data_vectorized)\n",
    "\n",
    "print(lda_model)  # Model attributes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Log Likelihood:  -23071.115748544937\n",
      "Perplexity:  879.7521705862422\n",
      "{'batch_size': 2,\n",
      " 'doc_topic_prior': None,\n",
      " 'evaluate_every': -1,\n",
      " 'learning_decay': 0.7,\n",
      " 'learning_method': 'online',\n",
      " 'learning_offset': 10.0,\n",
      " 'max_doc_update_iter': 100,\n",
      " 'max_iter': 20,\n",
      " 'mean_change_tol': 0.001,\n",
      " 'n_components': 5,\n",
      " 'n_jobs': -1,\n",
      " 'n_topics': None,\n",
      " 'perp_tol': 0.1,\n",
      " 'random_state': 100,\n",
      " 'topic_word_prior': None,\n",
      " 'total_samples': 1000000.0,\n",
      " 'verbose': 0}\n"
     ]
    }
   ],
   "source": [
    "#analyzing diagnosis parameters\n",
    "print(\"Log Likelihood: \", lda_model.score(data_vectorized))\n",
    "print(\"Perplexity: \", lda_model.perplexity(data_vectorized))\n",
    "\n",
    "pprint(lda_model.get_params())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=None, error_score='raise',\n",
       "       estimator=LatentDirichletAllocation(batch_size=128, doc_topic_prior=None,\n",
       "             evaluate_every=-1, learning_decay=0.7, learning_method=None,\n",
       "             learning_offset=10.0, max_doc_update_iter=100, max_iter=10,\n",
       "             mean_change_tol=0.001, n_components=10, n_jobs=1,\n",
       "             n_topics=None, perp_tol=0.1, random_state=None,\n",
       "             topic_word_prior=None, total_samples=1000000.0, verbose=0),\n",
       "       fit_params=None, iid=True, n_jobs=1,\n",
       "       param_grid={'n_components': [3, 5, 7, 9], 'learning_decay': [0.5, 0.7, 0.9]},\n",
       "       pre_dispatch='2*n_jobs', refit=True, return_train_score='warn',\n",
       "       scoring=None, verbose=0)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Applying Grid Search to find best LDA model\n",
    "\n",
    "\n",
    "# Define Search Param\n",
    "search_params = {'n_components': [3,5,7,9], 'learning_decay': [.5, .7, .9]}\n",
    "\n",
    "# Init the Model\n",
    "lda = LatentDirichletAllocation()\n",
    "\n",
    "# Init Grid Search Class\n",
    "model = GridSearchCV(lda, param_grid=search_params)\n",
    "\n",
    "# Do the Grid Search\n",
    "model.fit(data_vectorized)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Model's Params:  {'learning_decay': 0.9, 'n_components': 3}\n",
      "Best Log Likelihood Score:  -9603.82768045406\n",
      "Model Perplexity:  971.103617905369\n"
     ]
    }
   ],
   "source": [
    "# Printing params for best model among all the generated ones\n",
    "# Best Model\n",
    "best_lda_model = model.best_estimator_\n",
    "\n",
    "# Model Parameters\n",
    "print(\"Best Model's Params: \", model.best_params_)\n",
    "\n",
    "# Log Likelihood Score\n",
    "print(\"Best Log Likelihood Score: \", model.best_score_)\n",
    "\n",
    "# Perplexity\n",
    "print(\"Model Perplexity: \", best_lda_model.perplexity(data_vectorized))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#compare performance scores of models using matploitlib\n",
    "# Get Log Likelyhoods from Grid Search Output\n",
    "n_topics = [3,5,7,9]\n",
    "log_likelyhoods_5 = [round(gscore.mean_validation_score) for gscore in model.grid_scores_ if gscore.parameters['learning_decay']==0.5]\n",
    "log_likelyhoods_7 = [round(gscore.mean_validation_score) for gscore in model.grid_scores_ if gscore.parameters['learning_decay']==0.7]\n",
    "log_likelyhoods_9 = [round(gscore.mean_validation_score) for gscore in model.grid_scores_ if gscore.parameters['learning_decay']==0.9]\n",
    "\n",
    "# Show graph\n",
    "plt.figure(figsize=(12, 8))\n",
    "plt.plot(n_topics, log_likelyhoods_5, label='0.5')\n",
    "plt.plot(n_topics, log_likelyhoods_7, label='0.7')\n",
    "plt.plot(n_topics, log_likelyhoods_9, label='0.9')\n",
    "plt.title(\"Choosing Optimal LDA Model\")\n",
    "plt.xlabel(\"Num Topics\")\n",
    "plt.ylabel(\"Log Likelyhood Scores\")\n",
    "plt.legend(title='Learning decay', loc='best')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style  type=\"text/css\" >\n",
       "    #T_eace0fba_b97a_11e8_b225_f83441068983row0_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row0_col1 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row0_col2 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row0_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row1_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row1_col1 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row1_col2 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row1_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row2_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row2_col1 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row2_col2 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row2_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row3_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row3_col1 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row3_col2 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row3_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row4_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row4_col1 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row4_col2 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row4_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row5_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row5_col1 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row5_col2 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row5_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row6_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row6_col1 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row6_col2 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row6_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row7_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row7_col1 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row7_col2 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row7_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row8_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row8_col1 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row8_col2 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row8_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row9_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row9_col1 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row9_col2 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row9_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row10_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row10_col1 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row10_col2 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row10_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row11_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row11_col1 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row11_col2 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row11_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row12_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row12_col1 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row12_col2 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row12_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row13_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row13_col1 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row13_col2 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row13_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row14_col0 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row14_col1 {\n",
       "            color:  black;\n",
       "            font-weight:  400;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row14_col2 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }    #T_eace0fba_b97a_11e8_b225_f83441068983row14_col3 {\n",
       "            color:  green;\n",
       "            font-weight:  700;\n",
       "        }</style>  \n",
       "<table id=\"T_eace0fba_b97a_11e8_b225_f83441068983\" > \n",
       "<thead>    <tr> \n",
       "        <th class=\"blank level0\" ></th> \n",
       "        <th class=\"col_heading level0 col0\" >Topic0</th> \n",
       "        <th class=\"col_heading level0 col1\" >Topic1</th> \n",
       "        <th class=\"col_heading level0 col2\" >Topic2</th> \n",
       "        <th class=\"col_heading level0 col3\" >dominant_topic</th> \n",
       "    </tr></thead> \n",
       "<tbody>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row0\" class=\"row_heading level0 row0\" >Doc0</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row0_col0\" class=\"data row0 col0\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row0_col1\" class=\"data row0 col1\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row0_col2\" class=\"data row0 col2\" >0.99</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row0_col3\" class=\"data row0 col3\" >2</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row1\" class=\"row_heading level0 row1\" >Doc1</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row1_col0\" class=\"data row1 col0\" >0</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row1_col1\" class=\"data row1 col1\" >0</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row1_col2\" class=\"data row1 col2\" >0.99</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row1_col3\" class=\"data row1 col3\" >2</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row2\" class=\"row_heading level0 row2\" >Doc2</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row2_col0\" class=\"data row2 col0\" >0</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row2_col1\" class=\"data row2 col1\" >0</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row2_col2\" class=\"data row2 col2\" >0.99</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row2_col3\" class=\"data row2 col3\" >2</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row3\" class=\"row_heading level0 row3\" >Doc3</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row3_col0\" class=\"data row3 col0\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row3_col1\" class=\"data row3 col1\" >0.99</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row3_col2\" class=\"data row3 col2\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row3_col3\" class=\"data row3 col3\" >1</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row4\" class=\"row_heading level0 row4\" >Doc4</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row4_col0\" class=\"data row4 col0\" >0</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row4_col1\" class=\"data row4 col1\" >0.99</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row4_col2\" class=\"data row4 col2\" >0</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row4_col3\" class=\"data row4 col3\" >1</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row5\" class=\"row_heading level0 row5\" >Doc5</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row5_col0\" class=\"data row5 col0\" >0</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row5_col1\" class=\"data row5 col1\" >0</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row5_col2\" class=\"data row5 col2\" >0.99</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row5_col3\" class=\"data row5 col3\" >2</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row6\" class=\"row_heading level0 row6\" >Doc6</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row6_col0\" class=\"data row6 col0\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row6_col1\" class=\"data row6 col1\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row6_col2\" class=\"data row6 col2\" >0.98</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row6_col3\" class=\"data row6 col3\" >2</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row7\" class=\"row_heading level0 row7\" >Doc7</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row7_col0\" class=\"data row7 col0\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row7_col1\" class=\"data row7 col1\" >0.99</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row7_col2\" class=\"data row7 col2\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row7_col3\" class=\"data row7 col3\" >1</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row8\" class=\"row_heading level0 row8\" >Doc8</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row8_col0\" class=\"data row8 col0\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row8_col1\" class=\"data row8 col1\" >0.99</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row8_col2\" class=\"data row8 col2\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row8_col3\" class=\"data row8 col3\" >1</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row9\" class=\"row_heading level0 row9\" >Doc9</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row9_col0\" class=\"data row9 col0\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row9_col1\" class=\"data row9 col1\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row9_col2\" class=\"data row9 col2\" >0.98</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row9_col3\" class=\"data row9 col3\" >2</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row10\" class=\"row_heading level0 row10\" >Doc10</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row10_col0\" class=\"data row10 col0\" >0</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row10_col1\" class=\"data row10 col1\" >0.99</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row10_col2\" class=\"data row10 col2\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row10_col3\" class=\"data row10 col3\" >1</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row11\" class=\"row_heading level0 row11\" >Doc11</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row11_col0\" class=\"data row11 col0\" >0.02</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row11_col1\" class=\"data row11 col1\" >0.96</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row11_col2\" class=\"data row11 col2\" >0.02</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row11_col3\" class=\"data row11 col3\" >1</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row12\" class=\"row_heading level0 row12\" >Doc12</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row12_col0\" class=\"data row12 col0\" >0.02</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row12_col1\" class=\"data row12 col1\" >0.02</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row12_col2\" class=\"data row12 col2\" >0.97</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row12_col3\" class=\"data row12 col3\" >2</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row13\" class=\"row_heading level0 row13\" >Doc13</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row13_col0\" class=\"data row13 col0\" >0.02</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row13_col1\" class=\"data row13 col1\" >0.02</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row13_col2\" class=\"data row13 col2\" >0.97</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row13_col3\" class=\"data row13 col3\" >2</td> \n",
       "    </tr>    <tr> \n",
       "        <th id=\"T_eace0fba_b97a_11e8_b225_f83441068983level0_row14\" class=\"row_heading level0 row14\" >Doc14</th> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row14_col0\" class=\"data row14 col0\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row14_col1\" class=\"data row14 col1\" >0.01</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row14_col2\" class=\"data row14 col2\" >0.97</td> \n",
       "        <td id=\"T_eace0fba_b97a_11e8_b225_f83441068983row14_col3\" class=\"data row14 col3\" >2</td> \n",
       "    </tr></tbody> \n",
       "</table> "
      ],
      "text/plain": [
       "<pandas.io.formats.style.Styler at 0x2771a5e6128>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#dominant topic in each doc\n",
    "\n",
    "# Create Document - Topic Matrix\n",
    "lda_output = best_lda_model.transform(data_vectorized)\n",
    "\n",
    "# column names\n",
    "topicnames = [\"Topic\" + str(i) for i in range(best_lda_model.n_components)]\n",
    "\n",
    "# index names\n",
    "docnames = [\"Doc\" + str(i) for i in range(len(data))]\n",
    "\n",
    "# Make the pandas dataframe\n",
    "df_document_topic = pd.DataFrame(np.round(lda_output, 2), columns=topicnames, index=docnames)\n",
    "\n",
    "# Get dominant topic for each document\n",
    "dominant_topic = np.argmax(df_document_topic.values, axis=1)\n",
    "df_document_topic['dominant_topic'] = dominant_topic\n",
    "\n",
    "# Styling\n",
    "def color_green(val):\n",
    "    color = 'green' if val > .1 else 'black'\n",
    "    return 'color: {col}'.format(col=color)\n",
    "\n",
    "def make_bold(val):\n",
    "    weight = 700 if val > .1 else 400\n",
    "    return 'font-weight: {weight}'.format(weight=weight)\n",
    "\n",
    "# Apply Style\n",
    "df_document_topics = df_document_topic.head(15).style.applymap(color_green).applymap(make_bold)\n",
    "df_document_topics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>Num Documents</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Topic Num  Num Documents\n",
       "0          2             31\n",
       "1          1             19\n",
       "2          0              9"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_topic_distribution = df_document_topic['dominant_topic'].value_counts().reset_index(name=\"Num Documents\")\n",
    "df_topic_distribution.columns = ['Topic Num', 'Num Documents']\n",
    "df_topic_distribution\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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_el1672827105666678727519662294\"></div>\n",
       "<script type=\"text/javascript\">\n",
       "\n",
       "var ldavis_el1672827105666678727519662294_data = {\"mdsDat\": {\"x\": [-372.1424560546875, -437.7695617675781, -620.1272583007812], \"y\": [-195.24327087402344, -443.70355224609375, -262.63836669921875], \"topics\": [1, 2, 3], \"cluster\": [1, 1, 1], \"Freq\": [51.20624572029017, 31.432670969341753, 17.361083310368077]}, \"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\", \"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\", \"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\"], \"Freq\": [6.0, 7.0, 25.0, 6.0, 6.0, 10.0, 26.0, 14.0, 5.0, 3.0, 5.0, 36.0, 5.0, 7.0, 4.0, 2.0, 2.0, 4.0, 2.0, 3.0, 4.0, 5.0, 36.0, 4.0, 13.0, 5.0, 6.0, 2.0, 2.0, 3.0, 6.720132134058821, 6.085550265607596, 4.406022787738089, 4.7495874767860595, 4.368621094521085, 3.7307025754413567, 4.130099376191445, 3.6958064311659675, 5.68868347975015, 3.7769125492241646, 5.465356460036732, 7.43660992734177, 3.662354596036418, 3.6240183480629926, 3.502400953485589, 7.431721955146801, 4.351806228340089, 9.046311122720494, 3.0393942975051402, 3.013888608865884, 5.526059302435366, 7.036677364272344, 3.0567952125918225, 2.4509415552559997, 2.498068815446571, 2.413611003668929, 2.5033806720103673, 2.4908821354010664, 2.490451086586855, 3.643707954383911, 10.861173661055256, 26.160315998393802, 4.270823218048453, 3.0290054045378687, 25.54669495798955, 7.298542538573486, 12.308942262623585, 9.16907365816006, 9.823245781059248, 16.520427611452035, 9.618100918442986, 5.574684201712357, 19.548729020660407, 8.027381281826635, 14.556962274789177, 15.377477932757381, 7.910680691808881, 8.735196983915438, 7.867622821157432, 8.468817832460488, 6.699230566403847, 8.225240251384436, 5.603582747192316, 6.571263494682634, 6.839231652743237, 6.732703065886662, 6.217920336526026, 5.914139544129144, 3.515399852458243, 3.441645522756566, 3.3625422859371756, 4.386779629053275, 2.8939698118315387, 2.4366934981865223, 2.484194832414111, 3.8879508108166485, 2.4213086078680273, 2.475098862331443, 2.330288421471421, 1.9372614489258582, 3.4559151203800607, 1.9525886349092225, 1.9870779816243667, 2.0444443145372957, 2.007685714148989, 1.9688159463962644, 2.980674523602766, 4.338370808145411, 6.429985390007385, 3.4835841369461007, 1.4884289941406832, 2.5045042800510684, 1.4936336302088613, 1.4798753651334062, 1.4638721708427547, 1.5028439876307405, 1.504879113248653, 1.5022046925693788, 2.4000501609767224, 3.935930808733936, 3.1188608811048653, 2.961894199988038, 2.4156491831351685, 3.973068019346033, 4.907710407800607, 6.4076929788340955, 5.916051649947491, 11.432968857834158, 4.027289929817463, 4.948076271883291, 5.558647884473662, 9.924734360796236, 4.557590665337765, 3.988038552119973, 6.245236146798118, 4.738331458522681, 4.835858892570982, 3.9183732320857656, 6.045448930451533, 6.126253034022864, 4.0283122463186976, 3.38918961321255, 3.6094566511160058, 3.5799636014314116, 3.551635168779907, 3.5581282878830263, 3.463209552241569, 4.823714649774824, 1.959234167655566, 1.705200936447923, 1.6466262881883116, 1.4162372083966333, 1.3941153245769071, 1.3790117597625164, 1.6455972703926203, 1.3815474689127287, 1.360939421902811, 1.3598564042317935, 1.3500055714558818, 1.3444554255742727, 1.330439020404441, 1.3022759481681336, 1.2949362975843255, 1.3282332929435645, 1.0617444035154644, 1.0330521969710855, 3.0540798791212524, 1.0578008038131017, 3.141055974068694, 1.675160313733163, 1.3831743552110214, 1.021772320688308, 1.0594503245155638, 1.0276627741739326, 0.9928754979064744, 1.005432035095637, 3.465186889875641, 2.3375064917076847, 1.3386444661460404, 1.6837818716496722, 1.6065191895568176, 2.066635684849357, 5.358765127661253, 2.4065978876065492, 5.2685750176158495, 3.4301712524507915, 5.0605312598674566, 3.882177715629016, 2.0292593589860846, 3.6099072582301197, 1.7465172937819422, 2.067418787309386, 1.9576653695681798, 1.7094043117302617, 1.743723480900008, 1.7145297452204131, 1.7641173550540534, 1.5101679446236214, 1.4232654956610524, 1.4266611597859347], \"Term\": [\"cri\", \"studi\", \"help\", \"health\", \"exercis\", \"mani\", \"sad\", \"depress\", \"bodi\", \"profession\", \"someon\", \"fear\", \"effect\", \"shown\", \"suggest\", \"provid\", \"fatti\", \"past\", \"commun\", \"famili\", \"success\", \"usual\", \"feel\", \"woke\", \"deal\", \"ask\", \"lose\", \"chi\", \"interact\", \"mental\", \"control\", \"afraid\", \"talk\", \"express\", \"right\", \"tell\", \"medit\", \"existenti\", \"music\", \"difficult\", \"doctor\", \"loss\", \"away\", \"cours\", \"qualiti\", \"import\", \"everyon\", \"anger\", \"challeng\", \"creation\", \"physic\", \"symptom\", \"daili\", \"courag\", \"stay\", \"shower\", \"realli\", \"receiv\", \"tension\", \"antidepress\", \"thing\", \"fear\", \"reaction\", \"sleep\", \"feel\", \"emot\", \"life\", \"make\", \"time\", \"help\", \"like\", \"use\", \"tip\", \"way\", \"overcom\", \"sad\", \"mind\", \"don\", \"tri\", \"experi\", \"need\", \"joy\", \"medic\", \"think\", \"deal\", \"peopl\", \"just\", \"depress\", \"past\", \"success\", \"woke\", \"someon\", \"wake\", \"second\", \"failur\", \"ask\", \"word\", \"let\", \"lion\", \"flexibl\", \"remind\", \"hour\", \"skill\", \"chanc\", \"write\", \"dear\", \"isn\", \"lose\", \"mani\", \"neg\", \"repeat\", \"acknowledg\", \"insensit\", \"job\", \"bicycl\", \"perhap\", \"reason\", \"deliv\", \"aliv\", \"brain\", \"chang\", \"practic\", \"observ\", \"moment\", \"want\", \"deal\", \"peopl\", \"tip\", \"caus\", \"just\", \"experi\", \"fear\", \"think\", \"thought\", \"overcom\", \"depress\", \"joy\", \"someth\", \"sad\", \"feel\", \"life\", \"smile\", \"tri\", \"mind\", \"way\", \"help\", \"time\", \"cri\", \"profession\", \"provid\", \"fatti\", \"chi\", \"interact\", \"worsen\", \"commun\", \"tai\", \"servic\", \"carb\", \"degre\", \"veget\", \"acid\", \"psychotherapi\", \"licens\", \"fruit\", \"class\", \"boost\", \"health\", \"master\", \"exercis\", \"famili\", \"yoga\", \"power\", \"toxin\", \"expens\", \"incorpor\", \"howev\", \"studi\", \"bodi\", \"social\", \"suggest\", \"mental\", \"effect\", \"help\", \"shown\", \"sad\", \"depress\", \"feel\", \"overcom\", \"happi\", \"tip\", \"usual\", \"don\", \"like\", \"clinic\", \"natur\", \"mani\", \"joy\", \"symptom\", \"distract\", \"sever\"], \"Total\": [6.0, 7.0, 25.0, 6.0, 6.0, 10.0, 26.0, 14.0, 5.0, 3.0, 5.0, 36.0, 5.0, 7.0, 4.0, 2.0, 2.0, 4.0, 2.0, 3.0, 4.0, 5.0, 36.0, 4.0, 13.0, 5.0, 6.0, 2.0, 2.0, 3.0, 7.646640762716596, 6.961680203089529, 5.251481748748608, 5.719739518713011, 5.268349015969582, 4.5103701268682, 5.038905806565997, 4.559202041539707, 7.029663631754957, 4.677606367989901, 6.773101268351643, 9.228223975585426, 4.576552583119323, 4.533677911385831, 4.48762344252808, 9.594775571128114, 5.64175269975111, 11.759236767895192, 3.9521172273097798, 3.9336969935113015, 7.217536547960927, 9.198692436303666, 4.003791681114394, 3.214230195934186, 3.294495992316603, 3.1970951500953046, 3.3170992934691665, 3.300775261824702, 3.3125150676724266, 4.848942970237167, 14.574160291221576, 36.7647113640543, 5.729879456505104, 4.035085778679778, 36.73347925187987, 10.011675634989015, 17.423822568798194, 12.779636429581494, 14.043939583012792, 25.437321026996315, 14.082397096411487, 7.777475229268954, 34.59160513672468, 12.010704329016393, 24.68437613721631, 26.691501880824763, 11.93842024198578, 13.827366160593767, 12.204249878438999, 14.462374263153935, 10.5523437663302, 14.825216499009471, 7.983252664996387, 11.600944971754608, 13.58865677738803, 13.311968657566522, 11.888921193031708, 14.082642255102616, 4.434043157951289, 4.379869247584607, 4.371725470622549, 5.747669975254913, 3.9026599286548955, 3.3503102953994177, 3.473938791290778, 5.530981773761792, 3.4478341772333896, 3.545061284540969, 3.384372085555511, 2.849457313337683, 5.083385771534555, 2.879651097162423, 2.936637853358083, 3.0218636212016214, 2.9808232508945878, 2.9490323604815663, 4.521390404886445, 6.681273708626859, 10.03007230248889, 5.484196007709219, 2.3588220354247835, 4.015286524773584, 2.403727436727546, 2.39802876263899, 2.373111966161105, 2.4527683892775443, 2.4635337190549134, 2.4615841295916834, 3.938705132655507, 6.749767056082073, 5.383959624772723, 5.097258794135964, 4.125215422243672, 7.508276005336025, 9.69839589740599, 13.58865677738803, 13.311968657566522, 34.59160513672468, 8.598899347698634, 11.888921193031708, 14.462374263153935, 36.7647113640543, 11.600944971754608, 9.866506301532173, 24.68437613721631, 14.082642255102616, 14.825216499009471, 9.611473378954223, 26.691501880824763, 36.73347925187987, 17.423822568798194, 8.685694456858238, 12.204249878438999, 11.93842024198578, 12.010704329016393, 25.437321026996315, 14.043939583012792, 6.0609285662363686, 3.2486864159353637, 2.908969499596574, 2.846344857982401, 2.490083926792985, 2.4778500238901735, 2.4586194880637127, 2.9540587481479386, 2.4953813670804186, 2.471236734815702, 2.484421683503472, 2.4986958698293096, 2.5116441512369754, 2.4871730114725255, 2.477198314367909, 2.463255832177407, 2.5603418268368503, 2.1157625814253795, 2.072380785534775, 6.188847713719416, 2.189997292444839, 6.512184944943316, 3.4862632996231597, 2.899320249943063, 2.148038163004647, 2.2274852108724814, 2.165778232866164, 2.1012622230109885, 2.128643820955834, 7.347778370405951, 5.76721071470331, 3.0077846943050934, 4.231835225151932, 3.968801732979305, 5.657821707227348, 25.437321026996315, 7.5673336041349195, 26.691501880824763, 14.082642255102616, 36.73347925187987, 24.68437613721631, 7.244238622413871, 34.59160513672468, 5.987090243621179, 13.827366160593767, 14.082397096411487, 7.567304178211616, 10.139567203628705, 10.03007230248889, 14.825216499009471, 9.198692436303666, 3.5584820952695972, 5.75664734703568], \"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, 0.5402, 0.5348, 0.4938, 0.4834, 0.482, 0.4795, 0.4704, 0.4594, 0.4576, 0.4554, 0.4548, 0.4535, 0.4465, 0.4454, 0.4214, 0.4138, 0.4097, 0.407, 0.4067, 0.403, 0.4023, 0.4014, 0.3994, 0.3982, 0.3926, 0.3882, 0.3879, 0.3878, 0.3841, 0.3835, 0.3753, 0.329, 0.3754, 0.3825, 0.3061, 0.3532, 0.3218, 0.3373, 0.3119, 0.2377, 0.288, 0.3363, 0.0986, 0.2664, 0.1412, 0.1179, 0.2578, 0.21, 0.2303, 0.1341, 0.215, 0.0802, 0.3154, 0.1009, -0.0173, -0.0124, 0.0211, -0.1983, 0.9252, 0.9163, 0.8949, 0.8871, 0.8583, 0.8389, 0.822, 0.8048, 0.8039, 0.798, 0.7841, 0.7715, 0.7714, 0.7688, 0.7667, 0.7666, 0.7621, 0.7533, 0.7407, 0.7255, 0.7127, 0.7035, 0.6969, 0.6853, 0.6815, 0.6746, 0.6742, 0.6675, 0.6644, 0.6635, 0.662, 0.618, 0.6114, 0.6144, 0.6222, 0.5209, 0.4762, 0.4056, 0.3463, 0.0502, 0.3988, 0.2807, 0.2011, -0.1522, 0.223, 0.2515, -0.217, 0.0681, 0.0371, 0.26, -0.3277, -0.6338, -0.3072, 0.2162, -0.0609, -0.0471, -0.0611, -0.8097, -0.2427, 1.5226, 1.2452, 1.2168, 1.2036, 1.1866, 1.1758, 1.1727, 1.1659, 1.1597, 1.1544, 1.1483, 1.1353, 1.126, 1.1253, 1.1079, 1.1079, 1.0946, 1.0614, 1.0548, 1.0447, 1.0232, 1.0218, 1.018, 1.0108, 1.0079, 1.0078, 1.0054, 1.0013, 1.0009, 0.9993, 0.8478, 0.9414, 0.8293, 0.8465, 0.7438, 0.1935, 0.6053, 0.1284, 0.3386, -0.2313, -0.0988, 0.4784, -0.509, 0.519, -0.1494, -0.2222, 0.2632, -0.0095, -0.0155, -0.3777, -0.0559, 0.8346, 0.3559], \"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, -5.558, -5.6572, -5.9801, -5.905, -5.9887, -6.1465, -6.0448, -6.1559, -5.7246, -6.1342, -5.7647, -5.4567, -6.165, -6.1755, -6.2097, -5.4573, -5.9925, -5.2607, -6.3514, -6.3599, -5.7536, -5.512, -6.3457, -6.5666, -6.5476, -6.582, -6.5455, -6.5505, -6.5506, -6.1701, -5.0779, -4.1989, -6.0113, -6.3549, -4.2226, -5.4754, -4.9528, -5.2473, -5.1784, -4.6585, -5.1995, -5.7449, -4.4902, -5.3802, -4.785, -4.7302, -5.3949, -5.2957, -5.4003, -5.3267, -5.5611, -5.3559, -5.7397, -5.5804, -5.5404, -5.5561, -5.6357, -5.6858, -5.7179, -5.7391, -5.7624, -5.4965, -5.9125, -6.0844, -6.0651, -5.6172, -6.0908, -6.0688, -6.1291, -6.3138, -5.735, -6.3059, -6.2884, -6.26, -6.2781, -6.2977, -5.8829, -5.5076, -5.1141, -5.727, -6.5774, -6.057, -6.5739, -6.5831, -6.594, -6.5677, -6.5664, -6.5682, -6.0996, -5.6049, -5.8376, -5.8893, -6.0931, -5.5956, -5.3843, -5.1176, -5.1974, -4.5386, -5.582, -5.3761, -5.2597, -4.6801, -5.4583, -5.5918, -5.1433, -5.4194, -5.399, -5.6094, -5.1758, -5.1625, -5.5817, -5.7545, -5.6915, -5.6997, -5.7077, -5.7059, -5.7329, -4.8079, -5.7089, -5.8478, -5.8827, -6.0335, -6.0492, -6.0601, -5.8834, -6.0583, -6.0733, -6.0741, -6.0814, -6.0855, -6.096, -6.1174, -6.123, -6.0976, -6.3216, -6.349, -5.265, -6.3253, -5.2369, -5.8656, -6.0571, -6.3599, -6.3237, -6.3542, -6.3886, -6.3761, -5.1387, -5.5324, -6.0898, -5.8604, -5.9074, -5.6556, -4.7027, -5.5033, -4.7197, -5.1489, -4.76, -5.0251, -5.6738, -5.0978, -5.8238, -5.6552, -5.7097, -5.8453, -5.8255, -5.8423, -5.8138, -5.9693, -6.0285, -6.0261]}, \"token.table\": {\"Topic\": [1, 3, 1, 2, 1, 1, 2, 1, 2, 3, 1, 3, 1, 2, 1, 2, 1, 2, 1, 2, 3, 1, 3, 1, 2, 1, 3, 1, 2, 1, 2, 1, 2, 1, 2, 3, 1, 3, 1, 3, 1, 2, 3, 1, 2, 3, 1, 2, 1, 1, 2, 1, 2, 1, 2, 3, 1, 2, 1, 2, 1, 2, 1, 2, 3, 1, 2, 1, 2, 3, 1, 2, 1, 2, 3, 1, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 1, 2, 3, 1, 1, 2, 3, 1, 2, 1, 2, 1, 2, 1, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 1, 3, 1, 2, 3, 1, 3, 1, 2, 1, 2, 3, 1, 2, 1, 2, 1, 2, 3, 1, 2, 3, 1, 2, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 1, 2, 1, 2, 1, 2, 3, 1, 2, 3, 1, 3, 1, 2, 3, 1, 1, 2, 3, 1, 2, 1, 2, 1, 2, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 1, 2, 3, 1, 2, 1, 2, 3, 1, 2, 1, 2, 3, 1, 3, 1, 2, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 1, 2, 1, 1, 2, 1, 1, 2, 1, 2, 1, 2, 1, 2, 3, 1, 2, 1, 3, 1, 2, 3, 1, 1, 2, 3, 1, 2, 1, 2, 1, 2, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 1, 2, 3, 1, 2, 1, 2, 3, 1, 2, 3, 1, 3, 1, 1, 1, 1, 2, 3, 1, 2, 1, 2, 1, 2, 3, 1, 2, 3, 1, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 3, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 1, 2, 3, 1, 2, 1, 2, 3], \"Freq\": [0.40206290249505083, 0.40206290249505083, 0.24904822951741618, 0.7471446885522486, 0.8618609049776886, 0.2538905468472558, 0.5077810936945116, 0.7653557945675182, 0.1700790654594485, 0.08503953272972425, 0.8249220550854108, 0.2062305137713527, 0.1807997279513487, 0.7231989118053948, 0.8740203302271791, 0.21850508255679477, 0.42138761856132007, 0.42138761856132007, 0.3467880920149262, 0.3467880920149262, 0.3467880920149262, 0.48253680355463796, 0.48253680355463796, 0.29630652189661005, 0.5926130437932201, 0.40250815980233434, 0.40250815980233434, 0.46517581358485605, 0.46517581358485605, 0.7590867951156678, 0.2530289317052226, 0.33092161836289535, 0.6618432367257907, 0.37147381098431387, 0.5572107164764708, 0.18573690549215693, 0.40159288979786095, 0.40159288979786095, 0.47264282333904617, 0.47264282333904617, 0.528589826151976, 0.264294913075988, 0.264294913075988, 0.3385173028894415, 0.3385173028894415, 0.677034605778883, 0.9154346617315305, 0.13077638024736152, 0.6222329696640532, 0.8822858787463579, 0.22057146968658947, 0.7626413536549841, 0.2542137845516613, 0.16499122025141538, 0.16499122025141538, 0.824956101257077, 0.7492897330674797, 0.24976324435582656, 0.5151355365490009, 0.4415447456134293, 0.3390942783132782, 0.6781885566265564, 0.4002087697324732, 0.4002087697324732, 0.4002087697324732, 0.406242463127139, 0.812484926254278, 0.4260564098208202, 0.35504700818401685, 0.2130282049104101, 0.8551382235523407, 0.21378455588808518, 0.28101869651931977, 0.28101869651931977, 0.28101869651931977, 0.738214268752081, 0.1476428537504162, 0.6508831758320579, 0.21696105861068596, 0.14464070574045731, 0.5302394022363369, 0.17674646741211228, 0.35349293482422456, 0.6991836586810956, 0.19976675962317017, 0.09988337981158509, 0.7089995277844175, 0.1772498819461044, 0.3071165848188927, 0.15355829240944635, 0.46067487722833905, 0.8773465101031459, 0.46172779134298175, 0.46172779134298175, 0.46172779134298175, 0.5531595196220134, 0.41486963971651003, 0.8741656824828697, 0.17483313649657395, 0.2878576912486243, 0.5757153824972486, 0.28684006744645274, 0.5736801348929055, 0.3513277729490722, 0.3513277729490722, 0.7026555458981444, 0.7071998945548854, 0.2719999594441867, 0.027199995944418673, 0.7078011810893036, 0.1633387340975316, 0.13611561174794298, 0.35094401846948886, 0.7018880369389777, 0.39057284832761585, 0.39057284832761585, 0.39057284832761585, 0.41412219508036613, 0.27608146338691075, 0.27608146338691075, 0.4847429018732535, 0.16158096729108448, 0.4847429018732535, 0.668309370391564, 0.15724926362154448, 0.1965615795269306, 0.34726429218643506, 0.6945285843728701, 0.4697826804819633, 0.4697826804819633, 0.7295637035079674, 0.10422338621542392, 0.10422338621542392, 0.4759044297513031, 0.4759044297513031, 0.41602054572435554, 0.41602054572435554, 0.4035756766384192, 0.4035756766384192, 0.4035756766384192, 0.2211709032954245, 0.6635127098862734, 0.4170091766954109, 0.4170091766954109, 0.53962112462469, 0.3372632028904312, 0.1349052811561725, 0.5046715259174818, 0.42055960493123484, 0.08411192098624697, 0.2820825705780386, 0.5641651411560772, 0.4059667643681351, 0.4059667643681351, 0.4059667643681351, 0.688712247419752, 0.22957074913991732, 0.05739268728497933, 0.710106378306022, 0.21303191349180658, 0.14202127566120437, 0.29547578538069047, 0.5909515707613809, 0.299344120181396, 0.598688240362792, 0.7585424907890719, 0.10836321296986741, 0.7042453867597805, 0.1564989748355068, 0.0782494874177534, 0.19940035721414637, 0.5982010716424391, 0.19940035721414637, 0.4566215690995827, 0.4566215690995827, 0.7515733563473173, 0.1252622260578862, 0.1252622260578862, 0.7938231341391141, 0.2519652195498612, 0.2519652195498612, 0.5039304390997223, 0.6701054107531834, 0.3350527053765917, 0.399559099567989, 0.5327454660906521, 0.8535259031308869, 0.14225431718848114, 0.49311769423557067, 0.2958706165413424, 0.19724707769422828, 0.6633597383678107, 0.28429703072906176, 0.09476567690968725, 0.18234213339462788, 0.5470264001838836, 0.18234213339462788, 0.24241158282495412, 0.48482316564990824, 0.6076718292014962, 0.2430687316805985, 0.162045821120399, 0.22552780033427577, 0.9021112013371031, 0.5258425842237241, 0.4507222150489064, 0.07512036917481774, 0.40770257981616725, 0.8154051596323345, 0.831308571855462, 0.138551428642577, 0.138551428642577, 0.4655410770734229, 0.4655410770734229, 0.39236775701890175, 0.5885516355283525, 0.30781672096599677, 0.30781672096599677, 0.6156334419319935, 0.343764346837835, 0.343764346837835, 0.68752869367567, 0.40368185066166723, 0.40368185066166723, 0.40368185066166723, 0.8913403834405991, 0.6980949652368026, 0.17452374130920065, 0.9044046423049549, 0.4059209712719623, 0.8118419425439246, 0.6059182590014867, 0.19671928217600598, 0.590157846528018, 0.4239404181332896, 0.4239404181332896, 0.7592511407036772, 0.1898127851759193, 0.5619766196362309, 0.22479064785449238, 0.18732553987874365, 0.2984798158466638, 0.5969596316933276, 0.40465568753961456, 0.40465568753961456, 0.6948488866631294, 0.17371222166578235, 0.17371222166578235, 0.6255678689889416, 0.39644082802967073, 0.2642938853531138, 0.2642938853531138, 0.3405254750280111, 0.6810509500560222, 0.7434786184351097, 0.24782620614503656, 0.5756592089250839, 0.34539552535505036, 0.3324706059889822, 0.3324706059889822, 0.3324706059889822, 0.17398354538538885, 0.6959341815415554, 0.17398354538538885, 0.5202116057407241, 0.41616928459257935, 0.10404232114814484, 0.6070731318733985, 0.2721911167129424, 0.2721911167129424, 0.40828667506941363, 0.22831731804584715, 0.6849519541375414, 0.23630409663790675, 0.4726081932758135, 0.4726081932758135, 0.760977720308782, 0.10871110290125457, 0.21742220580250915, 0.4007403490272888, 0.4007403490272888, 0.7616897842124601, 0.8868451784415797, 0.60377083851435, 0.75476046511068, 0.1372291754746691, 0.06861458773733455, 0.60339912110981, 0.43099937222129286, 0.5067649933212477, 0.4054119946569982, 0.7120509128432706, 0.21361527385298118, 0.07120509128432706, 0.5781749624207727, 0.31799622933142496, 0.11563499248415454, 0.4489367629104532, 0.4489367629104532, 0.6555093577798204, 0.3277546788899102, 0.08193866972247756, 0.7714585804684038, 0.1285764300780673, 0.1285764300780673, 0.6681041770268482, 0.16702604425671205, 0.3340520885134241, 0.398145573093029, 0.398145573093029, 0.398145573093029, 0.25623549534962003, 0.7687064860488602, 0.4124393396922343, 0.5155491746152928, 0.6660725117238111, 0.33303625586190555, 0.2287426341658175, 0.6862279024974525, 0.29003715045322154, 0.5800743009064431, 0.40673231659265446, 0.40673231659265446, 0.40673231659265446, 0.3354777911437338, 0.6709555822874675, 0.3449084315606867, 0.3449084315606867, 0.3449084315606867], \"Term\": [\"acid\", \"acid\", \"acknowledg\", \"acknowledg\", \"afraid\", \"aliv\", \"aliv\", \"anger\", \"anger\", \"anger\", \"antidepress\", \"antidepress\", \"ask\", \"ask\", \"away\", \"away\", \"bicycl\", \"bicycl\", \"bodi\", \"bodi\", \"bodi\", \"boost\", \"boost\", \"brain\", \"brain\", \"carb\", \"carb\", \"caus\", \"caus\", \"challeng\", \"challeng\", \"chanc\", \"chanc\", \"chang\", \"chang\", \"chang\", \"chi\", \"chi\", \"class\", \"class\", \"clinic\", \"clinic\", \"clinic\", \"commun\", \"commun\", \"commun\", \"control\", \"control\", \"courag\", \"cours\", \"cours\", \"creation\", \"creation\", \"cri\", \"cri\", \"cri\", \"daili\", \"daili\", \"deal\", \"deal\", \"dear\", \"dear\", \"degre\", \"degre\", \"degre\", \"deliv\", \"deliv\", \"depress\", \"depress\", \"depress\", \"difficult\", \"difficult\", \"distract\", \"distract\", \"distract\", \"doctor\", \"doctor\", \"don\", \"don\", \"don\", \"effect\", \"effect\", \"effect\", \"emot\", \"emot\", \"emot\", \"everyon\", \"everyon\", \"exercis\", \"exercis\", \"exercis\", \"existenti\", \"expens\", \"expens\", \"expens\", \"experi\", \"experi\", \"express\", \"express\", \"failur\", \"failur\", \"famili\", \"famili\", \"fatti\", \"fatti\", \"fatti\", \"fear\", \"fear\", \"fear\", \"feel\", \"feel\", \"feel\", \"flexibl\", \"flexibl\", \"fruit\", \"fruit\", \"fruit\", \"happi\", \"happi\", \"happi\", \"health\", \"health\", \"health\", \"help\", \"help\", \"help\", \"hour\", \"hour\", \"howev\", \"howev\", \"import\", \"import\", \"import\", \"incorpor\", \"incorpor\", \"insensit\", \"insensit\", \"interact\", \"interact\", \"interact\", \"isn\", \"isn\", \"job\", \"job\", \"joy\", \"joy\", \"joy\", \"just\", \"just\", \"just\", \"let\", \"let\", \"licens\", \"licens\", \"licens\", \"life\", \"life\", \"life\", \"like\", \"like\", \"like\", \"lion\", \"lion\", \"lose\", \"lose\", \"loss\", \"loss\", \"make\", \"make\", \"make\", \"mani\", \"mani\", \"mani\", \"master\", \"master\", \"medic\", \"medic\", \"medic\", \"medit\", \"mental\", \"mental\", \"mental\", \"mind\", \"mind\", \"moment\", \"moment\", \"music\", \"music\", \"natur\", \"natur\", \"natur\", \"need\", \"need\", \"need\", \"neg\", \"neg\", \"neg\", \"observ\", \"observ\", \"overcom\", \"overcom\", \"overcom\", \"past\", \"past\", \"peopl\", \"peopl\", \"peopl\", \"perhap\", \"perhap\", \"physic\", \"physic\", \"physic\", \"power\", \"power\", \"practic\", \"practic\", \"profession\", \"profession\", \"profession\", \"provid\", \"provid\", \"provid\", \"psychotherapi\", \"psychotherapi\", \"psychotherapi\", \"qualiti\", \"reaction\", \"reaction\", \"realli\", \"reason\", \"reason\", \"receiv\", \"remind\", \"remind\", \"repeat\", \"repeat\", \"right\", \"right\", \"sad\", \"sad\", \"sad\", \"second\", \"second\", \"servic\", \"servic\", \"sever\", \"sever\", \"sever\", \"shower\", \"shown\", \"shown\", \"shown\", \"skill\", \"skill\", \"sleep\", \"sleep\", \"smile\", \"smile\", \"social\", \"social\", \"social\", \"someon\", \"someon\", \"someon\", \"someth\", \"someth\", \"someth\", \"stay\", \"studi\", \"studi\", \"studi\", \"success\", \"success\", \"suggest\", \"suggest\", \"suggest\", \"symptom\", \"symptom\", \"symptom\", \"tai\", \"tai\", \"talk\", \"tell\", \"tension\", \"thing\", \"thing\", \"thing\", \"think\", \"think\", \"thought\", \"thought\", \"time\", \"time\", \"time\", \"tip\", \"tip\", \"tip\", \"toxin\", \"toxin\", \"tri\", \"tri\", \"tri\", \"use\", \"use\", \"use\", \"usual\", \"usual\", \"usual\", \"veget\", \"veget\", \"veget\", \"wake\", \"wake\", \"want\", \"want\", \"way\", \"way\", \"woke\", \"woke\", \"word\", \"word\", \"worsen\", \"worsen\", \"worsen\", \"write\", \"write\", \"yoga\", \"yoga\", \"yoga\"]}, \"R\": 30, \"lambda.step\": 0.01, \"plot.opts\": {\"xlab\": \"PC1\", \"ylab\": \"PC2\"}, \"topic.order\": [3, 2, 1]};\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_el1672827105666678727519662294\", ldavis_el1672827105666678727519662294_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_el1672827105666678727519662294\", ldavis_el1672827105666678727519662294_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_el1672827105666678727519662294\", ldavis_el1672827105666678727519662294_data);\n",
       "            })\n",
       "         });\n",
       "}\n",
       "</script>"
      ],
      "text/plain": [
       "PreparedData(topic_coordinates=                x           y  topics  cluster       Freq\n",
       "topic                                                    \n",
       "2     -372.142456 -195.243271       1        1  51.206246\n",
       "1     -437.769562 -443.703552       2        1  31.432671\n",
       "0     -620.127258 -262.638367       3        1  17.361083, topic_info=     Category       Freq        Term      Total  loglift  logprob\n",
       "term                                                             \n",
       "214   Default   6.000000         cri   6.000000  30.0000  30.0000\n",
       "945   Default   7.000000       studi   7.000000  29.0000  29.0000\n",
       "457   Default  25.000000        help  25.000000  28.0000  28.0000\n",
       "450   Default   6.000000      health   6.000000  27.0000  27.0000\n",
       "334   Default   6.000000     exercis   6.000000  26.0000  26.0000\n",
       "594   Default  10.000000        mani  10.000000  25.0000  25.0000\n",
       "841   Default  26.000000         sad  26.000000  24.0000  24.0000\n",
       "242   Default  14.000000     depress  14.000000  23.0000  23.0000\n",
       "108   Default   5.000000        bodi   5.000000  22.0000  22.0000\n",
       "742   Default   3.000000  profession   3.000000  21.0000  21.0000\n",
       "901   Default   5.000000      someon   5.000000  20.0000  20.0000\n",
       "360   Default  36.000000        fear  36.000000  19.0000  19.0000\n",
       "299   Default   5.000000      effect   5.000000  18.0000  18.0000\n",
       "876   Default   7.000000       shown   7.000000  17.0000  17.0000\n",
       "954   Default   4.000000     suggest   4.000000  16.0000  16.0000\n",
       "747   Default   2.000000      provid   2.000000  15.0000  15.0000\n",
       "359   Default   2.000000       fatti   2.000000  14.0000  14.0000\n",
       "689   Default   4.000000        past   4.000000  13.0000  13.0000\n",
       "182   Default   2.000000      commun   2.000000  12.0000  12.0000\n",
       "353   Default   3.000000      famili   3.000000  11.0000  11.0000\n",
       "949   Default   4.000000     success   4.000000  10.0000  10.0000\n",
       "1035  Default   5.000000       usual   5.000000   9.0000   9.0000\n",
       "361   Default  36.000000        feel  36.000000   8.0000   8.0000\n",
       "1070  Default   4.000000        woke   4.000000   7.0000   7.0000\n",
       "230   Default  13.000000        deal  13.000000   6.0000   6.0000\n",
       "54    Default   5.000000         ask   5.000000   5.0000   5.0000\n",
       "579   Default   6.000000        lose   6.000000   4.0000   4.0000\n",
       "151   Default   2.000000         chi   2.000000   3.0000   3.0000\n",
       "513   Default   2.000000    interact   2.000000   2.0000   2.0000\n",
       "613   Default   3.000000      mental   3.000000   1.0000   1.0000\n",
       "...       ...        ...         ...        ...      ...      ...\n",
       "1088   Topic3   1.383174        yoga   2.899320   1.0108  -6.0571\n",
       "726    Topic3   1.021772       power   2.148038   1.0079  -6.3599\n",
       "1000   Topic3   1.059450       toxin   2.227485   1.0078  -6.3237\n",
       "340    Topic3   1.027663      expens   2.165778   1.0054  -6.3542\n",
       "498    Topic3   0.992875    incorpor   2.101262   1.0013  -6.3886\n",
       "476    Topic3   1.005432       howev   2.128644   1.0009  -6.3761\n",
       "945    Topic3   3.465187       studi   7.347778   0.9993  -5.1387\n",
       "108    Topic3   2.337506        bodi   5.767211   0.8478  -5.5324\n",
       "897    Topic3   1.338644      social   3.007785   0.9414  -6.0898\n",
       "954    Topic3   1.683782     suggest   4.231835   0.8293  -5.8604\n",
       "613    Topic3   1.606519      mental   3.968802   0.8465  -5.9074\n",
       "299    Topic3   2.066636      effect   5.657822   0.7438  -5.6556\n",
       "457    Topic3   5.358765        help  25.437321   0.1935  -4.7027\n",
       "876    Topic3   2.406598       shown   7.567334   0.6053  -5.5033\n",
       "841    Topic3   5.268575         sad  26.691502   0.1284  -4.7197\n",
       "242    Topic3   3.430171     depress  14.082642   0.3386  -5.1489\n",
       "361    Topic3   5.060531        feel  36.733479  -0.2313  -4.7600\n",
       "675    Topic3   3.882178     overcom  24.684376  -0.0988  -5.0251\n",
       "443    Topic3   2.029259       happi   7.244239   0.4784  -5.6738\n",
       "994    Topic3   3.609907         tip  34.591605  -0.5090  -5.0978\n",
       "1035   Topic3   1.746517       usual   5.987090   0.5190  -5.8238\n",
       "281    Topic3   2.067419         don  13.827366  -0.1494  -5.6552\n",
       "566    Topic3   1.957665        like  14.082397  -0.2222  -5.7097\n",
       "166    Topic3   1.709404      clinic   7.567304   0.2632  -5.8453\n",
       "639    Topic3   1.743723       natur  10.139567  -0.0095  -5.8255\n",
       "594    Topic3   1.714530        mani  10.030072  -0.0155  -5.8423\n",
       "528    Topic3   1.764117         joy  14.825216  -0.3777  -5.8138\n",
       "966    Topic3   1.510168     symptom   9.198692  -0.0559  -5.9693\n",
       "274    Topic3   1.423265    distract   3.558482   0.8346  -6.0285\n",
       "870    Topic3   1.426661       sever   5.756647   0.3559  -6.0261\n",
       "\n",
       "[200 rows x 6 columns], token_table=      Topic      Freq         Term\n",
       "term                              \n",
       "13        1  0.402063         acid\n",
       "13        3  0.402063         acid\n",
       "14        1  0.249048   acknowledg\n",
       "14        2  0.747145   acknowledg\n",
       "25        1  0.861861       afraid\n",
       "33        1  0.253891         aliv\n",
       "33        2  0.507781         aliv\n",
       "38        1  0.765356        anger\n",
       "38        2  0.170079        anger\n",
       "38        3  0.085040        anger\n",
       "42        1  0.824922  antidepress\n",
       "42        3  0.206231  antidepress\n",
       "54        1  0.180800          ask\n",
       "54        2  0.723199          ask\n",
       "64        1  0.874020         away\n",
       "64        2  0.218505         away\n",
       "92        1  0.421388       bicycl\n",
       "92        2  0.421388       bicycl\n",
       "108       1  0.346788         bodi\n",
       "108       2  0.346788         bodi\n",
       "108       3  0.346788         bodi\n",
       "110       1  0.482537        boost\n",
       "110       3  0.482537        boost\n",
       "111       1  0.296307        brain\n",
       "111       2  0.592613        brain\n",
       "129       1  0.402508         carb\n",
       "129       3  0.402508         carb\n",
       "135       1  0.465176         caus\n",
       "135       2  0.465176         caus\n",
       "143       1  0.759087     challeng\n",
       "...     ...       ...          ...\n",
       "1008      1  0.655509          tri\n",
       "1008      2  0.327755          tri\n",
       "1008      3  0.081939          tri\n",
       "1034      1  0.771459          use\n",
       "1034      2  0.128576          use\n",
       "1034      3  0.128576          use\n",
       "1035      1  0.668104        usual\n",
       "1035      2  0.167026        usual\n",
       "1035      3  0.334052        usual\n",
       "1040      1  0.398146        veget\n",
       "1040      2  0.398146        veget\n",
       "1040      3  0.398146        veget\n",
       "1046      1  0.256235         wake\n",
       "1046      2  0.768706         wake\n",
       "1048      1  0.412439         want\n",
       "1048      2  0.515549         want\n",
       "1054      1  0.666073          way\n",
       "1054      2  0.333036          way\n",
       "1070      1  0.228743         woke\n",
       "1070      2  0.686228         woke\n",
       "1073      1  0.290037         word\n",
       "1073      2  0.580074         word\n",
       "1080      1  0.406732       worsen\n",
       "1080      2  0.406732       worsen\n",
       "1080      3  0.406732       worsen\n",
       "1083      1  0.335478        write\n",
       "1083      2  0.670956        write\n",
       "1088      1  0.344908         yoga\n",
       "1088      2  0.344908         yoga\n",
       "1088      3  0.344908         yoga\n",
       "\n",
       "[329 rows x 3 columns], R=30, lambda_step=0.01, plot_opts={'xlab': 'PC1', 'ylab': 'PC2'}, topic_order=[3, 2, 1])"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pyLDAvis.enable_notebook()\n",
    "panel = pyLDAvis.sklearn.prepare(best_lda_model, data_vectorized, vectorizer,mds='tsne')\n",
    "panel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>abdomen</th>\n",
       "      <th>abil</th>\n",
       "      <th>abl</th>\n",
       "      <th>absolut</th>\n",
       "      <th>abus</th>\n",
       "      <th>academi</th>\n",
       "      <th>accept</th>\n",
       "      <th>accomplish</th>\n",
       "      <th>accord</th>\n",
       "      <th>account</th>\n",
       "      <th>...</th>\n",
       "      <th>wors</th>\n",
       "      <th>worsen</th>\n",
       "      <th>worthless</th>\n",
       "      <th>wow</th>\n",
       "      <th>write</th>\n",
       "      <th>wrong</th>\n",
       "      <th>www</th>\n",
       "      <th>year</th>\n",
       "      <th>yellow</th>\n",
       "      <th>yoga</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Topic0</th>\n",
       "      <td>0.605911</td>\n",
       "      <td>0.621080</td>\n",
       "      <td>0.699753</td>\n",
       "      <td>0.614075</td>\n",
       "      <td>0.546653</td>\n",
       "      <td>0.598405</td>\n",
       "      <td>0.641164</td>\n",
       "      <td>0.619712</td>\n",
       "      <td>0.594675</td>\n",
       "      <td>0.588146</td>\n",
       "      <td>...</td>\n",
       "      <td>0.637057</td>\n",
       "      <td>2.389329</td>\n",
       "      <td>0.703087</td>\n",
       "      <td>0.583808</td>\n",
       "      <td>0.556614</td>\n",
       "      <td>0.559208</td>\n",
       "      <td>0.570977</td>\n",
       "      <td>0.609959</td>\n",
       "      <td>0.611370</td>\n",
       "      <td>2.396541</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Topic1</th>\n",
       "      <td>0.622789</td>\n",
       "      <td>1.198479</td>\n",
       "      <td>3.534938</td>\n",
       "      <td>0.547543</td>\n",
       "      <td>1.182509</td>\n",
       "      <td>0.612230</td>\n",
       "      <td>2.857797</td>\n",
       "      <td>0.670544</td>\n",
       "      <td>0.547127</td>\n",
       "      <td>0.630968</td>\n",
       "      <td>...</td>\n",
       "      <td>0.624739</td>\n",
       "      <td>0.632156</td>\n",
       "      <td>0.579477</td>\n",
       "      <td>1.785210</td>\n",
       "      <td>2.437571</td>\n",
       "      <td>1.785731</td>\n",
       "      <td>2.441624</td>\n",
       "      <td>1.168159</td>\n",
       "      <td>0.664371</td>\n",
       "      <td>1.128858</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Topic2</th>\n",
       "      <td>1.137761</td>\n",
       "      <td>1.195480</td>\n",
       "      <td>2.405697</td>\n",
       "      <td>1.751575</td>\n",
       "      <td>0.629416</td>\n",
       "      <td>1.154473</td>\n",
       "      <td>5.381602</td>\n",
       "      <td>1.226811</td>\n",
       "      <td>1.227185</td>\n",
       "      <td>1.198032</td>\n",
       "      <td>...</td>\n",
       "      <td>2.423348</td>\n",
       "      <td>0.537768</td>\n",
       "      <td>1.198011</td>\n",
       "      <td>0.664543</td>\n",
       "      <td>0.627196</td>\n",
       "      <td>0.640176</td>\n",
       "      <td>1.830511</td>\n",
       "      <td>0.601302</td>\n",
       "      <td>1.147513</td>\n",
       "      <td>0.564163</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 1089 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         abdomen      abil       abl   absolut      abus   academi    accept  \\\n",
       "Topic0  0.605911  0.621080  0.699753  0.614075  0.546653  0.598405  0.641164   \n",
       "Topic1  0.622789  1.198479  3.534938  0.547543  1.182509  0.612230  2.857797   \n",
       "Topic2  1.137761  1.195480  2.405697  1.751575  0.629416  1.154473  5.381602   \n",
       "\n",
       "        accomplish    accord   account    ...         wors    worsen  \\\n",
       "Topic0    0.619712  0.594675  0.588146    ...     0.637057  2.389329   \n",
       "Topic1    0.670544  0.547127  0.630968    ...     0.624739  0.632156   \n",
       "Topic2    1.226811  1.227185  1.198032    ...     2.423348  0.537768   \n",
       "\n",
       "        worthless       wow     write     wrong       www      year    yellow  \\\n",
       "Topic0   0.703087  0.583808  0.556614  0.559208  0.570977  0.609959  0.611370   \n",
       "Topic1   0.579477  1.785210  2.437571  1.785731  2.441624  1.168159  0.664371   \n",
       "Topic2   1.198011  0.664543  0.627196  0.640176  1.830511  0.601302  1.147513   \n",
       "\n",
       "            yoga  \n",
       "Topic0  2.396541  \n",
       "Topic1  1.128858  \n",
       "Topic2  0.564163  \n",
       "\n",
       "[3 rows x 1089 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# defining topic keywords \n",
    "# Topic-Keyword Matrix\n",
    "df_topic_keywords = pd.DataFrame(best_lda_model.components_)\n",
    "\n",
    "# Assign Column and Index\n",
    "df_topic_keywords.columns = vectorizer.get_feature_names()\n",
    "df_topic_keywords.index = topicnames\n",
    "\n",
    "# View\n",
    "df_topic_keywords.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Word 0</th>\n",
       "      <th>Word 1</th>\n",
       "      <th>Word 2</th>\n",
       "      <th>Word 3</th>\n",
       "      <th>Word 4</th>\n",
       "      <th>Word 5</th>\n",
       "      <th>Word 6</th>\n",
       "      <th>Word 7</th>\n",
       "      <th>Word 8</th>\n",
       "      <th>Word 9</th>\n",
       "      <th>Word 10</th>\n",
       "      <th>Word 11</th>\n",
       "      <th>Word 12</th>\n",
       "      <th>Word 13</th>\n",
       "      <th>Word 14</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Topic 0</th>\n",
       "      <td>help</td>\n",
       "      <td>sad</td>\n",
       "      <td>feel</td>\n",
       "      <td>cri</td>\n",
       "      <td>overcom</td>\n",
       "      <td>tip</td>\n",
       "      <td>studi</td>\n",
       "      <td>depress</td>\n",
       "      <td>exercis</td>\n",
       "      <td>health</td>\n",
       "      <td>shown</td>\n",
       "      <td>bodi</td>\n",
       "      <td>don</td>\n",
       "      <td>effect</td>\n",
       "      <td>happi</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Topic 1</th>\n",
       "      <td>tip</td>\n",
       "      <td>fear</td>\n",
       "      <td>mani</td>\n",
       "      <td>deal</td>\n",
       "      <td>overcom</td>\n",
       "      <td>feel</td>\n",
       "      <td>sad</td>\n",
       "      <td>peopl</td>\n",
       "      <td>experi</td>\n",
       "      <td>just</td>\n",
       "      <td>want</td>\n",
       "      <td>joy</td>\n",
       "      <td>depress</td>\n",
       "      <td>think</td>\n",
       "      <td>someon</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Topic 2</th>\n",
       "      <td>fear</td>\n",
       "      <td>feel</td>\n",
       "      <td>tip</td>\n",
       "      <td>help</td>\n",
       "      <td>sad</td>\n",
       "      <td>overcom</td>\n",
       "      <td>life</td>\n",
       "      <td>thing</td>\n",
       "      <td>time</td>\n",
       "      <td>like</td>\n",
       "      <td>make</td>\n",
       "      <td>anger</td>\n",
       "      <td>don</td>\n",
       "      <td>experi</td>\n",
       "      <td>joy</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Word 0 Word 1 Word 2 Word 3   Word 4   Word 5 Word 6   Word 7  \\\n",
       "Topic 0   help    sad   feel    cri  overcom      tip  studi  depress   \n",
       "Topic 1    tip   fear   mani   deal  overcom     feel    sad    peopl   \n",
       "Topic 2   fear   feel    tip   help      sad  overcom   life    thing   \n",
       "\n",
       "          Word 8  Word 9 Word 10 Word 11  Word 12 Word 13 Word 14  \n",
       "Topic 0  exercis  health   shown    bodi      don  effect   happi  \n",
       "Topic 1   experi    just    want     joy  depress   think  someon  \n",
       "Topic 2     time    like    make   anger      don  experi     joy  "
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#get top 15 keywords for each doc\n",
    "\n",
    "\n",
    "# Show top n keywords for each topic\n",
    "def show_topics(vectorizer=vectorizer, lda_model=lda_model, n_words=20):\n",
    "    keywords = np.array(vectorizer.get_feature_names())\n",
    "    topic_keywords = []\n",
    "    for topic_weights in lda_model.components_:\n",
    "        top_keyword_locs = (-topic_weights).argsort()[:n_words]\n",
    "        topic_keywords.append(keywords.take(top_keyword_locs))\n",
    "    return topic_keywords\n",
    "\n",
    "topic_keywords = show_topics(vectorizer=vectorizer, lda_model=best_lda_model, n_words=15)        \n",
    "\n",
    "# Topic - Keywords Dataframe\n",
    "df_topic_keywords = pd.DataFrame(topic_keywords)\n",
    "df_topic_keywords.columns = ['Word '+str(i) for i in range(df_topic_keywords.shape[1])]\n",
    "df_topic_keywords.index = ['Topic '+str(i) for i in range(df_topic_keywords.shape[0])]\n",
    "df_topic_keywords"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['fear', 'feel', 'tip', 'help', 'sad', 'overcom', 'life', 'thing', 'time', 'like', 'make', 'anger', 'don', 'experi', 'joy']\n"
     ]
    }
   ],
   "source": [
    "#Given a piece of text, predicting the topic in document\n",
    "\n",
    "def predict_topic(text):\n",
    "    global sent_to_words\n",
    "    \n",
    "    mytext_2 = list(sent_to_words(text))\n",
    "    #print(mytext_2)\n",
    "    \n",
    "    mytext_3 =[]\n",
    "     \n",
    "    for i in mytext_2 :\n",
    "        \n",
    "        tokens=i\n",
    "        stopped_tokens = [i for i in tokens if not i in en_stop]\n",
    "        #print(stopped_tokens)\n",
    "        stemmed_tokens = [p_stemmer.stem(i) for i in stopped_tokens]\n",
    "        #print(stemmed_tokens)\n",
    "        mytext_3.append(' '.join(stemmed_tokens))\n",
    "        #print(mytext_3)\n",
    "    \n",
    "        mytext_4 = vectorizer.transform(mytext_3)\n",
    "    \n",
    "    topic_probability_scores = best_lda_model.transform(mytext_4)\n",
    "    topic = df_topic_keywords.iloc[np.argmax(topic_probability_scores), :].values.tolist()\n",
    "    return topic, topic_probability_scores\n",
    "\n",
    "\n",
    "mytext = [\"how to overcome Fear.Feeling worried\"]\n",
    "topic, prob_scores = predict_topic(text = mytext)\n",
    "print(topic)\n",
    "    \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Topic KeyWords:  ['fear', 'feel', 'tip', 'help', 'sad', 'overcom', 'life', 'thing', 'time', 'like', 'make', 'anger', 'don', 'experi', 'joy']\n",
      "Topic Prob Scores of text:  [[0.1 0.1 0.8]]\n",
      "Most Similar Doc's Probs:   [[0. 0. 1.]\n",
      " [0. 0. 1.]\n",
      " [0. 0. 1.]\n",
      " [0. 0. 1.]\n",
      " [0. 0. 1.]]\n",
      "\n",
      " Tip to Overcome Anger : Take a timeout Timeouts aren't just for kids. Give yourself short breaks during times of the day that tend to be stressful. A few moments of quiet time might help you feel better prepared to handle what's ahead without getting irritated or angry.\n",
      "\n",
      "\n",
      " Tip to Overcome Sadness : Shower or bathe in warm water. Research has shown that physical warmth has a comforting effect. Taking a warm bath or a hot shower will help you relax. It may also help you soothe your feelings of sadness\n"
     ]
    }
   ],
   "source": [
    "#Given a piece of Text, predicting the documents that are related to it most closely\n",
    "\n",
    "from sklearn.metrics.pairwise import euclidean_distances\n",
    "\n",
    "def similar_documents(text, doc_topic_probs, documents = data, top_n=5, verbose=False):\n",
    "    topic, x  = predict_topic(text)\n",
    "    dists = euclidean_distances(x.reshape(1, -1), doc_topic_probs)[0]\n",
    "    doc_ids = np.argsort(dists)[:top_n]\n",
    "    if verbose:        \n",
    "        print(\"Topic KeyWords: \", topic)\n",
    "        print(\"Topic Prob Scores of text: \", np.round(x, 1))\n",
    "        print(\"Most Similar Doc's Probs:  \", np.round(doc_topic_probs[doc_ids], 1))\n",
    "    return doc_ids, np.take(documents, doc_ids)\n",
    "\n",
    "# Get similar documents\n",
    "\n",
    "mytext = [\"how to maintain Joy.Feeling happy.Glad \"]\n",
    "doc_ids, docs = similar_documents(text=mytext, doc_topic_probs=lda_output, documents = data, top_n=5, verbose=True)\n",
    "print('\\n', docs[4][:700])\n",
    "print('\\n\\n', docs[3][:500])\n",
    "#print('\\n\\n', docs[2][:500])\n",
    "#print('\\n\\n', docs[1][:500])\n",
    "#print('\\n\\n', docs[0][:500])\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}