Company Clustering-Agglomerative.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package stopwords to C:\\Users\\Mr.\n",
      "[nltk_data]     Priyan\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n"
     ]
    }
   ],
   "source": [
    "# load required packages\n",
    "import nltk\n",
    "from nltk.corpus import PlaintextCorpusReader\n",
    "import pandas as pd\n",
    "import re\n",
    "import matplotlib.pyplot as plt\n",
    "nltk.download('stopwords')\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.stem.porter import PorterStemmer\n",
    "from __future__ import print_function\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.cluster import AgglomerativeClustering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Defining Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def find_number_clusters(tfidf):\n",
    "    import time\n",
    "    wcss = []\n",
    "    start = time.time()\n",
    "    for i in range(1, 10):\n",
    "        kmeans = KMeans(n_clusters = i, init = 'k-means++')\n",
    "        kmeans.fit(tfidf)\n",
    "        wcss.append(kmeans.inertia_)\n",
    "    end = time.time()\n",
    "    print(\"execution time:\",round(((end-start)/60),2),\"minutes\")\n",
    "    plt.plot(range(1, 10), wcss)\n",
    "    plt.title('The Elbow Method')\n",
    "    plt.xlabel('Number of clusters')\n",
    "    plt.ylabel('WCSS')\n",
    "    plt.show()  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def apply_kmeans(num_clusters,random_state,tfidf):\n",
    "    km = KMeans(n_clusters=num_clusters, random_state =random_state)\n",
    "    km.fit(tfidf)\n",
    "    clusters = km.labels_.tolist()\n",
    "    return km,clusters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def apply_agglomerative(num_clusters,tfidf):\n",
    "    hc = AgglomerativeClustering(n_clusters=num_clusters, affinity='euclidean',linkage='ward')\n",
    "    hc.fit(tfidf.todense())\n",
    "    clusters=hc.labels_.tolist()\n",
    "    return hc,clusters    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def freq_words(indexes,new_corpus,terms,X):\n",
    "    cluster_descriptions=\"\"\n",
    "    for i in (indexes):\n",
    "        cluster_descriptions = cluster_descriptions+new_corpus[i]+\" \"\n",
    "    cluster_descriptions=cluster_descriptions[:-1]\n",
    "    from collections import Counter\n",
    "    Counter = Counter(cluster_descriptions.split())\n",
    "    #freq_words has most commonly occuring words within parent cluster ranked in descending order\n",
    "    freq_words=Counter.most_common()\n",
    "    # freq_terms has most commonly occuring terms(from tfidf matrix) within parent cluster ranked in descending order along with \n",
    "    #frequencies\n",
    "    freq_terms_with_count = []\n",
    "    for word in freq_words:\n",
    "        if word[0] in set(terms):\n",
    "            freq_terms_with_count.append(word)\n",
    "    #sorted list of most occuring terms in cluster 3 , subcluster 1\n",
    "    freq_terms = []\n",
    "    for i in range(len(freq_terms_with_count)):\n",
    "        freq_terms.append(freq_terms_with_count[i][0])\n",
    "    # top X most occuring terms from that cluster\n",
    "    # here is where you can play around with the X value\n",
    "    X_freq_terms=freq_terms[0:X]\n",
    "    # freq terms after X terms removed\n",
    "    freq_terms_with_count = freq_terms_with_count[X:] \n",
    "    return X_freq_terms, freq_terms_with_count   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def print_agg_cluster(df, clusters, num_clusters, cluster_level, words, num_cluster_topics=3, num_cluster_companies=2) :\n",
    "    \n",
    "    print(\"Top terms per cluster:\")\n",
    "    print()\n",
    "    #sort cluster centers by proximity to centroid\n",
    "    #order_centroids = km.cluster_centers_.argsort()[:, ::-1] \n",
    "    for i in range(num_clusters):\n",
    "        X_freq_terms, freq_terms_with_count = freq_words(indexes=df[df[cluster_level]==i].index.values,new_corpus=new_corpus,\\\n",
    "                                                 terms=terms,X=0)\n",
    "        print(\"Cluster %d words:\" % int(i), end='')\n",
    "\n",
    "              \n",
    "        flag=0\n",
    "        ind=0\n",
    "        while flag < num_cluster_topics: # top n words per cluster\n",
    "            check_word = freq_terms_with_count[ind][0]\n",
    "            if check_word not in set(words):\n",
    "                print(' %s' % check_word, end=',')\n",
    "                flag=flag+1\n",
    "                ind = ind+1\n",
    "            else:\n",
    "                ind = ind+1\n",
    "        \n",
    "        \n",
    "        print() #add whitespace\n",
    "        print() #add whitespace\n",
    "\n",
    "        print(\"Cluster %d Companies:\" % int(i), end='')\n",
    "        #print()\n",
    "        for title in df[df[cluster_level]==i][\"Company\"].sample(num_cluster_companies).values.tolist():\n",
    "            print(' %s,' % title, end='')\n",
    "            #uncomment below to see the Sector and Industry of each Company\n",
    "            #print(\"Sector:\",merged_df.Sector[merged_df.Company==title].values[0],\"Industry:\",merged_df.Industry[merged_df.Company==title].values[0])\n",
    "        print() #add whitespace\n",
    "        print() #add whitespace\n",
    "\n",
    "    print(\"Split of classes within clusters:\")\n",
    "    print(df[cluster_level].value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def save_topics_words(df, clusters, num_clusters, cluster_level, words, num_cluster_topics=10) :\n",
    "    \n",
    "    if cluster_level=='Clusters':\n",
    "        topics_dict = {}\n",
    "        #order_centroids = km.cluster_centers_.argsort()[:, ::-1] \n",
    "        for i in range(num_clusters):\n",
    "            X_freq_terms, freq_terms_with_count = freq_words(indexes=df[df[cluster_level]==i].index.values,new_corpus=new_corpus,\\\n",
    "                                                 terms=terms,X=0)\n",
    "            topics_list = []\n",
    "            flag=0\n",
    "            ind=0\n",
    "            while flag<num_cluster_topics:\n",
    "                check_word = freq_terms_with_count[ind][0]\n",
    "                check_word = check_word.split(' ')            \n",
    "                if len(set(check_word).intersection(words))==0:\n",
    "                    topics_list.append(' '.join(check_word))\n",
    "                    flag = flag + 1\n",
    "                    ind = ind + 1\n",
    "                else:\n",
    "                    ind = ind + 1\n",
    "            topics_dict.update({str(i):topics_list})    \n",
    "        return topics_dict\n",
    "    \n",
    "    if cluster_level=='Subclusters_level_1':\n",
    "        level1= df['Clusters'].unique()[0]\n",
    "        topics_dict = {}\n",
    "        #order_centroids = km.cluster_centers_.argsort()[:, ::-1] \n",
    "        for i in range(num_clusters):\n",
    "            X_freq_terms, freq_terms_with_count = freq_words(indexes=df[df[cluster_level]==i].index.values,new_corpus=new_corpus,\\\n",
    "                                                 terms=terms,X=0)\n",
    "            topics_list = []\n",
    "            flag = 0\n",
    "            ind = 0\n",
    "            while flag<num_cluster_topics:                    \n",
    "                check_word = freq_terms_with_count[ind][0]\n",
    "                check_word = check_word.split(' ')            \n",
    "                if len(set(check_word).intersection(words))==0:\n",
    "                    topics_list.append(' '.join(check_word))\n",
    "                    flag = flag + 1\n",
    "                    ind = ind + 1\n",
    "                else:\n",
    "                    ind = ind + 1\n",
    "            \n",
    "            topics_dict.update({str(level1)+str(i):topics_list})\n",
    "        return topics_dict\n",
    "    \n",
    "    if cluster_level=='Subclusters_level_2':\n",
    "        level1= int(df['Clusters'].unique()[0])\n",
    "        level2 =int(df['Subclusters_level_1'].unique()[0]) \n",
    "        topics_dict = {}\n",
    "        #order_centroids = km.cluster_centers_.argsort()[:, ::-1] \n",
    "        for i in range(num_clusters):\n",
    "            X_freq_terms, freq_terms_with_count = freq_words(indexes=df[df[cluster_level]==i].index.values,new_corpus=new_corpus,\\\n",
    "                                                 terms=terms,X=0)\n",
    "            topics_list = []\n",
    "            flag = 0\n",
    "            ind = 0\n",
    "            while flag<num_cluster_topics:\n",
    "                check_word = freq_terms_with_count[ind][0]\n",
    "                check_word = check_word.split(' ')            \n",
    "                if len(set(check_word).intersection(words))==0:\n",
    "                    topics_list.append(' '.join(check_word))\n",
    "                    flag = flag + 1\n",
    "                    ind = ind + 1\n",
    "                else:\n",
    "                    ind = ind + 1\n",
    "            \n",
    "            topics_dict.update({str(level1)+str(level2)+str(i):topics_list})\n",
    "        return topics_dict\n",
    "    \n",
    "    if cluster_level=='Subclusters_level_3':\n",
    "        level1= int(df['Clusters'].unique()[0])\n",
    "        level2 =int(df['Subclusters_level_1'].unique()[0])\n",
    "        level3 =int(df['Subclusters_level_2'].unique()[0])\n",
    "        topics_dict = {}\n",
    "        #order_centroids = km.cluster_centers_.argsort()[:, ::-1] \n",
    "        for i in range(num_clusters):\n",
    "            X_freq_terms, freq_terms_with_count = freq_words(indexes=df[df[cluster_level]==i].index.values,new_corpus=new_corpus,\\\n",
    "                                                 terms=terms,X=0)\n",
    "            topics_list = []\n",
    "            flag = 0\n",
    "            ind = 0\n",
    "            while flag<num_cluster_topics:\n",
    "                check_word = freq_terms_with_count[ind][0]\n",
    "                check_word = check_word.split(' ')            \n",
    "                if len(set(check_word).intersection(words))==0:\n",
    "                    topics_list.append(' '.join(check_word)) \n",
    "                    flag = flag + 1\n",
    "                    ind = ind + 1\n",
    "                else:\n",
    "                    ind = ind + 1\n",
    "            \n",
    "            topics_dict.update({str(level1)+str(level2)+str(level3)+str(i):topics_list})\n",
    "        return topics_dict"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Reading from the text files to create corpus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Company Profile Corpus Statistics\n",
      "Number of profiles: 3558\n",
      "Number of paragraphs: 3558\n",
      "Number of sentences: 23988\n",
      "Number of words: 790836\n"
     ]
    }
   ],
   "source": [
    "#change path below to folder where you place the files\n",
    "corpus_path = \"C:\\\\Users\\\\Mr. Priyan\\\\Dropbox\\\\Data Science\\\\DDL\\\\company_profiles\" \n",
    "corpus = PlaintextCorpusReader(corpus_path, \".*\\.txt\")\n",
    "print(\"Company Profile Corpus Statistics\")\n",
    "print(\"Number of profiles: \" + str(len(corpus.fileids())))\n",
    "print(\"Number of paragraphs: \" + str(len(corpus.paras())))\n",
    "print(\"Number of sentences: \" + str(len(corpus.sents())))\n",
    "print(\"Number of words: \" + str(len(corpus.words())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Changing to list of lists"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "list_corpus = []\n",
    "#Converting corpus to list of lists\n",
    "for i in range(0,len(corpus.fileids())):\n",
    "    list_corpus.append(list(corpus.words(corpus.fileids()[i])))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing the corpus"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Creating list of common words to later remove from the corpus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "common_words = ['product','products','service','services','solution','solutions','segment','segments','offering','offerings',\\\n",
    "                'market','markets','provide','provides','provided','industry','industries','sells','sell','sold','state',\\\n",
    "                'states','headquarter','headquarters','headquartered', 'canada', 'california','texas','america','management',\\\n",
    "               'manages','manage', 'management','business','businesses','operation','operations','state','states',\\\n",
    "                'subsidiary','subsidiaries','approximate','approximately','group','groups','include','including','includes',\\\n",
    "                'base','bases','location','locations','use','uses','store','stores','unit','units','item','items',\\\n",
    "                'outlet','outlets','offer','offers','process','processes','processing','corporation','corporate','custom',\\\n",
    "               'offering','offerings','operate','operations','new york','africa','canada','america','international',\\\n",
    "                'manufacture','manufactures','manufacturer','manufacturers','brand','brands','benefit','benefits','customer',\\\n",
    "                'customers','industry','industries','firm','firms','commercial','commercials','com','new','speciality',\\\n",
    "                'specialities','used','unity','manufacturing','united','produce','produces','operates','operator','industrial',\\\n",
    "                'sales','company','companies','europe', 'asia','americas','middle east','sector','sectors','together',\\\n",
    "                'operators','production','productions','sale','private','public'\\\n",
    "               ]      \n",
    "               \n",
    "# to see the common words sorted in order\n",
    "common_words.sort() \n",
    "#common_words      \n",
    "               "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " Pre-Processing Steps :-\n",
    "- Removing punctuation\n",
    "- Converting to lowercase\n",
    "- Stemming\n",
    "- Removing stop words\n",
    "- Removing common words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "execution time: 10.41 minutes\n"
     ]
    }
   ],
   "source": [
    "# Preprocessing\n",
    "#commenting everything as I had already saved new_corpus after the first time. \n",
    "\n",
    "import time\n",
    "start = time.time()\n",
    "\n",
    "new_corpus = []    \n",
    "for i in range(0,len(list_corpus)):\n",
    "    #remove punctuation\n",
    "    review = re.sub('[^a-zA-Z]', ' ', str(list_corpus[i]))\n",
    "    #convert all to lowercase\n",
    "    review = review.lower()\n",
    "    review = review.split()\n",
    "    #stemming\n",
    "    ps = PorterStemmer()\n",
    "    review = [ps.stem(word) for word in review if not word in set(stopwords.words('english')) if not word in set(common_words)]\n",
    "    review = ' '.join(review)\n",
    "    new_corpus.append(review)\n",
    "\n",
    "end = time.time()\n",
    "print(\"execution time:\",round(((end-start)/60),2),\"minutes\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Stored 'new_corpus' (list)\n"
     ]
    }
   ],
   "source": [
    "# to store the new_corpus list instead of having to run it each time \n",
    "#%store new_corpus \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#restoring new_corpus which was stored already\n",
    "%store -r "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original description:\n",
      "United States Steel Corporation produces and sells flat - rolled and tubular steel products primarily in North America and Europe . It operates through three segments : Flat - Rolled Products ( Flat - Rolled ), U . S . Steel Europe ( USSE ), and Tubular Products ( Tubular ). The Flat - Rolled segment offers slabs , rounds , strip mill plates , sheets , and tin mill products . This segment serves customers in the service center , conversion , automotive , construction , container , and appliance and electrical markets . The USSE segment provides slabs , sheets , strip mill plates , tin mill products , and spiral welded pipes , as well as heating radiators and refractory ceramic materials . This segment serves customers in the construction , service center , conversion , container , transportation , appliance and electrical , oil , gas , and petrochemical markets . The Tubular segment offers seamless and electric resistance welded steel casing and tubing products ; and standard and line pipe and mechanical tubing products primarily to customers in the oil , gas , and petrochemical markets . The company also provides railroad services ; and owns , develops , and manages various real estate assets . United States Steel Corporation was founded in 1901 and is headquartered in Pittsburgh , Pennsylvania .\n",
      "\n",
      "After pre-processing:\n",
      "steel flat roll tubular steel primarili north three flat roll flat roll u steel uss tubular tubular flat roll slab round strip mill plate sheet tin mill serv center convers automot construct contain applianc electr uss slab sheet strip mill plate tin mill spiral weld pipe well heat radiat refractori ceram materi serv construct center convers contain transport applianc electr oil ga petrochem tubular seamless electr resist weld steel case tube standard line pipe mechan tube primarili oil ga petrochem also railroad own develop variou real estat asset steel found pittsburgh pennsylvania\n",
      "\n"
     ]
    }
   ],
   "source": [
    "#checking the content of a random document from new_corpus and comparing it to original description\n",
    "import random\n",
    "rand_num = random.randrange(1,3559,1)\n",
    "print(\"Original description:\")\n",
    "print(' '.join(corpus.words(corpus.fileids()[rand_num])))\n",
    "print()\n",
    "print(\"After pre-processing:\")\n",
    "print(new_corpus[rand_num])\n",
    "print()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Extraction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using TF-IDF "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "\n",
    "#define vectorizer parameters\n",
    "#original: max_df=0.8, min_df=0.05, max_features=200000\n",
    "tfidf_vectorizer = TfidfVectorizer(max_df=0.8, min_df=0.02, max_features=200000,\n",
    "                                 stop_words='english',\n",
    "                                 use_idf=True, tokenizer=None, ngram_range=(1,3))\n",
    "#fit the vectorizer to corpus\n",
    "tfidf_matrix = tfidf_vectorizer.fit_transform(new_corpus) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3558, 735)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# to see shape of tfidf matrix\n",
    "tfidf_matrix.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#tdidf feature names\n",
    "terms = tfidf_vectorizer.get_feature_names()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#terms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "execution time: 0.63 minutes\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEWCAYAAABxMXBSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x21b72b43588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "find_number_clusters(tfidf=tfidf_matrix.todense())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Apply Agglomerative Clustering\n",
    "hc, clusters = apply_agglomerative(num_clusters=5,tfidf= tfidf_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# saving the clustering model\n",
    "#from sklearn.externals import joblib\n",
    "\n",
    "#uncomment the below to save your model \n",
    "#joblib.dump(km,  '5clusters.pkl')\n",
    "#since I've already run my model I am loading from the pickle\n",
    "#Uncomment to load the model with 5 clusters\n",
    "num_clusters = 5\n",
    "#km = joblib.load('5clusters.pkl')\n",
    "#clusters = km.labels_.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Ticker</th>\n",
       "      <th>Company</th>\n",
       "      <th>Sector</th>\n",
       "      <th>Industry</th>\n",
       "      <th>Country</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5575</th>\n",
       "      <td>RWXL</td>\n",
       "      <td>UBS ETRACS M Py 2xLvg DJ Intl RelEst ETN</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Exchange Traded Fund</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4432</th>\n",
       "      <td>NBR</td>\n",
       "      <td>Nabors Industries Ltd.</td>\n",
       "      <td>Basic Materials</td>\n",
       "      <td>Oil &amp; Gas Drilling &amp; Exploration</td>\n",
       "      <td>Bermuda</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3247</th>\n",
       "      <td>IDN</td>\n",
       "      <td>Intellicheck Mobilisa, Inc.</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Application Software</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3897</th>\n",
       "      <td>LIQT</td>\n",
       "      <td>LiqTech International, Inc.</td>\n",
       "      <td>Industrial Goods</td>\n",
       "      <td>Pollution &amp; Treatment Controls</td>\n",
       "      <td>Denmark</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1446</th>\n",
       "      <td>CUR</td>\n",
       "      <td>Neuralstem, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Ticker                                   Company            Sector  \\\n",
       "5575   RWXL  UBS ETRACS M Py 2xLvg DJ Intl RelEst ETN         Financial   \n",
       "4432    NBR                    Nabors Industries Ltd.   Basic Materials   \n",
       "3247    IDN               Intellicheck Mobilisa, Inc.        Technology   \n",
       "3897   LIQT               LiqTech International, Inc.  Industrial Goods   \n",
       "1446    CUR                          Neuralstem, Inc.        Healthcare   \n",
       "\n",
       "                              Industry  Country  \n",
       "5575              Exchange Traded Fund      USA  \n",
       "4432  Oil & Gas Drilling & Exploration  Bermuda  \n",
       "3247              Application Software      USA  \n",
       "3897    Pollution & Treatment Controls  Denmark  \n",
       "1446                     Biotechnology      USA  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# creating a dataframe which contains additional information about each company\n",
    "df = pd.read_csv(\"C:\\\\Users\\\\Mr. Priyan\\\\Dropbox\\\\Data Science\\\\DDL\\\\company_info.csv\")\n",
    "df.sample(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Clusters</th>\n",
       "      <th>Ticker</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>A-profile.txt</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>AA-profile.txt</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>AAC-profile.txt</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>AAL-profile.txt</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>AAMC-profile.txt</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Clusters            Ticker\n",
       "0         1     A-profile.txt\n",
       "1         1    AA-profile.txt\n",
       "2         1   AAC-profile.txt\n",
       "3         1   AAL-profile.txt\n",
       "4         1  AAMC-profile.txt"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Creating a dataframe that contains the company ticker along with the cluster number\n",
    "df_corpus = pd.DataFrame({'Ticker':corpus.fileids(), 'Clusters':clusters})\n",
    "df_corpus.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Clusters</th>\n",
       "      <th>Ticker</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>AA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>AAC</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>AAL</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>AAMC</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Clusters Ticker\n",
       "0         1      A\n",
       "1         1     AA\n",
       "2         1    AAC\n",
       "3         1    AAL\n",
       "4         1   AAMC"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Removing the \"-profile.txt\" from the Ticker\n",
    "import re\n",
    "for i in range(0,len(df_corpus)):\n",
    "    df_corpus.loc[i,'Ticker']=re.sub(\"-profile.txt\",\"\",df_corpus.loc[i,'Ticker'])\n",
    "df_corpus.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# merging the two dataframe on the Ticker\n",
    "merged_df = pd.merge(left=df, right=df_corpus, how='inner', on=\"Ticker\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Ticker</th>\n",
       "      <th>Company</th>\n",
       "      <th>Sector</th>\n",
       "      <th>Industry</th>\n",
       "      <th>Country</th>\n",
       "      <th>Clusters</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>768</th>\n",
       "      <td>CRAY</td>\n",
       "      <td>Cray Inc.</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Diversified Computer Systems</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>AGIO</td>\n",
       "      <td>Agios Pharmaceuticals, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>598</th>\n",
       "      <td>CECE</td>\n",
       "      <td>CECO Environmental Corp.</td>\n",
       "      <td>Industrial Goods</td>\n",
       "      <td>Pollution &amp; Treatment Controls</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1998</th>\n",
       "      <td>MBWM</td>\n",
       "      <td>Mercantile Bank Corp.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Regional - Midwest Banks</td>\n",
       "      <td>USA</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3315</th>\n",
       "      <td>UWN</td>\n",
       "      <td>Nevada Gold &amp; Casinos Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Resorts &amp; Casinos</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Ticker                      Company            Sector  \\\n",
       "768    CRAY                    Cray Inc.        Technology   \n",
       "92     AGIO  Agios Pharmaceuticals, Inc.        Healthcare   \n",
       "598    CECE     CECO Environmental Corp.  Industrial Goods   \n",
       "1998   MBWM        Mercantile Bank Corp.         Financial   \n",
       "3315    UWN   Nevada Gold & Casinos Inc.          Services   \n",
       "\n",
       "                            Industry Country  Clusters  \n",
       "768     Diversified Computer Systems     USA         1  \n",
       "92                     Biotechnology     USA         2  \n",
       "598   Pollution & Treatment Controls     USA         1  \n",
       "1998        Regional - Midwest Banks     USA         3  \n",
       "3315               Resorts & Casinos     USA         1  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "merged_df.sample(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cluster 0 : Financial :- % correct categories is  89.68\n",
      "Cluster 1 : Services :- % correct categories is  25.3\n",
      "Cluster 2 : Healthcare :- % correct categories is  98.47\n",
      "Cluster 3 : Financial :- % correct categories is  99.72\n",
      "Cluster 4 : Basic Materials :- % correct categories is  71.12\n"
     ]
    }
   ],
   "source": [
    "# Printing the most recurring Sector(taken from previous dataframe) in each cluster\n",
    "for i in range(num_clusters):\n",
    "    print(\"Cluster\",i,\":\",merged_df[merged_df.Clusters==i].Sector.mode()[0],\":- % correct categories is \", \\\n",
    "    round(len(merged_df[merged_df.Clusters==i][merged_df[merged_df.Clusters==i].Sector==\\\n",
    "    merged_df[merged_df.Clusters==i].Sector.mode()[0]])/len(merged_df[merged_df.Clusters==i])*100,2))   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top terms per cluster:\n",
      "\n",
      "Cluster 0 words: invest, fund, real, estat,\n",
      "\n",
      "Cluster 0 Companies: Solar Capital Ltd., Ashford Hospitality Prime, Inc., National Holdings Corporation, CBL & Associates Properties Inc., Tri-Continental Corporation,\n",
      "\n",
      "Cluster 1 words: equip, applic, technolog, base,\n",
      "\n",
      "Cluster 1 Companies: Altisource Residential Corporation, Towerstream Corporation, Selective Insurance Group Inc., The Sherwin-Williams Company, CONMED Corporation,\n",
      "\n",
      "Cluster 2 words: develop, treatment, clinic, phase,\n",
      "\n",
      "Cluster 2 Companies: Catalyst Biosciences, Inc., Oragenics Inc., Cymabay Therapeutics, Inc., Syndax Pharmaceuticals, Inc., Celgene Corporation,\n",
      "\n",
      "Cluster 3 words: loan, bank, deposit, account,\n",
      "\n",
      "Cluster 3 Companies: Pacific Mercantile Bancorp, Trustmark Corporation, Territorial Bancorp Inc., United Security Bancshares, Fidelity Southern Corporation,\n",
      "\n",
      "Cluster 4 words: ga, natur, oil, energi,\n",
      "\n",
      "Cluster 4 Companies: Isramco Inc., Genie Energy Ltd., Atmos Energy Corporation, Vaalco Energy Inc., Black Stone Minerals, L.P.,\n",
      "\n",
      "Split of classes within clusters:\n",
      "1    2241\n",
      "0     407\n",
      "3     352\n",
      "2     326\n",
      "4     232\n",
      "Name: Clusters, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "print_agg_cluster(df=merged_df,clusters=clusters,num_clusters=num_clusters,cluster_level='Clusters',words=[],num_cluster_topics=4, num_cluster_companies=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "topics = save_topics_words(df=merged_df,clusters=clusters,num_clusters=num_clusters,cluster_level='Clusters',words=[],num_cluster_topics=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Splitting clusters 1, 0 & 4 into sub-clusters"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Splitting the biggest cluster , cluster 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "execution time: 0.42 minutes\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY8AAAEWCAYAAACe8xtsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJzt3Xd8VuX9//HXJwmEETYBkRWEgDIU\nJCJDwYp7guMrLlBbUetC66/Vtnbbamtbq7gQBakI4q4VVxUZskGRpbIhzCCyISHJ5/fHfaK3GJLc\nkDsn4/18PPLIfV/3GZ8TMe+c61zXOebuiIiIxCIh7AJERKTiUXiIiEjMFB4iIhIzhYeIiMRM4SEi\nIjFTeIiISMwUHlIhmdnvzOyFMthPmpm5mSUF7z82s5/Ee79loTSPxcxGm9mfSmNbUjEoPKRcMrPd\nUV/5ZrYv6v3Vpbyv0WaWc9A+F5TmPg5XVHjNP6i9cVDz6hJup0zCVqoOhYeUS+6eUvAFrAUujGob\nG4dd/jV6n+5+Qhz2cSRqm1nnqPdXAavCKkZE4SEVWXUzG2Nmu8xssZllFHxgZkeb2atmlmVmq8zs\njlLcb1szm21mO8zsTTNrGLXfi4JatgfdQscF7deb2VtRyy03swlR79eZWdci9vlvYEjU+8HAmOgF\nDnXMZnYO8EvgikLOqlqb2SfBz/B9M2tc3LEEn3Uzs/nBei8BNUr2o5PKQuEhFdlFwHigPvAfYDiA\nmSUAbwELgOZAf2CYmZ1dSvsdDNwAHA3kAo8G+20PjAOGAanAROAtM6sOTAZONbMEM2sGVAP6BOsd\nA6QAnxexzxeAQWaWGPwSrwPMKviwqGN293eBPwMvFXJWdRVwPdAEqA7cU9yxBMfzBpFAawi8DFwa\n009QKjyFh1Rk09x9orvnEflFVvBL8SQg1d3/4O457r4SeAYYVMS27gn+wi74er6IZf/t7ovcfQ9w\nP/B/ZpYIXAG87e4fuPsB4GGgJtA7qGEX0BXoB7wHrDezY4P3U909v4h9ZgJfAmcQOQMZc9Dnh3PM\nAKPc/St33wdMCOqjqGMBehIJv0fc/YC7vwLMKWY/UskkhV2AyBHYFPV6L1AjGBXVGjjazLZHfZ4I\nTC1iWw+7+69LuN91Ua/XEPlF2pjImciagg/cPd/M1hE5E4DI2cdpQLvg9XYiwdEreF+cMcB1RH6B\n9wXSoz47nGOGH/4MU4LXRR1LHrDev39X1TVIlaLwkMpoHbDK3dOLXfLwtIx63Qo4AGwFNgBdCj4w\nMwuWXR80TQYuBNoQ6UbaDlxNJDyGl2C/rwbLzXP3NWYWfXzFHXOst88u6lgcaG5mFhUgrYAVMe5D\nKjB1W0llNBvYaWa/MLOawXWCzmZ2Uilt/xoz62hmtYA/AK8EXWcTgPPNrL+ZVQN+BmQD04P1JgM/\nAmq6eyaRs4JzgEbAp8XtNOgmOx0obG5Gcce8GUgLro2URFHHMoPItZ47zCzJzC4BepRwu1JJKDyk\n0gl+kV9IpP9+FZGzgpFAvSJW+/lB8zy2FrHsv4HRRLp8agB3BPv9ErgGeCzY54VEhhjnBJ9/Bewm\n6Epy953ASuCToOaSHNtcd//BX/glOOaXg+9fHzxn5BD7OeSxBMdzCZEutG+IXB95rST1S+VhehiU\niIjESmceIiISM4WHiIjETOEhIiIxU3iIiEjMKu08j8aNG3taWlrYZYiIVBjz5s3b6u6pJVm20oZH\nWloac+fODbsMEZEKw8xKfKcAdVuJiEjMFB4iIhIzhYeIiMRM4SEiIjFTeIiISMwUHiIiEjOFh4iI\nxEzhEcXdeezDZSzesCPsUkREyjWFR5Ttew8wbvZarhwxkwXrthe/gohIFaXwiNKgdnVeuqkX9WtV\n5+qRs5i7elvYJYmIlEsKj4O0bFiLl27qSZM6yQx+bjbTVxT1QDkRkapJ4VGIZvVqMv6mnrRoUJPr\nR81h8ldZYZckIlKuxC08zKylmU0ys6VmttjM7gzaLw/e55tZxkHr3Gdmy83sSzM7O6r9nKBtuZnd\nG6+aozWpU4PxQ3vRNjWFG5+fywdLNpfFbkVEKoR4nnnkAj9z9+OAnsCtZtYRWARcAkyJXjj4bBDQ\nCTgHeMLMEs0sEXgcOBfoCFwZLBt3DWtXZ9yNPTmuWR1ueWEeb3++sSx2KyJS7sUtPNx9o7vPD17v\nApYCzd19qbt/WcgqFwPj3T3b3VcBy4Eewddyd1/p7jnA+GDZMlGvVjVe+MnJdG1Zn9vHzef1TzPL\natciIuVWmVzzMLM0oBswq4jFmgProt5nBm2Hai9sP0PNbK6Zzc3KKr3rFHVqVOP5G3pwcptG3D1h\nAS/NWVtq2xYRqYjiHh5mlgK8Cgxz951FLVpImxfR/sNG9xHunuHuGampJXoYVonVTk5i1PUn0Tc9\nlV+8upAxM1aX6vZFRCqSuIaHmVUjEhxj3f21YhbPBFpGvW8BbCiivczVqJbIiMHdOeO4pvzmzcU8\nM2VlGGWIiIQunqOtDHgWWOru/yjBKv8BBplZspm1AdKB2cAcIN3M2phZdSIX1f8Tr7qLk5yUyJPX\nnMj5XZrxwMSlDP9oWViliIiEJp7PMO8DXAssNLPPgrZfAsnAY0Aq8LaZfebuZ7v7YjObACwhMlLr\nVnfPAzCz24D3gETgOXdfHMe6i1UtMYF/DepKclICD7//Fdm5+dx9ZnsieSkiUvnFLTzcfRqFX68A\neP0Q6zwAPFBI+0RgYulVd+SSEhP42+UnUD0pgcc+Ws7+A3n88rzjFCAiUiXE88yj0ktMMP48sAvV\nkxJ4ZuoqsnPz+d2FnUhIUICISOWm8DhCCQnG7y/qFLmYPmUlObn5PDCwC4kKEBGpxBQepcDMuO/c\nY6mRlMCjHy0nOzefv112PEmJunWYiFROCo9SYmbcfVYHqgcX0XNy83lkUFeqKUBEpBJSeJSy205P\np0a1RP709lKyc/N5/OpuJCclhl2WiEip0p/FcfCTU4/hjxd34n9LNzN0zDz2H8gLuyQRkVKl8IiT\na3ul8dClXZiyLIvrR81hT3Zu2CWJiJQahUccXXFSK/7xfycwa9XXDHluNjv3Hwi7JBGRUqHwiLOB\n3Vow/KoT+Wzddq4dOYsdexUgIlLxKTzKwHldmvHUNd1ZunEXVz4zk693Z4ddkojIEVF4lJEzOjbl\nmSEZrMjazaARM9mya3/YJYmIHDaFRxnq1z6VUdefxPrt+xj09Ew27tgXdkkiIodF4VHGerdtzJgb\nerBlVzb/9/QM1m3bG3ZJIiIxU3iEICOtIWN/cjI79h7giqdnsGrrnrBLEhGJicIjJCe0rM+4oT3Z\nn5vPFU/PYNnmXWGXJCJSYgqPEHU6uh7jh/bEgUEjZrJ0Y1GPeBcRKT8UHiFr37QOLw3tSbXEBK58\nZiYLM3eEXZKISLEUHuXAMakpTLipFynJSVz1zEzmrfkm7JJERIqk8CgnWjWqxYSbetEopTrXPjuL\nmSu/DrskEZFDUniUI0fXr8mEm3pxdP2aXDdqNlOXZYVdkohIoRQe5UyTujUYP7QnaY1q8+Pn5/Lh\n0s1hlyQi8gMKj3KocUoy44f2pEPTOtz8wjzeXbQx7JJERL5H4VFO1a9VnbE3nkyX5vW49cVPefOz\n9WGXJCLyLYVHOVa3RjXG/PhkMlo3YNhLnzFu9tqwSxIRARQe5V5KchKjr+9B3/RU7nttIQ++8wX5\n+R52WSJSxSk8KoCa1RN5dkgGV53ciqcmr+C2cfP1XHQRCZXCo4JISkzggQGd+dV5x/HOok0MGjGT\nrF16qJSIhEPhUYGYGTf2PYanrunOF5t2MvCJT3RDRREJhcKjAjq701FMuKkX2bn5XPLEdKYt2xp2\nSSJSxSg8KqjjW9TnjVv7fDsbfbxGYolIGVJ4VGDN69fklVt60btdY+7VSCwRKUMKjwquTo1qPDck\ng6s1EktEypDCoxJISkzgTwM68+vzIyOxrtBILBGJM4VHJWFm/OTUyEisLzftZMDjn/CVRmKJSJwo\nPCqZgpFYOXn5XPrEdN3WXUTiQuFRCRWMxGreoCbXjZqje2KJSKlTeFRSzevX5OWbe3FKu8bc99pC\n/vLOUo3EEpFSo/CoxOrUqMazQzK4pmcrnp68kltfnM++HI3EEpEjF7fwMLOWZjbJzJaa2WIzuzNo\nb2hmH5jZsuB7g6D9NDPbYWafBV+/idrWOWb2pZktN7N741VzZZSUmMAfL46MxHp38SYGPTOTLbv2\nh12WiFRw8TzzyAV+5u7HAT2BW82sI3Av8KG7pwMfBu8LTHX3rsHXHwDMLBF4HDgX6AhcGWxHSqhg\nJNbT13Tnq027GPj4dI3EEpEjErfwcPeN7j4/eL0LWAo0By4Gng8Wex4YUMymegDL3X2lu+cA44Nt\nSIzO0kgsESklZXLNw8zSgG7ALKCpu2+ESMAATaIW7WVmC8zsHTPrFLQ1B9ZFLZMZtBW2n6FmNtfM\n5mZl6RdjYbq0qKeRWCJyxOIeHmaWArwKDHP3nUUsOh9o7e4nAI8BbxRsopBlCx025O4j3D3D3TNS\nU1OPpOxK7QcjsSZqJJaIxCau4WFm1YgEx1h3fy1o3mxmzYLPmwFbANx9p7vvDl5PBKqZWWMiZxot\nozbbAtgQz7qrgoKRWNf2bM3TU1by07EaiSUiJRfP0VYGPAssdfd/RH30H2BI8HoI8Gaw/FHBOphZ\nj6C2r4E5QLqZtTGz6sCgYBtyhJISE/jDxZ24/4KOvLdkE4NGzNBILBEpkXieefQBrgVOjxp+ex7w\nIHCmmS0DzgzeA1wGLDKzBcCjwCCPyAVuA94jctF9grsvjmPdVYqZ8eNT2kRGYm3ezcDHp/PlJo3E\nEpGimXvl7OvOyMjwuXPnhl1GhbIwcwc/fn4O+3LyeOKaEzk1XdeNRKoSM5vn7hklWVYzzOVbB4/E\nenGWRmKJSOEUHvI9R9evySu39ObU9Mb88vWF/FkjsUSkEAoP+YGU5CRGDs5gcK/WjJiyklvGztNI\nLBH5HoWHFCopMYHfX9SJ31zQkfeXbNZILBH5HoWHHJKZccMpbRhxbYZGYonI9yg8pFhndmzKhJt6\ncSAvn8uenM6Ur3TrF5GqTuEhJRI9Euv60XMYO2tN2CWJSIgUHlJiBSOx+qY35levL+KBt5doJJZI\nFaXwkJikJCfxzOAMhvRqzTNTV2kklkgVpfCQmCUlJvD7izvz2wsjI7EufXI6a77eE3ZZIlKGFB5y\n2K7v04bnrjuJ9dv3ccFj03hv8aawSxKRMqLwkCPyow5N+O/tp9CmcW1u+vc8/vLOUnLz8sMuS0Ti\nTOEhR6xlw1q8fHMvrj65FU9PXsnVI2dpQqFIJafwkFKRnJTIAwO78M8rTuDzzB2c/+g0Zq78Ouyy\nRCROFB5SqgZ2a8Ebt/ahTo0krh45i6cmr6Cy3vZfpCpTeEip63BUHf5z2ymc0+koHnznC4b+ex47\n9h0IuywRKUUKD4mLlOQkhl/Vjd9c0JFJX2zhouHTWLxhR9hliUgpUXhI3BTcWPGlm3qSfSCfgU9M\n56U5esCUSGWg8JC46966If+94xROSmvAL15dyM9fWcD+A5qVLlKRKTykTDROSWbMDSdz++ntmDA3\nk0ue0Kx0kYpM4SFlJjHB+NlZHRh1/Uls2KFZ6SIVmcJDytwPZqVP1Kx0kYpG4SGhaNEgMiv9mp6t\neHrKSq4aOYstOzUrXaSiUHhIaJKTEvnTgMis9IWZOzhPs9JFKgyFh4SuYFZ63ZqalS5SUSg8pFzQ\nrHSRikXhIeWGZqWLVBwKDylXvpuV3kuz0kXKsSLDw8xOMrOjot4PNrM3zexRM2sY//KkqureugFv\n33EKPdIa8otXF/L/XtasdJHypLgzj6eBHAAz6ws8CIwBdgAj4luaVHWNUpJ5/oYe3HF6O16el8nA\nJ6azeqtmpYuUB8WFR6K7bwteXwGMcPdX3f1+oF18SxOJzEq/O5iVvnHHPi7UrHSRcqHY8DCzpOB1\nf+CjqM+SClleJC6+nZWeqlnpIuVBceExDphsZm8C+4CpAGbWjkjXlUiZ0ax0kfKjyPBw9weAnwGj\ngVP8u5lbCcDt8S1N5IcKZqU/ckVXzUoXCVFxo61qAfPc/XV332NmHczsLqCzu88vmxJFfmhAt+a8\neVtkVvpVz8zkyY81K12kLBXXbfUukAbfdlXNAI4BbjWzv8S3NJGitW8amZV+bpdmPPTuF9w4RrPS\nRcpKceHRwN2XBa+HAOPc/XbgXOCCuFYmUgIpyUkMv7Ibv72wIx9/uYULH5vGovW6HCcSb8WFR3Q/\nwOnABwDungNoqIuUC2bG9X0is9JzcvO55EnNSheJt+LC43Mzezi4ztEOeB/AzOrHvTKRGB08K/2O\ncZ+StSs77LJEKqXiwuNGYCuR6x5nufveoL0j8HBRK5pZSzObZGZLzWyxmd0ZtDc0sw/MbFnwvUHQ\nbsFtT5ab2edmdmLUtoYEyy8zsyGHeaxSBRTMSr/7zPa8u2gT/f/+MS/MXEN+vi6mi5QmK2qEipml\nAk3cffFB7Z2Bze6eVcS6zYBm7j7fzOoA84ABwHXANnd/0MzuJXJd5Rdmdh6R4b/nAScD/3L3k4N7\naM0FMoh0o80Durv7N0UdWEZGhs+dO7eYw5fKbEXWbu5/YxHTV3xN15b1eWBgZzodXS/sskTKLTOb\n5+4ZJVm2uDOPx4DGhbQ3B/5V1IruvrFgOK+77wKWButdDDwfLPY8kUAhaB/jETOB+kEAnQ184O7b\ngsD4ADin2COTKq9tagpjf3Iyj1zRlcxv9nLhY9P443+XsDs7N+zSRCq84sKji7tPPrjR3d8Dji/p\nTswsDegGzAKauvvGYDsbgSbBYs2BdVGrZQZth2ovbD9DzWyumc3NyjrkSZFUIWbGgG7N+fDu07iy\nRyue+2QVZ/x9Mu8s3Kh5ISJHoLjwqHaYn33LzFKAV4Fh7r6zqEULafMi2n/Y6D7C3TPcPSM1NbUk\n5UkVUa9WNR4Y2IVXb+lNg9rVuWXsfG4YPYd12/YWv7KI/EBx4bEsuBbxPWZ2LrCyuI2bWTUiwTHW\n3V8LmjcH3VEF10W2BO2ZQMuo1VsAG4poF4nZia0a8NZtfbj/go7MXrWNM/85mccnLScnVyPPRWJR\n3AXzdOBtYDqRC9UQuXDdC7jA3b8qYl0jck1jm7sPi2r/G/B11AXzhu7+czM7H7iN7y6YP+ruPYIL\n5vOAgtFX84lcMN9GEXTBXIqzccc+/vDWEt5ZtIl2TVJ4YEBnTj6mUdhliYQmlgvmxYXHMCK/uI8F\nOgTNS4AX3b3I25ma2SlE7sK7kO8mFP6SyHWPCUArYC1wubtvC8JmOJGL4XuB6919brCtG4J1AR5w\n91HFHZjCQ0pq0hdbuP/NRWR+s4/LurfgvnOPpVFKcthliZS50gyPh4HewHHAAiJnIJ8AM4r7yz9s\nCg+Jxb6cPIZPWsaIKSupnZzEveccy/9ltCQhobBLbiKVU6mFR9QGqxPprupNpMuqF7Dd3TseSaHx\npPCQw7Fs8y5+9cYiZq/aRvfWDXhgYGeOPapu2GWJlInSnOdRoCZQF6gXfG0g0v0kUqmkN63DS0N7\n8vDlJ7Bq6x7Of3Qaf5m4lL05mhsiEq24bqsRQCdgF5GwmAnMLG52d3mgMw85Ut/syeGhd79g/Jx1\nNK9fk99d1IkzOzYNuyyRuCnNM49WQDKwCVhPZNjs9iMrT6RiaFC7Og9eejyv3NyLlOQkbhwzlxvH\nzGX99n1hlyYSumKveQSjoDoRud7RG+gMbCNy0fy3ca/wMOnMQ0rTgbx8Rn2yin9+EHm8zbAz0rnh\nlDZUSyxpz69I+VfqF8yDjbYA+hAJkAuARu5ebm/NrvCQeFi/fR+/+89iPliymQ5N6/DAwM5kpDUM\nuyyRUlFq3VZmdoeZjTezdcAUIqHxJXAJoP9jpMppXr8mzwzOYMS13dm1/wCXPTWDe1/9nG/25IRd\nmkiZSirm8zTgFeCugpsZigic1eko+rRrzKMfLmPktFW8v2Qz9517LJd1b0Gkp1ekcitxt1VFo24r\nKStfbNrJr15fxLw139CjTUMeGNCZ9KZ1wi5LJGbxmOchIodw7FF1efmmXjx0aRe+2ryLc/81lb++\n+wX7cvLCLk0kbhQeIqUgIcG44qRWfHh3PwZ0a84TH6/gzH9OZtIXW4pfWaQCUniIlKJGKck8fPkJ\njB/akxrVErl+9BxueWEeG3dobohULgoPkTjoeUwjJt5xKj8/pwOTvtzCGX+fzMipK8nN03NDpHJQ\neIjESfWkBH56Wjs+uKsfPdo05E9vL+XC4Z8wZ3W5viG1SIkoPETirGXDWjx33Uk8dc2JfLMnh8uf\nmsE1I2cpRKRC01BdkTK0NyeXsTPX8vSUFWzdnUPvto24s3+6nmAo5UJcbk9S0Sg8pDzbl5PH2Flr\neGrySrbuzqbXMY2484x0eipEJEQKDxQeUjHsy8njxdlreWryCrJ2ZXNym4YMO6M9vdoqRKTsKTxQ\neEjFsv9AHi/OioTIll3Z9GjTkGH90+nVtpFudyJlRuGBwkMqpv0H8hg/ey1PTl7B5p3ZnJTWgDv7\nt6dPO4WIxJ/CA4WHVGz7D+QxYe46npi0gk0799O9dQOGnZHOKe0aK0QkbhQeKDykcsjOzWPCnHU8\n8fEKNu7Yz4mt6nPnGe3pm64QkdKn8EDhIZVLdm4eL8/N5IlJy9mwYz9dW9Zn2Bnp9GufqhCRUqPw\nQOEhlVN2bh6vzMvkiUkrWL99Hye0rM+w/umc1kEhIkdO4YHCQyq3nNx8Xp2fyeOTlpP5zT5OaFGP\nO89I50cdmihE5LApPFB4SNVwIC+f1+Zn8thHkRA5vkU97jg9nf7HKUQkdgoPFB5StRzIy+f1+esZ\nPmk5a7ftpXPzutxxejpndmyqEJESU3ig8JCq6UBePm98GgmRNV/vpWOzutx5RjpnKUSkBBQeKDyk\nasvNy+eNzzYw/KNlrP56L8c1q8ud/dtxVsejSEhQiEjhFB4oPEQgEiL/WbCB4R8tZ+XWPRx7VB3u\n7J/O2Z0UIvJDCg8UHiLR8vKdtxZs4NGPlrEyaw8dmtbhjv7pnNtZISLfUXig8BApTF6+89/PN/Do\nh8tYkbWH9k1TuP30dM7r0oxEhUiVF0t46EmCIlVIYoJxcdfmvH9XPx69shvucPu4Tzn7kSm8s3Aj\nlfWPSSl9Cg+RKigxwbjohKN5b1hfhl/VDQNuGTufa56dxfItu8IuTyoAhYdIFZaQYFxw/NG8O6wv\nf7y4Ewszd3DOI1P588Sl7M7ODbs8KccUHiJCYoJxba80Jt1zGpee2IIRU1Zy+sMf8+Zn69WVJYVS\neIjItxqlJPPQZcfzxq19OKpeDe4c/xlXjJjJF5t2hl2alDMKDxH5ga4t6/P6T/vwl0u6sGzzLs5/\ndBq/f2sxO/YdCLs0KScUHiJSqMQE48oerfjoZ6dxZY+WjJ6+mv5//5hX5mWSn6+urKoubuFhZs+Z\n2RYzWxTVdoKZzTCzhWb2lpnVDdrTzGyfmX0WfD0VtU73YPnlZvao6QY9ImWqQe3q/GlAF9667RRa\nNqzFPS8v4PKnZ7Bo/Y6wS5MQxfPMYzRwzkFtI4F73b0L8Drw/6I+W+HuXYOvm6PanwSGAunB18Hb\nFJEy0Ll5PV69uTd/u+x4Vm/dw0XDp3H/G4vYvjcn7NIkBHELD3efAmw7qLkDMCV4/QFwaVHbMLNm\nQF13n+GRIR9jgAGlXauIlExCgnF5Rks+uuc0BvdKY+ysNZz+98mMn71WXVlVTFlf81gEXBS8vhxo\nGfVZGzP71Mwmm9mpQVtzIDNqmcygrVBmNtTM5prZ3KysrNKsW0Si1KtZjd9d1Im37ziVdqkp3Pva\nQgY+8QkL1m0PuzQpI2UdHjcAt5rZPKAOUHC+uxFo5e7dgLuBF4PrIYVd3zjknzfuPsLdM9w9IzU1\ntZRLF5GDHdesLi/d1JNHrujKhh37GfDEJ9z32uds26OurMouqSx35u5fAGcBmFl74PygPRvIDl7P\nM7MVQHsiZxotojbRAthQljWLSNHMjAHdmtP/uCY8+uEyRn2ymokLN3HP2R24qkcr3XCxkirTMw8z\naxJ8TwB+DTwVvE81s8Tg9TFELoyvdPeNwC4z6xmMshoMvFmWNYtIydSpUY1fnd+Rd+48lY7N6nL/\nG4u4aPg05q35JuzSJA7iOVR3HDAD6GBmmWb2Y+BKM/sK+ILIGcSoYPG+wOdmtgB4BbjZ3Qsutt9C\nZJTWcmAF8E68ahaRI5fetA4v3ngyw6/qxte7c7j0yenc8/ICsnZlh12alCI9z0NE4mZPdi7DJy1n\n5NSV1KiWyN1ntufanq1JStT85PJIz/MQkXKhdnISvzjnWN4d1peuLevz+7eWcMFj05i18uuwS5Mj\npPAQkbhrm5rCmBt68NQ13dm1P5crRsxk2PhP2bxzf9ilyWFSeIhImTAzzul8FP+7ux93nN6OiYs2\ncfrDH/PMlJUcyMsPuzyJkcJDRMpUzeqJ3H1WBz64qy8nH9OIByYu5dx/TWX68q1hlyYxUHiISCha\nN6rNc9edxLNDMsjJzeeqkbO49cX5bNi+L+zSpAQUHiISqv7HNeX9u/py95nt+d+SzfT/+2Se+Hg5\n2bl5YZcmRVB4iEjoalRL5I7+6fzv7n70bd+Yv777Jec+MpXJX+kedeWVwkNEyo2WDWvx9LUZPH9D\nDxwY8txsrhs1m6nLsvQs9XJGkwRFpFzKzs3juWmrGTl1JV/vyaFdkxSG9GrNJSe2oHZymd6Wr8qI\nZZKgwkNEyrXs3Dze/nwjo6ev5vPMHdRJTuLyjJYM7tWatMa1wy6vUlF4oPAQqWzcnU/XbWf0J6uZ\nuHAjee78qEMThvRO49R2jUnQ3XuPmMIDhYdIZbZl537GzlrL2Flr2bo7m2NSazOkVxqXdm9Birq0\nDpvCA4WHSFWQnZvHOws3MWr6ahas205KchKXdW/BkN5ptFGXVswUHig8RKqaz9Zt5/npq/nv5xs4\nkOec1iGVIb3T6Jeeqi6tElJ4oPAQqaq27NrPuFnreGHWGrJ2ZdOmcW0G92rNZd1bUKdGtbDLK9cU\nHig8RKq6nNx83lkUGaX16dp/6s4tAAANGUlEQVTt1K6eyGXdWzC4dxptU1PCLq9cUnig8BCR7yz4\ntktrIzl5+Zya3pjr+6RxWvsm6tKKovBA4SEiP5S1K5txs9fywsw1bNmVTetGtRjcK43LM1pQV11a\nCg9QeIjIoR3Iy+fdRZsYPX0189Z8Q63qiVxyYnOu651GuyZ1wi4vNAoPFB4iUjILM3cwevpq3lqw\ngZy8fE5p15jreqfxo2ObkFjFurQUHig8RCQ2X+/OZvycdfx7xho27dxPy4Y1GdIrjcszWlKvZtXo\n0lJ4oPAQkcNzIC+f9xdvZvT0VcxZ/Q01qyUyMOjSat+0cndpKTxQeIjIkVu0fgdjZqzmjc82kJOb\nT++2jbiudxr9j2taKbu0FB4oPESk9Gzbk8P4OWv594w1bNyxnxYNIl1ag3q0rFQTDxUeKDxEpPTl\n5uXzwZLNjJq+mtmrtlGnRhKDe7Xmut5tSK2THHZ5R0zhgcJDROJrwbrtPDV5Be8u3kS1xAQu796C\noX2PoXWjintDRoUHCg8RKRsrs3YzYspKXpu/ntz8fM7r0oyb+7Wlc/N6YZcWM4UHCg8RKVubd+7n\nuU9WMXbmWnZn53JqemNu7teW3m0bYVYxLq4rPFB4iEg4duw7wNhZa3hu2mq27s7m+Bb1uLlfW87u\ndFS5H6Gl8EDhISLh2n8gj9fmr2fElBWs/novbRrX5sZTj+GSE5tTo1pi2OUVSuGBwkNEyoe8fOfd\nRZt4avIKFq7fQWqdZG7o04are7YqdzdjVHig8BCR8sXdmb7ia56avIKpy7ZSJzmJq3q24sd92tCk\nbo2wywMUHoDCQ0TKr0Xrd/Dk5BW8s3AjSQkJXNq9OUP7tg39uesKDxQeIlL+rd66h2emruTleZkc\nyMvnnE5HcXO/tpzQsn4o9Sg8UHiISMWRtSubUZ+s4t8z17Brfy692zbi5n5tOTW9cZkO81V4oPAQ\nkYpn1/4DjJu9lmenrWLzzmw6HV2Xm/u15dzOR5GUmBD3/Ss8UHiISMWVnZvHG5+u5+kpK1mZtYdW\nDWtxY99juLx7i7gO81V4oPAQkYovP995f8lmnpy8ggXrttM4pTrX92nDNSe3pl6t0h/mG0t4xO08\nyMyeM7MtZrYoqu0EM5thZgvN7C0zqxv12X1mttzMvjSzs6PazwnalpvZvfGqV0SkvElIMM7pfBRv\n/LQ3427sSaej6/G3976k94Mf8sDbS9i0Y39otcXtzMPM+gK7gTHu3jlomwPc4+6TzewGoI27329m\nHYFxQA/gaOB/QPtgU18BZwKZwBzgSndfUtz+deYhIpXRkg07eXrKCv77+UYSDAZ2iwzzbdck5Yi3\nXS7OPNx9CrDtoOYOwJTg9QfApcHri4Hx7p7t7quA5USCpAew3N1XunsOMD5YVkSkSup4dF3+Nagb\nH99zGlf2aMWbn23gzH9OZuiYucxf+02Z1RH/y/fftwi4KHh9OdAyeN0cWBe1XGbQdqj2QpnZUDOb\na2Zzs7KySq1oEZHypmXDWvzh4s5Mv/d0bv9RO2at2sYlT0zniqdnsP9AXtz3X9bhcQNwq5nNA+oA\nOUF7YQOZvYj2Qrn7CHfPcPeM1NTUIy5WRKS8a5SSzN1ndWD6vafz6/OPo03j2mVy48WkuO8hirt/\nAZwFYGbtgfODjzL57iwEoAWwIXh9qHYREQnUTk7iJ6ceU2b7K9MzDzNrEnxPAH4NPBV89B9gkJkl\nm1kbIB2YTeQCebqZtTGz6sCgYFkREQlR3M48zGwccBrQ2Mwygd8CKWZ2a7DIa8AoAHdfbGYTgCVA\nLnCru+cF27kNeA9IBJ5z98XxqllEREpGkwRFRAQoJ0N1RUSk8lJ4iIhIzBQeIiISM4WHiIjETOEh\nIiIxq7SjrcwsC1hzmKs3BraWYjmlRXXFRnXFRnXFpjLW1drdS3R7jkobHkfCzOaWdLhaWVJdsVFd\nsVFdsanqdanbSkREYqbwEBGRmCk8Cjci7AIOQXXFRnXFRnXFpkrXpWseIiISM515iIhIzBQeIiIS\nM4VHFDN7zsy2mNmisGspYGYtzWySmS01s8VmdmfYNQGYWQ0zm21mC4K6fh92TdHMLNHMPjWz/4Zd\nSzQzW21mC83sMzMrN7d9NrP6ZvaKmX0R/FvrVQ5q6hD8nAq+dprZsLDrAjCzu4J/94vMbJyZ1Qi7\nJgAzuzOoaXG8f1a65hHFzPoCu4Ex7t457HoAzKwZ0Mzd55tZHWAeMMDdl4RclwG13X23mVUDpgF3\nuvvMMOsqYGZ3AxlAXXe/IOx6CpjZaiDD3cvV5DIzex6Y6u4jgwev1XL37WHXVcDMEoH1wMnufriT\nf0urluZE/r13dPd9wbOIJrr76JDr6gyMB3oQecT3u8At7r4sHvvTmUcUd58CbAu7jmjuvtHd5wev\ndwFLgebhVgUesTt4Wy34Khd/iZhZCyKPOB4Zdi0VgZnVBfoCzwK4e055Co5Af2BF2MERJQmoaWZJ\nQC3Kx+OxjwNmuvted88FJgMD47UzhUcFYmZpQDdgVriVRARdQ58BW4AP3L1c1AU8AvwcyA+7kEI4\n8L6ZzTOzoWEXEzgGyAJGBV19I82sdthFHWQQMC7sIgDcfT3wMLAW2AjscPf3w60KgEVAXzNrZGa1\ngPOAlvHamcKjgjCzFOBVYJi77wy7HgB3z3P3rkALoEdw2hwqM7sA2OLu88Ku5RD6uPuJwLnArUFX\nadiSgBOBJ929G7AHuDfckr4TdKNdBLwcdi0AZtYAuBhoAxwN1Daza8KtCtx9KfAQ8AGRLqsFRB7r\nHRcKjwoguKbwKjDW3V8Lu56DBV0cHwPnhFwKQB/gouDawnjgdDN7IdySvuPuG4LvW4DXifRPhy0T\nyIw6c3yFSJiUF+cC8919c9iFBM4AVrl7lrsfAF4DeodcEwDu/qy7n+jufYl0wcflegcoPMq94ML0\ns8BSd/9H2PUUMLNUM6sfvK5J5H+oL8KtCtz9Pndv4e5pRLo6PnL30P8qBDCz2sGgB4JuobOIdDWE\nyt03AevMrEPQ1B8IdUDGQa6knHRZBdYCPc2sVvD/Z38i1yJDZ2ZNgu+tgEuI488tKV4brojMbBxw\nGtDYzDKB37r7s+FWRR/gWmBhcH0B4JfuPjHEmgCaAc8Ho2ASgAnuXq6GxZZDTYHXI79vSAJedPd3\nwy3pW7cDY4MuopXA9SHXA0DQd38mcFPYtRRw91lm9gown0i30KeUn1uVvGpmjYADwK3u/k28dqSh\nuiIiEjN1W4mISMwUHiIiEjOFh4iIxEzhISIiMVN4iIhIzBQeUiGZmZvZ36Pe32NmvyulbY82s8tK\nY1vF7Ofy4A62k+JZl5mlmdlVsVcocmgKD6mosoFLzKxx2IVEC+a9lNSPgZ+6+4/iVU8gDYgpPGI8\nDqmCFB5SUeUSmZh118EfHPwXupntDr6fZmaTzWyCmX1lZg+a2dXBc0kWmlnbqM2cYWZTg+UuCNZP\nNLO/mdkcM/vczG6K2u4kM3sRWFhIPVcG219kZg8Fbb8BTgGeMrO/FbLOz4N1FpjZg4V8vrogOM0s\nw8w+Dl73s++ef/FpMKP9QeDUoO2ukh5HMCP+7aCGRWZ2RUn+w0jVoBnmUpE9DnxuZn+NYZ0TiNy6\nehuRmdQj3b2HRR6ydTtQ8ACdNKAf0BaYZGbtgMFE7qB6kpklA5+YWcHdVHsAnd19VfTOzOxoIjer\n6w58Q+SOugPc/Q9mdjpwj7vPPWidc4EBRJ5dsdfMGsZwfPcQmVn8SXAzzf1EbnJ4T8FzTYK7+RZ7\nHGZ2KbDB3c8P1qsXQx1SyenMQyqs4O7CY4A7YlhtTvCMlGxgBVDwS3MhkcAoMMHd84MH6awEjiVy\nL6rBwW1iZgGNgPRg+dkHB0fgJODj4CZ6ucBYIs/OKMoZwCh33xscZyzPmPkE+IeZ3QHUD/Z5sJIe\nx0IiZ2APmdmp7r4jhjqkklN4SEX3CJFrB9HPn8gl+Lcd3LiuetRn2VGv86Pe5/P9M/GD79vjgAG3\nu3vX4KtN1HMc9hyiPivpgRy0TnH3Dfr2GIFvH4Hq7g8CPwFqAjPN7NhDbL/Y43D3r4icMS0E/hJ0\ntYkACg+p4IK/yicQCZACq4n80oPIcxeqHcamLzezhOA6yDHAl8B7wC3BLfIxs/ZW/EOTZgH9zKxx\ncBH6SiJPeCvK+8ANwU0BOUS31Wq+O8ZLCxrNrK27L3T3h4C5RM6YdgF1otYt0XEEXW573f0FIg8/\nKk+3aZeQ6ZqHVAZ/B26Lev8M8KaZzQY+5NBnBUX5ksgv+abAze6+38xGEunamh+c0WQRuTZxSO6+\n0czuAyYR+Yt/oru/Wcw675pZV2CumeUAE4FfHrTY74FnzeyXfP/JksPM7EdAHpHbqr9D5Kwq18wW\nAKOBf5XwOLoAfzOzfCJ3ab2lqLqlatFddUVEJGbqthIRkZgpPEREJGYKDxERiZnCQ0REYqbwEBGR\nmCk8REQkZgoPERGJ2f8HxLXeMYh9gi8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x21b730d37f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "find_number_clusters(tfidf=cluster1_tfidif.todense())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top terms per cluster:\n",
      "\n",
      "Cluster 0 words: equip, retail, variou, design, develop,\n",
      "\n",
      "Cluster 0 Companies: Skyworks Solutions Inc., Interface Inc., Nordson Corporation, Genco Shipping & Trading Ltd., Kirkland's Inc.,\n",
      "\n",
      "Cluster 1 words: insur, life, properti, casualti, health,\n",
      "\n",
      "Cluster 1 Companies: Independence Holding Co., American International Group, Inc., Amerisafe, Inc., Fidelity National Financial, Inc., The Navigators Group, Inc.,\n",
      "\n",
      "Cluster 2 words: softwar, network, data, applic, technolog,\n",
      "\n",
      "Cluster 2 Companies: Datawatch Corporation, Mannatech, Incorporated, Iridium Communications Inc., HealthStream, Inc., Resonant Inc.,\n",
      "\n",
      "Cluster 3 words: restaur, franchis, oper, decemb, develop,\n",
      "\n",
      "Cluster 3 Companies: Luby's, Inc., Papa John's International Inc., Biglari Holdings Inc., Dunkin' Brands Group, Inc., Noodles & Company,\n",
      "\n",
      "Split of classes within clusters:\n",
      "0    1521\n",
      "2     591\n",
      "1      89\n",
      "3      40\n",
      "Name: Subclusters_level_1, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# trying to find subclusters for any cluster \n",
    "Subcluster1_df=merged_df[merged_df['Clusters']==1]\n",
    "# to get only the indices for this cluster\n",
    "cluster1_tfidif=tfidf_matrix[Subcluster1_df.index.values]\n",
    "hc,clusters = apply_agglomerative(num_clusters = 4,tfidf = cluster1_tfidif)\n",
    "# add new column which includes the numbers of the new clusters\n",
    "Subcluster1_df = Subcluster1_df.assign(Subclusters_level_1 = clusters)\n",
    "# Printing terms per cluster\n",
    "print_agg_cluster(df=Subcluster1_df, clusters=clusters, num_clusters=4, num_cluster_topics=5, \\\n",
    "              num_cluster_companies=5, words=[], cluster_level='Subclusters_level_1')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "topics.update(save_topics_words(df=Subcluster1_df, clusters=clusters, num_clusters=4, num_cluster_topics=10, \\\n",
    "              words=[], cluster_level='Subclusters_level_1'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Splitting cluster 4\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "execution time: 0.03 minutes\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEWCAYAAAB8LwAVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngBARkbgUECIiEpcCQkRE4lJAiIhIXP8f97aVkTTTl+AAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x21b72889d68>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "find_number_clusters(tfidf=cluster4_tfidif.todense())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top terms per cluster:\n",
      "\n",
      "Cluster 0 words: electr, distribut, energi,\n",
      "\n",
      "Cluster 0 Companies: Northwestern Corporation, American Midstream Partners, LP, Shell Midstream Partners, L.P., Alliant Energy Corporation, NextEra Energy, Inc.,\n",
      "\n",
      "Cluster 1 words: explor, develop, energi,\n",
      "\n",
      "Cluster 1 Companies: US Energy Corp., Northern Technologies International Corp., Glori Energy Inc., Atlas Energy Group, LLC, Murphy Oil Corporation,\n",
      "\n",
      "Cluster 2 words: crude, storag, refin,\n",
      "\n",
      "Cluster 2 Companies: Matrix Service Company, Plains All American Pipeline, L.P., CHS, Inc., PBF Logistics LP, Martin Midstream Partners LP,\n",
      "\n",
      "Split of classes within clusters:\n",
      "1    112\n",
      "0     88\n",
      "2     32\n",
      "Name: Subclusters_level_1, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# trying to find subclusters for any cluster \n",
    "Subcluster4_df=merged_df[merged_df['Clusters']==4]\n",
    "# to get only the indices for this cluster\n",
    "cluster4_tfidif=tfidf_matrix[Subcluster4_df.index.values]\n",
    "hc,clusters = apply_agglomerative(num_clusters = 3,tfidf = cluster4_tfidif)\n",
    "# add new column which includes the numbers of the new clusters\n",
    "Subcluster4_df = Subcluster4_df.assign(Subclusters_level_1 = clusters)\n",
    "# Printing terms per cluster\n",
    "X_freq_terms, freq_terms_with_count = freq_words(indexes=Subcluster4_df.index.values,new_corpus=new_corpus,\\\n",
    "                                                terms=terms,X=3)\n",
    "print_agg_cluster(df=Subcluster4_df, clusters=clusters, num_clusters=3, num_cluster_topics=3, \\\n",
    "              num_cluster_companies=5, words=X_freq_terms, cluster_level='Subclusters_level_1')\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "topics.update(save_topics_words(df=Subcluster4_df, clusters=clusters, num_clusters=3, num_cluster_topics=10, \\\n",
    "              words=X_freq_terms, cluster_level='Subclusters_level_1'))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Splitting cluster 0\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "execution time: 0.04 minutes\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEWCAYAAAB8LwAVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nBYSIiBRKASEiIoVSQIiISKH+H/fIC1TwjribAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x21b73af01d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "find_number_clusters(tfidf=cluster0_tfidif.todense())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top terms per cluster:\n",
      "\n",
      "Cluster 0 words: invest, fund, equiti,\n",
      "\n",
      "Cluster 0 Companies: Virtu Financial, Inc., Oxford Lane Capital Corp., Nuveen Tax-Advantaged Dividend Growth Fund,\n",
      "\n",
      "Cluster 1 words: real, invest, estat,\n",
      "\n",
      "Cluster 1 Companies: PennyMac Mortgage Investment Trust, CatchMark Timber Trust, Inc., ARMOUR Residential REIT, Inc.,\n",
      "\n",
      "Split of classes within clusters:\n",
      "1    244\n",
      "0    163\n",
      "Name: Subclusters_level_1, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# trying to find subclusters for any cluster \n",
    "Subcluster0_df=merged_df[merged_df['Clusters']==0]\n",
    "# to get only the indices for this cluster\n",
    "cluster0_tfidif=tfidf_matrix[Subcluster0_df.index.values]\n",
    "hc,clusters = apply_agglomerative(num_clusters = 2,tfidf = cluster0_tfidif)\n",
    "#len(clusters1)\n",
    "# add new column which includes the numbers of the new clusters\n",
    "Subcluster0_df = Subcluster0_df.assign(Subclusters_level_1 = clusters)\n",
    "# Printing terms per cluster\n",
    "#print_cluster(df=Subcluster1_df, km=km, num_clusters=2, num_cluster_topics=6, \\\n",
    " #             num_cluster_companies=5, cluster_level='Subclusters_level_1')\n",
    "X_freq_terms, freq_terms_with_count = freq_words(indexes=Subcluster1_df.index.values,new_corpus=new_corpus,\\\n",
    "                                                terms=terms,X=2)\n",
    "print_agg_cluster(df=Subcluster0_df, clusters=clusters, num_clusters=2, cluster_level='Subclusters_level_1',\\\n",
    "                    words=X_freq_terms , num_cluster_topics=3, num_cluster_companies=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "topics.update(save_topics_words(df=Subcluster0_df, clusters=clusters, num_clusters=2, cluster_level='Subclusters_level_1',\\\n",
    "                    words=X_freq_terms,num_cluster_topics=10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Combining Subclusters and adding into original dataframe\n",
    "test=pd.concat([Subcluster1_df,Subcluster4_df,Subcluster0_df])\n",
    "test.drop(['Country','Company','Sector','Industry','Clusters'], axis=1, inplace=True)\n",
    "merged_df = pd.merge(left=merged_df, right=test, how='left', on=\"Ticker\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Creating one more level of sub clusters"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Splitting cluster 1, subcluster 0 into more clusters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "execution time: 0.26 minutes\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEWCAYAAABxMXBSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x21b728895c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "find_number_clusters(tfidf=Subcluster1_level1_0_tfidif.todense())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top terms per cluster:\n",
      "\n",
      "Cluster 0 words: power, energi, construct, engin, distribut,\n",
      "\n",
      "Cluster 0 Companies: AutoNation, Inc., Titan International Inc., Olympic Steel Inc., The Timken Company, Stepan Company,\n",
      "\n",
      "Cluster 1 words: home, accessori, apparel, oper, variou,\n",
      "\n",
      "Cluster 1 Companies: Koss Corporation, The Children's Place, Inc., Avon Products Inc., Guess', Inc., Restoration Hardware Holdings, Inc.,\n",
      "\n",
      "Cluster 2 words: care, health, test, medic, develop,\n",
      "\n",
      "Cluster 2 Companies: RadNet, Inc., Hill-Rom Holdings, Inc., Medtronic plc, Vermillion, Inc., UnitedHealth Group Incorporated,\n",
      "\n",
      "Cluster 3 words: food, beverag, distribut, natur, specialti,\n",
      "\n",
      "Cluster 3 Companies: AMCON Distributing Co., Tyson Foods, Inc., Dean Foods Company, United Natural Foods, Inc., Sealed Air Corporation,\n",
      "\n",
      "Cluster 4 words: compon, applic, electron, control, design,\n",
      "\n",
      "Cluster 4 Companies: Science Applications International Corporation, Honeywell International Inc., Leidos Holdings, Inc., SIFCO Industries Inc., Shiloh Industries Inc.,\n",
      "\n",
      "Split of classes within clusters:\n",
      "0    734\n",
      "1    246\n",
      "2    242\n",
      "4    237\n",
      "3     62\n",
      "Name: Subclusters_level_2, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# trying to find subclusters for any cluster \n",
    "Subcluster1_level1_0_df = merged_df[(merged_df['Clusters']==1) & (merged_df['Subclusters_level_1']==0)]\n",
    "# to get only the indices for this cluster\n",
    "Subcluster1_level1_0_tfidif=tfidf_matrix[Subcluster1_level1_0_df.index.values]\n",
    "hc,clusters = apply_agglomerative(num_clusters = 5,tfidf = Subcluster1_level1_0_tfidif)\n",
    "# add new column which includes the numbers of the new clusters\n",
    "Subcluster1_level1_0_df = Subcluster1_level1_0_df.assign(Subclusters_level_2 = clusters)\n",
    "# Printing terms per cluster\n",
    "X_freq_terms, freq_terms_with_count = freq_words(indexes=Subcluster1_level1_0_df.index.values,new_corpus=new_corpus,\\\n",
    "                                                terms=terms,X=2)\n",
    "print_agg_cluster(df=Subcluster1_level1_0_df, clusters=clusters, num_clusters=5, num_cluster_topics=5, \\\n",
    "              num_cluster_companies=5, words = X_freq_terms,cluster_level='Subclusters_level_2')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "topics.update(save_topics_words(df=Subcluster1_level1_0_df, clusters=clusters, num_clusters=5, num_cluster_topics=10, \\\n",
    "              words = X_freq_terms,cluster_level='Subclusters_level_2'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### Splitting cluster 1, subcluster 2 into more clusters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "execution time: 0.08 minutes\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEWCAYAAAB8LwAVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x21b73b18710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "find_number_clusters(tfidf=Subcluster1_level1_2_tfidif.todense())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top terms per cluster:\n",
      "\n",
      "Cluster 0 words: payment, financi, consum,\n",
      "\n",
      "Cluster 0 Companies: FactSet Research Systems Inc., Moody's Corporation, Stein Mart Inc.,\n",
      "\n",
      "Cluster 1 words: digit, video, content,\n",
      "\n",
      "Cluster 1 Companies: Tribune Media Company, IAC/InterActiveCorp, Glowpoint, Inc.,\n",
      "\n",
      "Cluster 2 words: softwar, network, data,\n",
      "\n",
      "Cluster 2 Companies: Intellicheck Mobilisa, Inc., IDT Corporation, Upland Software, Inc.,\n",
      "\n",
      "Cluster 3 words: game, entertain, park,\n",
      "\n",
      "Cluster 3 Companies: Gaming Partners International Corporation, Dover Downs Gaming & Entertainment Inc., Century Casinos Inc.,\n",
      "\n",
      "Split of classes within clusters:\n",
      "2    246\n",
      "0    210\n",
      "1     96\n",
      "3     39\n",
      "Name: Subclusters_level_2, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# trying to find subclusters for any cluster \n",
    "Subcluster1_level1_2_df = merged_df[(merged_df['Clusters']==1) & (merged_df['Subclusters_level_1']==2)]\n",
    "Subcluster1_level1_2_tfidif=tfidf_matrix[Subcluster1_level1_2_df.index.values]\n",
    "hc,clusters = apply_agglomerative(num_clusters = 4,tfidf = Subcluster1_level1_2_tfidif)\n",
    "# add new column which includes the numbers of the new clusters\n",
    "Subcluster1_level1_2_df = Subcluster1_level1_2_df.assign(Subclusters_level_2 = clusters)\n",
    "# Printing terms per cluster\n",
    "#print_cluster(df=Subcluster3_level1_2_df, km=km, num_clusters=3, num_cluster_topics=5, \\\n",
    "#              num_cluster_companies=5, cluster_level='Subclusters_level_2')\n",
    "#removing X most frequent terms from parent cluster\n",
    "X_freq_terms, freq_terms_with_count = freq_words(indexes=Subcluster1_level1_2_df.index.values,new_corpus=new_corpus,\\\n",
    "                                                 terms=terms,X=0)\n",
    "print_agg_cluster(df=Subcluster1_level1_2_df, clusters=clusters, num_clusters=4, cluster_level='Subclusters_level_2',\\\n",
    "                    words=X_freq_terms , num_cluster_topics=3, num_cluster_companies=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "topics.update(save_topics_words(df=Subcluster1_level1_2_df, clusters=clusters, num_clusters=4, cluster_level='Subclusters_level_2',\\\n",
    "                    words=X_freq_terms , num_cluster_topics=10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "test=pd.concat([Subcluster1_level1_2_df,Subcluster1_level1_0_df])\n",
    "test.drop(['Country','Company','Sector','Industry','Clusters','Subclusters_level_1'], axis=1, inplace=True)\n",
    "merged_df = pd.merge(left=merged_df, right=test, how='left', on=\"Ticker\")\n",
    "\n",
    " #check if new_merged_df is correct, if so then rename it to merged_df\n",
    "merged_df.columns = ['Ticker', 'Company', 'Sector', 'Industry', 'Country', 'Clusters Level 0','Clusters Level 1', \\\n",
    "                         'Clusters Level 2']\n",
    "#temp = merged_df[merged_df['Clusters Level 0']==3]\n",
    "#temp = temp[temp['Clusters Level 1']==3]\n",
    "#temp = temp[temp['Clusters Level 2']==3].info()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Ticker</th>\n",
       "      <th>Company</th>\n",
       "      <th>Sector</th>\n",
       "      <th>Industry</th>\n",
       "      <th>Country</th>\n",
       "      <th>Clusters Level 0</th>\n",
       "      <th>Clusters Level 1</th>\n",
       "      <th>Clusters Level 2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A</td>\n",
       "      <td>Agilent Technologies, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Medical Laboratories &amp; Research</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AA</td>\n",
       "      <td>Alcoa Inc.</td>\n",
       "      <td>Basic Materials</td>\n",
       "      <td>Aluminum</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AAC</td>\n",
       "      <td>AAC Holdings, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Specialized Health Services</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>AAL</td>\n",
       "      <td>American Airlines Group Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Major Airlines</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>AAMC</td>\n",
       "      <td>Altisource Asset Management Corporation</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Asset Management</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>AAME</td>\n",
       "      <td>Atlantic American Corp.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Life Insurance</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>AAN</td>\n",
       "      <td>Aaron's, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Rental &amp; Leasing Services</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>AAOI</td>\n",
       "      <td>Applied Optoelectronics, Inc.</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Semiconductor - Integrated Circuits</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>AAON</td>\n",
       "      <td>AAON Inc.</td>\n",
       "      <td>Industrial Goods</td>\n",
       "      <td>General Building Materials</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>AAP</td>\n",
       "      <td>Advance Auto Parts Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Auto Parts Stores</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>AAPL</td>\n",
       "      <td>Apple Inc.</td>\n",
       "      <td>Consumer Goods</td>\n",
       "      <td>Electronic Equipment</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>AAT</td>\n",
       "      <td>American Assets Trust, Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>REIT - Retail</td>\n",
       "      <td>USA</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>AAWW</td>\n",
       "      <td>Atlas Air Worldwide Holdings, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Air Services, Other</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>AB</td>\n",
       "      <td>AllianceBernstein Holding L.P.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Asset Management</td>\n",
       "      <td>USA</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>ABAX</td>\n",
       "      <td>Abaxis, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Medical Laboratories &amp; Research</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>ABBV</td>\n",
       "      <td>AbbVie Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Drug Manufacturers - Major</td>\n",
       "      <td>USA</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>ABC</td>\n",
       "      <td>AmerisourceBergen Corporation</td>\n",
       "      <td>Services</td>\n",
       "      <td>Drugs Wholesale</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>ABCB</td>\n",
       "      <td>Ameris Bancorp</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Regional - Mid-Atlantic Banks</td>\n",
       "      <td>USA</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>ABCD</td>\n",
       "      <td>Cambium Learning Group, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Education &amp; Training Services</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>ABDC</td>\n",
       "      <td>Alcentra Capital Corporation</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Asset Management</td>\n",
       "      <td>USA</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>ABEO</td>\n",
       "      <td>Abeona Therapeutics Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>ABG</td>\n",
       "      <td>Asbury Automotive Group, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Auto Dealerships</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>ABIO</td>\n",
       "      <td>ARCA biopharma, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>ABM</td>\n",
       "      <td>ABM Industries Incorporated</td>\n",
       "      <td>Services</td>\n",
       "      <td>Business Services</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>ABMD</td>\n",
       "      <td>ABIOMED, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Medical Appliances &amp; Equipment</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>ABR</td>\n",
       "      <td>Arbor Realty Trust Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>REIT - Diversified</td>\n",
       "      <td>USA</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>ABT</td>\n",
       "      <td>Abbott Laboratories</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Medical Appliances &amp; Equipment</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>ABTX</td>\n",
       "      <td>Allegiance Bancshares, Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Regional - Southwest Banks</td>\n",
       "      <td>USA</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>ACAD</td>\n",
       "      <td>ACADIA Pharmaceuticals Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>ACC</td>\n",
       "      <td>American Campus Communities, Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>REIT - Residential</td>\n",
       "      <td>USA</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3528</th>\n",
       "      <td>Y</td>\n",
       "      <td>Alleghany Corporation</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Property &amp; Casualty Insurance</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3529</th>\n",
       "      <td>YELP</td>\n",
       "      <td>Yelp Inc.</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Internet Information Providers</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3530</th>\n",
       "      <td>YORW</td>\n",
       "      <td>The York Water Company</td>\n",
       "      <td>Utilities</td>\n",
       "      <td>Water Utilities</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3531</th>\n",
       "      <td>YRCW</td>\n",
       "      <td>YRC Worldwide Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Trucking</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3532</th>\n",
       "      <td>YUM</td>\n",
       "      <td>Yum! Brands, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Restaurants</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3533</th>\n",
       "      <td>YUMA</td>\n",
       "      <td>Yuma Energy, Inc.</td>\n",
       "      <td>Basic Materials</td>\n",
       "      <td>Independent Oil &amp; Gas</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3534</th>\n",
       "      <td>YUME</td>\n",
       "      <td>YuMe, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Advertising Agencies</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3535</th>\n",
       "      <td>Z</td>\n",
       "      <td>Zillow Group, Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Property Management</td>\n",
       "      <td>USA</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3536</th>\n",
       "      <td>ZAGG</td>\n",
       "      <td>ZAGG Inc</td>\n",
       "      <td>Services</td>\n",
       "      <td>Specialty Retail, Other</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3537</th>\n",
       "      <td>ZAIS</td>\n",
       "      <td>ZAIS Group Holdings, Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Asset Management</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3538</th>\n",
       "      <td>ZAYO</td>\n",
       "      <td>Zayo Group Holdings, Inc.</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Networking &amp; Communication Devices</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3539</th>\n",
       "      <td>ZAZA</td>\n",
       "      <td>ZaZa Energy Corporation</td>\n",
       "      <td>Basic Materials</td>\n",
       "      <td>Independent Oil &amp; Gas</td>\n",
       "      <td>USA</td>\n",
       "      <td>4</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3540</th>\n",
       "      <td>ZBH</td>\n",
       "      <td>Zimmer Biomet Holdings, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Medical Appliances &amp; Equipment</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3541</th>\n",
       "      <td>ZBK</td>\n",
       "      <td>Zions Bancorporation</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Regional - Pacific Banks</td>\n",
       "      <td>USA</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3542</th>\n",
       "      <td>ZBRA</td>\n",
       "      <td>Zebra Technologies Corporation</td>\n",
       "      <td>Industrial Goods</td>\n",
       "      <td>Diversified Machinery</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3543</th>\n",
       "      <td>ZEN</td>\n",
       "      <td>Zendesk, Inc.</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Application Software</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3544</th>\n",
       "      <td>ZEUS</td>\n",
       "      <td>Olympic Steel Inc.</td>\n",
       "      <td>Basic Materials</td>\n",
       "      <td>Steel &amp; Iron</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3545</th>\n",
       "      <td>ZFGN</td>\n",
       "      <td>Zafgen, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3546</th>\n",
       "      <td>ZG</td>\n",
       "      <td>Zillow Group, Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Property Management</td>\n",
       "      <td>USA</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3547</th>\n",
       "      <td>ZGNX</td>\n",
       "      <td>Zogenix, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Drug Manufacturers - Major</td>\n",
       "      <td>USA</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3548</th>\n",
       "      <td>ZION</td>\n",
       "      <td>Zions Bancorporation</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Regional - Pacific Banks</td>\n",
       "      <td>USA</td>\n",
       "      <td>3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3549</th>\n",
       "      <td>ZIOP</td>\n",
       "      <td>ZIOPHARM Oncology, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3550</th>\n",
       "      <td>ZIXI</td>\n",
       "      <td>Zix Corporation</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Application Software</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3551</th>\n",
       "      <td>ZN</td>\n",
       "      <td>Zion Oil &amp; Gas, Inc.</td>\n",
       "      <td>Basic Materials</td>\n",
       "      <td>Independent Oil &amp; Gas</td>\n",
       "      <td>USA</td>\n",
       "      <td>4</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3552</th>\n",
       "      <td>ZNGA</td>\n",
       "      <td>Zynga, Inc.</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Multimedia &amp; Graphics Software</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3553</th>\n",
       "      <td>ZOES</td>\n",
       "      <td>Zoe's Kitchen, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Restaurants</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>3.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3554</th>\n",
       "      <td>ZSAN</td>\n",
       "      <td>Zosano Pharma Corporation</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3555</th>\n",
       "      <td>ZTS</td>\n",
       "      <td>Zoetis Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Drugs - Generic</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3556</th>\n",
       "      <td>ZUMZ</td>\n",
       "      <td>Zumiez, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Specialty Retail, Other</td>\n",
       "      <td>USA</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3557</th>\n",
       "      <td>ZYNE</td>\n",
       "      <td>Zynerba Pharmaceuticals, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Drugs - Generic</td>\n",
       "      <td>USA</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3558 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Ticker                                  Company            Sector  \\\n",
       "0         A               Agilent Technologies, Inc.        Healthcare   \n",
       "1        AA                               Alcoa Inc.   Basic Materials   \n",
       "2       AAC                       AAC Holdings, Inc.        Healthcare   \n",
       "3       AAL             American Airlines Group Inc.          Services   \n",
       "4      AAMC  Altisource Asset Management Corporation         Financial   \n",
       "5      AAME                  Atlantic American Corp.         Financial   \n",
       "6       AAN                            Aaron's, Inc.          Services   \n",
       "7      AAOI            Applied Optoelectronics, Inc.        Technology   \n",
       "8      AAON                                AAON Inc.  Industrial Goods   \n",
       "9       AAP                  Advance Auto Parts Inc.          Services   \n",
       "10     AAPL                               Apple Inc.    Consumer Goods   \n",
       "11      AAT              American Assets Trust, Inc.         Financial   \n",
       "12     AAWW       Atlas Air Worldwide Holdings, Inc.          Services   \n",
       "13       AB           AllianceBernstein Holding L.P.         Financial   \n",
       "14     ABAX                             Abaxis, Inc.        Healthcare   \n",
       "15     ABBV                              AbbVie Inc.        Healthcare   \n",
       "16      ABC            AmerisourceBergen Corporation          Services   \n",
       "17     ABCB                           Ameris Bancorp         Financial   \n",
       "18     ABCD             Cambium Learning Group, Inc.          Services   \n",
       "19     ABDC             Alcentra Capital Corporation         Financial   \n",
       "20     ABEO                 Abeona Therapeutics Inc.        Healthcare   \n",
       "21      ABG            Asbury Automotive Group, Inc.          Services   \n",
       "22     ABIO                     ARCA biopharma, Inc.        Healthcare   \n",
       "23      ABM              ABM Industries Incorporated          Services   \n",
       "24     ABMD                            ABIOMED, Inc.        Healthcare   \n",
       "25      ABR                  Arbor Realty Trust Inc.         Financial   \n",
       "26      ABT                      Abbott Laboratories        Healthcare   \n",
       "27     ABTX              Allegiance Bancshares, Inc.         Financial   \n",
       "28     ACAD              ACADIA Pharmaceuticals Inc.        Healthcare   \n",
       "29      ACC        American Campus Communities, Inc.         Financial   \n",
       "...     ...                                      ...               ...   \n",
       "3528      Y                    Alleghany Corporation         Financial   \n",
       "3529   YELP                                Yelp Inc.        Technology   \n",
       "3530   YORW                   The York Water Company         Utilities   \n",
       "3531   YRCW                       YRC Worldwide Inc.          Services   \n",
       "3532    YUM                        Yum! Brands, Inc.          Services   \n",
       "3533   YUMA                        Yuma Energy, Inc.   Basic Materials   \n",
       "3534   YUME                               YuMe, Inc.          Services   \n",
       "3535      Z                       Zillow Group, Inc.         Financial   \n",
       "3536   ZAGG                                 ZAGG Inc          Services   \n",
       "3537   ZAIS                ZAIS Group Holdings, Inc.         Financial   \n",
       "3538   ZAYO                Zayo Group Holdings, Inc.        Technology   \n",
       "3539   ZAZA                  ZaZa Energy Corporation   Basic Materials   \n",
       "3540    ZBH             Zimmer Biomet Holdings, Inc.        Healthcare   \n",
       "3541    ZBK                     Zions Bancorporation         Financial   \n",
       "3542   ZBRA           Zebra Technologies Corporation  Industrial Goods   \n",
       "3543    ZEN                            Zendesk, Inc.        Technology   \n",
       "3544   ZEUS                       Olympic Steel Inc.   Basic Materials   \n",
       "3545   ZFGN                             Zafgen, Inc.        Healthcare   \n",
       "3546     ZG                       Zillow Group, Inc.         Financial   \n",
       "3547   ZGNX                            Zogenix, Inc.        Healthcare   \n",
       "3548   ZION                     Zions Bancorporation         Financial   \n",
       "3549   ZIOP                  ZIOPHARM Oncology, Inc.        Healthcare   \n",
       "3550   ZIXI                          Zix Corporation        Technology   \n",
       "3551     ZN                     Zion Oil & Gas, Inc.   Basic Materials   \n",
       "3552   ZNGA                              Zynga, Inc.        Technology   \n",
       "3553   ZOES                      Zoe's Kitchen, Inc.          Services   \n",
       "3554   ZSAN                Zosano Pharma Corporation        Healthcare   \n",
       "3555    ZTS                              Zoetis Inc.        Healthcare   \n",
       "3556   ZUMZ                             Zumiez, Inc.          Services   \n",
       "3557   ZYNE            Zynerba Pharmaceuticals, Inc.        Healthcare   \n",
       "\n",
       "                                 Industry Country  Clusters Level 0  \\\n",
       "0         Medical Laboratories & Research     USA                 1   \n",
       "1                                Aluminum     USA                 1   \n",
       "2             Specialized Health Services     USA                 1   \n",
       "3                          Major Airlines     USA                 1   \n",
       "4                        Asset Management     USA                 1   \n",
       "5                          Life Insurance     USA                 1   \n",
       "6               Rental & Leasing Services     USA                 1   \n",
       "7     Semiconductor - Integrated Circuits     USA                 1   \n",
       "8              General Building Materials     USA                 1   \n",
       "9                       Auto Parts Stores     USA                 1   \n",
       "10                   Electronic Equipment     USA                 1   \n",
       "11                          REIT - Retail     USA                 0   \n",
       "12                    Air Services, Other     USA                 1   \n",
       "13                       Asset Management     USA                 0   \n",
       "14        Medical Laboratories & Research     USA                 1   \n",
       "15             Drug Manufacturers - Major     USA                 2   \n",
       "16                        Drugs Wholesale     USA                 1   \n",
       "17          Regional - Mid-Atlantic Banks     USA                 3   \n",
       "18          Education & Training Services     USA                 1   \n",
       "19                       Asset Management     USA                 0   \n",
       "20                          Biotechnology     USA                 2   \n",
       "21                       Auto Dealerships     USA                 1   \n",
       "22                          Biotechnology     USA                 2   \n",
       "23                      Business Services     USA                 1   \n",
       "24         Medical Appliances & Equipment     USA                 1   \n",
       "25                     REIT - Diversified     USA                 0   \n",
       "26         Medical Appliances & Equipment     USA                 1   \n",
       "27             Regional - Southwest Banks     USA                 3   \n",
       "28                          Biotechnology     USA                 2   \n",
       "29                     REIT - Residential     USA                 0   \n",
       "...                                   ...     ...               ...   \n",
       "3528        Property & Casualty Insurance     USA                 1   \n",
       "3529       Internet Information Providers     USA                 1   \n",
       "3530                      Water Utilities     USA                 1   \n",
       "3531                             Trucking     USA                 1   \n",
       "3532                          Restaurants     USA                 1   \n",
       "3533                Independent Oil & Gas     USA                 1   \n",
       "3534                 Advertising Agencies     USA                 1   \n",
       "3535                  Property Management     USA                 0   \n",
       "3536              Specialty Retail, Other     USA                 1   \n",
       "3537                     Asset Management     USA                 1   \n",
       "3538   Networking & Communication Devices     USA                 1   \n",
       "3539                Independent Oil & Gas     USA                 4   \n",
       "3540       Medical Appliances & Equipment     USA                 1   \n",
       "3541             Regional - Pacific Banks     USA                 3   \n",
       "3542                Diversified Machinery     USA                 1   \n",
       "3543                 Application Software     USA                 1   \n",
       "3544                         Steel & Iron     USA                 1   \n",
       "3545                        Biotechnology     USA                 2   \n",
       "3546                  Property Management     USA                 0   \n",
       "3547           Drug Manufacturers - Major     USA                 2   \n",
       "3548             Regional - Pacific Banks     USA                 3   \n",
       "3549                        Biotechnology     USA                 2   \n",
       "3550                 Application Software     USA                 1   \n",
       "3551                Independent Oil & Gas     USA                 4   \n",
       "3552       Multimedia & Graphics Software     USA                 1   \n",
       "3553                          Restaurants     USA                 1   \n",
       "3554                        Biotechnology     USA                 2   \n",
       "3555                      Drugs - Generic     USA                 1   \n",
       "3556              Specialty Retail, Other     USA                 1   \n",
       "3557                      Drugs - Generic     USA                 2   \n",
       "\n",
       "      Clusters Level 1  Clusters Level 2  \n",
       "0                  0.0               2.0  \n",
       "1                  0.0               3.0  \n",
       "2                  0.0               2.0  \n",
       "3                  0.0               0.0  \n",
       "4                  0.0               0.0  \n",
       "5                  1.0               NaN  \n",
       "6                  0.0               1.0  \n",
       "7                  0.0               4.0  \n",
       "8                  0.0               0.0  \n",
       "9                  0.0               0.0  \n",
       "10                 2.0               1.0  \n",
       "11                 1.0               NaN  \n",
       "12                 0.0               4.0  \n",
       "13                 0.0               NaN  \n",
       "14                 0.0               2.0  \n",
       "15                 NaN               NaN  \n",
       "16                 0.0               2.0  \n",
       "17                 NaN               NaN  \n",
       "18                 2.0               0.0  \n",
       "19                 0.0               NaN  \n",
       "20                 NaN               NaN  \n",
       "21                 0.0               0.0  \n",
       "22                 NaN               NaN  \n",
       "23                 0.0               0.0  \n",
       "24                 0.0               0.0  \n",
       "25                 1.0               NaN  \n",
       "26                 0.0               2.0  \n",
       "27                 NaN               NaN  \n",
       "28                 NaN               NaN  \n",
       "29                 1.0               NaN  \n",
       "...                ...               ...  \n",
       "3528               1.0               NaN  \n",
       "3529               0.0               0.0  \n",
       "3530               0.0               0.0  \n",
       "3531               0.0               0.0  \n",
       "3532               3.0               NaN  \n",
       "3533               0.0               0.0  \n",
       "3534               2.0               1.0  \n",
       "3535               1.0               NaN  \n",
       "3536               2.0               2.0  \n",
       "3537               2.0               0.0  \n",
       "3538               2.0               2.0  \n",
       "3539               1.0               NaN  \n",
       "3540               0.0               2.0  \n",
       "3541               NaN               NaN  \n",
       "3542               0.0               2.0  \n",
       "3543               2.0               2.0  \n",
       "3544               0.0               0.0  \n",
       "3545               NaN               NaN  \n",
       "3546               1.0               NaN  \n",
       "3547               NaN               NaN  \n",
       "3548               NaN               NaN  \n",
       "3549               NaN               NaN  \n",
       "3550               2.0               2.0  \n",
       "3551               1.0               NaN  \n",
       "3552               2.0               3.0  \n",
       "3553               3.0               NaN  \n",
       "3554               NaN               NaN  \n",
       "3555               0.0               2.0  \n",
       "3556               0.0               1.0  \n",
       "3557               NaN               NaN  \n",
       "\n",
       "[3558 rows x 8 columns]"
      ]
     },
     "execution_count": 131,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# See the dataframe with the 3 levels, level 0, level 1 and level 2\n",
    "merged_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# copy the dataframe to a new one (has its own values)\n",
    "temp_df = merged_df.copy(deep=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## adding topics into the dataframe\n",
    "temp_df['Clusters Level 0'] = temp_df['Clusters Level 0'].astype(object)\n",
    "temp_df['Clusters Level 1'] = temp_df['Clusters Level 1'].astype(object)\n",
    "temp_df['Clusters Level 2'] = temp_df['Clusters Level 2'].astype(object)\n",
    "#temp_df['Clusters Level 3'] = temp_df['Clusters Level 3'].astype(object)\n",
    "\n",
    "#for key in topics:\n",
    " #   if len(key)==4:\n",
    "  #      l0=key[0]\n",
    "   #     l1=key[1]\n",
    "    #    l2=key[2]\n",
    "     #   l3=key[3]\n",
    "      #  test = temp_df[(temp_df['Clusters Level 0']==int(l0)) & (temp_df['Clusters Level 1']==int(l1)) & (temp_df['Clusters Level 2']==int(l2)) & (temp_df['Clusters Level 3']==int(l3))]                   \n",
    "       # for ind in test.index.values:\n",
    "        #    temp_df.at[ind,'Clusters Level 3'] =topics[key]\n",
    "\n",
    "for key in topics:    \n",
    "    if len(key)==3:\n",
    "        l0=key[0]\n",
    "        l1=key[1]\n",
    "        l2=key[2]\n",
    "        \n",
    "        test = temp_df[(temp_df['Clusters Level 0']==int(l0)) & (temp_df['Clusters Level 1']==int(l1)) & (temp_df['Clusters Level 2']==int(l2))]                   \n",
    "        for ind in test.index.values:\n",
    "            temp_df.at[ind,'Clusters Level 2'] =topics[key]\n",
    "for key in topics:\n",
    "    if len(key)==2:\n",
    "        l0=key[0]\n",
    "        l1=key[1]\n",
    "               \n",
    "        test = temp_df[(temp_df['Clusters Level 0']==int(l0)) & (temp_df['Clusters Level 1']==int(l1))]                   \n",
    "        for ind in test.index.values:\n",
    "            temp_df.at[ind,'Clusters Level 1'] =topics[key]\n",
    "for key in topics:\n",
    "    if len(key)==1:        \n",
    "        test = temp_df[temp_df['Clusters Level 0']==int(key)]\n",
    "        for i in test.index.values:\n",
    "            temp_df.at[i,'Clusters Level 0'] = topics[key]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Dataframe showing all hierarchical topics\n",
    "merged_df = temp_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Ticker</th>\n",
       "      <th>Company</th>\n",
       "      <th>Sector</th>\n",
       "      <th>Industry</th>\n",
       "      <th>Country</th>\n",
       "      <th>Clusters Level 0</th>\n",
       "      <th>Clusters Level 1</th>\n",
       "      <th>Clusters Level 2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A</td>\n",
       "      <td>Agilent Technologies, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Medical Laboratories &amp; Research</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[care, health, test, medic, develop, devic, di...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AA</td>\n",
       "      <td>Alcoa Inc.</td>\n",
       "      <td>Basic Materials</td>\n",
       "      <td>Aluminum</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[food, beverag, distribut, natur, specialti, g...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AAC</td>\n",
       "      <td>AAC Holdings, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Specialized Health Services</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[care, health, test, medic, develop, devic, di...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>AAL</td>\n",
       "      <td>American Airlines Group Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Major Airlines</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[power, energi, construct, engin, distribut, c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>AAMC</td>\n",
       "      <td>Altisource Asset Management Corporation</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Asset Management</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[power, energi, construct, engin, distribut, c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>AAME</td>\n",
       "      <td>Atlantic American Corp.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Life Insurance</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[insur, life, properti, casualti, health, pers...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>AAN</td>\n",
       "      <td>Aaron's, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Rental &amp; Leasing Services</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[home, accessori, apparel, oper, variou, desig...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>AAOI</td>\n",
       "      <td>Applied Optoelectronics, Inc.</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Semiconductor - Integrated Circuits</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[compon, applic, electron, control, design, de...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>AAON</td>\n",
       "      <td>AAON Inc.</td>\n",
       "      <td>Industrial Goods</td>\n",
       "      <td>General Building Materials</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[power, energi, construct, engin, distribut, c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>AAP</td>\n",
       "      <td>Advance Auto Parts Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Auto Parts Stores</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[power, energi, construct, engin, distribut, c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>AAPL</td>\n",
       "      <td>Apple Inc.</td>\n",
       "      <td>Consumer Goods</td>\n",
       "      <td>Electronic Equipment</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[softwar, network, data, applic, technolog, mo...</td>\n",
       "      <td>[digit, video, content, network, media, televi...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>AAT</td>\n",
       "      <td>American Assets Trust, Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>REIT - Retail</td>\n",
       "      <td>USA</td>\n",
       "      <td>[invest, fund, real, estat, equiti, properti, ...</td>\n",
       "      <td>[real, invest, estat, properti, trust, mortgag...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>AAWW</td>\n",
       "      <td>Atlas Air Worldwide Holdings, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Air Services, Other</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[compon, applic, electron, control, design, de...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>AB</td>\n",
       "      <td>AllianceBernstein Holding L.P.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Asset Management</td>\n",
       "      <td>USA</td>\n",
       "      <td>[invest, fund, real, estat, equiti, properti, ...</td>\n",
       "      <td>[invest, fund, equiti, secur, financi, incom, ...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>ABAX</td>\n",
       "      <td>Abaxis, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Medical Laboratories &amp; Research</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[care, health, test, medic, develop, devic, di...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>ABBV</td>\n",
       "      <td>AbbVie Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Drug Manufacturers - Major</td>\n",
       "      <td>USA</td>\n",
       "      <td>[develop, treatment, clinic, phase, trial, pha...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>ABC</td>\n",
       "      <td>AmerisourceBergen Corporation</td>\n",
       "      <td>Services</td>\n",
       "      <td>Drugs Wholesale</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[care, health, test, medic, develop, devic, di...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>ABCB</td>\n",
       "      <td>Ameris Bancorp</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Regional - Mid-Atlantic Banks</td>\n",
       "      <td>USA</td>\n",
       "      <td>[loan, bank, deposit, account, estat, real, cr...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>ABCD</td>\n",
       "      <td>Cambium Learning Group, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Education &amp; Training Services</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[softwar, network, data, applic, technolog, mo...</td>\n",
       "      <td>[payment, financi, consum, loan, technolog, cr...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>ABDC</td>\n",
       "      <td>Alcentra Capital Corporation</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Asset Management</td>\n",
       "      <td>USA</td>\n",
       "      <td>[invest, fund, real, estat, equiti, properti, ...</td>\n",
       "      <td>[invest, fund, equiti, secur, financi, incom, ...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>ABEO</td>\n",
       "      <td>Abeona Therapeutics Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>[develop, treatment, clinic, phase, trial, pha...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>ABG</td>\n",
       "      <td>Asbury Automotive Group, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Auto Dealerships</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[power, energi, construct, engin, distribut, c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>ABIO</td>\n",
       "      <td>ARCA biopharma, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>[develop, treatment, clinic, phase, trial, pha...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>ABM</td>\n",
       "      <td>ABM Industries Incorporated</td>\n",
       "      <td>Services</td>\n",
       "      <td>Business Services</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[power, energi, construct, engin, distribut, c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>ABMD</td>\n",
       "      <td>ABIOMED, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Medical Appliances &amp; Equipment</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[power, energi, construct, engin, distribut, c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>ABR</td>\n",
       "      <td>Arbor Realty Trust Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>REIT - Diversified</td>\n",
       "      <td>USA</td>\n",
       "      <td>[invest, fund, real, estat, equiti, properti, ...</td>\n",
       "      <td>[real, invest, estat, properti, trust, mortgag...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>ABT</td>\n",
       "      <td>Abbott Laboratories</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Medical Appliances &amp; Equipment</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[care, health, test, medic, develop, devic, di...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>ABTX</td>\n",
       "      <td>Allegiance Bancshares, Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Regional - Southwest Banks</td>\n",
       "      <td>USA</td>\n",
       "      <td>[loan, bank, deposit, account, estat, real, cr...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>ACAD</td>\n",
       "      <td>ACADIA Pharmaceuticals Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>[develop, treatment, clinic, phase, trial, pha...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>ACC</td>\n",
       "      <td>American Campus Communities, Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>REIT - Residential</td>\n",
       "      <td>USA</td>\n",
       "      <td>[invest, fund, real, estat, equiti, properti, ...</td>\n",
       "      <td>[real, invest, estat, properti, trust, mortgag...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3528</th>\n",
       "      <td>Y</td>\n",
       "      <td>Alleghany Corporation</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Property &amp; Casualty Insurance</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[insur, life, properti, casualti, health, pers...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3529</th>\n",
       "      <td>YELP</td>\n",
       "      <td>Yelp Inc.</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Internet Information Providers</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[power, energi, construct, engin, distribut, c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3530</th>\n",
       "      <td>YORW</td>\n",
       "      <td>The York Water Company</td>\n",
       "      <td>Utilities</td>\n",
       "      <td>Water Utilities</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[power, energi, construct, engin, distribut, c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3531</th>\n",
       "      <td>YRCW</td>\n",
       "      <td>YRC Worldwide Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Trucking</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[power, energi, construct, engin, distribut, c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3532</th>\n",
       "      <td>YUM</td>\n",
       "      <td>Yum! Brands, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Restaurants</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[restaur, franchis, oper, decemb, develop, bas...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3533</th>\n",
       "      <td>YUMA</td>\n",
       "      <td>Yuma Energy, Inc.</td>\n",
       "      <td>Basic Materials</td>\n",
       "      <td>Independent Oil &amp; Gas</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[power, energi, construct, engin, distribut, c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3534</th>\n",
       "      <td>YUME</td>\n",
       "      <td>YuMe, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Advertising Agencies</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[softwar, network, data, applic, technolog, mo...</td>\n",
       "      <td>[digit, video, content, network, media, televi...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3535</th>\n",
       "      <td>Z</td>\n",
       "      <td>Zillow Group, Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Property Management</td>\n",
       "      <td>USA</td>\n",
       "      <td>[invest, fund, real, estat, equiti, properti, ...</td>\n",
       "      <td>[real, invest, estat, properti, trust, mortgag...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3536</th>\n",
       "      <td>ZAGG</td>\n",
       "      <td>ZAGG Inc</td>\n",
       "      <td>Services</td>\n",
       "      <td>Specialty Retail, Other</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[softwar, network, data, applic, technolog, mo...</td>\n",
       "      <td>[softwar, network, data, applic, cloud, mobil,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3537</th>\n",
       "      <td>ZAIS</td>\n",
       "      <td>ZAIS Group Holdings, Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Asset Management</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[softwar, network, data, applic, technolog, mo...</td>\n",
       "      <td>[payment, financi, consum, loan, technolog, cr...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3538</th>\n",
       "      <td>ZAYO</td>\n",
       "      <td>Zayo Group Holdings, Inc.</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Networking &amp; Communication Devices</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[softwar, network, data, applic, technolog, mo...</td>\n",
       "      <td>[softwar, network, data, applic, cloud, mobil,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3539</th>\n",
       "      <td>ZAZA</td>\n",
       "      <td>ZaZa Energy Corporation</td>\n",
       "      <td>Basic Materials</td>\n",
       "      <td>Independent Oil &amp; Gas</td>\n",
       "      <td>USA</td>\n",
       "      <td>[ga, natur, oil, energi, pipelin, crude, elect...</td>\n",
       "      <td>[explor, develop, energi, reserv, basin, decem...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3540</th>\n",
       "      <td>ZBH</td>\n",
       "      <td>Zimmer Biomet Holdings, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Medical Appliances &amp; Equipment</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[care, health, test, medic, develop, devic, di...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3541</th>\n",
       "      <td>ZBK</td>\n",
       "      <td>Zions Bancorporation</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Regional - Pacific Banks</td>\n",
       "      <td>USA</td>\n",
       "      <td>[loan, bank, deposit, account, estat, real, cr...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3542</th>\n",
       "      <td>ZBRA</td>\n",
       "      <td>Zebra Technologies Corporation</td>\n",
       "      <td>Industrial Goods</td>\n",
       "      <td>Diversified Machinery</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[care, health, test, medic, develop, devic, di...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3543</th>\n",
       "      <td>ZEN</td>\n",
       "      <td>Zendesk, Inc.</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Application Software</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[softwar, network, data, applic, technolog, mo...</td>\n",
       "      <td>[softwar, network, data, applic, cloud, mobil,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3544</th>\n",
       "      <td>ZEUS</td>\n",
       "      <td>Olympic Steel Inc.</td>\n",
       "      <td>Basic Materials</td>\n",
       "      <td>Steel &amp; Iron</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[power, energi, construct, engin, distribut, c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3545</th>\n",
       "      <td>ZFGN</td>\n",
       "      <td>Zafgen, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>[develop, treatment, clinic, phase, trial, pha...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3546</th>\n",
       "      <td>ZG</td>\n",
       "      <td>Zillow Group, Inc.</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Property Management</td>\n",
       "      <td>USA</td>\n",
       "      <td>[invest, fund, real, estat, equiti, properti, ...</td>\n",
       "      <td>[real, invest, estat, properti, trust, mortgag...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3547</th>\n",
       "      <td>ZGNX</td>\n",
       "      <td>Zogenix, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Drug Manufacturers - Major</td>\n",
       "      <td>USA</td>\n",
       "      <td>[develop, treatment, clinic, phase, trial, pha...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3548</th>\n",
       "      <td>ZION</td>\n",
       "      <td>Zions Bancorporation</td>\n",
       "      <td>Financial</td>\n",
       "      <td>Regional - Pacific Banks</td>\n",
       "      <td>USA</td>\n",
       "      <td>[loan, bank, deposit, account, estat, real, cr...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3549</th>\n",
       "      <td>ZIOP</td>\n",
       "      <td>ZIOPHARM Oncology, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>[develop, treatment, clinic, phase, trial, pha...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3550</th>\n",
       "      <td>ZIXI</td>\n",
       "      <td>Zix Corporation</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Application Software</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[softwar, network, data, applic, technolog, mo...</td>\n",
       "      <td>[softwar, network, data, applic, cloud, mobil,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3551</th>\n",
       "      <td>ZN</td>\n",
       "      <td>Zion Oil &amp; Gas, Inc.</td>\n",
       "      <td>Basic Materials</td>\n",
       "      <td>Independent Oil &amp; Gas</td>\n",
       "      <td>USA</td>\n",
       "      <td>[ga, natur, oil, energi, pipelin, crude, elect...</td>\n",
       "      <td>[explor, develop, energi, reserv, basin, decem...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3552</th>\n",
       "      <td>ZNGA</td>\n",
       "      <td>Zynga, Inc.</td>\n",
       "      <td>Technology</td>\n",
       "      <td>Multimedia &amp; Graphics Software</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[softwar, network, data, applic, technolog, mo...</td>\n",
       "      <td>[game, entertain, park, hotel, base, sport, ro...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3553</th>\n",
       "      <td>ZOES</td>\n",
       "      <td>Zoe's Kitchen, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Restaurants</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[restaur, franchis, oper, decemb, develop, bas...</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3554</th>\n",
       "      <td>ZSAN</td>\n",
       "      <td>Zosano Pharma Corporation</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Biotechnology</td>\n",
       "      <td>USA</td>\n",
       "      <td>[develop, treatment, clinic, phase, trial, pha...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3555</th>\n",
       "      <td>ZTS</td>\n",
       "      <td>Zoetis Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Drugs - Generic</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[care, health, test, medic, develop, devic, di...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3556</th>\n",
       "      <td>ZUMZ</td>\n",
       "      <td>Zumiez, Inc.</td>\n",
       "      <td>Services</td>\n",
       "      <td>Specialty Retail, Other</td>\n",
       "      <td>USA</td>\n",
       "      <td>[equip, applic, technolog, base, variou, retai...</td>\n",
       "      <td>[equip, retail, variou, design, develop, distr...</td>\n",
       "      <td>[home, accessori, apparel, oper, variou, desig...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3557</th>\n",
       "      <td>ZYNE</td>\n",
       "      <td>Zynerba Pharmaceuticals, Inc.</td>\n",
       "      <td>Healthcare</td>\n",
       "      <td>Drugs - Generic</td>\n",
       "      <td>USA</td>\n",
       "      <td>[develop, treatment, clinic, phase, trial, pha...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3558 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Ticker                                  Company            Sector  \\\n",
       "0         A               Agilent Technologies, Inc.        Healthcare   \n",
       "1        AA                               Alcoa Inc.   Basic Materials   \n",
       "2       AAC                       AAC Holdings, Inc.        Healthcare   \n",
       "3       AAL             American Airlines Group Inc.          Services   \n",
       "4      AAMC  Altisource Asset Management Corporation         Financial   \n",
       "5      AAME                  Atlantic American Corp.         Financial   \n",
       "6       AAN                            Aaron's, Inc.          Services   \n",
       "7      AAOI            Applied Optoelectronics, Inc.        Technology   \n",
       "8      AAON                                AAON Inc.  Industrial Goods   \n",
       "9       AAP                  Advance Auto Parts Inc.          Services   \n",
       "10     AAPL                               Apple Inc.    Consumer Goods   \n",
       "11      AAT              American Assets Trust, Inc.         Financial   \n",
       "12     AAWW       Atlas Air Worldwide Holdings, Inc.          Services   \n",
       "13       AB           AllianceBernstein Holding L.P.         Financial   \n",
       "14     ABAX                             Abaxis, Inc.        Healthcare   \n",
       "15     ABBV                              AbbVie Inc.        Healthcare   \n",
       "16      ABC            AmerisourceBergen Corporation          Services   \n",
       "17     ABCB                           Ameris Bancorp         Financial   \n",
       "18     ABCD             Cambium Learning Group, Inc.          Services   \n",
       "19     ABDC             Alcentra Capital Corporation         Financial   \n",
       "20     ABEO                 Abeona Therapeutics Inc.        Healthcare   \n",
       "21      ABG            Asbury Automotive Group, Inc.          Services   \n",
       "22     ABIO                     ARCA biopharma, Inc.        Healthcare   \n",
       "23      ABM              ABM Industries Incorporated          Services   \n",
       "24     ABMD                            ABIOMED, Inc.        Healthcare   \n",
       "25      ABR                  Arbor Realty Trust Inc.         Financial   \n",
       "26      ABT                      Abbott Laboratories        Healthcare   \n",
       "27     ABTX              Allegiance Bancshares, Inc.         Financial   \n",
       "28     ACAD              ACADIA Pharmaceuticals Inc.        Healthcare   \n",
       "29      ACC        American Campus Communities, Inc.         Financial   \n",
       "...     ...                                      ...               ...   \n",
       "3528      Y                    Alleghany Corporation         Financial   \n",
       "3529   YELP                                Yelp Inc.        Technology   \n",
       "3530   YORW                   The York Water Company         Utilities   \n",
       "3531   YRCW                       YRC Worldwide Inc.          Services   \n",
       "3532    YUM                        Yum! Brands, Inc.          Services   \n",
       "3533   YUMA                        Yuma Energy, Inc.   Basic Materials   \n",
       "3534   YUME                               YuMe, Inc.          Services   \n",
       "3535      Z                       Zillow Group, Inc.         Financial   \n",
       "3536   ZAGG                                 ZAGG Inc          Services   \n",
       "3537   ZAIS                ZAIS Group Holdings, Inc.         Financial   \n",
       "3538   ZAYO                Zayo Group Holdings, Inc.        Technology   \n",
       "3539   ZAZA                  ZaZa Energy Corporation   Basic Materials   \n",
       "3540    ZBH             Zimmer Biomet Holdings, Inc.        Healthcare   \n",
       "3541    ZBK                     Zions Bancorporation         Financial   \n",
       "3542   ZBRA           Zebra Technologies Corporation  Industrial Goods   \n",
       "3543    ZEN                            Zendesk, Inc.        Technology   \n",
       "3544   ZEUS                       Olympic Steel Inc.   Basic Materials   \n",
       "3545   ZFGN                             Zafgen, Inc.        Healthcare   \n",
       "3546     ZG                       Zillow Group, Inc.         Financial   \n",
       "3547   ZGNX                            Zogenix, Inc.        Healthcare   \n",
       "3548   ZION                     Zions Bancorporation         Financial   \n",
       "3549   ZIOP                  ZIOPHARM Oncology, Inc.        Healthcare   \n",
       "3550   ZIXI                          Zix Corporation        Technology   \n",
       "3551     ZN                     Zion Oil & Gas, Inc.   Basic Materials   \n",
       "3552   ZNGA                              Zynga, Inc.        Technology   \n",
       "3553   ZOES                      Zoe's Kitchen, Inc.          Services   \n",
       "3554   ZSAN                Zosano Pharma Corporation        Healthcare   \n",
       "3555    ZTS                              Zoetis Inc.        Healthcare   \n",
       "3556   ZUMZ                             Zumiez, Inc.          Services   \n",
       "3557   ZYNE            Zynerba Pharmaceuticals, Inc.        Healthcare   \n",
       "\n",
       "                                 Industry Country  \\\n",
       "0         Medical Laboratories & Research     USA   \n",
       "1                                Aluminum     USA   \n",
       "2             Specialized Health Services     USA   \n",
       "3                          Major Airlines     USA   \n",
       "4                        Asset Management     USA   \n",
       "5                          Life Insurance     USA   \n",
       "6               Rental & Leasing Services     USA   \n",
       "7     Semiconductor - Integrated Circuits     USA   \n",
       "8              General Building Materials     USA   \n",
       "9                       Auto Parts Stores     USA   \n",
       "10                   Electronic Equipment     USA   \n",
       "11                          REIT - Retail     USA   \n",
       "12                    Air Services, Other     USA   \n",
       "13                       Asset Management     USA   \n",
       "14        Medical Laboratories & Research     USA   \n",
       "15             Drug Manufacturers - Major     USA   \n",
       "16                        Drugs Wholesale     USA   \n",
       "17          Regional - Mid-Atlantic Banks     USA   \n",
       "18          Education & Training Services     USA   \n",
       "19                       Asset Management     USA   \n",
       "20                          Biotechnology     USA   \n",
       "21                       Auto Dealerships     USA   \n",
       "22                          Biotechnology     USA   \n",
       "23                      Business Services     USA   \n",
       "24         Medical Appliances & Equipment     USA   \n",
       "25                     REIT - Diversified     USA   \n",
       "26         Medical Appliances & Equipment     USA   \n",
       "27             Regional - Southwest Banks     USA   \n",
       "28                          Biotechnology     USA   \n",
       "29                     REIT - Residential     USA   \n",
       "...                                   ...     ...   \n",
       "3528        Property & Casualty Insurance     USA   \n",
       "3529       Internet Information Providers     USA   \n",
       "3530                      Water Utilities     USA   \n",
       "3531                             Trucking     USA   \n",
       "3532                          Restaurants     USA   \n",
       "3533                Independent Oil & Gas     USA   \n",
       "3534                 Advertising Agencies     USA   \n",
       "3535                  Property Management     USA   \n",
       "3536              Specialty Retail, Other     USA   \n",
       "3537                     Asset Management     USA   \n",
       "3538   Networking & Communication Devices     USA   \n",
       "3539                Independent Oil & Gas     USA   \n",
       "3540       Medical Appliances & Equipment     USA   \n",
       "3541             Regional - Pacific Banks     USA   \n",
       "3542                Diversified Machinery     USA   \n",
       "3543                 Application Software     USA   \n",
       "3544                         Steel & Iron     USA   \n",
       "3545                        Biotechnology     USA   \n",
       "3546                  Property Management     USA   \n",
       "3547           Drug Manufacturers - Major     USA   \n",
       "3548             Regional - Pacific Banks     USA   \n",
       "3549                        Biotechnology     USA   \n",
       "3550                 Application Software     USA   \n",
       "3551                Independent Oil & Gas     USA   \n",
       "3552       Multimedia & Graphics Software     USA   \n",
       "3553                          Restaurants     USA   \n",
       "3554                        Biotechnology     USA   \n",
       "3555                      Drugs - Generic     USA   \n",
       "3556              Specialty Retail, Other     USA   \n",
       "3557                      Drugs - Generic     USA   \n",
       "\n",
       "                                       Clusters Level 0  \\\n",
       "0     [equip, applic, technolog, base, variou, retai...   \n",
       "1     [equip, applic, technolog, base, variou, retai...   \n",
       "2     [equip, applic, technolog, base, variou, retai...   \n",
       "3     [equip, applic, technolog, base, variou, retai...   \n",
       "4     [equip, applic, technolog, base, variou, retai...   \n",
       "5     [equip, applic, technolog, base, variou, retai...   \n",
       "6     [equip, applic, technolog, base, variou, retai...   \n",
       "7     [equip, applic, technolog, base, variou, retai...   \n",
       "8     [equip, applic, technolog, base, variou, retai...   \n",
       "9     [equip, applic, technolog, base, variou, retai...   \n",
       "10    [equip, applic, technolog, base, variou, retai...   \n",
       "11    [invest, fund, real, estat, equiti, properti, ...   \n",
       "12    [equip, applic, technolog, base, variou, retai...   \n",
       "13    [invest, fund, real, estat, equiti, properti, ...   \n",
       "14    [equip, applic, technolog, base, variou, retai...   \n",
       "15    [develop, treatment, clinic, phase, trial, pha...   \n",
       "16    [equip, applic, technolog, base, variou, retai...   \n",
       "17    [loan, bank, deposit, account, estat, real, cr...   \n",
       "18    [equip, applic, technolog, base, variou, retai...   \n",
       "19    [invest, fund, real, estat, equiti, properti, ...   \n",
       "20    [develop, treatment, clinic, phase, trial, pha...   \n",
       "21    [equip, applic, technolog, base, variou, retai...   \n",
       "22    [develop, treatment, clinic, phase, trial, pha...   \n",
       "23    [equip, applic, technolog, base, variou, retai...   \n",
       "24    [equip, applic, technolog, base, variou, retai...   \n",
       "25    [invest, fund, real, estat, equiti, properti, ...   \n",
       "26    [equip, applic, technolog, base, variou, retai...   \n",
       "27    [loan, bank, deposit, account, estat, real, cr...   \n",
       "28    [develop, treatment, clinic, phase, trial, pha...   \n",
       "29    [invest, fund, real, estat, equiti, properti, ...   \n",
       "...                                                 ...   \n",
       "3528  [equip, applic, technolog, base, variou, retai...   \n",
       "3529  [equip, applic, technolog, base, variou, retai...   \n",
       "3530  [equip, applic, technolog, base, variou, retai...   \n",
       "3531  [equip, applic, technolog, base, variou, retai...   \n",
       "3532  [equip, applic, technolog, base, variou, retai...   \n",
       "3533  [equip, applic, technolog, base, variou, retai...   \n",
       "3534  [equip, applic, technolog, base, variou, retai...   \n",
       "3535  [invest, fund, real, estat, equiti, properti, ...   \n",
       "3536  [equip, applic, technolog, base, variou, retai...   \n",
       "3537  [equip, applic, technolog, base, variou, retai...   \n",
       "3538  [equip, applic, technolog, base, variou, retai...   \n",
       "3539  [ga, natur, oil, energi, pipelin, crude, elect...   \n",
       "3540  [equip, applic, technolog, base, variou, retai...   \n",
       "3541  [loan, bank, deposit, account, estat, real, cr...   \n",
       "3542  [equip, applic, technolog, base, variou, retai...   \n",
       "3543  [equip, applic, technolog, base, variou, retai...   \n",
       "3544  [equip, applic, technolog, base, variou, retai...   \n",
       "3545  [develop, treatment, clinic, phase, trial, pha...   \n",
       "3546  [invest, fund, real, estat, equiti, properti, ...   \n",
       "3547  [develop, treatment, clinic, phase, trial, pha...   \n",
       "3548  [loan, bank, deposit, account, estat, real, cr...   \n",
       "3549  [develop, treatment, clinic, phase, trial, pha...   \n",
       "3550  [equip, applic, technolog, base, variou, retai...   \n",
       "3551  [ga, natur, oil, energi, pipelin, crude, elect...   \n",
       "3552  [equip, applic, technolog, base, variou, retai...   \n",
       "3553  [equip, applic, technolog, base, variou, retai...   \n",
       "3554  [develop, treatment, clinic, phase, trial, pha...   \n",
       "3555  [equip, applic, technolog, base, variou, retai...   \n",
       "3556  [equip, applic, technolog, base, variou, retai...   \n",
       "3557  [develop, treatment, clinic, phase, trial, pha...   \n",
       "\n",
       "                                       Clusters Level 1  \\\n",
       "0     [equip, retail, variou, design, develop, distr...   \n",
       "1     [equip, retail, variou, design, develop, distr...   \n",
       "2     [equip, retail, variou, design, develop, distr...   \n",
       "3     [equip, retail, variou, design, develop, distr...   \n",
       "4     [equip, retail, variou, design, develop, distr...   \n",
       "5     [insur, life, properti, casualti, health, pers...   \n",
       "6     [equip, retail, variou, design, develop, distr...   \n",
       "7     [equip, retail, variou, design, develop, distr...   \n",
       "8     [equip, retail, variou, design, develop, distr...   \n",
       "9     [equip, retail, variou, design, develop, distr...   \n",
       "10    [softwar, network, data, applic, technolog, mo...   \n",
       "11    [real, invest, estat, properti, trust, mortgag...   \n",
       "12    [equip, retail, variou, design, develop, distr...   \n",
       "13    [invest, fund, equiti, secur, financi, incom, ...   \n",
       "14    [equip, retail, variou, design, develop, distr...   \n",
       "15                                                  NaN   \n",
       "16    [equip, retail, variou, design, develop, distr...   \n",
       "17                                                  NaN   \n",
       "18    [softwar, network, data, applic, technolog, mo...   \n",
       "19    [invest, fund, equiti, secur, financi, incom, ...   \n",
       "20                                                  NaN   \n",
       "21    [equip, retail, variou, design, develop, distr...   \n",
       "22                                                  NaN   \n",
       "23    [equip, retail, variou, design, develop, distr...   \n",
       "24    [equip, retail, variou, design, develop, distr...   \n",
       "25    [real, invest, estat, properti, trust, mortgag...   \n",
       "26    [equip, retail, variou, design, develop, distr...   \n",
       "27                                                  NaN   \n",
       "28                                                  NaN   \n",
       "29    [real, invest, estat, properti, trust, mortgag...   \n",
       "...                                                 ...   \n",
       "3528  [insur, life, properti, casualti, health, pers...   \n",
       "3529  [equip, retail, variou, design, develop, distr...   \n",
       "3530  [equip, retail, variou, design, develop, distr...   \n",
       "3531  [equip, retail, variou, design, develop, distr...   \n",
       "3532  [restaur, franchis, oper, decemb, develop, bas...   \n",
       "3533  [equip, retail, variou, design, develop, distr...   \n",
       "3534  [softwar, network, data, applic, technolog, mo...   \n",
       "3535  [real, invest, estat, properti, trust, mortgag...   \n",
       "3536  [softwar, network, data, applic, technolog, mo...   \n",
       "3537  [softwar, network, data, applic, technolog, mo...   \n",
       "3538  [softwar, network, data, applic, technolog, mo...   \n",
       "3539  [explor, develop, energi, reserv, basin, decem...   \n",
       "3540  [equip, retail, variou, design, develop, distr...   \n",
       "3541                                                NaN   \n",
       "3542  [equip, retail, variou, design, develop, distr...   \n",
       "3543  [softwar, network, data, applic, technolog, mo...   \n",
       "3544  [equip, retail, variou, design, develop, distr...   \n",
       "3545                                                NaN   \n",
       "3546  [real, invest, estat, properti, trust, mortgag...   \n",
       "3547                                                NaN   \n",
       "3548                                                NaN   \n",
       "3549                                                NaN   \n",
       "3550  [softwar, network, data, applic, technolog, mo...   \n",
       "3551  [explor, develop, energi, reserv, basin, decem...   \n",
       "3552  [softwar, network, data, applic, technolog, mo...   \n",
       "3553  [restaur, franchis, oper, decemb, develop, bas...   \n",
       "3554                                                NaN   \n",
       "3555  [equip, retail, variou, design, develop, distr...   \n",
       "3556  [equip, retail, variou, design, develop, distr...   \n",
       "3557                                                NaN   \n",
       "\n",
       "                                       Clusters Level 2  \n",
       "0     [care, health, test, medic, develop, devic, di...  \n",
       "1     [food, beverag, distribut, natur, specialti, g...  \n",
       "2     [care, health, test, medic, develop, devic, di...  \n",
       "3     [power, energi, construct, engin, distribut, c...  \n",
       "4     [power, energi, construct, engin, distribut, c...  \n",
       "5                                                   NaN  \n",
       "6     [home, accessori, apparel, oper, variou, desig...  \n",
       "7     [compon, applic, electron, control, design, de...  \n",
       "8     [power, energi, construct, engin, distribut, c...  \n",
       "9     [power, energi, construct, engin, distribut, c...  \n",
       "10    [digit, video, content, network, media, televi...  \n",
       "11                                                  NaN  \n",
       "12    [compon, applic, electron, control, design, de...  \n",
       "13                                                  NaN  \n",
       "14    [care, health, test, medic, develop, devic, di...  \n",
       "15                                                  NaN  \n",
       "16    [care, health, test, medic, develop, devic, di...  \n",
       "17                                                  NaN  \n",
       "18    [payment, financi, consum, loan, technolog, cr...  \n",
       "19                                                  NaN  \n",
       "20                                                  NaN  \n",
       "21    [power, energi, construct, engin, distribut, c...  \n",
       "22                                                  NaN  \n",
       "23    [power, energi, construct, engin, distribut, c...  \n",
       "24    [power, energi, construct, engin, distribut, c...  \n",
       "25                                                  NaN  \n",
       "26    [care, health, test, medic, develop, devic, di...  \n",
       "27                                                  NaN  \n",
       "28                                                  NaN  \n",
       "29                                                  NaN  \n",
       "...                                                 ...  \n",
       "3528                                                NaN  \n",
       "3529  [power, energi, construct, engin, distribut, c...  \n",
       "3530  [power, energi, construct, engin, distribut, c...  \n",
       "3531  [power, energi, construct, engin, distribut, c...  \n",
       "3532                                                NaN  \n",
       "3533  [power, energi, construct, engin, distribut, c...  \n",
       "3534  [digit, video, content, network, media, televi...  \n",
       "3535                                                NaN  \n",
       "3536  [softwar, network, data, applic, cloud, mobil,...  \n",
       "3537  [payment, financi, consum, loan, technolog, cr...  \n",
       "3538  [softwar, network, data, applic, cloud, mobil,...  \n",
       "3539                                                NaN  \n",
       "3540  [care, health, test, medic, develop, devic, di...  \n",
       "3541                                                NaN  \n",
       "3542  [care, health, test, medic, develop, devic, di...  \n",
       "3543  [softwar, network, data, applic, cloud, mobil,...  \n",
       "3544  [power, energi, construct, engin, distribut, c...  \n",
       "3545                                                NaN  \n",
       "3546                                                NaN  \n",
       "3547                                                NaN  \n",
       "3548                                                NaN  \n",
       "3549                                                NaN  \n",
       "3550  [softwar, network, data, applic, cloud, mobil,...  \n",
       "3551                                                NaN  \n",
       "3552  [game, entertain, park, hotel, base, sport, ro...  \n",
       "3553                                                NaN  \n",
       "3554                                                NaN  \n",
       "3555  [care, health, test, medic, develop, devic, di...  \n",
       "3556  [home, accessori, apparel, oper, variou, desig...  \n",
       "3557                                                NaN  \n",
       "\n",
       "[3558 rows x 8 columns]"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "merged_df"
   ]
  }
 ],
 "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
}