LDA Topic Modeling and Clustering.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LDA Topic Modeling and Clustering"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use the LDA model to create topics, or clusters, present in the dataset using the abstracts of the papers. Each topic corresponds to a set of word-probability pairs, and I chose to use the top 15 highest probability words of each topic to represent the given topic. Each paper is assigned a topic based on how many of the top 15 words of each topic are included in the abstract, and how often."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# download gensim and wordcloud libraries with\n",
    "# pip install --upgrade gensim\n",
    "# pip install wordcloud"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/jayfeng/opt/anaconda3/lib/python3.8/site-packages/gensim/similarities/__init__.py:15: UserWarning: The gensim.similarities.levenshtein submodule is disabled, because the optional Levenshtein package <https://pypi.org/project/python-Levenshtein/> is unavailable. Install Levenhstein (e.g. `pip install python-Levenshtein`) to suppress this warning.\n",
      "  warnings.warn(msg)\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import json\n",
    "import itertools\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "from sklearn.cluster import KMeans\n",
    "\n",
    "from sklearn.cluster import MiniBatchKMeans\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.manifold import TSNE\n",
    "\n",
    "import nltk\n",
    "from nltk.stem.snowball import SnowballStemmer\n",
    "#from nltk.stem.porter import PorterStemmer\n",
    "from nltk.corpus import stopwords \n",
    "#nltk.download('stopwords')\n",
    "from nltk.tokenize import word_tokenize \n",
    "#nltk.download('punkt')\n",
    "from nltk.stem import WordNetLemmatizer \n",
    "#nltk.download('wordnet')\n",
    "\n",
    "import re\n",
    "\n",
    "import gensim\n",
    "from gensim import corpora, models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Read in abstracts.csv and filters out rows with missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"abstracts.csv\")\n",
    "df = df[df[\"abstract\"] != \"NaN\"]\n",
    "df = df.dropna()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Set up stop words, stemmer, and lemmatizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "stop_words = set(stopwords.words('english')) \n",
    "snowBallStemmer = SnowballStemmer(\"english\")\n",
    "lemmatizer = WordNetLemmatizer()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Tokenize and clean the abstracts of every paper"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tokenize_clean(abstract):\n",
    "    #tokenizes abstract string\n",
    "    tokens = word_tokenize(abstract.lower())\n",
    "    \n",
    "    #lemmatizes tokens\n",
    "    counter = 0\n",
    "    while counter < len(tokens):\n",
    "        tokens[counter] = lemmatizer.lemmatize(tokens[counter])\n",
    "        counter += 1\n",
    "    \n",
    "    #filters, stems, and lowercases tokens\n",
    "    filtered_tokens = []\n",
    "    for i in tokens:\n",
    "        if i not in stop_words and len(i) > 3 and i != \"abstract\":\n",
    "            stemmed_word = snowBallStemmer.stem(i)\n",
    "            filtered_tokens.append(stemmed_word)\n",
    "    \n",
    "    return filtered_tokens\n",
    "\n",
    "df[\"abstract tokens\"] = df.apply(lambda row: tokenize_clean(row.abstract), axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Perform LDA topic modelling on a sample of the papers for time/speed purposes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "partial_df = df.sample(10000)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create a list of lists of cleaned tokens of abstracts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "partial_texts = []\n",
    "for index, row in partial_df.iterrows():\n",
    "    partial_texts.append(row[3])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use Gensim library to perform LDA topic modelling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "dictionary = corpora.Dictionary(partial_texts)\n",
    "dictionary.filter_extremes(no_below=3)\n",
    "\n",
    "corpus = [dictionary.doc2bow(text) for text in partial_texts]\n",
    "\n",
    "ldamodel = gensim.models.ldamodel.LdaModel(corpus, num_topics=10, id2word = dictionary, passes=50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(0,\n",
       "  [('virus', 0.036254164),\n",
       "   ('sequenc', 0.023398373),\n",
       "   ('strain', 0.02198435),\n",
       "   ('genom', 0.01556951),\n",
       "   ('gene', 0.014512897),\n",
       "   ('coronavirus', 0.013465737),\n",
       "   ('speci', 0.01161349),\n",
       "   ('isol', 0.0113983005),\n",
       "   ('human', 0.010600142),\n",
       "   ('analysi', 0.009933317),\n",
       "   ('pedv', 0.009383051),\n",
       "   ('studi', 0.0090853665),\n",
       "   ('identifi', 0.007882741),\n",
       "   ('differ', 0.007333822),\n",
       "   ('viral', 0.0071400437),\n",
       "   ('infect', 0.007107439),\n",
       "   ('region', 0.0070415894),\n",
       "   ('anim', 0.006940225),\n",
       "   ('sampl', 0.0068891817),\n",
       "   ('porcin', 0.0068152435)]),\n",
       " (1,\n",
       "  [('patient', 0.035610527),\n",
       "   ('respiratori', 0.019537201),\n",
       "   ('infect', 0.014949),\n",
       "   ('studi', 0.013843928),\n",
       "   ('child', 0.010920468),\n",
       "   ('hospit', 0.010721948),\n",
       "   ('clinic', 0.010582626),\n",
       "   ('group', 0.010419497),\n",
       "   ('associ', 0.009747383),\n",
       "   ('pneumonia', 0.007978254),\n",
       "   ('signific', 0.007691999),\n",
       "   ('sever', 0.007560507),\n",
       "   ('result', 0.0072342437),\n",
       "   ('acut', 0.0069785556),\n",
       "   ('calf', 0.0064666444),\n",
       "   ('year', 0.00639693),\n",
       "   ('conclus', 0.0062917825),\n",
       "   ('treatment', 0.005994524),\n",
       "   ('common', 0.0053888727),\n",
       "   ('case', 0.005385293)]),\n",
       " (2,\n",
       "  [('influenza', 0.031880673),\n",
       "   ('infect', 0.025836844),\n",
       "   ('virus', 0.023291491),\n",
       "   ('case', 0.019353988),\n",
       "   ('outbreak', 0.0127011705),\n",
       "   ('mers-cov', 0.012422747),\n",
       "   ('transmiss', 0.012350284),\n",
       "   ('diseas', 0.011869485),\n",
       "   ('respiratori', 0.010769872),\n",
       "   ('human', 0.008457823),\n",
       "   ('epidem', 0.008350471),\n",
       "   ('vaccin', 0.008074394),\n",
       "   ('pandem', 0.007000548),\n",
       "   ('h1n1', 0.00656803),\n",
       "   ('caus', 0.006529917),\n",
       "   ('contact', 0.006455609),\n",
       "   ('spread', 0.0064371694),\n",
       "   ('rate', 0.0062637194),\n",
       "   ('coronavirus', 0.0062131975),\n",
       "   ('report', 0.0062009813)]),\n",
       " (3,\n",
       "  [('vaccin', 0.041530434),\n",
       "   ('protein', 0.037913714),\n",
       "   ('antibodi', 0.023627322),\n",
       "   ('immun', 0.016009578),\n",
       "   ('sars-cov', 0.011680685),\n",
       "   ('virus', 0.010427077),\n",
       "   ('antigen', 0.009986016),\n",
       "   ('peptid', 0.009654134),\n",
       "   ('protect', 0.009472112),\n",
       "   ('respons', 0.009250863),\n",
       "   ('neutral', 0.008865088),\n",
       "   ('bind', 0.0088245915),\n",
       "   ('fusion', 0.008635012),\n",
       "   ('recombin', 0.008609768),\n",
       "   ('cell', 0.008098558),\n",
       "   ('develop', 0.008069051),\n",
       "   ('domain', 0.007825358),\n",
       "   ('epitop', 0.007511896),\n",
       "   ('sar', 0.007104304),\n",
       "   ('express', 0.0068325093)]),\n",
       " (4,\n",
       "  [('detect', 0.03986719),\n",
       "   ('virus', 0.031416066),\n",
       "   ('assay', 0.022378182),\n",
       "   ('sampl', 0.022093296),\n",
       "   ('use', 0.022067238),\n",
       "   ('test', 0.020186752),\n",
       "   ('method', 0.015457835),\n",
       "   ('sensit', 0.012525741),\n",
       "   ('result', 0.011306924),\n",
       "   ('pathogen', 0.011162823),\n",
       "   ('posit', 0.010365505),\n",
       "   ('specif', 0.010237087),\n",
       "   ('viral', 0.0091813905),\n",
       "   ('diagnost', 0.008664505),\n",
       "   ('clinic', 0.007903918),\n",
       "   ('develop', 0.0072801854),\n",
       "   ('real-tim', 0.0070777996),\n",
       "   ('rt-pcr', 0.006883347),\n",
       "   ('reaction', 0.006789166),\n",
       "   ('infect', 0.0067124637)]),\n",
       " (5,\n",
       "  [('activ', 0.021063887),\n",
       "   ('effect', 0.019686667),\n",
       "   ('drug', 0.016215045),\n",
       "   ('treatment', 0.013041537),\n",
       "   ('antivir', 0.012491824),\n",
       "   ('compound', 0.009755181),\n",
       "   ('develop', 0.0073152184),\n",
       "   ('agent', 0.0071348897),\n",
       "   ('potenti', 0.0070345527),\n",
       "   ('inhibitor', 0.007025907),\n",
       "   ('use', 0.0069039804),\n",
       "   ('therapeut', 0.0065759183),\n",
       "   ('concentr', 0.0063114646),\n",
       "   ('therapi', 0.0062725116),\n",
       "   ('resist', 0.006147317),\n",
       "   ('target', 0.006127591),\n",
       "   ('inhibit', 0.0058866534),\n",
       "   ('cancer', 0.005753958),\n",
       "   ('product', 0.005384012),\n",
       "   ('studi', 0.0053805877)]),\n",
       " (6,\n",
       "  [('cell', 0.040377468),\n",
       "   ('virus', 0.024000071),\n",
       "   ('infect', 0.020937432),\n",
       "   ('protein', 0.018826133),\n",
       "   ('viral', 0.017808903),\n",
       "   ('express', 0.015658928),\n",
       "   ('replic', 0.01240874),\n",
       "   ('activ', 0.012384279),\n",
       "   ('gene', 0.00976809),\n",
       "   ('host', 0.009707637),\n",
       "   ('respons', 0.008857327),\n",
       "   ('function', 0.007584089),\n",
       "   ('role', 0.0075139087),\n",
       "   ('receptor', 0.007351183),\n",
       "   ('mechan', 0.0070785726),\n",
       "   ('immun', 0.0066829287),\n",
       "   ('signal', 0.006510655),\n",
       "   ('pathway', 0.0062218937),\n",
       "   ('inhibit', 0.0061945347),\n",
       "   ('cellular', 0.0059689754)]),\n",
       " (7,\n",
       "  [('model', 0.015524733),\n",
       "   ('use', 0.010979432),\n",
       "   ('structur', 0.009470307),\n",
       "   ('data', 0.0091685355),\n",
       "   ('studi', 0.008712917),\n",
       "   ('analysi', 0.0075916564),\n",
       "   ('provid', 0.0063696),\n",
       "   ('method', 0.006287391),\n",
       "   ('diseas', 0.006167199),\n",
       "   ('approach', 0.006137775),\n",
       "   ('predict', 0.0058526746),\n",
       "   ('differ', 0.0057759723),\n",
       "   ('develop', 0.0052746986),\n",
       "   ('function', 0.0051407916),\n",
       "   ('sequenc', 0.005126796),\n",
       "   ('biolog', 0.005049438),\n",
       "   ('also', 0.004855113),\n",
       "   ('base', 0.0048175976),\n",
       "   ('process', 0.004631778),\n",
       "   ('identifi', 0.0046274033)]),\n",
       " (8,\n",
       "  [('health', 0.024147201),\n",
       "   ('diseas', 0.016767701),\n",
       "   ('public', 0.011118835),\n",
       "   ('emerg', 0.008565125),\n",
       "   ('develop', 0.00776195),\n",
       "   ('system', 0.0074603437),\n",
       "   ('infecti', 0.0073489933),\n",
       "   ('control', 0.0068114423),\n",
       "   ('risk', 0.006475862),\n",
       "   ('research', 0.0062215715),\n",
       "   ('studi', 0.005246385),\n",
       "   ('countri', 0.00509172),\n",
       "   ('care', 0.0050506056),\n",
       "   ('prevent', 0.0049747336),\n",
       "   ('global', 0.004940657),\n",
       "   ('manag', 0.004848498),\n",
       "   ('outbreak', 0.0048375153),\n",
       "   ('respons', 0.004777985),\n",
       "   ('need', 0.0047036395),\n",
       "   ('provid', 0.004683478)]),\n",
       " (9,\n",
       "  [('infect', 0.026039844),\n",
       "   ('mous', 0.014685364),\n",
       "   ('diseas', 0.014007277),\n",
       "   ('cell', 0.012678494),\n",
       "   ('respons', 0.011093459),\n",
       "   ('immun', 0.010074337),\n",
       "   ('lung', 0.010015441),\n",
       "   ('virus', 0.0071676187),\n",
       "   ('increas', 0.006878616),\n",
       "   ('level', 0.0059505557),\n",
       "   ('tissu', 0.0059184646),\n",
       "   ('studi', 0.005779709),\n",
       "   ('signific', 0.005660177),\n",
       "   ('blood', 0.0055147274),\n",
       "   ('anim', 0.005356044),\n",
       "   ('group', 0.005209725),\n",
       "   ('result', 0.005159749),\n",
       "   ('system', 0.005117491),\n",
       "   ('cytokin', 0.004862093),\n",
       "   ('caus', 0.004798188)])]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lda_results = ldamodel.show_topics(num_topics=10, num_words=20, formatted=False)\n",
    "lda_results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create topic_words, a list of buckets of words that most represent each topic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[['virus',\n",
       "  'sequenc',\n",
       "  'strain',\n",
       "  'genom',\n",
       "  'gene',\n",
       "  'coronavirus',\n",
       "  'speci',\n",
       "  'isol',\n",
       "  'human',\n",
       "  'analysi',\n",
       "  'pedv',\n",
       "  'studi',\n",
       "  'identifi',\n",
       "  'differ',\n",
       "  'viral',\n",
       "  'infect',\n",
       "  'region',\n",
       "  'anim',\n",
       "  'sampl',\n",
       "  'porcin'],\n",
       " ['patient',\n",
       "  'respiratori',\n",
       "  'infect',\n",
       "  'studi',\n",
       "  'child',\n",
       "  'hospit',\n",
       "  'clinic',\n",
       "  'group',\n",
       "  'associ',\n",
       "  'pneumonia',\n",
       "  'signific',\n",
       "  'sever',\n",
       "  'result',\n",
       "  'acut',\n",
       "  'calf',\n",
       "  'year',\n",
       "  'conclus',\n",
       "  'treatment',\n",
       "  'common',\n",
       "  'case'],\n",
       " ['influenza',\n",
       "  'infect',\n",
       "  'virus',\n",
       "  'case',\n",
       "  'outbreak',\n",
       "  'mers-cov',\n",
       "  'transmiss',\n",
       "  'diseas',\n",
       "  'respiratori',\n",
       "  'human',\n",
       "  'epidem',\n",
       "  'vaccin',\n",
       "  'pandem',\n",
       "  'h1n1',\n",
       "  'caus',\n",
       "  'contact',\n",
       "  'spread',\n",
       "  'rate',\n",
       "  'coronavirus',\n",
       "  'report'],\n",
       " ['vaccin',\n",
       "  'protein',\n",
       "  'antibodi',\n",
       "  'immun',\n",
       "  'sars-cov',\n",
       "  'virus',\n",
       "  'antigen',\n",
       "  'peptid',\n",
       "  'protect',\n",
       "  'respons',\n",
       "  'neutral',\n",
       "  'bind',\n",
       "  'fusion',\n",
       "  'recombin',\n",
       "  'cell',\n",
       "  'develop',\n",
       "  'domain',\n",
       "  'epitop',\n",
       "  'sar',\n",
       "  'express'],\n",
       " ['detect',\n",
       "  'virus',\n",
       "  'assay',\n",
       "  'sampl',\n",
       "  'use',\n",
       "  'test',\n",
       "  'method',\n",
       "  'sensit',\n",
       "  'result',\n",
       "  'pathogen',\n",
       "  'posit',\n",
       "  'specif',\n",
       "  'viral',\n",
       "  'diagnost',\n",
       "  'clinic',\n",
       "  'develop',\n",
       "  'real-tim',\n",
       "  'rt-pcr',\n",
       "  'reaction',\n",
       "  'infect'],\n",
       " ['activ',\n",
       "  'effect',\n",
       "  'drug',\n",
       "  'treatment',\n",
       "  'antivir',\n",
       "  'compound',\n",
       "  'develop',\n",
       "  'agent',\n",
       "  'potenti',\n",
       "  'inhibitor',\n",
       "  'use',\n",
       "  'therapeut',\n",
       "  'concentr',\n",
       "  'therapi',\n",
       "  'resist',\n",
       "  'target',\n",
       "  'inhibit',\n",
       "  'cancer',\n",
       "  'product',\n",
       "  'studi'],\n",
       " ['cell',\n",
       "  'virus',\n",
       "  'infect',\n",
       "  'protein',\n",
       "  'viral',\n",
       "  'express',\n",
       "  'replic',\n",
       "  'activ',\n",
       "  'gene',\n",
       "  'host',\n",
       "  'respons',\n",
       "  'function',\n",
       "  'role',\n",
       "  'receptor',\n",
       "  'mechan',\n",
       "  'immun',\n",
       "  'signal',\n",
       "  'pathway',\n",
       "  'inhibit',\n",
       "  'cellular'],\n",
       " ['model',\n",
       "  'use',\n",
       "  'structur',\n",
       "  'data',\n",
       "  'studi',\n",
       "  'analysi',\n",
       "  'provid',\n",
       "  'method',\n",
       "  'diseas',\n",
       "  'approach',\n",
       "  'predict',\n",
       "  'differ',\n",
       "  'develop',\n",
       "  'function',\n",
       "  'sequenc',\n",
       "  'biolog',\n",
       "  'also',\n",
       "  'base',\n",
       "  'process',\n",
       "  'identifi'],\n",
       " ['health',\n",
       "  'diseas',\n",
       "  'public',\n",
       "  'emerg',\n",
       "  'develop',\n",
       "  'system',\n",
       "  'infecti',\n",
       "  'control',\n",
       "  'risk',\n",
       "  'research',\n",
       "  'studi',\n",
       "  'countri',\n",
       "  'care',\n",
       "  'prevent',\n",
       "  'global',\n",
       "  'manag',\n",
       "  'outbreak',\n",
       "  'respons',\n",
       "  'need',\n",
       "  'provid'],\n",
       " ['infect',\n",
       "  'mous',\n",
       "  'diseas',\n",
       "  'cell',\n",
       "  'respons',\n",
       "  'immun',\n",
       "  'lung',\n",
       "  'virus',\n",
       "  'increas',\n",
       "  'level',\n",
       "  'tissu',\n",
       "  'studi',\n",
       "  'signific',\n",
       "  'blood',\n",
       "  'anim',\n",
       "  'group',\n",
       "  'result',\n",
       "  'system',\n",
       "  'cytokin',\n",
       "  'caus']]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "topic_words = []\n",
    "#i is each topic\n",
    "for i in lda_results:\n",
    "    topic_i_words = []\n",
    "    word_pairs = i[1]\n",
    "    #j is the list of word-probability pairs (we don't care about the probabilities here)\n",
    "    for j in word_pairs:\n",
    "        topic_i_words.append(j[0])\n",
    "    topic_words.append(topic_i_words)\n",
    "        \n",
    "topic_words   "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Assign a topic to each of the papers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>sha</th>\n",
       "      <th>abstract</th>\n",
       "      <th>abstract tokens</th>\n",
       "      <th>assigned topic</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>244</th>\n",
       "      <td>328</td>\n",
       "      <td>87d5b80231b8956e498791ab3507f0f1ca529be8</td>\n",
       "      <td>Abstract Human coronaviruses (HCV) are ubiquit...</td>\n",
       "      <td>[human, coronavirus, ubiquit, pathogen, caus, ...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26113</th>\n",
       "      <td>42029</td>\n",
       "      <td>bb5d7caba7ff8afec3c1fde62cadf65db745ce35</td>\n",
       "      <td>A disease with a sudden drop in egg production...</td>\n",
       "      <td>[diseas, sudden, drop, product, shell-less, ca...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15829</th>\n",
       "      <td>41702</td>\n",
       "      <td>c48208dd7beb82687c017d5d1b44bde08de0990a</td>\n",
       "      <td>Abstract The plausible release of deeply engin...</td>\n",
       "      <td>[plausibl, releas, deepli, engin, even, entir,...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23467</th>\n",
       "      <td>24422</td>\n",
       "      <td>e1e50e72368b173717a1b7f886c65c61ce5ade59</td>\n",
       "      <td>Here, we report the genome sequence of a diver...</td>\n",
       "      <td>[report, genom, sequenc, diverg, human, rhinov...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12551</th>\n",
       "      <td>15916</td>\n",
       "      <td>dfe64ba8bf59ab09b956b293c47a3798df11b31a</td>\n",
       "      <td>Abstract Background Despite recent discovery o...</td>\n",
       "      <td>[background, despit, recent, discoveri, novel,...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Unnamed: 0                                       sha  \\\n",
       "244           328  87d5b80231b8956e498791ab3507f0f1ca529be8   \n",
       "26113       42029  bb5d7caba7ff8afec3c1fde62cadf65db745ce35   \n",
       "15829       41702  c48208dd7beb82687c017d5d1b44bde08de0990a   \n",
       "23467       24422  e1e50e72368b173717a1b7f886c65c61ce5ade59   \n",
       "12551       15916  dfe64ba8bf59ab09b956b293c47a3798df11b31a   \n",
       "\n",
       "                                                abstract  \\\n",
       "244    Abstract Human coronaviruses (HCV) are ubiquit...   \n",
       "26113  A disease with a sudden drop in egg production...   \n",
       "15829  Abstract The plausible release of deeply engin...   \n",
       "23467  Here, we report the genome sequence of a diver...   \n",
       "12551  Abstract Background Despite recent discovery o...   \n",
       "\n",
       "                                         abstract tokens  assigned topic  \n",
       "244    [human, coronavirus, ubiquit, pathogen, caus, ...               0  \n",
       "26113  [diseas, sudden, drop, product, shell-less, ca...               0  \n",
       "15829  [plausibl, releas, deepli, engin, even, entir,...               0  \n",
       "23467  [report, genom, sequenc, diverg, human, rhinov...               0  \n",
       "12551  [background, despit, recent, discoveri, novel,...               0  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "assigned_topic = []\n",
    "for index, row in partial_df.iterrows():\n",
    "    tokens = row[\"abstract tokens\"]\n",
    "    counter_array = [0] * 15\n",
    "    for i in tokens:\n",
    "        for j in np.arange(10):\n",
    "            if i in topic_words[j]:\n",
    "                counter_array[j] += 1\n",
    "    max_topic = counter_array.index(max(counter_array))\n",
    "    assigned_topic.append(max_topic)\n",
    "                \n",
    "partial_df[\"assigned topic\"] = assigned_topic\n",
    "partial_df = partial_df.sort_values(\"assigned topic\")\n",
    "partial_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "cluster3 = partial_df[partial_df[\"assigned topic\"] == 9]\n",
    "abstractStrings = cluster3.sample(100)[\"abstract\"].values\n",
    "wordCloudString = \"\"\n",
    "for i in np.arange(100):\n",
    "    wordCloudString += abstractStrings[i]\n",
    "    \n",
    "from PIL import Image\n",
    "from wordcloud import WordCloud, STOPWORDS, ImageColorGenerator\n",
    "    \n",
    "# lower max_font_size, change the maximum number of word and lighten the background:\n",
    "wordcloud = WordCloud(max_font_size=50, max_words=20, background_color=\"white\").generate(wordCloudString)\n",
    "plt.figure()\n",
    "plt.imshow(wordcloud, interpolation=\"bilinear\")\n",
    "plt.axis(\"off\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "partial_df.to_csv(\"document_clusters.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}