DIP Presentation Modeling DEMO.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Topic Modeling Demo: Latent Dirichlet Analysis w/ Gensim"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Install Libraries; A lot of these are for pre-processing, but some of them are for the actual model. We will be using gensim to create the actual model.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package wordnet to\n",
      "[nltk_data]     C:\\Users\\iyerr\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Package wordnet is already up-to-date!\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd \n",
    "from textblob import TextBlob\n",
    "import re\n",
    "import string\n",
    "import nltk\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from nltk import word_tokenize, pos_tag\n",
    "from nltk.stem.snowball import SnowballStemmer\n",
    "from nltk.corpus import stopwords\n",
    "import gensim\n",
    "from gensim.utils import simple_preprocess\n",
    "from gensim.parsing.preprocessing import STOPWORDS\n",
    "from nltk.stem import WordNetLemmatizer, SnowballStemmer\n",
    "from nltk.stem.porter import *\n",
    "import numpy as np\n",
    "np.random.seed(2018)\n",
    "from gensim import corpora, models\n",
    "from gensim import matutils, models\n",
    "import scipy.sparse\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.decomposition import LatentDirichletAllocation\n",
    "from sklearn.feature_extraction import text\n",
    "import numpy as np\n",
    "from helper import *\n",
    "import pickle\n",
    "nltk.download('wordnet');"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Read in the data "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Anaconda3\\lib\\site-packages\\ipykernel_launcher.py:7: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  import sys\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>y/m/d</th>\n",
       "      <th>president</th>\n",
       "      <th>speech</th>\n",
       "      <th>word count</th>\n",
       "      <th>type</th>\n",
       "      <th>party</th>\n",
       "      <th>home state</th>\n",
       "      <th>year</th>\n",
       "      <th>polarity</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1790-01-08</td>\n",
       "      <td>George Washington</td>\n",
       "      <td>Fellow-Citizens of the Senate and House of Rep...</td>\n",
       "      <td>1008</td>\n",
       "      <td>spoken</td>\n",
       "      <td>Independent</td>\n",
       "      <td>Virginia</td>\n",
       "      <td>1790</td>\n",
       "      <td>0.235029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1791-10-25</td>\n",
       "      <td>George Washington</td>\n",
       "      <td>Fellow-Citizens of the Senate and House of Rep...</td>\n",
       "      <td>2148</td>\n",
       "      <td>spoken</td>\n",
       "      <td>Independent</td>\n",
       "      <td>Virginia</td>\n",
       "      <td>1791</td>\n",
       "      <td>0.131196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1792-11-06</td>\n",
       "      <td>George Washington</td>\n",
       "      <td>Fellow-Citizens of the Senate and House of Rep...</td>\n",
       "      <td>1959</td>\n",
       "      <td>spoken</td>\n",
       "      <td>Independent</td>\n",
       "      <td>Virginia</td>\n",
       "      <td>1792</td>\n",
       "      <td>0.078748</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1793-12-03</td>\n",
       "      <td>George Washington</td>\n",
       "      <td>Fellow-Citizens of the Senate and House of Rep...</td>\n",
       "      <td>1830</td>\n",
       "      <td>spoken</td>\n",
       "      <td>Independent</td>\n",
       "      <td>Virginia</td>\n",
       "      <td>1793</td>\n",
       "      <td>0.093337</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1794-11-19</td>\n",
       "      <td>George Washington</td>\n",
       "      <td>Fellow-Citizens of the Senate and House of Rep...</td>\n",
       "      <td>2706</td>\n",
       "      <td>spoken</td>\n",
       "      <td>Independent</td>\n",
       "      <td>Virginia</td>\n",
       "      <td>1794</td>\n",
       "      <td>0.115751</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       y/m/d          president  \\\n",
       "0 1790-01-08  George Washington   \n",
       "1 1791-10-25  George Washington   \n",
       "2 1792-11-06  George Washington   \n",
       "3 1793-12-03  George Washington   \n",
       "4 1794-11-19  George Washington   \n",
       "\n",
       "                                              speech  word count    type  \\\n",
       "0  Fellow-Citizens of the Senate and House of Rep...        1008  spoken   \n",
       "1  Fellow-Citizens of the Senate and House of Rep...        2148  spoken   \n",
       "2  Fellow-Citizens of the Senate and House of Rep...        1959  spoken   \n",
       "3  Fellow-Citizens of the Senate and House of Rep...        1830  spoken   \n",
       "4  Fellow-Citizens of the Senate and House of Rep...        2706  spoken   \n",
       "\n",
       "         party home state  year  polarity  \n",
       "0  Independent   Virginia  1790  0.235029  \n",
       "1  Independent   Virginia  1791  0.131196  \n",
       "2  Independent   Virginia  1792  0.078748  \n",
       "3  Independent   Virginia  1793  0.093337  \n",
       "4  Independent   Virginia  1794  0.115751  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sotu = pd.read_csv('stateunion.tsv', sep='\\t',  dtype={'texts':str, 'date':str, 'name':str}, index_col=0)\n",
    "sotu = sotu.T # We want each speech to be a row instead of a column \n",
    "sotu = sotu.reset_index().drop(columns = ['index']) # get rid of row index\n",
    "sotu = sotu.rename(columns = {'date':'y/m/d', 'name':'president', 'texts':'speech'}) # rename column for clarity purposes\n",
    "sotu['y/m/d'] = pd.to_datetime(sotu['y/m/d']) # re-format structure of time data\n",
    "for i in range(len(sotu['speech'])): # get rid of HTML tags and newlines \n",
    "    sotu['speech'][i] = re.sub('(?<! )(?<=[.,!:?();])(?! )', r' ', re.sub('<[^<]+?>', '', sotu['speech'][i]).replace('\\n', ''))\n",
    "sotu['word count'] = [len(speech.split(' ')) for speech in sotu['speech'].values] #count number of words per speech \n",
    "sotu['type'] = ['spoken' if count <= 12000 else 'written' for count in sotu['word count']] #classify speech type \n",
    "presidents = pd.read_csv('revised_presidents.csv') # Data File containing information about every president in SOTU dataset\n",
    "presidents = presidents.drop(columns = [\"Presidency \", \"Wikipedia Entry\",'Took office ', \"Left office \", \"Portrait\", 'Thumbnail'])\n",
    "# line above drops unnecessary columns \n",
    "sotu = sotu.join(presidents.set_index('President '), on='president') # merge our datasets\n",
    "sotu.iloc[184] = sotu.iloc[184].fillna({'Party ': 'Republican', 'Home State':'Michigan'}) #fix nulls after merge \n",
    "sotu.iloc[185] = sotu.iloc[185].fillna({'Party ': 'Republican', 'Home State':'Michigan'}) \n",
    "sotu.iloc[186] = sotu.iloc[186].fillna({'Party ': 'Republican', 'Home State':'Michigan'})\n",
    "sotu.iloc[218] = sotu.iloc[218].fillna({'Party ': 'Democratic', 'Home State':'Illinois'})\n",
    "sotu.iloc[219] = sotu.iloc[219].fillna({'Party ': 'Democratic', 'Home State':'Illinois'})\n",
    "sotu.iloc[220] = sotu.iloc[220].fillna({'Party ': 'Democratic', 'Home State':'Illinois'})\n",
    "sotu.iloc[221] = sotu.iloc[221].fillna({'Party ': 'Democratic', 'Home State':'Illinois'})\n",
    "sotu.iloc[222] = sotu.iloc[222].fillna({'Party ': 'Democratic', 'Home State':'Illinois'})\n",
    "sotu.iloc[223] = sotu.iloc[223].fillna({'Party ': 'Democratic', 'Home State':'Illinois'})\n",
    "sotu.iloc[224] = sotu.iloc[224].fillna({'Party ': 'Democratic', 'Home State':'Illinois'})\n",
    "sotu.iloc[225] = sotu.iloc[225].fillna({'Party ': 'Democratic', 'Home State':'Illinois'})\n",
    "sotu['year'] = sotu['y/m/d'].dt.year # for ease of formatting\n",
    "sotu = sotu.rename(columns = {'Party ': 'party', 'Home State':'home state'}) #fix formatting issues\n",
    "sotu['party'] = sotu['party'].str.strip(' ') #fix formatting issues\n",
    "\n",
    "def get_polarity(text): # add polarity score of speech to dataframe\n",
    "    \"\"\"Return the polarity of TEXT\"\"\"   \n",
    "    blob = TextBlob(text)\n",
    "    return blob.sentiment.polarity\n",
    "\n",
    "polarities = sotu['speech'].apply(get_polarity)\n",
    "sotu['polarity'] = polarities #update dataframe with polarity scores \n",
    "sotu.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Text Pre-Processing; this is very time intensive and also depends a lot on the dataset that you are working with. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>speech</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1790</th>\n",
       "      <td>fellow citizen senat hous repres embrac great satisfact opportun present itselfof congratul present favor prospect publicaffair recent access impo...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1791</th>\n",
       "      <td>fellow citizen senat hous repres vain expect peac indian frontier long alawless unprincipl wretch violat right hospit infring solemn treati receiv...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1792</th>\n",
       "      <td>fellow citizen senat hous repres abat satisfact meet thepres occas felicit continu general abl inform theindian hostil time past distress ournorth...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1793</th>\n",
       "      <td>fellow citizen senat hous repres commenc term call intooffic occas aris express fellow citizen atlarg deep respect sens feel renew testimonyof pub...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1794</th>\n",
       "      <td>fellow citizen senat hous repres mind gracious indulg heaven theamerican peopl nation survey general prosper ofour countri look forward rich power...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                                                                                                     speech\n",
       "1790  fellow citizen senat hous repres embrac great satisfact opportun present itselfof congratul present favor prospect publicaffair recent access impo...\n",
       "1791  fellow citizen senat hous repres vain expect peac indian frontier long alawless unprincipl wretch violat right hospit infring solemn treati receiv...\n",
       "1792  fellow citizen senat hous repres abat satisfact meet thepres occas felicit continu general abl inform theindian hostil time past distress ournorth...\n",
       "1793  fellow citizen senat hous repres commenc term call intooffic occas aris express fellow citizen atlarg deep respect sens feel renew testimonyof pub...\n",
       "1794  fellow citizen senat hous repres mind gracious indulg heaven theamerican peopl nation survey general prosper ofour countri look forward rich power..."
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"\"\"!mkdir transcripts\n",
    "\n",
    "for i, c in enumerate(year):\n",
    "    with open(\"transcripts/\" + c + \".txt\", \"wb\") as file:\n",
    "        pickle.dump(documents[i], file)\n",
    "\"\"\"\n",
    "documents = sotu['speech'].values\n",
    "year = sotu['y/m/d'].dt.year.astype(str)\n",
    "stemmer = SnowballStemmer('english')\n",
    "data = {} # data dictionary\n",
    "for i, c in enumerate(year):\n",
    "    with open(\"transcripts/\" + c + \".txt\", \"rb\") as file:\n",
    "        data[c] = pickle.load(file)\n",
    "        \n",
    "        \n",
    "def combine_text(list_of_text): \n",
    "    '''Takes a list of text and combines them into one large chunk of text.'''\n",
    "    combined_text = ''.join(list_of_text)\n",
    "    return combined_text\n",
    "\n",
    "def combine_text_new(list_of_text): \n",
    "    '''Takes a list of text and combines them into one large chunk of text.'''\n",
    "    combined_text = ' '.join(list_of_text)\n",
    "    return combined_text\n",
    "\n",
    "data_combined = {key: [combine_text(value)] for (key, value) in data.items()}\n",
    "pd.set_option('max_colwidth',150)\n",
    "\n",
    "frame = pd.DataFrame.from_dict(data_combined).transpose()\n",
    "frame.columns = ['speech']\n",
    "frame = frame.sort_index()\n",
    "\n",
    "def basic_processing(text):\n",
    "    '''Make text lowercase, remove text in square brackets, remove punctuation and remove words containing numbers.'''\n",
    "    text = text.lower()\n",
    "    text = re.sub('\\[.*?\\]', '', text)\n",
    "    text = re.sub('[%s]' % re.escape(string.punctuation), ' ', text)\n",
    "    text = re.sub('\\w*\\d\\w*', '', text)\n",
    "    text = re.sub('[‘’“”…]', '', text)\n",
    "    text = re.sub('\\n', '', text)\n",
    "    return text\n",
    "\n",
    "basic = lambda x: basic_processing(x)\n",
    "\n",
    "basic_frame = pd.DataFrame(frame.speech.apply(basic))\n",
    "\n",
    "def lemmatize_stemming(text):\n",
    "    return stemmer.stem(WordNetLemmatizer().lemmatize(text, pos='v'))\n",
    "\n",
    "def preprocess(text):\n",
    "    result = []\n",
    "    for token in gensim.utils.simple_preprocess(text):\n",
    "        if token not in gensim.parsing.preprocessing.STOPWORDS and len(token) > 3:\n",
    "            result.append(lemmatize_stemming(token))\n",
    "    return result\n",
    "\n",
    "f1 = lambda x: lemmatize_stemming(x)\n",
    "preprocessor = lambda x: preprocess(x)\n",
    "\n",
    "inter = pd.DataFrame(basic_frame.speech.apply(f1))\n",
    "inter_1 = pd.DataFrame(inter.speech.apply(preprocess))\n",
    "final = pd.DataFrame(inter_1.speech.apply(combine_text_new)) \n",
    "\n",
    "final.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Convert dataframe to document-term matrix for model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "cv = CountVectorizer(stop_words='english')\n",
    "data_cv = cv.fit_transform(final.speech)\n",
    "data_dtm = pd.DataFrame(data_cv.toarray(), columns=cv.get_feature_names())\n",
    "data_dtm.index = final.index\n",
    "#data_dtm.to_pickle(\"dtm.pkl\") #save to local directory\n",
    "\n",
    "#final.to_pickle('final.pkl') #save to local directory\n",
    "#pickle.dump(cv, open(\"cv.pkl\", \"wb\")) # read into memory"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Now, we can start creating models. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model #1: Include All Words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(0,\n",
       "  '0.009*\"year\" + 0.009*\"nation\" + 0.007*\"peopl\" + 0.006*\"american\" + 0.006*\"world\" + 0.006*\"work\" + 0.005*\"govern\" + 0.005*\"congress\" + 0.005*\"america\" + 0.005*\"need\"'),\n",
       " (1,\n",
       "  '0.013*\"state\" + 0.010*\"govern\" + 0.007*\"unit\" + 0.006*\"year\" + 0.006*\"congress\" + 0.005*\"countri\" + 0.005*\"nation\" + 0.004*\"public\" + 0.004*\"great\" + 0.004*\"power\"')]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_pickle('dtm.pkl') #read in our document-term matrix\n",
    "tdm = data.transpose() # change document-term matrix to term-document matrix for model\n",
    "sparse_counts = scipy.sparse.csr_matrix(tdm) # construct sparse csr matrix\n",
    "corpus = matutils.Sparse2Corpus(sparse_counts) # create gensim corpus for model\n",
    "cv = pickle.load(open(\"cv.pkl\", \"rb\"))\n",
    "id2word = dict((v, k) for k, v in cv.vocabulary_.items())\n",
    "lda = models.LdaModel(corpus=corpus, id2word=id2word, num_topics=2, passes=20)\n",
    "lda.print_topics()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model #2: Include only Nouns "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(0,\n",
       "  '0.021*\"state\" + 0.011*\"unit\" + 0.009*\"year\" + 0.008*\"congress\" + 0.007*\"countri\" + 0.007*\"nation\" + 0.007*\"power\" + 0.005*\"import\" + 0.004*\"duti\" + 0.004*\"peopl\"'),\n",
       " (1,\n",
       "  '0.015*\"year\" + 0.015*\"nation\" + 0.009*\"world\" + 0.009*\"work\" + 0.008*\"peopl\" + 0.007*\"state\" + 0.007*\"congress\" + 0.005*\"countri\" + 0.005*\"program\" + 0.004*\"unit\"')]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def nouns(text):\n",
    "    '''Given a string of text, tokenize the text and pull out only the nouns.'''\n",
    "    is_noun = lambda pos: pos[:2] == 'NN'\n",
    "    tokenized = word_tokenize(text)\n",
    "    all_nouns = [word for (word, pos) in pos_tag(tokenized) if is_noun(pos)] \n",
    "    return ' '.join(all_nouns)\n",
    "\n",
    "data_clean = pd.read_pickle('final.pkl')\n",
    "data_nouns = pd.DataFrame(data_clean.speech.apply(nouns))\n",
    "\n",
    "# Re-add the additional stop words since we are recreating the document-term matrix\n",
    "add_stop_words = ['like', 'im', 'know', 'just', 'dont', 'thats', 'right', 'people',\n",
    "                  'youre', 'got', 'gonna', 'time', 'think', 'yeah', 'said']\n",
    "stop_words = text.ENGLISH_STOP_WORDS.union(add_stop_words)\n",
    "\n",
    "# Recreate a document-term matrix with only nouns\n",
    "cvn = CountVectorizer(stop_words=stop_words)\n",
    "data_cvn = cvn.fit_transform(data_nouns.speech)\n",
    "data_dtmn = pd.DataFrame(data_cvn.toarray(), columns=cvn.get_feature_names())\n",
    "data_dtmn.index = data_nouns.index\n",
    "#data_dtmn.head()\n",
    "\n",
    "# Create the gensim corpus\n",
    "corpusn = matutils.Sparse2Corpus(scipy.sparse.csr_matrix(data_dtmn.transpose()))\n",
    "\n",
    "# Create the vocabulary dictionary\n",
    "id2wordn = dict((v, k) for k, v in cvn.vocabulary_.items())\n",
    "\n",
    "ldnoun = models.LdaModel(corpus=corpusn, num_topics=2, id2word=id2wordn, passes=20)\n",
    "ldnoun.print_topics()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model #3: Nouns and Adjectives"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Topic: 0 \n",
      "Words: 0.005*\"america\" + 0.005*\"program\" + 0.004*\"help\" + 0.004*\"feder\" + 0.003*\"econom\" + 0.003*\"busi\" + 0.003*\"problem\" + 0.003*\"budget\" + 0.003*\"plan\" + 0.003*\"live\"\n",
      "\n",
      "\n",
      "Topic: 1 \n",
      "Words: 0.004*\"subject\" + 0.004*\"general\" + 0.003*\"treati\" + 0.003*\"depart\" + 0.003*\"condit\" + 0.003*\"report\" + 0.003*\"constitut\" + 0.003*\"question\" + 0.003*\"claim\" + 0.002*\"attent\"\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "def nouns_adj(text):\n",
    "    '''Given a string of text, tokenize the text and pull out only the nouns and adjectives.'''\n",
    "    is_noun_adj = lambda pos: pos[:2] == 'NN' or pos[:2] == 'JJ'\n",
    "    tokenized = word_tokenize(text)\n",
    "    nouns_adj = [word for (word, pos) in pos_tag(tokenized) if is_noun_adj(pos)] \n",
    "    return ' '.join(nouns_adj)\n",
    "\n",
    "data_nouns_adj = pd.DataFrame(data_clean.speech.apply(nouns_adj))\n",
    "cvna = CountVectorizer(stop_words=stop_words, max_df=.8)\n",
    "data_cvna = cvna.fit_transform(data_nouns_adj.speech)\n",
    "data_dtmna = pd.DataFrame(data_cvna.toarray(), columns=cvna.get_feature_names())\n",
    "data_dtmna.index = data_nouns_adj.index\n",
    "data_dtmna.head()\n",
    "\n",
    "# Create the gensim corpus\n",
    "corpusna = matutils.Sparse2Corpus(scipy.sparse.csr_matrix(data_dtmna.transpose()))\n",
    "\n",
    "# Create the vocabulary dictionary\n",
    "id2wordna = dict((v, k) for k, v in cvna.vocabulary_.items())\n",
    "\n",
    "lda_noun_adj = models.LdaModel(corpus=corpusna, num_topics=2, id2word=id2wordna, passes=10)\n",
    "for idx, topic in lda_noun_adj.print_topics(-1):\n",
    "    print(\"Topic: {} \\nWords: {}\".format(idx, topic ))\n",
    "    print(\"\\n\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Looking at the sets of words each model outputs, we get the most distinct sets in the third model. It looks like the noun and adjective LDA model is the most useful for us. We can try to interpret the topic that these sets of words pertain to. Looking at the words, we can see that the topics are roughly: \n",
    "\n",
    "Topic 0: Domestic Wellness\n",
    "Topic 1: Legal Issues"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Now, let's try to look at the most prevalent topic in each SOTU speech to get an idea of how topics have changed over time. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(1, '1790'),\n",
       " (1, '1791'),\n",
       " (1, '1792'),\n",
       " (1, '1793'),\n",
       " (1, '1794'),\n",
       " (1, '1795'),\n",
       " (1, '1796'),\n",
       " (1, '1797'),\n",
       " (1, '1798'),\n",
       " (1, '1799'),\n",
       " (1, '1800'),\n",
       " (1, '1801'),\n",
       " (1, '1802'),\n",
       " (1, '1803'),\n",
       " (1, '1804'),\n",
       " (1, '1805'),\n",
       " (1, '1806'),\n",
       " (1, '1807'),\n",
       " (1, '1808'),\n",
       " (1, '1809'),\n",
       " (1, '1810'),\n",
       " (1, '1811'),\n",
       " (1, '1812'),\n",
       " (1, '1813'),\n",
       " (1, '1814'),\n",
       " (1, '1815'),\n",
       " (1, '1816'),\n",
       " (1, '1817'),\n",
       " (1, '1818'),\n",
       " (1, '1819'),\n",
       " (1, '1820'),\n",
       " (1, '1821'),\n",
       " (1, '1822'),\n",
       " (1, '1823'),\n",
       " (1, '1824'),\n",
       " (1, '1825'),\n",
       " (1, '1826'),\n",
       " (1, '1827'),\n",
       " (1, '1828'),\n",
       " (1, '1829'),\n",
       " (1, '1830'),\n",
       " (1, '1831'),\n",
       " (1, '1832'),\n",
       " (1, '1833'),\n",
       " (1, '1834'),\n",
       " (1, '1835'),\n",
       " (1, '1836'),\n",
       " (1, '1837'),\n",
       " (1, '1838'),\n",
       " (1, '1839'),\n",
       " (1, '1840'),\n",
       " (1, '1841'),\n",
       " (1, '1842'),\n",
       " (1, '1843'),\n",
       " (1, '1844'),\n",
       " (1, '1845'),\n",
       " (1, '1846'),\n",
       " (1, '1847'),\n",
       " (1, '1848'),\n",
       " (1, '1849'),\n",
       " (1, '1850'),\n",
       " (1, '1851'),\n",
       " (1, '1852'),\n",
       " (1, '1853'),\n",
       " (1, '1854'),\n",
       " (1, '1855'),\n",
       " (1, '1856'),\n",
       " (1, '1857'),\n",
       " (1, '1858'),\n",
       " (1, '1859'),\n",
       " (1, '1860'),\n",
       " (1, '1861'),\n",
       " (1, '1862'),\n",
       " (1, '1863'),\n",
       " (1, '1864'),\n",
       " (1, '1865'),\n",
       " (1, '1866'),\n",
       " (1, '1867'),\n",
       " (1, '1868'),\n",
       " (1, '1869'),\n",
       " (1, '1870'),\n",
       " (1, '1871'),\n",
       " (1, '1872'),\n",
       " (1, '1873'),\n",
       " (1, '1874'),\n",
       " (1, '1875'),\n",
       " (1, '1876'),\n",
       " (1, '1877'),\n",
       " (1, '1878'),\n",
       " (1, '1879'),\n",
       " (1, '1880'),\n",
       " (1, '1881'),\n",
       " (1, '1882'),\n",
       " (1, '1883'),\n",
       " (1, '1884'),\n",
       " (1, '1885'),\n",
       " (1, '1886'),\n",
       " (1, '1887'),\n",
       " (1, '1888'),\n",
       " (1, '1889'),\n",
       " (1, '1890'),\n",
       " (1, '1891'),\n",
       " (1, '1892'),\n",
       " (1, '1893'),\n",
       " (1, '1894'),\n",
       " (1, '1895'),\n",
       " (1, '1896'),\n",
       " (1, '1897'),\n",
       " (1, '1898'),\n",
       " (1, '1899'),\n",
       " (1, '1900'),\n",
       " (1, '1901'),\n",
       " (1, '1902'),\n",
       " (1, '1903'),\n",
       " (1, '1904'),\n",
       " (1, '1905'),\n",
       " (1, '1906'),\n",
       " (1, '1907'),\n",
       " (1, '1908'),\n",
       " (1, '1909'),\n",
       " (1, '1910'),\n",
       " (1, '1911'),\n",
       " (1, '1912'),\n",
       " (0, '1913'),\n",
       " (1, '1914'),\n",
       " (1, '1915'),\n",
       " (1, '1916'),\n",
       " (0, '1917'),\n",
       " (1, '1918'),\n",
       " (0, '1919'),\n",
       " (0, '1920'),\n",
       " (0, '1921'),\n",
       " (0, '1922'),\n",
       " (0, '1923'),\n",
       " (0, '1924'),\n",
       " (0, '1925'),\n",
       " (0, '1926'),\n",
       " (0, '1927'),\n",
       " (0, '1928'),\n",
       " (0, '1929'),\n",
       " (0, '1930'),\n",
       " (0, '1931'),\n",
       " (0, '1932'),\n",
       " (0, '1934'),\n",
       " (0, '1935'),\n",
       " (0, '1936'),\n",
       " (0, '1937'),\n",
       " (0, '1938'),\n",
       " (0, '1939'),\n",
       " (0, '1940'),\n",
       " (0, '1941'),\n",
       " (0, '1942'),\n",
       " (0, '1943'),\n",
       " (0, '1944'),\n",
       " (0, '1945'),\n",
       " (0, '1946'),\n",
       " (0, '1947'),\n",
       " (0, '1948'),\n",
       " (0, '1949'),\n",
       " (0, '1950'),\n",
       " (0, '1951'),\n",
       " (0, '1952'),\n",
       " (0, '1953'),\n",
       " (0, '1954'),\n",
       " (0, '1955'),\n",
       " (0, '1956'),\n",
       " (0, '1957'),\n",
       " (0, '1958'),\n",
       " (0, '1959'),\n",
       " (0, '1960'),\n",
       " (0, '1961'),\n",
       " (0, '1962'),\n",
       " (0, '1963'),\n",
       " (0, '1964'),\n",
       " (0, '1965'),\n",
       " (0, '1966'),\n",
       " (0, '1967'),\n",
       " (0, '1968'),\n",
       " (0, '1969'),\n",
       " (0, '1970'),\n",
       " (0, '1971'),\n",
       " (0, '1972'),\n",
       " (0, '1973'),\n",
       " (0, '1974'),\n",
       " (0, '1975'),\n",
       " (0, '1976'),\n",
       " (0, '1977'),\n",
       " (0, '1978'),\n",
       " (0, '1979'),\n",
       " (0, '1980'),\n",
       " (0, '1981'),\n",
       " (0, '1982'),\n",
       " (0, '1983'),\n",
       " (0, '1984'),\n",
       " (0, '1985'),\n",
       " (0, '1986'),\n",
       " (0, '1987'),\n",
       " (0, '1988'),\n",
       " (0, '1989'),\n",
       " (0, '1990'),\n",
       " (0, '1991'),\n",
       " (0, '1992'),\n",
       " (0, '1993'),\n",
       " (0, '1994'),\n",
       " (0, '1995'),\n",
       " (0, '1996'),\n",
       " (0, '1997'),\n",
       " (0, '1998'),\n",
       " (0, '1999'),\n",
       " (0, '2000'),\n",
       " (0, '2001'),\n",
       " (0, '2002'),\n",
       " (0, '2003'),\n",
       " (0, '2004'),\n",
       " (0, '2005'),\n",
       " (0, '2006'),\n",
       " (0, '2007'),\n",
       " (0, '2008'),\n",
       " (0, '2009'),\n",
       " (0, '2010'),\n",
       " (0, '2011'),\n",
       " (0, '2012'),\n",
       " (0, '2013'),\n",
       " (0, '2014'),\n",
       " (0, '2015'),\n",
       " (0, '2016'),\n",
       " (0, '2017'),\n",
       " (0, '2018')]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corpus_transformed = lda_noun_adj[corpusna]\n",
    "a = []\n",
    "for topic in corpus_transformed:\n",
    "    if len(topic) > 1:\n",
    "        val1 = topic[0][1]\n",
    "        val2 = topic[1][1]\n",
    "        if val1 > val2:\n",
    "            a.append(topic[0][0])\n",
    "        else:\n",
    "            a.append(topic[1][0])\n",
    "    else:\n",
    "        a.append(topic[0][0])\n",
    "topic_trend = list(zip(a, data_dtmna.index))\n",
    "topic_trend"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "topics = []\n",
    "years = []\n",
    "for item in topic_trend:\n",
    "    topics.append(item[0])\n",
    "    years.append(item[1])\n",
    "years = [int(x) for x in years]\n",
    "frame = pd.DataFrame({'Topics':topics, 'Years':years})\n",
    "plt.figure(figsize = (20, 10))\n",
    "plt.xticks(np.arange(1790, 2020, 20))\n",
    "plt.plot(years, topics);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Let's Focus in on the spike in the middle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "new_frame = frame[(frame['Years'] >= 1900) & (frame['Years'] <= 1935)]\n",
    "plt.figure(figsize = (20, 10))\n",
    "plt.plot(new_frame['Years'], new_frame['Topics']);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Explanations?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Possibly correlated with switch in speech formats (written to spoken)\n",
    "### Could be in related to a change in intent (ie. speaking about legislation vs. vision of the country)"
   ]
  }
 ],
 "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}