Topicmodeling-wealthmanagement .ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import the libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package stopwords to\n",
      "[nltk_data]     C:\\Users\\aditya.baser\\AppData\\Roaming\\nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from gensim.corpora import Dictionary\n",
    "from gensim.models.ldamodel import LdaModel\n",
    "from gensim.models import CoherenceModel\n",
    "from nltk.corpus import stopwords\n",
    "from nltk.tokenize import RegexpTokenizer\n",
    "from datetime import datetime\n",
    "import nltk\n",
    "nltk.download('stopwords')\n",
    "import pandas as pd\n",
    "import re\n",
    "import math"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Importing the dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>date</th>\n",
       "      <th>username</th>\n",
       "      <th>to</th>\n",
       "      <th>replies</th>\n",
       "      <th>retweets</th>\n",
       "      <th>favorites</th>\n",
       "      <th>text</th>\n",
       "      <th>geo</th>\n",
       "      <th>mentions</th>\n",
       "      <th>hashtags</th>\n",
       "      <th>id</th>\n",
       "      <th>permalink</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12/31/2019 23:45</td>\n",
       "      <td>HT_Great_Lakes</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Thinking about transitioning your business? Yo...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#wealthmanagement #hightowergreatlakes #busine...</td>\n",
       "      <td>1.212160e+18</td>\n",
       "      <td>https://twitter.com/HT_Great_Lakes/status/1212...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12/31/2019 23:31</td>\n",
       "      <td>ripsawwealth</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>http://RipsawWealth.com #2020WillBeTheYearFor ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#2020WillBeTheYearFor #HappyNewYear #happynewy...</td>\n",
       "      <td>1.212150e+18</td>\n",
       "      <td>https://twitter.com/ripsawwealth/status/121215...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12/31/2019 23:00</td>\n",
       "      <td>TedHunter1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>#wealthmanagement #maxous</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#wealthmanagement #maxous</td>\n",
       "      <td>1.212150e+18</td>\n",
       "      <td>https://twitter.com/TedHunter1/status/12121465...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12/31/2019 22:45</td>\n",
       "      <td>RAlexJimenez</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>#WealthManagement and private banking trends: ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>@FinTechMagazine</td>\n",
       "      <td>#WealthManagement #wealthtech #fintech #finserv</td>\n",
       "      <td>1.212140e+18</td>\n",
       "      <td>https://twitter.com/RAlexJimenez/status/121214...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12/31/2019 22:43</td>\n",
       "      <td>findadvisor</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Is Your Money Ready for Retirement? #Retiremen...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#RetirementPlan #FinancialPlan #WealthManageme...</td>\n",
       "      <td>1.212140e+18</td>\n",
       "      <td>https://twitter.com/findadvisor/status/1212142...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               date        username   to  replies  retweets  favorites  \\\n",
       "0  12/31/2019 23:45  HT_Great_Lakes  NaN        0         1          1   \n",
       "1  12/31/2019 23:31    ripsawwealth  NaN        0         2          3   \n",
       "2  12/31/2019 23:00      TedHunter1  NaN        0         0          0   \n",
       "3  12/31/2019 22:45    RAlexJimenez  NaN        0         0          0   \n",
       "4  12/31/2019 22:43     findadvisor  NaN        0         0          0   \n",
       "\n",
       "                                                text  geo          mentions  \\\n",
       "0  Thinking about transitioning your business? Yo...  NaN               NaN   \n",
       "1  http://RipsawWealth.com #2020WillBeTheYearFor ...  NaN               NaN   \n",
       "2                          #wealthmanagement #maxous  NaN               NaN   \n",
       "3  #WealthManagement and private banking trends: ...  NaN  @FinTechMagazine   \n",
       "4  Is Your Money Ready for Retirement? #Retiremen...  NaN               NaN   \n",
       "\n",
       "                                            hashtags            id  \\\n",
       "0  #wealthmanagement #hightowergreatlakes #busine...  1.212160e+18   \n",
       "1  #2020WillBeTheYearFor #HappyNewYear #happynewy...  1.212150e+18   \n",
       "2                          #wealthmanagement #maxous  1.212150e+18   \n",
       "3    #WealthManagement #wealthtech #fintech #finserv  1.212140e+18   \n",
       "4  #RetirementPlan #FinancialPlan #WealthManageme...  1.212140e+18   \n",
       "\n",
       "                                           permalink  \n",
       "0  https://twitter.com/HT_Great_Lakes/status/1212...  \n",
       "1  https://twitter.com/ripsawwealth/status/121215...  \n",
       "2  https://twitter.com/TedHunter1/status/12121465...  \n",
       "3  https://twitter.com/RAlexJimenez/status/121214...  \n",
       "4  https://twitter.com/findadvisor/status/1212142...  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv('combined_csv_wealthmanagement.csv')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Factoring in the number of retweets and weighing tweets accordingly"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>date</th>\n",
       "      <th>text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12/31/2019 23:45</td>\n",
       "      <td>Thinking about transitioning your business? Yo...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12/31/2019 23:31</td>\n",
       "      <td>http://RipsawWealth.com #2020WillBeTheYearFor ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12/31/2019 23:00</td>\n",
       "      <td>#wealthmanagement #maxous</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12/31/2019 22:45</td>\n",
       "      <td>#WealthManagement and private banking trends: ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12/31/2019 22:43</td>\n",
       "      <td>Is Your Money Ready for Retirement? #Retiremen...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               date                                               text\n",
       "0  12/31/2019 23:45  Thinking about transitioning your business? Yo...\n",
       "1  12/31/2019 23:31  http://RipsawWealth.com #2020WillBeTheYearFor ...\n",
       "2  12/31/2019 23:00                         #wealthmanagement #maxous \n",
       "3  12/31/2019 22:45  #WealthManagement and private banking trends: ...\n",
       "4  12/31/2019 22:43  Is Your Money Ready for Retirement? #Retiremen..."
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_tweets = df[[\"date\",\"text\",\"retweets\"]]\n",
    "df_tweets = df_tweets.dropna()\n",
    "l1= df_tweets[\"text\"].values\n",
    "for i in range(len(l1)):\n",
    "    l1[i] = (l1[i]+' ')*(int(df_tweets[\"retweets\"].values[i]+1))\n",
    "df_tweets[\"text\"] =  l1\n",
    "df_tweets = df_tweets[[\"date\",\"text\"]]\n",
    "df_tweets.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cleaning and tokenizing the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "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>date</th>\n",
       "      <th>text</th>\n",
       "      <th>preprocessed_text</th>\n",
       "      <th>tokenized_text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12/31/2019 23:45</td>\n",
       "      <td>Thinking about transitioning your business? Yo...</td>\n",
       "      <td>thinking transitioning business? you’ll surpri...</td>\n",
       "      <td>[thinking, transitioning, business, you, ll, s...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12/31/2019 23:31</td>\n",
       "      <td>http://RipsawWealth.com #2020WillBeTheYearFor ...</td>\n",
       "      <td>#2020willbetheyearfor financial control! #happ...</td>\n",
       "      <td>[willbetheyearfor, financial, control, happyne...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12/31/2019 23:00</td>\n",
       "      <td>#wealthmanagement #maxous</td>\n",
       "      <td>#maxous</td>\n",
       "      <td>[maxous]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12/31/2019 22:45</td>\n",
       "      <td>#WealthManagement and private banking trends: ...</td>\n",
       "      <td>private banking trends: 2020 | @fintechmagazin...</td>\n",
       "      <td>[private, banking, trends, fintechmagazine, we...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12/31/2019 22:43</td>\n",
       "      <td>Is Your Money Ready for Retirement? #Retiremen...</td>\n",
       "      <td>money ready retirement? #retirementplan #finan...</td>\n",
       "      <td>[money, ready, retirement, retirementplan, fin...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26951</th>\n",
       "      <td>3/19/2019 17:11</td>\n",
       "      <td>.@Praemium said its major technology upgrade w...</td>\n",
       "      <td>.@praemium said major technology upgrade expan...</td>\n",
       "      <td>[praemium, said, major, technology, upgrade, e...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26952</th>\n",
       "      <td>3/19/2019 16:59</td>\n",
       "      <td>A Seismic Shift in the Estate-Planning Landsca...</td>\n",
       "      <td>seismic shift estate-planning landscape #estat...</td>\n",
       "      <td>[seismic, shift, estate, planning, landscape, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26953</th>\n",
       "      <td>3/19/2019 16:58</td>\n",
       "      <td>Interview: Scott Cooper, @CybiWealth – investi...</td>\n",
       "      <td>interview: scott cooper, @cybiwealth – investi...</td>\n",
       "      <td>[interview, scott, cooper, cybiwealth, investi...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26954</th>\n",
       "      <td>3/19/2019 16:18</td>\n",
       "      <td>How much are you worth? . . . . #wealth #Wealt...</td>\n",
       "      <td>much worth? . . . . #wealth #wealthy #wealthbu...</td>\n",
       "      <td>[much, worth, wealth, wealthy, wealthbuilding,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26955</th>\n",
       "      <td>3/19/2019 16:15</td>\n",
       "      <td>RRSP Investors: 3 Undervalued Stocks to Buy To...</td>\n",
       "      <td>rrsp investors: 3 undervalued stocks buy today...</td>\n",
       "      <td>[rrsp, investors, undervalued, stocks, buy, to...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>26956 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   date                                               text  \\\n",
       "0      12/31/2019 23:45  Thinking about transitioning your business? Yo...   \n",
       "1      12/31/2019 23:31  http://RipsawWealth.com #2020WillBeTheYearFor ...   \n",
       "2      12/31/2019 23:00                         #wealthmanagement #maxous    \n",
       "3      12/31/2019 22:45  #WealthManagement and private banking trends: ...   \n",
       "4      12/31/2019 22:43  Is Your Money Ready for Retirement? #Retiremen...   \n",
       "...                 ...                                                ...   \n",
       "26951   3/19/2019 17:11  .@Praemium said its major technology upgrade w...   \n",
       "26952   3/19/2019 16:59  A Seismic Shift in the Estate-Planning Landsca...   \n",
       "26953   3/19/2019 16:58  Interview: Scott Cooper, @CybiWealth – investi...   \n",
       "26954   3/19/2019 16:18  How much are you worth? . . . . #wealth #Wealt...   \n",
       "26955   3/19/2019 16:15  RRSP Investors: 3 Undervalued Stocks to Buy To...   \n",
       "\n",
       "                                       preprocessed_text  \\\n",
       "0      thinking transitioning business? you’ll surpri...   \n",
       "1      #2020willbetheyearfor financial control! #happ...   \n",
       "2                                                #maxous   \n",
       "3      private banking trends: 2020 | @fintechmagazin...   \n",
       "4      money ready retirement? #retirementplan #finan...   \n",
       "...                                                  ...   \n",
       "26951  .@praemium said major technology upgrade expan...   \n",
       "26952  seismic shift estate-planning landscape #estat...   \n",
       "26953  interview: scott cooper, @cybiwealth – investi...   \n",
       "26954  much worth? . . . . #wealth #wealthy #wealthbu...   \n",
       "26955  rrsp investors: 3 undervalued stocks buy today...   \n",
       "\n",
       "                                          tokenized_text  \n",
       "0      [thinking, transitioning, business, you, ll, s...  \n",
       "1      [willbetheyearfor, financial, control, happyne...  \n",
       "2                                               [maxous]  \n",
       "3      [private, banking, trends, fintechmagazine, we...  \n",
       "4      [money, ready, retirement, retirementplan, fin...  \n",
       "...                                                  ...  \n",
       "26951  [praemium, said, major, technology, upgrade, e...  \n",
       "26952  [seismic, shift, estate, planning, landscape, ...  \n",
       "26953  [interview, scott, cooper, cybiwealth, investi...  \n",
       "26954  [much, worth, wealth, wealthy, wealthbuilding,...  \n",
       "26955  [rrsp, investors, undervalued, stocks, buy, to...  \n",
       "\n",
       "[26956 rows x 4 columns]"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def clean_tweets(df=df_tweets, \n",
    "                 tweet_col='text', \n",
    "                 date_col='date'\n",
    "                ):\n",
    "    \n",
    "    df_copy = df.copy()\n",
    "    \n",
    "    # drop rows with empty values\n",
    "    df_copy.dropna(inplace=True)\n",
    "    \n",
    "    # lower the tweets\n",
    "    df_copy['preprocessed_' + tweet_col] = df_copy[tweet_col].str.lower()\n",
    "    \n",
    "    # filter out stop words and URLs\n",
    "    en_stop_words = stopwords.words('english')\n",
    "    extended_words = ['&amp;', 'rt','th','co', 're', 've', 'kim', 'daca', '#wealthmanagement', 'wealthmanagement', 'de', 'la', 'en', 'day', 'di', 'el', 'ask', 'un', 'con', 'per', 'del', 'help', 'you', 'via', 'us', 'gt', 'ai', 'cx', 'you', 'time', 'make', 'get', 'one', 'uk']\n",
    "    en_stop_words.extend(extended_words)\n",
    "    en_stop_words = set(en_stop_words)\n",
    "    url_re = '(https?:\\/\\/(?:www\\.|(?!www))[a-zA-Z0-9][a-zA-Z0-9-]+[a-zA-Z0-9]\\.[^\\s]{2,}|www\\.[a-zA-Z0-9][a-zA-Z0-9-]+[a-zA-Z0-9]\\.[^\\s]{2,}|https?:\\/\\/(?:www\\.|(?!www))[a-zA-Z0-9]+\\.[^\\s]{2,}|www\\.[a-zA-Z0-9]+\\.[^\\s]{2,})'        \n",
    "    df_copy['preprocessed_' + tweet_col] = df_copy['preprocessed_' + tweet_col].apply(lambda row: ' '.join([word.strip() for word in row.split() if (not word.strip() in en_stop_words) and (not re.match(url_re, word.strip()))]))\n",
    "    \n",
    "    # tokenize the tweets\n",
    "    tokenizer = RegexpTokenizer('[a-zA-Z]\\w+\\'?\\w*')\n",
    "    df_copy['tokenized_' + tweet_col] = df_copy['preprocessed_' + tweet_col].apply(lambda row: tokenizer.tokenize(row))\n",
    "    \n",
    "    return df_copy\n",
    "  \n",
    "df_tweets_clean = clean_tweets(df_tweets)\n",
    "df_tweets_clean"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Most frequent words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('wealth', 9035),\n",
       " ('investing', 5677),\n",
       " ('financialplanning', 4843),\n",
       " ('investment', 4549),\n",
       " ('fintech', 4428),\n",
       " ('financial', 4288),\n",
       " ('money', 3916),\n",
       " ('retirement', 3600),\n",
       " ('finance', 3435),\n",
       " ('management', 3323)]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_most_freq_words(str, n=None):\n",
    "    vect = CountVectorizer().fit(str)\n",
    "    bag_of_words = vect.transform(str)\n",
    "    sum_words = bag_of_words.sum(axis=0) \n",
    "    freq = [(word, sum_words[0, idx]) for word, idx in vect.vocabulary_.items()]\n",
    "    freq =sorted(freq, key = lambda x: x[1], reverse=True)\n",
    "    return freq[:n]\n",
    "  \n",
    "get_most_freq_words([ word for tweet in df_tweets_clean.tokenized_text for word in tweet],10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# build a dictionary where for each tweet, each word has its own id.\n",
    "tweets_dictionary = Dictionary(df_tweets_clean.tokenized_text)\n",
    "\n",
    "# build the corpus i.e. vectors with the number of occurence of each word per tweet\n",
    "tweets_corpus = [tweets_dictionary.doc2bow(tweet) for tweet in df_tweets_clean.tokenized_text]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Do not Run the below Cell - Takes 45 mins to run"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# compute coherence\n",
    "tweets_coherence = []\n",
    "for nb_topics in range(1,25):\n",
    "    lda = LdaModel(tweets_corpus, num_topics = nb_topics, id2word = tweets_dictionary, passes=10)\n",
    "    cohm = CoherenceModel(model=lda, corpus=tweets_corpus, dictionary=tweets_dictionary, coherence='u_mass')\n",
    "    coh = cohm.get_coherence()\n",
    "    tweets_coherence.append(coh)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualize coherence\n",
    "plt.figure(figsize=(10,5))\n",
    "plt.plot(range(1,25),tweets_coherence)\n",
    "plt.xlabel(\"Number of Topics\")\n",
    "plt.ylabel(\"Coherence Score\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "k = 10\n",
    "tweets_lda = LdaModel(tweets_corpus, num_topics = k, id2word = tweets_dictionary, passes=10)\n",
    "\n",
    "def plot_top_words(lda=tweets_lda, nb_topics=k, nb_words=10):\n",
    "    top_words = [[word for word,_ in lda.show_topic(topic_id, topn=50)] for topic_id in range(lda.num_topics)]\n",
    "    top_betas = [[beta for _,beta in lda.show_topic(topic_id, topn=50)] for topic_id in range(lda.num_topics)]\n",
    "\n",
    "    gs  = gridspec.GridSpec(round(math.sqrt(k))+1,round(math.sqrt(k))+1)\n",
    "    gs.update(wspace=0.5, hspace=0.5)\n",
    "    plt.figure(figsize=(20,15))\n",
    "    for i in range(nb_topics):\n",
    "        ax = plt.subplot(gs[i])\n",
    "        plt.barh(range(nb_words), top_betas[i][:nb_words], align='center',color='blue', ecolor='black')\n",
    "        ax.invert_yaxis()\n",
    "        ax.set_yticks(range(nb_words))\n",
    "        ax.set_yticklabels(top_words[i][:nb_words])\n",
    "        plt.title(\"Topic \"+str(i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1080 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.gridspec as gridspec\n",
    "plot_top_words()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "list1 = []\n",
    "list2 = []\n",
    "for i in range(len(df_tweets_clean)):\n",
    "    x = tweets_lda[tweets_corpus[i]]\n",
    "    x.sort(key=lambda x:x[1], reverse = True)\n",
    "    list1.append('Topic'+str(x[0][0]))\n",
    "    try:\n",
    "        if x[1][1] > 0.25:\n",
    "            list2.append('Topic'+str(x[1][0]))\n",
    "        else: \n",
    "            list2.append(None)\n",
    "    except: \n",
    "            list2.append(None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "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>date</th>\n",
       "      <th>text</th>\n",
       "      <th>preprocessed_text</th>\n",
       "      <th>tokenized_text</th>\n",
       "      <th>Label1</th>\n",
       "      <th>Label2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12/31/2019 23:45</td>\n",
       "      <td>Thinking about transitioning your business? Yo...</td>\n",
       "      <td>thinking transitioning business? you’ll surpri...</td>\n",
       "      <td>[thinking, transitioning, business, you, ll, s...</td>\n",
       "      <td>Topic5</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12/31/2019 23:31</td>\n",
       "      <td>http://RipsawWealth.com #2020WillBeTheYearFor ...</td>\n",
       "      <td>#2020willbetheyearfor financial control! #happ...</td>\n",
       "      <td>[willbetheyearfor, financial, control, happyne...</td>\n",
       "      <td>Topic5</td>\n",
       "      <td>Topic6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12/31/2019 23:00</td>\n",
       "      <td>#wealthmanagement #maxous</td>\n",
       "      <td>#maxous</td>\n",
       "      <td>[maxous]</td>\n",
       "      <td>Topic8</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12/31/2019 22:45</td>\n",
       "      <td>#WealthManagement and private banking trends: ...</td>\n",
       "      <td>private banking trends: 2020 | @fintechmagazin...</td>\n",
       "      <td>[private, banking, trends, fintechmagazine, we...</td>\n",
       "      <td>Topic1</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12/31/2019 22:43</td>\n",
       "      <td>Is Your Money Ready for Retirement? #Retiremen...</td>\n",
       "      <td>money ready retirement? #retirementplan #finan...</td>\n",
       "      <td>[money, ready, retirement, retirementplan, fin...</td>\n",
       "      <td>Topic5</td>\n",
       "      <td>None</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26951</th>\n",
       "      <td>3/19/2019 17:11</td>\n",
       "      <td>.@Praemium said its major technology upgrade w...</td>\n",
       "      <td>.@praemium said major technology upgrade expan...</td>\n",
       "      <td>[praemium, said, major, technology, upgrade, e...</td>\n",
       "      <td>Topic1</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26952</th>\n",
       "      <td>3/19/2019 16:59</td>\n",
       "      <td>A Seismic Shift in the Estate-Planning Landsca...</td>\n",
       "      <td>seismic shift estate-planning landscape #estat...</td>\n",
       "      <td>[seismic, shift, estate, planning, landscape, ...</td>\n",
       "      <td>Topic5</td>\n",
       "      <td>Topic6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26953</th>\n",
       "      <td>3/19/2019 16:58</td>\n",
       "      <td>Interview: Scott Cooper, @CybiWealth – investi...</td>\n",
       "      <td>interview: scott cooper, @cybiwealth – investi...</td>\n",
       "      <td>[interview, scott, cooper, cybiwealth, investi...</td>\n",
       "      <td>Topic1</td>\n",
       "      <td>Topic0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26954</th>\n",
       "      <td>3/19/2019 16:18</td>\n",
       "      <td>How much are you worth? . . . . #wealth #Wealt...</td>\n",
       "      <td>much worth? . . . . #wealth #wealthy #wealthbu...</td>\n",
       "      <td>[much, worth, wealth, wealthy, wealthbuilding,...</td>\n",
       "      <td>Topic7</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26955</th>\n",
       "      <td>3/19/2019 16:15</td>\n",
       "      <td>RRSP Investors: 3 Undervalued Stocks to Buy To...</td>\n",
       "      <td>rrsp investors: 3 undervalued stocks buy today...</td>\n",
       "      <td>[rrsp, investors, undervalued, stocks, buy, to...</td>\n",
       "      <td>Topic9</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>26956 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   date                                               text  \\\n",
       "0      12/31/2019 23:45  Thinking about transitioning your business? Yo...   \n",
       "1      12/31/2019 23:31  http://RipsawWealth.com #2020WillBeTheYearFor ...   \n",
       "2      12/31/2019 23:00                         #wealthmanagement #maxous    \n",
       "3      12/31/2019 22:45  #WealthManagement and private banking trends: ...   \n",
       "4      12/31/2019 22:43  Is Your Money Ready for Retirement? #Retiremen...   \n",
       "...                 ...                                                ...   \n",
       "26951   3/19/2019 17:11  .@Praemium said its major technology upgrade w...   \n",
       "26952   3/19/2019 16:59  A Seismic Shift in the Estate-Planning Landsca...   \n",
       "26953   3/19/2019 16:58  Interview: Scott Cooper, @CybiWealth – investi...   \n",
       "26954   3/19/2019 16:18  How much are you worth? . . . . #wealth #Wealt...   \n",
       "26955   3/19/2019 16:15  RRSP Investors: 3 Undervalued Stocks to Buy To...   \n",
       "\n",
       "                                       preprocessed_text  \\\n",
       "0      thinking transitioning business? you’ll surpri...   \n",
       "1      #2020willbetheyearfor financial control! #happ...   \n",
       "2                                                #maxous   \n",
       "3      private banking trends: 2020 | @fintechmagazin...   \n",
       "4      money ready retirement? #retirementplan #finan...   \n",
       "...                                                  ...   \n",
       "26951  .@praemium said major technology upgrade expan...   \n",
       "26952  seismic shift estate-planning landscape #estat...   \n",
       "26953  interview: scott cooper, @cybiwealth – investi...   \n",
       "26954  much worth? . . . . #wealth #wealthy #wealthbu...   \n",
       "26955  rrsp investors: 3 undervalued stocks buy today...   \n",
       "\n",
       "                                          tokenized_text  Label1  Label2  \n",
       "0      [thinking, transitioning, business, you, ll, s...  Topic5    None  \n",
       "1      [willbetheyearfor, financial, control, happyne...  Topic5  Topic6  \n",
       "2                                               [maxous]  Topic8    None  \n",
       "3      [private, banking, trends, fintechmagazine, we...  Topic1    None  \n",
       "4      [money, ready, retirement, retirementplan, fin...  Topic5    None  \n",
       "...                                                  ...     ...     ...  \n",
       "26951  [praemium, said, major, technology, upgrade, e...  Topic1    None  \n",
       "26952  [seismic, shift, estate, planning, landscape, ...  Topic5  Topic6  \n",
       "26953  [interview, scott, cooper, cybiwealth, investi...  Topic1  Topic0  \n",
       "26954  [much, worth, wealth, wealthy, wealthbuilding,...  Topic7    None  \n",
       "26955  [rrsp, investors, undervalued, stocks, buy, to...  Topic9    None  \n",
       "\n",
       "[26956 rows x 6 columns]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_tweets_clean[\"Label1\"] = list1\n",
    "df_tweets_clean[\"Label2\"] = list2\n",
    "df_tweets_clean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "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>date</th>\n",
       "      <th>username</th>\n",
       "      <th>to</th>\n",
       "      <th>replies</th>\n",
       "      <th>retweets</th>\n",
       "      <th>favorites</th>\n",
       "      <th>text</th>\n",
       "      <th>geo</th>\n",
       "      <th>mentions</th>\n",
       "      <th>hashtags</th>\n",
       "      <th>id</th>\n",
       "      <th>permalink</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>12/31/2019 23:45</td>\n",
       "      <td>HT_Great_Lakes</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Thinking about transitioning your business? Yo...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#wealthmanagement #hightowergreatlakes #busine...</td>\n",
       "      <td>1.212160e+18</td>\n",
       "      <td>https://twitter.com/HT_Great_Lakes/status/1212...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12/31/2019 23:31</td>\n",
       "      <td>ripsawwealth</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>http://RipsawWealth.com #2020WillBeTheYearFor ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#2020WillBeTheYearFor #HappyNewYear #happynewy...</td>\n",
       "      <td>1.212150e+18</td>\n",
       "      <td>https://twitter.com/ripsawwealth/status/121215...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12/31/2019 23:00</td>\n",
       "      <td>TedHunter1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>#wealthmanagement #maxous</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#wealthmanagement #maxous</td>\n",
       "      <td>1.212150e+18</td>\n",
       "      <td>https://twitter.com/TedHunter1/status/12121465...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>12/31/2019 22:45</td>\n",
       "      <td>RAlexJimenez</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>#WealthManagement and private banking trends: ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>@FinTechMagazine</td>\n",
       "      <td>#WealthManagement #wealthtech #fintech #finserv</td>\n",
       "      <td>1.212140e+18</td>\n",
       "      <td>https://twitter.com/RAlexJimenez/status/121214...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>12/31/2019 22:43</td>\n",
       "      <td>findadvisor</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Is Your Money Ready for Retirement? #Retiremen...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#RetirementPlan #FinancialPlan #WealthManageme...</td>\n",
       "      <td>1.212140e+18</td>\n",
       "      <td>https://twitter.com/findadvisor/status/1212142...</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",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26951</th>\n",
       "      <td>3/19/2019 17:11</td>\n",
       "      <td>haiyencutelam</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>.@Praemium said its major technology upgrade w...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#Exchanges #Explained #healthinsurance #insura...</td>\n",
       "      <td>1.108050e+18</td>\n",
       "      <td>https://twitter.com/haiyencutelam/status/11080...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26952</th>\n",
       "      <td>3/19/2019 16:59</td>\n",
       "      <td>FFPWealthMgt</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>A Seismic Shift in the Estate-Planning Landsca...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#WealthManagement #economicUpdate #FinancialAd...</td>\n",
       "      <td>1.108050e+18</td>\n",
       "      <td>https://twitter.com/FFPWealthMgt/status/110805...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26953</th>\n",
       "      <td>3/19/2019 16:58</td>\n",
       "      <td>PaulBeachAL</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>Interview: Scott Cooper, @CybiWealth – investi...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#winner #inspiringinnovator #wealthmanagement ...</td>\n",
       "      <td>1.108050e+18</td>\n",
       "      <td>https://twitter.com/PaulBeachAL/status/1108050...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26954</th>\n",
       "      <td>3/19/2019 16:18</td>\n",
       "      <td>danwoodruff</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>How much are you worth? . . . . #wealth #Wealt...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#FinancialPlanning #WealthManagement #Inherita...</td>\n",
       "      <td>1.108040e+18</td>\n",
       "      <td>https://twitter.com/danwoodruff/status/1108039...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26955</th>\n",
       "      <td>3/19/2019 16:15</td>\n",
       "      <td>davidwerdiger</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>RRSP Investors: 3 Undervalued Stocks to Buy To...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>#familybusiness #familylaw #familypractice #su...</td>\n",
       "      <td>1.108040e+18</td>\n",
       "      <td>https://twitter.com/davidwerdiger/status/11080...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>26956 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   date        username   to  replies  retweets  favorites  \\\n",
       "0      12/31/2019 23:45  HT_Great_Lakes  NaN        0         1          1   \n",
       "1      12/31/2019 23:31    ripsawwealth  NaN        0         2          3   \n",
       "2      12/31/2019 23:00      TedHunter1  NaN        0         0          0   \n",
       "3      12/31/2019 22:45    RAlexJimenez  NaN        0         0          0   \n",
       "4      12/31/2019 22:43     findadvisor  NaN        0         0          0   \n",
       "...                 ...             ...  ...      ...       ...        ...   \n",
       "26951   3/19/2019 17:11   haiyencutelam  NaN        0         0          0   \n",
       "26952   3/19/2019 16:59    FFPWealthMgt  NaN        0         0          0   \n",
       "26953   3/19/2019 16:58     PaulBeachAL  NaN        0         0          0   \n",
       "26954   3/19/2019 16:18     danwoodruff  NaN        0         0          1   \n",
       "26955   3/19/2019 16:15   davidwerdiger  NaN        0         0          0   \n",
       "\n",
       "                                                    text  geo  \\\n",
       "0      Thinking about transitioning your business? Yo...  NaN   \n",
       "1      http://RipsawWealth.com #2020WillBeTheYearFor ...  NaN   \n",
       "2                              #wealthmanagement #maxous  NaN   \n",
       "3      #WealthManagement and private banking trends: ...  NaN   \n",
       "4      Is Your Money Ready for Retirement? #Retiremen...  NaN   \n",
       "...                                                  ...  ...   \n",
       "26951  .@Praemium said its major technology upgrade w...  NaN   \n",
       "26952  A Seismic Shift in the Estate-Planning Landsca...  NaN   \n",
       "26953  Interview: Scott Cooper, @CybiWealth – investi...  NaN   \n",
       "26954  How much are you worth? . . . . #wealth #Wealt...  NaN   \n",
       "26955  RRSP Investors: 3 Undervalued Stocks to Buy To...  NaN   \n",
       "\n",
       "               mentions                                           hashtags  \\\n",
       "0                   NaN  #wealthmanagement #hightowergreatlakes #busine...   \n",
       "1                   NaN  #2020WillBeTheYearFor #HappyNewYear #happynewy...   \n",
       "2                   NaN                          #wealthmanagement #maxous   \n",
       "3      @FinTechMagazine    #WealthManagement #wealthtech #fintech #finserv   \n",
       "4                   NaN  #RetirementPlan #FinancialPlan #WealthManageme...   \n",
       "...                 ...                                                ...   \n",
       "26951               NaN  #Exchanges #Explained #healthinsurance #insura...   \n",
       "26952               NaN  #WealthManagement #economicUpdate #FinancialAd...   \n",
       "26953               NaN  #winner #inspiringinnovator #wealthmanagement ...   \n",
       "26954               NaN  #FinancialPlanning #WealthManagement #Inherita...   \n",
       "26955               NaN  #familybusiness #familylaw #familypractice #su...   \n",
       "\n",
       "                 id                                          permalink  \n",
       "0      1.212160e+18  https://twitter.com/HT_Great_Lakes/status/1212...  \n",
       "1      1.212150e+18  https://twitter.com/ripsawwealth/status/121215...  \n",
       "2      1.212150e+18  https://twitter.com/TedHunter1/status/12121465...  \n",
       "3      1.212140e+18  https://twitter.com/RAlexJimenez/status/121214...  \n",
       "4      1.212140e+18  https://twitter.com/findadvisor/status/1212142...  \n",
       "...             ...                                                ...  \n",
       "26951  1.108050e+18  https://twitter.com/haiyencutelam/status/11080...  \n",
       "26952  1.108050e+18  https://twitter.com/FFPWealthMgt/status/110805...  \n",
       "26953  1.108050e+18  https://twitter.com/PaulBeachAL/status/1108050...  \n",
       "26954  1.108040e+18  https://twitter.com/danwoodruff/status/1108039...  \n",
       "26955  1.108040e+18  https://twitter.com/davidwerdiger/status/11080...  \n",
       "\n",
       "[26956 rows x 12 columns]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df.dropna(subset = [\"text\"])\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_tweets_merge = df_tweets_clean.join(df, lsuffix = [\"date\", \"text\"], rsuffix = [\"date\", \"text\"])\n",
    "df_tweets_merge = df_tweets_merge.loc[:,~df_tweets_merge.columns.duplicated()]\n",
    "df_tweets_merge = df_tweets_merge.rename(columns = {\"date['date', 'text']\":\"date\", \"text['date', 'text']\": \"text\"})\n",
    "df_tweets_merge.to_csv(\"labeled_data_wealthmanagement.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "documents = list(df_tweets_clean[\"preprocessed_text\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "import gensim\n",
    "list_tfidf = []\n",
    "tfidf = gensim.models.TfidfModel(tweets_corpus)\n",
    "for doc in tfidf[tweets_corpus]:\n",
    "    list_tfidf.append([[tweets_dictionary[id], np.around(freq, decimals=2)] for id, freq in doc])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'business'"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "list_tfidf[0][1][0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "list_1 = []\n",
    "list_3 = []\n",
    "for i in range(len(list_tfidf)):\n",
    "    list_2 = []\n",
    "    list_4 = []\n",
    "    for j in range(len(list_tfidf[i])):\n",
    "        list_2.append(list_tfidf[i][j][0])\n",
    "        list_4.append(list_tfidf[i][j][1])\n",
    "    list_1.append(list_2)\n",
    "    list_3.append(list_4)   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "tfidf_df = pd.DataFrame({\"Tweets\":list_1, \"tfidf_score\":list_3})\n",
    "tfidf_df.to_csv(\"tfidf.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "list_1 = []\n",
    "list_2 = []\n",
    "for i in range(len(weighted_phrases)):\n",
    "    list_1.append(weighted_phrases[i][0])\n",
    "    list_2.append(weighted_phrases[i][1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "wphrases = pd.DataFrame({\"Phrases\":list_1, \"Score\":list_2})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "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>Phrases</th>\n",
       "      <th>Score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>maxous</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>financialadvisors</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>increasewealth</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>axaexclusiv</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ripsawwealth</td>\n",
       "      <td>1.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41030</th>\n",
       "      <td>clients</td>\n",
       "      <td>0.099</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41031</th>\n",
       "      <td>top</td>\n",
       "      <td>0.095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41032</th>\n",
       "      <td>management</td>\n",
       "      <td>0.089</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41033</th>\n",
       "      <td>private</td>\n",
       "      <td>0.073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41034</th>\n",
       "      <td>wealth</td>\n",
       "      <td>0.051</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>41035 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                 Phrases  Score\n",
       "0                 maxous  1.000\n",
       "1      financialadvisors  1.000\n",
       "2         increasewealth  1.000\n",
       "3            axaexclusiv  1.000\n",
       "4           ripsawwealth  1.000\n",
       "...                  ...    ...\n",
       "41030            clients  0.099\n",
       "41031                top  0.095\n",
       "41032         management  0.089\n",
       "41033            private  0.073\n",
       "41034             wealth  0.051\n",
       "\n",
       "[41035 rows x 2 columns]"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wphrases"
   ]
  },
  {
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}