TSwiftLyricsJN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "import re\n",
    "import string\n",
    "from collections import Counter\n",
    "import nltk\n",
    "\n",
    "from sklearn.feature_extraction import text\n",
    "from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer\n",
    "from sklearn.decomposition import TruncatedSVD, NMF, PCA\n",
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "\n",
    "from sklearn.cluster import KMeans, DBSCAN\n",
    "from sklearn.manifold import TSNE\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "spotify = pd.read_csv('LyricsFeaturesGenre.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>track_album</th>\n",
       "      <th>track_artist</th>\n",
       "      <th>track_title</th>\n",
       "      <th>danceability</th>\n",
       "      <th>energy</th>\n",
       "      <th>key</th>\n",
       "      <th>loudness</th>\n",
       "      <th>mode</th>\n",
       "      <th>speechiness</th>\n",
       "      <th>acousticness</th>\n",
       "      <th>instrumentalness</th>\n",
       "      <th>liveness</th>\n",
       "      <th>valence</th>\n",
       "      <th>tempo</th>\n",
       "      <th>duration_ms</th>\n",
       "      <th>time_signature</th>\n",
       "      <th>track_uri</th>\n",
       "      <th>track_id</th>\n",
       "      <th>track_lyric</th>\n",
       "      <th>genres</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Tim McGraw</td>\n",
       "      <td>0.580</td>\n",
       "      <td>0.491</td>\n",
       "      <td>0</td>\n",
       "      <td>-6.462</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0251</td>\n",
       "      <td>0.575</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.1210</td>\n",
       "      <td>0.425</td>\n",
       "      <td>76.009</td>\n",
       "      <td>232107</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:0Om9WAB5RS09L80DyOfTNa</td>\n",
       "      <td>0Om9WAB5RS09L80DyOfTNa</td>\n",
       "      <td>He said the way my blue eyes shined\\nPut those...</td>\n",
       "      <td>Country</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Picture To Burn</td>\n",
       "      <td>0.658</td>\n",
       "      <td>0.877</td>\n",
       "      <td>7</td>\n",
       "      <td>-2.098</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0323</td>\n",
       "      <td>0.173</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0962</td>\n",
       "      <td>0.821</td>\n",
       "      <td>105.586</td>\n",
       "      <td>173067</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:32mVHdy0bi1XKgr0ajsBlG</td>\n",
       "      <td>32mVHdy0bi1XKgr0ajsBlG</td>\n",
       "      <td>State the obvious, I didn't get my perfect fan...</td>\n",
       "      <td>Country Rock</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Teardrops On My Guitar</td>\n",
       "      <td>0.621</td>\n",
       "      <td>0.417</td>\n",
       "      <td>10</td>\n",
       "      <td>-6.941</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0231</td>\n",
       "      <td>0.288</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.1190</td>\n",
       "      <td>0.289</td>\n",
       "      <td>99.953</td>\n",
       "      <td>203040</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:7zMcNqs55Mxer82bvZFkpg</td>\n",
       "      <td>7zMcNqs55Mxer82bvZFkpg</td>\n",
       "      <td>Drew looks at me\\nI fake a smile so he won't s...</td>\n",
       "      <td>Country Pop</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>A Place In This World</td>\n",
       "      <td>0.576</td>\n",
       "      <td>0.777</td>\n",
       "      <td>9</td>\n",
       "      <td>-2.881</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0324</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.3200</td>\n",
       "      <td>0.428</td>\n",
       "      <td>115.028</td>\n",
       "      <td>199200</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:73OX8GdpOeGzKC6OvGSbsv</td>\n",
       "      <td>73OX8GdpOeGzKC6OvGSbsv</td>\n",
       "      <td>I don't know what I want, so don't ask me\\nCau...</td>\n",
       "      <td>Country Pop</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Cold as You</td>\n",
       "      <td>0.418</td>\n",
       "      <td>0.482</td>\n",
       "      <td>5</td>\n",
       "      <td>-5.769</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0266</td>\n",
       "      <td>0.217</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.1230</td>\n",
       "      <td>0.261</td>\n",
       "      <td>175.558</td>\n",
       "      <td>239013</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:7an1exwMnfYRcdVQm0yDev</td>\n",
       "      <td>7an1exwMnfYRcdVQm0yDev</td>\n",
       "      <td>You have a way of coming easily to me\\nAnd whe...</td>\n",
       "      <td>Country</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    track_album  track_artist             track_title  danceability  energy  \\\n",
       "0  Taylor Swift  Taylor Swift              Tim McGraw         0.580   0.491   \n",
       "1  Taylor Swift  Taylor Swift         Picture To Burn         0.658   0.877   \n",
       "2  Taylor Swift  Taylor Swift  Teardrops On My Guitar         0.621   0.417   \n",
       "3  Taylor Swift  Taylor Swift   A Place In This World         0.576   0.777   \n",
       "4  Taylor Swift  Taylor Swift             Cold as You         0.418   0.482   \n",
       "\n",
       "   key  loudness  mode  speechiness  acousticness  instrumentalness  liveness  \\\n",
       "0    0    -6.462     1       0.0251         0.575               0.0    0.1210   \n",
       "1    7    -2.098     1       0.0323         0.173               0.0    0.0962   \n",
       "2   10    -6.941     1       0.0231         0.288               0.0    0.1190   \n",
       "3    9    -2.881     1       0.0324         0.051               0.0    0.3200   \n",
       "4    5    -5.769     1       0.0266         0.217               0.0    0.1230   \n",
       "\n",
       "   valence    tempo  duration_ms  time_signature  \\\n",
       "0    0.425   76.009       232107               4   \n",
       "1    0.821  105.586       173067               4   \n",
       "2    0.289   99.953       203040               4   \n",
       "3    0.428  115.028       199200               4   \n",
       "4    0.261  175.558       239013               4   \n",
       "\n",
       "                              track_uri                track_id  \\\n",
       "0  spotify:track:0Om9WAB5RS09L80DyOfTNa  0Om9WAB5RS09L80DyOfTNa   \n",
       "1  spotify:track:32mVHdy0bi1XKgr0ajsBlG  32mVHdy0bi1XKgr0ajsBlG   \n",
       "2  spotify:track:7zMcNqs55Mxer82bvZFkpg  7zMcNqs55Mxer82bvZFkpg   \n",
       "3  spotify:track:73OX8GdpOeGzKC6OvGSbsv  73OX8GdpOeGzKC6OvGSbsv   \n",
       "4  spotify:track:7an1exwMnfYRcdVQm0yDev  7an1exwMnfYRcdVQm0yDev   \n",
       "\n",
       "                                         track_lyric        genres  \n",
       "0  He said the way my blue eyes shined\\nPut those...       Country  \n",
       "1  State the obvious, I didn't get my perfect fan...  Country Rock  \n",
       "2  Drew looks at me\\nI fake a smile so he won't s...   Country Pop  \n",
       "3  I don't know what I want, so don't ask me\\nCau...   Country Pop  \n",
       "4  You have a way of coming easily to me\\nAnd whe...       Country  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>track_album</th>\n",
       "      <th>track_artist</th>\n",
       "      <th>track_title</th>\n",
       "      <th>danceability</th>\n",
       "      <th>energy</th>\n",
       "      <th>key</th>\n",
       "      <th>loudness</th>\n",
       "      <th>mode</th>\n",
       "      <th>speechiness</th>\n",
       "      <th>acousticness</th>\n",
       "      <th>instrumentalness</th>\n",
       "      <th>liveness</th>\n",
       "      <th>valence</th>\n",
       "      <th>tempo</th>\n",
       "      <th>duration_ms</th>\n",
       "      <th>time_signature</th>\n",
       "      <th>track_uri</th>\n",
       "      <th>track_id</th>\n",
       "      <th>track_lyric</th>\n",
       "      <th>genres</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>Lover</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Miss Americana &amp; the Heartbreak Prince</td>\n",
       "      <td>0.662</td>\n",
       "      <td>0.747</td>\n",
       "      <td>11</td>\n",
       "      <td>-6.926</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0736</td>\n",
       "      <td>0.02800</td>\n",
       "      <td>0.006150</td>\n",
       "      <td>0.1380</td>\n",
       "      <td>0.487</td>\n",
       "      <td>150.088</td>\n",
       "      <td>234147</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:214nt20w5wOxJnY462klLw</td>\n",
       "      <td>214nt20w5wOxJnY462klLw</td>\n",
       "      <td>You know I adore you, I'm crazier for you\\nTha...</td>\n",
       "      <td>Electropop</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>Lover</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Paper Rings</td>\n",
       "      <td>0.811</td>\n",
       "      <td>0.719</td>\n",
       "      <td>9</td>\n",
       "      <td>-6.553</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0497</td>\n",
       "      <td>0.01290</td>\n",
       "      <td>0.000014</td>\n",
       "      <td>0.0742</td>\n",
       "      <td>0.865</td>\n",
       "      <td>103.979</td>\n",
       "      <td>222400</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:4y5bvROuBDPr5fuwXbIBZR</td>\n",
       "      <td>4y5bvROuBDPr5fuwXbIBZR</td>\n",
       "      <td>The moon is high\\nLike your friends were the n...</td>\n",
       "      <td>Pop Punk</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>Lover</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Soon You'll Get Better (feat. Dixie Chicks)</td>\n",
       "      <td>0.433</td>\n",
       "      <td>0.182</td>\n",
       "      <td>0</td>\n",
       "      <td>-12.566</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0641</td>\n",
       "      <td>0.90700</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.1230</td>\n",
       "      <td>0.421</td>\n",
       "      <td>207.476</td>\n",
       "      <td>201587</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:4AYtqFyFbX0Xkc2wtcygTr</td>\n",
       "      <td>4AYtqFyFbX0Xkc2wtcygTr</td>\n",
       "      <td>The buttons of my coat were tangled in my hair...</td>\n",
       "      <td>Country</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>Lover</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>The Archer</td>\n",
       "      <td>0.292</td>\n",
       "      <td>0.574</td>\n",
       "      <td>0</td>\n",
       "      <td>-9.375</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0401</td>\n",
       "      <td>0.12000</td>\n",
       "      <td>0.005690</td>\n",
       "      <td>0.0663</td>\n",
       "      <td>0.166</td>\n",
       "      <td>124.344</td>\n",
       "      <td>211240</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:3pHkh7d0lzM2AldUtz2x37</td>\n",
       "      <td>3pHkh7d0lzM2AldUtz2x37</td>\n",
       "      <td>Combat\\nI'm ready for combat\\nI say I don't wa...</td>\n",
       "      <td>Dream Pop</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>Lover</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>You Need to Calm Down</td>\n",
       "      <td>0.771</td>\n",
       "      <td>0.671</td>\n",
       "      <td>2</td>\n",
       "      <td>-5.617</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0553</td>\n",
       "      <td>0.00929</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0637</td>\n",
       "      <td>0.714</td>\n",
       "      <td>85.026</td>\n",
       "      <td>171360</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:6RRNNciQGZEXnqk8SQ9yv5</td>\n",
       "      <td>6RRNNciQGZEXnqk8SQ9yv5</td>\n",
       "      <td>You are somebody that I don't know\\nBut you're...</td>\n",
       "      <td>Pop</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    track_album  track_artist                                  track_title  \\\n",
       "106       Lover  Taylor Swift       Miss Americana & the Heartbreak Prince   \n",
       "107       Lover  Taylor Swift                                  Paper Rings   \n",
       "108       Lover  Taylor Swift  Soon You'll Get Better (feat. Dixie Chicks)   \n",
       "109       Lover  Taylor Swift                                   The Archer   \n",
       "110       Lover  Taylor Swift                        You Need to Calm Down   \n",
       "\n",
       "     danceability  energy  key  loudness  mode  speechiness  acousticness  \\\n",
       "106         0.662   0.747   11    -6.926     0       0.0736       0.02800   \n",
       "107         0.811   0.719    9    -6.553     1       0.0497       0.01290   \n",
       "108         0.433   0.182    0   -12.566     1       0.0641       0.90700   \n",
       "109         0.292   0.574    0    -9.375     1       0.0401       0.12000   \n",
       "110         0.771   0.671    2    -5.617     1       0.0553       0.00929   \n",
       "\n",
       "     instrumentalness  liveness  valence    tempo  duration_ms  \\\n",
       "106          0.006150    0.1380    0.487  150.088       234147   \n",
       "107          0.000014    0.0742    0.865  103.979       222400   \n",
       "108          0.000000    0.1230    0.421  207.476       201587   \n",
       "109          0.005690    0.0663    0.166  124.344       211240   \n",
       "110          0.000000    0.0637    0.714   85.026       171360   \n",
       "\n",
       "     time_signature                             track_uri  \\\n",
       "106               4  spotify:track:214nt20w5wOxJnY462klLw   \n",
       "107               4  spotify:track:4y5bvROuBDPr5fuwXbIBZR   \n",
       "108               4  spotify:track:4AYtqFyFbX0Xkc2wtcygTr   \n",
       "109               4  spotify:track:3pHkh7d0lzM2AldUtz2x37   \n",
       "110               4  spotify:track:6RRNNciQGZEXnqk8SQ9yv5   \n",
       "\n",
       "                   track_id  \\\n",
       "106  214nt20w5wOxJnY462klLw   \n",
       "107  4y5bvROuBDPr5fuwXbIBZR   \n",
       "108  4AYtqFyFbX0Xkc2wtcygTr   \n",
       "109  3pHkh7d0lzM2AldUtz2x37   \n",
       "110  6RRNNciQGZEXnqk8SQ9yv5   \n",
       "\n",
       "                                           track_lyric      genres  \n",
       "106  You know I adore you, I'm crazier for you\\nTha...  Electropop  \n",
       "107  The moon is high\\nLike your friends were the n...    Pop Punk  \n",
       "108  The buttons of my coat were tangled in my hair...     Country  \n",
       "109  Combat\\nI'm ready for combat\\nI say I don't wa...   Dream Pop  \n",
       "110  You are somebody that I don't know\\nBut you're...         Pop  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "spotify['year_released'] = spotify['track_album']\n",
    "spotify['year_released'] = spotify['year_released'].replace(['Taylor Swift', 'Fearless', 'Speak Now', 'Red', '1989',\n",
    "       'reputation', 'Lover'],[2006,2008,2010,2012,2014,2017,2019])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "spotify['world_sales_USD'] = spotify['track_album'].replace(['Taylor Swift', 'Fearless', 'Speak Now', 'Red', '1989',\n",
    "       'reputation', 'Lover'],[7000000,12000000,5500000,6000000,10500000,5000000,4000000])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>track_album</th>\n",
       "      <th>track_artist</th>\n",
       "      <th>track_title</th>\n",
       "      <th>danceability</th>\n",
       "      <th>energy</th>\n",
       "      <th>key</th>\n",
       "      <th>loudness</th>\n",
       "      <th>mode</th>\n",
       "      <th>speechiness</th>\n",
       "      <th>acousticness</th>\n",
       "      <th>...</th>\n",
       "      <th>valence</th>\n",
       "      <th>tempo</th>\n",
       "      <th>duration_ms</th>\n",
       "      <th>time_signature</th>\n",
       "      <th>track_uri</th>\n",
       "      <th>track_id</th>\n",
       "      <th>track_lyric</th>\n",
       "      <th>genres</th>\n",
       "      <th>year_released</th>\n",
       "      <th>world_sales_USD</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>111</td>\n",
       "      <td>111</td>\n",
       "      <td>111</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>111</td>\n",
       "      <td>111</td>\n",
       "      <td>111</td>\n",
       "      <td>111</td>\n",
       "      <td>111.000000</td>\n",
       "      <td>1.110000e+02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>7</td>\n",
       "      <td>5</td>\n",
       "      <td>111</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>111</td>\n",
       "      <td>111</td>\n",
       "      <td>111</td>\n",
       "      <td>16</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>Red</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Bad Blood</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>spotify:track:1dGr1c8CrMLDpV6mPbImSI</td>\n",
       "      <td>7vvIpJZye5cRR6De1LKM0m</td>\n",
       "      <td>You should've been there\\nShould've burst thro...</td>\n",
       "      <td>Country Pop</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>19</td>\n",
       "      <td>107</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>22</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.613595</td>\n",
       "      <td>0.616441</td>\n",
       "      <td>4.639640</td>\n",
       "      <td>-6.453261</td>\n",
       "      <td>0.882883</td>\n",
       "      <td>0.073693</td>\n",
       "      <td>0.197342</td>\n",
       "      <td>...</td>\n",
       "      <td>0.423593</td>\n",
       "      <td>122.176027</td>\n",
       "      <td>231977.243243</td>\n",
       "      <td>3.963964</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2012.504505</td>\n",
       "      <td>6.959459e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.112205</td>\n",
       "      <td>0.176523</td>\n",
       "      <td>3.257977</td>\n",
       "      <td>2.266960</td>\n",
       "      <td>0.323018</td>\n",
       "      <td>0.132049</td>\n",
       "      <td>0.253976</td>\n",
       "      <td>...</td>\n",
       "      <td>0.198856</td>\n",
       "      <td>31.460427</td>\n",
       "      <td>45882.035346</td>\n",
       "      <td>0.328299</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.289255</td>\n",
       "      <td>2.681702e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.292000</td>\n",
       "      <td>0.151000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-12.864000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.023100</td>\n",
       "      <td>0.000197</td>\n",
       "      <td>...</td>\n",
       "      <td>0.049900</td>\n",
       "      <td>68.534000</td>\n",
       "      <td>51920.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2006.000000</td>\n",
       "      <td>4.000000e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.552500</td>\n",
       "      <td>0.484000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>-7.499500</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.029500</td>\n",
       "      <td>0.022950</td>\n",
       "      <td>...</td>\n",
       "      <td>0.264500</td>\n",
       "      <td>96.005000</td>\n",
       "      <td>207813.500000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2010.000000</td>\n",
       "      <td>5.000000e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.612000</td>\n",
       "      <td>0.651000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>-6.237000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.037200</td>\n",
       "      <td>0.082600</td>\n",
       "      <td>...</td>\n",
       "      <td>0.416000</td>\n",
       "      <td>118.984000</td>\n",
       "      <td>231000.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2012.000000</td>\n",
       "      <td>6.000000e+06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.691000</td>\n",
       "      <td>0.753000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>-4.947000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.063500</td>\n",
       "      <td>0.238500</td>\n",
       "      <td>...</td>\n",
       "      <td>0.558000</td>\n",
       "      <td>146.015000</td>\n",
       "      <td>245080.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017.000000</td>\n",
       "      <td>1.050000e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.897000</td>\n",
       "      <td>0.950000</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>-2.098000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.941000</td>\n",
       "      <td>0.971000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.943000</td>\n",
       "      <td>207.476000</td>\n",
       "      <td>403920.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2019.000000</td>\n",
       "      <td>1.200000e+07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>11 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       track_album  track_artist track_title  danceability      energy  \\\n",
       "count          111           111         111    111.000000  111.000000   \n",
       "unique           7             5         111           NaN         NaN   \n",
       "top            Red  Taylor Swift   Bad Blood           NaN         NaN   \n",
       "freq            19           107           1           NaN         NaN   \n",
       "mean           NaN           NaN         NaN      0.613595    0.616441   \n",
       "std            NaN           NaN         NaN      0.112205    0.176523   \n",
       "min            NaN           NaN         NaN      0.292000    0.151000   \n",
       "25%            NaN           NaN         NaN      0.552500    0.484000   \n",
       "50%            NaN           NaN         NaN      0.612000    0.651000   \n",
       "75%            NaN           NaN         NaN      0.691000    0.753000   \n",
       "max            NaN           NaN         NaN      0.897000    0.950000   \n",
       "\n",
       "               key    loudness        mode  speechiness  acousticness  ...  \\\n",
       "count   111.000000  111.000000  111.000000   111.000000    111.000000  ...   \n",
       "unique         NaN         NaN         NaN          NaN           NaN  ...   \n",
       "top            NaN         NaN         NaN          NaN           NaN  ...   \n",
       "freq           NaN         NaN         NaN          NaN           NaN  ...   \n",
       "mean      4.639640   -6.453261    0.882883     0.073693      0.197342  ...   \n",
       "std       3.257977    2.266960    0.323018     0.132049      0.253976  ...   \n",
       "min       0.000000  -12.864000    0.000000     0.023100      0.000197  ...   \n",
       "25%       2.000000   -7.499500    1.000000     0.029500      0.022950  ...   \n",
       "50%       5.000000   -6.237000    1.000000     0.037200      0.082600  ...   \n",
       "75%       7.000000   -4.947000    1.000000     0.063500      0.238500  ...   \n",
       "max      11.000000   -2.098000    1.000000     0.941000      0.971000  ...   \n",
       "\n",
       "           valence       tempo    duration_ms  time_signature  \\\n",
       "count   111.000000  111.000000     111.000000      111.000000   \n",
       "unique         NaN         NaN            NaN             NaN   \n",
       "top            NaN         NaN            NaN             NaN   \n",
       "freq           NaN         NaN            NaN             NaN   \n",
       "mean      0.423593  122.176027  231977.243243        3.963964   \n",
       "std       0.198856   31.460427   45882.035346        0.328299   \n",
       "min       0.049900   68.534000   51920.000000        1.000000   \n",
       "25%       0.264500   96.005000  207813.500000        4.000000   \n",
       "50%       0.416000  118.984000  231000.000000        4.000000   \n",
       "75%       0.558000  146.015000  245080.000000        4.000000   \n",
       "max       0.943000  207.476000  403920.000000        5.000000   \n",
       "\n",
       "                                   track_uri                track_id  \\\n",
       "count                                    111                     111   \n",
       "unique                                   111                     111   \n",
       "top     spotify:track:1dGr1c8CrMLDpV6mPbImSI  7vvIpJZye5cRR6De1LKM0m   \n",
       "freq                                       1                       1   \n",
       "mean                                     NaN                     NaN   \n",
       "std                                      NaN                     NaN   \n",
       "min                                      NaN                     NaN   \n",
       "25%                                      NaN                     NaN   \n",
       "50%                                      NaN                     NaN   \n",
       "75%                                      NaN                     NaN   \n",
       "max                                      NaN                     NaN   \n",
       "\n",
       "                                              track_lyric       genres  \\\n",
       "count                                                 111          111   \n",
       "unique                                                111           16   \n",
       "top     You should've been there\\nShould've burst thro...  Country Pop   \n",
       "freq                                                    1           22   \n",
       "mean                                                  NaN          NaN   \n",
       "std                                                   NaN          NaN   \n",
       "min                                                   NaN          NaN   \n",
       "25%                                                   NaN          NaN   \n",
       "50%                                                   NaN          NaN   \n",
       "75%                                                   NaN          NaN   \n",
       "max                                                   NaN          NaN   \n",
       "\n",
       "       year_released world_sales_USD  \n",
       "count     111.000000    1.110000e+02  \n",
       "unique           NaN             NaN  \n",
       "top              NaN             NaN  \n",
       "freq             NaN             NaN  \n",
       "mean     2012.504505    6.959459e+06  \n",
       "std         4.289255    2.681702e+06  \n",
       "min      2006.000000    4.000000e+06  \n",
       "25%      2010.000000    5.000000e+06  \n",
       "50%      2012.000000    6.000000e+06  \n",
       "75%      2017.000000    1.050000e+07  \n",
       "max      2019.000000    1.200000e+07  \n",
       "\n",
       "[11 rows x 22 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify.describe(include='all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Country', 'Country Rock', 'Country Pop', 'Contemporary Country',\n",
       "       'Pop Rock', 'Folk Pop', 'Blue grass', 'Pop Punk', 'Soft Rock',\n",
       "       'Alternative Rock', 'Pop', 'Electropop', 'Synth Pop', 'Dance Pop',\n",
       "       'Dream Pop', 'R&B'], dtype=object)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify['genres'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'He said the way my blue eyes shined\\nPut those Georgia stars to shame that night\\nI said, \"That\\'s a lie\"\\nJust a boy in a Chevy truck\\nThat had a tendency of gettin\\' stuck\\nOn backroads at night\\nAnd I was right there beside him all summer long\\nAnd then the time we woke up to find that summer gone\\nBut when you think Tim McGraw\\nI hope you think my favorite song\\nThe one we danced to all night long\\nThe moon like a spotlight on the lake\\nWhen you think happiness\\nI hope you think that little black dress\\nThink of my head on your chest\\nAnd my old faded blue jeans\\nWhen you think Tim McGraw\\nI hope you think of me\\nSeptember saw a month of tears\\nAnd thankin\\' God that you weren\\'t here\\nTo see me like that\\nBut in a box beneath my bed\\nIs a letter that you never read\\nFrom three summers back\\nIt\\'s hard not to find it all a little bittersweet\\nAnd lookin\\' back on all of that, it\\'s nice to believe\\nWhen you think Tim McGraw\\nI hope you think my favorite song\\nThe one we danced to all night long\\nThe moon like a spotlight on the lake\\nWhen you think happiness\\nI hope you think that little black dress\\nThink of my head on your chest\\nAnd my old faded blue jeans\\nWhen you think Tim McGraw\\nI hope you think of me\\nAnd I\\'m back for the first time since then\\nI\\'m standin\\' on your street\\nAnd there\\'s a letter left on your doorstep\\nAnd the first thing that you\\'ll read is:\\n\"When you think Tim McGraw\\nI hope you think my favorite song\\nSomeday you\\'ll turn your radio on\\nI hope it takes you back to that place\"\\nWhen you think happiness\\nI hope you think that little black dress\\nThink of my head on your chest\\nAnd my old faded blue jeans\\nWhen you think Tim McGraw\\nI hope you think of me\\nOh, think of me\\nMmmm\\nHe said the way my blue eyes shine\\nPut those Georgia stars to shame that night\\nI said, \"That\\'s a lie\"'"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify.track_lyric[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cleaning for NLP"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "remove_n = lambda x: re.sub('\\n',' ',x)\n",
    "alphanumeric = lambda x: re.sub('\\w*\\d\\w*', ' ', x)\n",
    "punc_lower = lambda x: re.sub('[%s]' % re.escape(string.punctuation), ' ', x.lower())\n",
    "\n",
    "spotify['track_lyric'] = spotify.track_lyric.map(remove_n).map(punc_lower).map(alphanumeric)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'don t look at me  you got a girl at home and everybody knows that  everybody knows that don t look at me  you got a girl at home and everybody knows that  everybody knows that i don t even know her but i feel a responsibility to do what s upstanding and right it s kinda like a code  yeah and you ve been getting closer and closer and crossing so many lines and it would be a fine proposition if i was a stupid girl but honey i am no one s exception this i have previously learned so don t look at me  you got a girl at home and everybody knows that  everybody knows that don t look at me  you got a girl at home and everybody knows that  everybody knows that i see you turn off your phone now you ve got me alone  and i say don t look at me  you got a girl at home and everybody knows that  everybody knows that i just want to make sure you understand perfectly you re the kind of man who makes me sad while she waits up  you chase down the newest thing and take for granted what you have and it would be a fine proposition if i was a stupid girl and yeah  i might go with it if i hadn t once been just like her so don t look at me  you got a girl at home and everybody knows that  everybody knows that don t look at me  you got a girl at home and everybody knows that  everybody knows that i see you turn off your phone now you ve got me alone  and i say don t look at me  you got a girl at home and everybody knows that  everybody knows that call a cab  lose my number you re about to lose your girl call a cab  lose my number let s consider this lesson learned want to see you pick up your phone and tell her you re coming home don t look at me  you got a girl at home and everybody knows that  everybody knows that it would be a fine proposition if i hadn t once been just like her'"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify.track_lyric[62]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "song_titles = spotify['track_title'].values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# NLP Round 1:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Document-term Matrix with just Unigrams"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "vectorizer = CountVectorizer(stop_words='english', min_df=0.1,max_df=0.7)\n",
    "\n",
    "doc_matrix = vectorizer.fit_transform(spotify.track_lyric)\n",
    "\n",
    "doc_matrix = pd.DataFrame(doc_matrix.toarray(), columns=vectorizer.get_feature_names())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>ain</th>\n",
       "      <th>away</th>\n",
       "      <th>baby</th>\n",
       "      <th>bad</th>\n",
       "      <th>beautiful</th>\n",
       "      <th>best</th>\n",
       "      <th>better</th>\n",
       "      <th>blue</th>\n",
       "      <th>break</th>\n",
       "      <th>car</th>\n",
       "      <th>...</th>\n",
       "      <th>walk</th>\n",
       "      <th>wanna</th>\n",
       "      <th>want</th>\n",
       "      <th>wanted</th>\n",
       "      <th>way</th>\n",
       "      <th>won</th>\n",
       "      <th>world</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yeah</th>\n",
       "      <th>years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 122 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   ain  away  baby  bad  beautiful  best  better  blue  break  car  ...  walk  \\\n",
       "0    0     0     0    0          0     0       0     5      0    0  ...     0   \n",
       "1    0     0     2    5          0     1       1     0      0    0  ...     0   \n",
       "2    0     0     0    0          2     0       1     0      1    3  ...     0   \n",
       "3    0     0     0    0          0     0       0     1      0    0  ...     0   \n",
       "4    0     4     0    0          0     1       0     0      0    0  ...     2   \n",
       "\n",
       "   wanna  want  wanted  way  won  world  wrong  yeah  years  \n",
       "0      0     0       0    2    0      0      0     0      0  \n",
       "1      0     0       0    1    0      0      0     0      0  \n",
       "2      0     1       0    0    2      0      0     0      0  \n",
       "3      0     1       0    1    0      3      3     1      0  \n",
       "4      0     1       1    1    0      0      0     0      0  \n",
       "\n",
       "[5 rows x 122 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc_matrix.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Document-term Matrix with Bigrams and Unigrams"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "vectorizer_bi = CountVectorizer(ngram_range=(1,2),binary=True,stop_words='english',min_df=0.1,max_df=0.7)\n",
    "\n",
    "doc_matrix_bi = vectorizer_bi.fit_transform(spotify.track_lyric)\n",
    "\n",
    "bi_doc_matrix = pd.DataFrame(doc_matrix_bi.toarray(), columns=vectorizer_bi.get_feature_names())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>ain</th>\n",
       "      <th>away</th>\n",
       "      <th>baby</th>\n",
       "      <th>bad</th>\n",
       "      <th>beautiful</th>\n",
       "      <th>best</th>\n",
       "      <th>better</th>\n",
       "      <th>blue</th>\n",
       "      <th>break</th>\n",
       "      <th>car</th>\n",
       "      <th>...</th>\n",
       "      <th>walk</th>\n",
       "      <th>wanna</th>\n",
       "      <th>want</th>\n",
       "      <th>wanted</th>\n",
       "      <th>way</th>\n",
       "      <th>won</th>\n",
       "      <th>world</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yeah</th>\n",
       "      <th>years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 125 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   ain  away  baby  bad  beautiful  best  better  blue  break  car  ...  walk  \\\n",
       "0    0     0     0    0          0     0       0     1      0    0  ...     0   \n",
       "1    0     0     1    1          0     1       1     0      0    0  ...     0   \n",
       "2    0     0     0    0          1     0       1     0      1    1  ...     0   \n",
       "3    0     0     0    0          0     0       0     1      0    0  ...     0   \n",
       "4    0     1     0    0          0     1       0     0      0    0  ...     1   \n",
       "\n",
       "   wanna  want  wanted  way  won  world  wrong  yeah  years  \n",
       "0      0     0       0    1    0      0      0     0      0  \n",
       "1      0     0       0    1    0      0      0     0      0  \n",
       "2      0     1       0    0    1      0      0     0      0  \n",
       "3      0     1       0    1    0      1      1     1      0  \n",
       "4      0     1       1    1    0      0      0     0      0  \n",
       "\n",
       "[5 rows x 125 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bi_doc_matrix.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Scree Plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Scree plot')"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "lsascree = TruncatedSVD(8)\n",
    "lsascree.fit(doc_matrix)\n",
    "lsa_features = lsascree.transform(doc_matrix)\n",
    "plt.plot(lsascree.explained_variance_ratio_)\n",
    "plt.xlabel('Number of Components')\n",
    "plt.ylabel('Explained Variance')\n",
    "plt.title('Scree plot')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Scree plot')"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "lsascree = TruncatedSVD(12)\n",
    "lsascree.fit(doc_matrix_bi)\n",
    "lsa_features = lsascree.transform(doc_matrix_bi)\n",
    "plt.plot(lsascree.explained_variance_ratio_)\n",
    "plt.xlabel('Number of Components')\n",
    "plt.ylabel('Explained Variance')\n",
    "plt.title('Scree plot')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LSA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def display_topics(model, feature_names, no_top_words, topic_names=None):\n",
    "    for ix, topic in enumerate(model.components_):\n",
    "        if not topic_names or not topic_names[ix]:\n",
    "            print(\"\\nTopic \", ix)\n",
    "        else:\n",
    "            print(\"\\nTopic: '\",topic_names[ix],\"'\")\n",
    "        print(\", \".join([feature_names[i]\n",
    "                        for i in topic.argsort()[:-no_top_words - 1:-1]]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def lsa_function(num_components, doc_text_matrix, vectorizer):\n",
    "    lsa = TruncatedSVD(num_components)\n",
    "    doc_topic = lsa.fit_transform(doc_text_matrix)\n",
    "    print('Explained Variance:',lsa.explained_variance_ratio_)\n",
    "    \n",
    "    index = []\n",
    "    for i in range(num_components):\n",
    "        index.append(i)\n",
    "    topic_word = pd.DataFrame(lsa.components_.round(3),\n",
    "             index = index,\n",
    "             columns = vectorizer.get_feature_names())\n",
    "    \n",
    "    print(display_topics(lsa, vectorizer.get_feature_names(), 15))\n",
    "    \n",
    "    return topic_word"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance: [0.05980753 0.09185649 0.0768749 ]\n",
      "\n",
      "Topic  0\n",
      "oh, don, just, ll, got, ve, ooh, time, love, look, cause, say, baby, want, wanna\n",
      "\n",
      "Topic  1\n",
      "ooh, getting, wanna, friends, look, waiting, end, miss, yeah, say, remember, night, heart, change, day\n",
      "\n",
      "Topic  2\n",
      "got, ve, wanna, don, look, love, baby, bad, cause, come, end, did, think, ll, good\n",
      "None\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>ain</th>\n",
       "      <th>away</th>\n",
       "      <th>baby</th>\n",
       "      <th>bad</th>\n",
       "      <th>beautiful</th>\n",
       "      <th>best</th>\n",
       "      <th>better</th>\n",
       "      <th>blue</th>\n",
       "      <th>break</th>\n",
       "      <th>car</th>\n",
       "      <th>...</th>\n",
       "      <th>walk</th>\n",
       "      <th>wanna</th>\n",
       "      <th>want</th>\n",
       "      <th>wanted</th>\n",
       "      <th>way</th>\n",
       "      <th>won</th>\n",
       "      <th>world</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yeah</th>\n",
       "      <th>years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.038</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.159</td>\n",
       "      <td>0.106</td>\n",
       "      <td>0.030</td>\n",
       "      <td>0.038</td>\n",
       "      <td>0.061</td>\n",
       "      <td>0.017</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.049</td>\n",
       "      <td>...</td>\n",
       "      <td>0.018</td>\n",
       "      <td>0.131</td>\n",
       "      <td>0.142</td>\n",
       "      <td>0.017</td>\n",
       "      <td>0.062</td>\n",
       "      <td>0.035</td>\n",
       "      <td>0.033</td>\n",
       "      <td>0.025</td>\n",
       "      <td>0.110</td>\n",
       "      <td>0.009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.010</td>\n",
       "      <td>0.006</td>\n",
       "      <td>-0.016</td>\n",
       "      <td>-0.001</td>\n",
       "      <td>-0.011</td>\n",
       "      <td>-0.013</td>\n",
       "      <td>-0.022</td>\n",
       "      <td>-0.006</td>\n",
       "      <td>0.000</td>\n",
       "      <td>-0.049</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.005</td>\n",
       "      <td>0.119</td>\n",
       "      <td>-0.049</td>\n",
       "      <td>-0.010</td>\n",
       "      <td>-0.013</td>\n",
       "      <td>-0.012</td>\n",
       "      <td>-0.006</td>\n",
       "      <td>-0.013</td>\n",
       "      <td>0.033</td>\n",
       "      <td>-0.004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.002</td>\n",
       "      <td>0.033</td>\n",
       "      <td>0.164</td>\n",
       "      <td>0.160</td>\n",
       "      <td>0.009</td>\n",
       "      <td>-0.004</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.003</td>\n",
       "      <td>0.013</td>\n",
       "      <td>-0.109</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.003</td>\n",
       "      <td>0.197</td>\n",
       "      <td>-0.016</td>\n",
       "      <td>-0.011</td>\n",
       "      <td>0.058</td>\n",
       "      <td>0.012</td>\n",
       "      <td>0.010</td>\n",
       "      <td>-0.007</td>\n",
       "      <td>-0.031</td>\n",
       "      <td>-0.001</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 122 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     ain   away   baby    bad  beautiful   best  better   blue  break    car  \\\n",
       "0  0.038  0.055  0.159  0.106      0.030  0.038   0.061  0.017  0.055  0.049   \n",
       "1 -0.010  0.006 -0.016 -0.001     -0.011 -0.013  -0.022 -0.006  0.000 -0.049   \n",
       "2  0.002  0.033  0.164  0.160      0.009 -0.004   0.008  0.003  0.013 -0.109   \n",
       "\n",
       "   ...   walk  wanna   want  wanted    way    won  world  wrong   yeah  years  \n",
       "0  ...  0.018  0.131  0.142   0.017  0.062  0.035  0.033  0.025  0.110  0.009  \n",
       "1  ... -0.005  0.119 -0.049  -0.010 -0.013 -0.012 -0.006 -0.013  0.033 -0.004  \n",
       "2  ... -0.003  0.197 -0.016  -0.011  0.058  0.012  0.010 -0.007 -0.031 -0.001  \n",
       "\n",
       "[3 rows x 122 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lsa_function(3,doc_matrix,vectorizer)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Top words in each song:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "word_doc = doc_matrix.transpose()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "top_songs = {}\n",
    "for song in word_doc.columns:\n",
    "    top = word_doc[song].sort_values(ascending=False).head(30)\n",
    "    top_songs[song] = list(zip(top.index,top.values))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "think, night, blue, little, said, long, head, dress, old, time, ll, eyes, way, turn\n",
      "---\n",
      "1\n",
      "really, just, time, bad, old, drive, let, sorry, tell, baby, cause, ll, love, friends\n",
      "---\n",
      "2\n",
      "don, got, car, smile, won, beautiful, ll, love, thing, cause, light, night, drive, eyes\n",
      "---\n",
      "3\n",
      "oh, just, girl, ll, trying, life, don, place, wrong, world, got, cause, tell, need\n",
      "---\n",
      "4\n",
      "away, oh, day, thinking, cold, ve, walk, say, cause, just, best, way, fight, feel\n",
      "---\n",
      "5\n",
      "ve, looking, ain, best, better, try, didn, got, right, saw, times, knew, baby, wanted\n",
      "---\n",
      "6\n",
      "baby, smile, coming, tell, hold, don, high, cause, oh, face, looking, rain, end, thinking\n",
      "---\n",
      "7\n",
      "beautiful, oh, don, stay, door, gonna, little, love, life, really, way, ll, day, hard\n",
      "---\n",
      "8\n",
      "ve, said, gone, baby, thought, mind, say, did, let, home, right, just, tell, looking\n",
      "---\n",
      "9\n",
      "oh, said, ll, did, eyes, time, years, love, town, lights, little, need, met, look\n",
      "---\n",
      "10\n",
      "got, didn, cause, play, way, don, said, phone, home, door, kiss, wrong, car, lost\n",
      "---\n",
      "11\n",
      "don, wanna, trying, feel, drive, let, time, just, say, hard, night, need, girl, tell\n",
      "---\n",
      "12\n",
      "just, want, light, love, right, knew, way, beautiful, going, eyes, don, smile, think, make\n",
      "---\n",
      "13\n",
      "wanna, good, heart, break, make, love, ve, eyes, say, don, walk, fall, way, away\n",
      "---\n",
      "14\n",
      "don, oh, better, yeah, head, hand, dress, best, right, hands, run, just, wanna, way\n",
      "---\n",
      "15\n",
      "going, love, friends, say, night, day, didn, wanted, walk, ll, feeling, life, kiss, look\n",
      "---\n",
      "16\n",
      "ll, oh, said, love, say, baby, don, just, waiting, eyes, town, little, saw, run\n",
      "---\n",
      "17\n",
      "come, ve, hey, feeling, wanna, cause, look, rain, met, feel, way, kiss, say, waiting\n",
      "---\n",
      "18\n",
      "ain, town, cause, come, just, oh, ll, let, sorry, love, got, gonna, need, girl\n",
      "---\n",
      "19\n",
      "time, night, baby, hey, think, thinking, looking, waiting, door, day, dreams, say, said, room\n",
      "---\n",
      "20\n",
      "sorry, people, want, end, cause, feeling, just, lost, don, gonna, mind, going, thing, hand\n",
      "---\n",
      "21\n",
      "tell, got, took, remember, night, said, feel, need, run, don, love, feeling, want, thing\n",
      "---\n",
      "22\n",
      "don, sorry, baby, time, ve, did, won, phone, tell, left, wanna, thinking, long, ll\n",
      "---\n",
      "23\n",
      "way, loved, feel, oh, knew, love, rain, miss, coming, look, fine, smile, got, need\n",
      "---\n",
      "24\n",
      "said, forever, feel, cause, baby, wrong, oh, gone, did, don, phone, didn, knew, thought\n",
      "---\n",
      "25\n",
      "don, best, home, day, run, fall, drive, change, ve, going, look, away, got, love\n",
      "---\n",
      "26\n",
      "ll, things, oh, fight, time, hold, getting, cause, change, come, fall, feel, ve, night\n",
      "---\n",
      "27\n",
      "time, thing, best, remember, hold, yeah, make, say, love, got, ll, said, street, lights\n",
      "---\n",
      "28\n",
      "smile, cause, baby, eyes, lights, rain, just, away, ll, kiss, really, won, good, thinking\n",
      "---\n",
      "29\n",
      "time, night, turn, mind, change, sorry, ain, love, right, loved, make, good, saw, door\n",
      "---\n",
      "30\n",
      "say, don, ll, away, door, said, run, need, girl, ooh, wrong, room, looking, didn\n",
      "---\n",
      "31\n",
      "don, think, oh, gone, girl, dress, way, home, look, long, run, sorry, ll, said\n",
      "---\n",
      "32\n",
      "ll, gonna, old, don, cause, just, going, got, years, feeling, right, feel, things, cold\n",
      "---\n",
      "33\n",
      "say, don, yeah, room, tell, things, end, think, day, oh, knew, break, way, fight\n",
      "---\n",
      "34\n",
      "don, oh, just, little, stay, let, won, try, remember, light, getting, night, break, heart\n",
      "---\n",
      "35\n",
      "don, forever, knew, let, night, ll, love, way, home, waiting, say, eyes, door, saw\n",
      "---\n",
      "36\n",
      "better, think, mind, things, just, people, cause, don, gonna, friends, won, oh, make, eyes\n",
      "---\n",
      "37\n",
      "did, just, lost, ve, lights, life, time, new, oh, beautiful, left, trying, change, ll\n",
      "---\n",
      "38\n",
      "oh, thought, gone, leave, don, turn, won, wrong, time, walk, break, away, just, wanted\n",
      "---\n",
      "39\n",
      "forever, kiss, just, ll, don, miss, thought, end, feel, life, remember, love, away, day\n",
      "---\n",
      "40\n",
      "long, day, time, just, life, lights, tell, said, face, look, remember, forever, fall, feeling\n",
      "---\n",
      "41\n",
      "love, don, people, mind, things, little, look, cause, ll, hard, life, high, right, gonna\n",
      "---\n",
      "42\n",
      "come, just, said, ve, rain, want, sorry, baby, thought, change, gone, night, people, say\n",
      "---\n",
      "43\n",
      "ll, away, forget, day, come, got, world, don, say, right, love, ve, loved, forever\n",
      "---\n",
      "44\n",
      "coming, ll, love, saw, good, right, room, play, fight, just, lights, time, hands, fall\n",
      "---\n",
      "45\n",
      "trying, dark, blue, met, right, end, new, yeah, street, head, love, time, old, just\n",
      "---\n",
      "46\n",
      "home, think, mind, just, drive, night, away, say, ll, hands, getting, going, stay, don\n",
      "---\n",
      "47\n",
      "oh, knew, hard, cold, gone, ll, long, time, heard, new, didn, met, loved, took\n",
      "---\n",
      "48\n",
      "remember, cause, long, lost, thing, yeah, little, gone, ve, forget, old, home, got, loved\n",
      "---\n",
      "49\n",
      "yeah, don, oh, night, just, feeling, wanted, time, look, forget, bad, won, best, fall\n",
      "---\n",
      "50\n",
      "want, just, tell, time, run, don, night, looking, hard, long, try, dreams, say, think\n",
      "---\n",
      "51\n",
      "ooh, getting, friends, say, time, yeah, oh, remember, night, miss, love, gonna, day, mind\n",
      "---\n",
      "52\n",
      "stay, time, think, best, ve, said, love, took, leave, times, dreams, room, night, didn\n",
      "---\n",
      "53\n",
      "time, oh, heart, break, just, times, door, eyes, got, let, better, right, home, say\n",
      "---\n",
      "54\n",
      "don, gonna, wanna, looking, right, good, day, time, took, ve, room, new, need, girl\n",
      "---\n",
      "55\n",
      "beautiful, love, time, little, long, try, look, right, face, good, dreams, got, kiss, life\n",
      "---\n",
      "56\n",
      "oh, tell, ll, lights, let, yeah, cause, look, took, make, got, feel, think, say\n",
      "---\n",
      "57\n",
      "better, want, just, ll, said, door, eyes, home, look, feel, cause, coming, ve, mind\n",
      "---\n",
      "58\n",
      "ooh, night, don, forget, best, place, oh, things, said, dreams, change, thinking, look, ll\n",
      "---\n",
      "59\n",
      "think, little, cause, head, ve, did, thinking, end, break, love, don, said, didn, got\n",
      "---\n",
      "60\n",
      "oh, ve, said, knew, thing, just, dress, sorry, say, didn, door, feeling, friends, right\n",
      "---\n",
      "61\n",
      "come, don, wanna, way, feeling, miss, need, new, took, said, fall, kiss, right, mind\n",
      "---\n",
      "62\n",
      "girl, got, home, don, look, just, ve, fine, phone, yeah, turn, want, say, thing\n",
      "---\n",
      "63\n",
      "new, waiting, lights, love, want, said, heard, wanted, took, change, good, forever, gonna, going\n",
      "---\n",
      "64\n",
      "ll, got, tell, love, cause, gonna, long, didn, leave, baby, say, way, want, oh\n",
      "---\n",
      "65\n",
      "got, cause, come, time, home, girl, heard, just, oh, thing, ve, look, long, good\n",
      "---\n",
      "66\n",
      "good, remember, looking, thinking, baby, fall, room, took, night, oh, ooh, said, forget, world\n",
      "---\n",
      "67\n",
      "stay, oh, hey, say, want, ve, let, just, people, hand, love, wanted, gone, forever\n",
      "---\n",
      "68\n",
      "gonna, break, play, just, baby, oh, got, say, don, people, cause, won, hey, ve\n",
      "---\n",
      "69\n",
      "right, knew, come, good, forget, love, phone, did, long, run, miss, turn, remember, thinking\n",
      "---\n",
      "70\n",
      "ve, got, baby, bad, hey, think, cause, don, love, look, did, really, just, sorry\n",
      "---\n",
      "71\n",
      "ll, dreams, say, just, dress, remember, forever, said, leave, bad, night, thought, gonna, good\n",
      "---\n",
      "72\n",
      "oh, girl, want, forever, say, yeah, ll, better, heart, rain, lost, tell, ooh, left\n",
      "---\n",
      "73\n",
      "love, gone, hands, let, just, good, bad, left, dark, kiss, run, smile, night, new\n",
      "---\n",
      "74\n",
      "cause, run, won, ll, trying, love, hand, baby, got, don, time, bad, just, little\n",
      "---\n",
      "75\n",
      "think, rain, gone, don, lost, room, dress, gonna, thing, miss, head, heard, left, away\n",
      "---\n",
      "76\n",
      "got, lost, didn, forever, better, life, eyes, new, don, feel, oh, things, tell, mind\n",
      "---\n",
      "77\n",
      "love, home, way, lights, feel, look, night, face, room, fight, play, place, lost, saw\n",
      "---\n",
      "78\n",
      "baby, new, come, day, life, night, cause, people, best, play, just, honey, face, need\n",
      "---\n",
      "79\n",
      "let, night, baby, dreams, time, ll, lights, gonna, things, saw, knew, loved, hard, love\n",
      "---\n",
      "80\n",
      "wanna, end, don, ooh, yeah, got, love, just, say, bad, heard, miss, ve, play\n",
      "---\n",
      "81\n",
      "good, light, feel, bad, did, say, just, world, tell, don, oh, cause, love, look\n",
      "---\n",
      "82\n",
      "time, baby, life, ll, oh, ain, right, don, love, just, high, mind, say, face\n",
      "---\n",
      "83\n",
      "cause, said, make, yeah, head, want, ain, best, look, oh, stay, eyes, honey, phone\n",
      "---\n",
      "84\n",
      "look, just, ooh, don, got, ll, oh, bad, dreams, time, ve, honey, world, play\n",
      "---\n",
      "85\n",
      "baby, bad, eyes, cause, fall, face, things, make, place, did, come, getting, right, girl\n",
      "---\n",
      "86\n",
      "say, face, make, way, feel, cause, got, look, think, ve, eyes, just, room, hand\n",
      "---\n",
      "87\n",
      "oh, car, met, place, think, leave, heart, good, don, didn, old, sorry, knew, said\n",
      "---\n",
      "88\n",
      "waiting, ooh, heart, cause, room, let, want, ll, ve, took, baby, girl, fine, change\n",
      "---\n",
      "89\n",
      "hands, yeah, oh, feeling, bad, time, world, face, love, room, forever, loved, ve, say\n",
      "---\n",
      "90\n",
      "dress, best, don, want, oh, just, hands, say, cause, time, waiting, saw, room, face\n",
      "---\n",
      "91\n",
      "things, said, friends, did, think, away, break, honey, baby, don, phone, feeling, rain, lost\n",
      "---\n",
      "92\n",
      "want, yeah, baby, right, new, thing, head, did, high, cause, away, run, heard, look\n",
      "---\n",
      "93\n",
      "hold, don, ll, stay, day, want, new, hard, wrong, town, gonna, hand, times, lost\n",
      "---\n",
      "94\n",
      "don, wanna, just, say, tell, need, love, break, oh, ooh, face, sorry, hey, ll\n",
      "---\n",
      "95\n",
      "street, walk, hand, baby, wanna, don, say, just, place, oh, new, gone, years, head\n",
      "---\n",
      "96\n",
      "oh, yeah, want, just, right, waiting, night, bad, ll, ooh, trying, new, ain, said\n",
      "---\n",
      "97\n",
      "cause, coming, oh, bad, say, good, let, yeah, getting, play, just, make, hey, gonna\n",
      "---\n",
      "98\n",
      "love, cause, didn, time, baby, way, home, ll, long, right, said, say, dress, hand\n",
      "---\n",
      "99\n",
      "just, love, away, remember, think, honey, lost, said, say, new, thing, good, make, fight\n",
      "---\n",
      "100\n",
      "did, just, gone, wrong, yeah, ve, night, going, got, thinking, thought, love, hard, town\n",
      "---\n",
      "101\n",
      "want, got, baby, ll, wanna, oh, gonna, think, make, ain, stay, smile, drive, night\n",
      "---\n",
      "102\n",
      "ooh, yeah, home, walk, hand, tell, stay, lost, ve, light, wanna, day, good, girl\n",
      "---\n",
      "103\n",
      "love, wanna, oh, just, yeah, ooh, heard, friends, smile, say, saw, end, good, met\n",
      "---\n",
      "104\n",
      "home, ll, make, place, forever, years, end, leave, honey, let, lights, heart, loved, hand\n",
      "---\n",
      "105\n",
      "baby, ooh, things, want, ll, gonna, ain, cause, make, love, think, girl, phone, just\n",
      "---\n",
      "106\n",
      "bad, gonna, cause, just, thought, girl, blue, feeling, coming, fight, lost, town, don, long\n",
      "---\n",
      "107\n",
      "want, cause, kiss, night, oh, gonna, right, times, long, friends, dreams, life, ve, met\n",
      "---\n",
      "108\n",
      "ll, better, cause, won, make, night, didn, day, time, say, try, years, best, ooh\n",
      "---\n",
      "109\n",
      "right, stay, ve, cause, years, dark, leave, want, don, just, hold, got, night, getting\n",
      "---\n",
      "110\n",
      "oh, just, need, say, street, don, ve, ain, try, friends, people, night, got, coming\n",
      "---\n"
     ]
    }
   ],
   "source": [
    "for song, top_words in top_songs.items():\n",
    "    print(song)\n",
    "    print(', '.join([word for word, count in top_words[0:14]]))\n",
    "    print('---')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Add stop words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(111, 122)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc_matrix.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "words = []\n",
    "for song in word_doc.columns:\n",
    "    top = [word for (word,count) in top_songs[song]]\n",
    "    for t in top:\n",
    "        words.append(t)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('just', 77),\n",
       " ('got', 74),\n",
       " ('girl', 68),\n",
       " ('don', 68),\n",
       " ('gonna', 67),\n",
       " ('gone', 66),\n",
       " ('cause', 65),\n",
       " ('going', 64),\n",
       " ('oh', 62),\n",
       " ('good', 62),\n",
       " ('friends', 61),\n",
       " ('say', 59),\n",
       " ('getting', 54),\n",
       " ('ll', 53),\n",
       " ('love', 53),\n",
       " ('time', 51),\n",
       " ('ve', 49),\n",
       " ('home', 45),\n",
       " ('hands', 45),\n",
       " ('hand', 44),\n",
       " ('night', 39),\n",
       " ('said', 38),\n",
       " ('hard', 38),\n",
       " ('baby', 38),\n",
       " ('right', 37),\n",
       " ('away', 37),\n",
       " ('think', 36),\n",
       " ('let', 36),\n",
       " ('eyes', 34),\n",
       " ('want', 34),\n",
       " ('forget', 33),\n",
       " ('tell', 33),\n",
       " ('way', 32),\n",
       " ('forever', 32),\n",
       " ('heart', 32),\n",
       " ('hold', 32),\n",
       " ('yeah', 32),\n",
       " ('come', 31),\n",
       " ('high', 31),\n",
       " ('look', 30),\n",
       " ('mind', 30),\n",
       " ('make', 30),\n",
       " ('head', 29),\n",
       " ('better', 29),\n",
       " ('years', 29),\n",
       " ('hey', 29),\n",
       " ('didn', 27),\n",
       " ('heard', 27),\n",
       " ('honey', 26),\n",
       " ('face', 26),\n",
       " ('little', 24),\n",
       " ('things', 24),\n",
       " ('long', 23),\n",
       " ('thing', 22),\n",
       " ('best', 22),\n",
       " ('break', 22),\n",
       " ('life', 22),\n",
       " ('day', 22),\n",
       " ('feel', 22),\n",
       " ('lights', 22),\n",
       " ('room', 22),\n",
       " ('need', 21),\n",
       " ('did', 21),\n",
       " ('bad', 20),\n",
       " ('knew', 20),\n",
       " ('fight', 19),\n",
       " ('kiss', 19),\n",
       " ('wanna', 19),\n",
       " ('new', 19),\n",
       " ('fall', 18),\n",
       " ('feeling', 18),\n",
       " ('end', 18),\n",
       " ('coming', 17),\n",
       " ('fine', 17),\n",
       " ('walk', 17),\n",
       " ('leave', 17),\n",
       " ('blue', 16),\n",
       " ('place', 16),\n",
       " ('saw', 16),\n",
       " ('smile', 16),\n",
       " ('won', 16),\n",
       " ('wrong', 16),\n",
       " ('world', 16),\n",
       " ('change', 16),\n",
       " ('remember', 16),\n",
       " ('took', 16),\n",
       " ('street', 15),\n",
       " ('car', 15),\n",
       " ('beautiful', 15),\n",
       " ('thinking', 15),\n",
       " ('wanted', 15),\n",
       " ('ain', 15),\n",
       " ('try', 15),\n",
       " ('stay', 15),\n",
       " ('door', 15),\n",
       " ('thought', 15),\n",
       " ('met', 15),\n",
       " ('lost', 15),\n",
       " ('dark', 15),\n",
       " ('ooh', 15),\n",
       " ('old', 14),\n",
       " ('left', 14),\n",
       " ('really', 14),\n",
       " ('looking', 14),\n",
       " ('town', 14),\n",
       " ('run', 14),\n",
       " ('sorry', 13),\n",
       " ('light', 13),\n",
       " ('trying', 13),\n",
       " ('rain', 13),\n",
       " ('dreams', 13),\n",
       " ('people', 13),\n",
       " ('loved', 13),\n",
       " ('dress', 12),\n",
       " ('drive', 12),\n",
       " ('play', 12),\n",
       " ('phone', 12),\n",
       " ('miss', 12),\n",
       " ('turn', 11),\n",
       " ('cold', 11),\n",
       " ('times', 11),\n",
       " ('waiting', 11)]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Counter(words).most_common()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_stops = ['just','don','gonna','cause','ll','ve','got','oh','eh','aah','want','way','away','ooh','wanna']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "stop_words = text.ENGLISH_STOP_WORDS.union(all_stops)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "vectorizer2 = CountVectorizer(stop_words=stop_words,min_df=0.1,max_df=0.7)\n",
    "\n",
    "doc_matrix2 = vectorizer2.fit_transform(spotify.track_lyric)\n",
    "\n",
    "doc_matrix2 = pd.DataFrame(doc_matrix2.toarray(), columns=vectorizer2.get_feature_names())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>ain</th>\n",
       "      <th>baby</th>\n",
       "      <th>bad</th>\n",
       "      <th>beautiful</th>\n",
       "      <th>best</th>\n",
       "      <th>better</th>\n",
       "      <th>blue</th>\n",
       "      <th>break</th>\n",
       "      <th>car</th>\n",
       "      <th>change</th>\n",
       "      <th>...</th>\n",
       "      <th>trying</th>\n",
       "      <th>turn</th>\n",
       "      <th>waiting</th>\n",
       "      <th>walk</th>\n",
       "      <th>wanted</th>\n",
       "      <th>won</th>\n",
       "      <th>world</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yeah</th>\n",
       "      <th>years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 109 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   ain  baby  bad  beautiful  best  better  blue  break  car  change  ...  \\\n",
       "0    0     0    0          0     0       0     5      0    0       0  ...   \n",
       "1    0     2    5          0     1       1     0      0    0       0  ...   \n",
       "2    0     0    0          2     0       1     0      1    3       0  ...   \n",
       "3    0     0    0          0     0       0     1      0    0       0  ...   \n",
       "4    0     0    0          0     1       0     0      0    0       0  ...   \n",
       "\n",
       "   trying  turn  waiting  walk  wanted  won  world  wrong  yeah  years  \n",
       "0       0     1        0     0       0    0      0      0     0      0  \n",
       "1       0     0        0     0       0    0      0      0     0      0  \n",
       "2       0     1        0     0       0    2      0      0     0      0  \n",
       "3       5     0        0     0       0    0      3      3     1      0  \n",
       "4       0     0        0     2       1    0      0      0     0      0  \n",
       "\n",
       "[5 rows x 109 columns]"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc_matrix2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "vectorizer_bi2 = CountVectorizer(ngram_range=(1,2),binary=True,stop_words=stop_words,min_df=0.1,max_df=0.7)\n",
    "\n",
    "doc_matrix_bi2 = vectorizer_bi2.fit_transform(spotify.track_lyric)\n",
    "\n",
    "bi_doc_matrix2 = pd.DataFrame(doc_matrix_bi2.toarray(), columns=vectorizer_bi2.get_feature_names())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LSA 2 (after removing additional stop words)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance: [0.0450345  0.08486639 0.06568329 0.05213635]\n",
      "\n",
      "Topic  0\n",
      "time, love, stay, baby, say, think, look, bad, said, right, come, yeah, let, good, night\n",
      "\n",
      "Topic  1\n",
      "stay, time, think, hey, best, hand, let, people, wanted, took, hold, times, leave, turn, try\n",
      "\n",
      "Topic  2\n",
      "look, stay, time, dreams, home, girl, honey, phone, world, think, thing, bad, old, fine, place\n",
      "\n",
      "Topic  3\n",
      "time, life, right, ain, better, night, baby, new, best, long, heart, times, eyes, sorry, door\n",
      "None\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>ain</th>\n",
       "      <th>baby</th>\n",
       "      <th>bad</th>\n",
       "      <th>beautiful</th>\n",
       "      <th>best</th>\n",
       "      <th>better</th>\n",
       "      <th>blue</th>\n",
       "      <th>break</th>\n",
       "      <th>car</th>\n",
       "      <th>change</th>\n",
       "      <th>...</th>\n",
       "      <th>trying</th>\n",
       "      <th>turn</th>\n",
       "      <th>waiting</th>\n",
       "      <th>walk</th>\n",
       "      <th>wanted</th>\n",
       "      <th>won</th>\n",
       "      <th>world</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yeah</th>\n",
       "      <th>years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.057</td>\n",
       "      <td>0.256</td>\n",
       "      <td>0.172</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.074</td>\n",
       "      <td>0.086</td>\n",
       "      <td>0.032</td>\n",
       "      <td>0.081</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.024</td>\n",
       "      <td>...</td>\n",
       "      <td>0.024</td>\n",
       "      <td>0.037</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.023</td>\n",
       "      <td>0.027</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.031</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.016</td>\n",
       "      <td>-0.142</td>\n",
       "      <td>-0.156</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.053</td>\n",
       "      <td>-0.046</td>\n",
       "      <td>-0.012</td>\n",
       "      <td>-0.010</td>\n",
       "      <td>-0.016</td>\n",
       "      <td>-0.013</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.014</td>\n",
       "      <td>0.006</td>\n",
       "      <td>-0.033</td>\n",
       "      <td>-0.008</td>\n",
       "      <td>0.015</td>\n",
       "      <td>-0.003</td>\n",
       "      <td>-0.040</td>\n",
       "      <td>-0.014</td>\n",
       "      <td>-0.090</td>\n",
       "      <td>-0.004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.034</td>\n",
       "      <td>-0.153</td>\n",
       "      <td>0.012</td>\n",
       "      <td>-0.018</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.004</td>\n",
       "      <td>-0.010</td>\n",
       "      <td>-0.054</td>\n",
       "      <td>-0.031</td>\n",
       "      <td>-0.014</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.016</td>\n",
       "      <td>0.002</td>\n",
       "      <td>-0.075</td>\n",
       "      <td>-0.013</td>\n",
       "      <td>-0.004</td>\n",
       "      <td>-0.015</td>\n",
       "      <td>0.017</td>\n",
       "      <td>-0.016</td>\n",
       "      <td>-0.055</td>\n",
       "      <td>-0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.087</td>\n",
       "      <td>0.064</td>\n",
       "      <td>-0.206</td>\n",
       "      <td>-0.011</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.084</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.029</td>\n",
       "      <td>-0.019</td>\n",
       "      <td>0.013</td>\n",
       "      <td>...</td>\n",
       "      <td>0.007</td>\n",
       "      <td>0.023</td>\n",
       "      <td>0.020</td>\n",
       "      <td>0.012</td>\n",
       "      <td>-0.006</td>\n",
       "      <td>0.019</td>\n",
       "      <td>-0.034</td>\n",
       "      <td>0.007</td>\n",
       "      <td>0.020</td>\n",
       "      <td>0.010</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 109 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     ain   baby    bad  beautiful   best  better   blue  break    car  change  \\\n",
       "0  0.057  0.256  0.172      0.047  0.074   0.086  0.032  0.081  0.048   0.024   \n",
       "1 -0.016 -0.142 -0.156      0.001  0.053  -0.046 -0.012 -0.010 -0.016  -0.013   \n",
       "2 -0.034 -0.153  0.012     -0.018  0.005   0.004 -0.010 -0.054 -0.031  -0.014   \n",
       "3  0.087  0.064 -0.206     -0.011  0.054   0.084  0.002  0.029 -0.019   0.013   \n",
       "\n",
       "   ...  trying   turn  waiting   walk  wanted    won  world  wrong   yeah  \\\n",
       "0  ...   0.024  0.037    0.055  0.023   0.027  0.045  0.047  0.031  0.135   \n",
       "1  ...  -0.014  0.006   -0.033 -0.008   0.015 -0.003 -0.040 -0.014 -0.090   \n",
       "2  ...  -0.016  0.002   -0.075 -0.013  -0.004 -0.015  0.017 -0.016 -0.055   \n",
       "3  ...   0.007  0.023    0.020  0.012  -0.006  0.019 -0.034  0.007  0.020   \n",
       "\n",
       "   years  \n",
       "0  0.014  \n",
       "1 -0.004  \n",
       "2 -0.002  \n",
       "3  0.010  \n",
       "\n",
       "[4 rows x 109 columns]"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lsa_function(4,doc_matrix2,vectorizer2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance: [0.04503452 0.08486836 0.06568804 0.05213944 0.0471926 ]\n",
      "\n",
      "Topic  0\n",
      "time, love, stay, baby, say, think, look, bad, said, right, come, yeah, let, good, night\n",
      "\n",
      "Topic  1\n",
      "stay, time, think, hey, best, hand, let, people, wanted, took, hold, times, leave, turn, try\n",
      "\n",
      "Topic  2\n",
      "look, stay, time, dreams, home, girl, honey, phone, world, think, thing, bad, old, fine, place\n",
      "\n",
      "Topic  3\n",
      "time, life, right, ain, better, night, baby, best, long, new, heart, times, eyes, hands, door\n",
      "\n",
      "Topic  4\n",
      "new, love, waiting, look, stay, lights, beautiful, hold, end, let, life, home, day, heard, high\n",
      "None\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>ain</th>\n",
       "      <th>baby</th>\n",
       "      <th>bad</th>\n",
       "      <th>beautiful</th>\n",
       "      <th>best</th>\n",
       "      <th>better</th>\n",
       "      <th>blue</th>\n",
       "      <th>break</th>\n",
       "      <th>car</th>\n",
       "      <th>change</th>\n",
       "      <th>...</th>\n",
       "      <th>trying</th>\n",
       "      <th>turn</th>\n",
       "      <th>waiting</th>\n",
       "      <th>walk</th>\n",
       "      <th>wanted</th>\n",
       "      <th>won</th>\n",
       "      <th>world</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yeah</th>\n",
       "      <th>years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.057</td>\n",
       "      <td>0.256</td>\n",
       "      <td>0.172</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.074</td>\n",
       "      <td>0.086</td>\n",
       "      <td>0.032</td>\n",
       "      <td>0.081</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.024</td>\n",
       "      <td>...</td>\n",
       "      <td>0.024</td>\n",
       "      <td>0.037</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.023</td>\n",
       "      <td>0.027</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.031</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.016</td>\n",
       "      <td>-0.142</td>\n",
       "      <td>-0.156</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.053</td>\n",
       "      <td>-0.046</td>\n",
       "      <td>-0.012</td>\n",
       "      <td>-0.010</td>\n",
       "      <td>-0.016</td>\n",
       "      <td>-0.013</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.014</td>\n",
       "      <td>0.006</td>\n",
       "      <td>-0.033</td>\n",
       "      <td>-0.008</td>\n",
       "      <td>0.015</td>\n",
       "      <td>-0.003</td>\n",
       "      <td>-0.040</td>\n",
       "      <td>-0.014</td>\n",
       "      <td>-0.090</td>\n",
       "      <td>-0.004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.034</td>\n",
       "      <td>-0.153</td>\n",
       "      <td>0.011</td>\n",
       "      <td>-0.018</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.003</td>\n",
       "      <td>-0.010</td>\n",
       "      <td>-0.054</td>\n",
       "      <td>-0.030</td>\n",
       "      <td>-0.014</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.016</td>\n",
       "      <td>0.002</td>\n",
       "      <td>-0.075</td>\n",
       "      <td>-0.013</td>\n",
       "      <td>-0.004</td>\n",
       "      <td>-0.015</td>\n",
       "      <td>0.017</td>\n",
       "      <td>-0.016</td>\n",
       "      <td>-0.055</td>\n",
       "      <td>-0.002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.087</td>\n",
       "      <td>0.064</td>\n",
       "      <td>-0.206</td>\n",
       "      <td>-0.012</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.083</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.025</td>\n",
       "      <td>-0.025</td>\n",
       "      <td>0.012</td>\n",
       "      <td>...</td>\n",
       "      <td>0.007</td>\n",
       "      <td>0.023</td>\n",
       "      <td>0.020</td>\n",
       "      <td>0.012</td>\n",
       "      <td>-0.006</td>\n",
       "      <td>0.019</td>\n",
       "      <td>-0.034</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.017</td>\n",
       "      <td>0.010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.008</td>\n",
       "      <td>-0.062</td>\n",
       "      <td>-0.091</td>\n",
       "      <td>0.070</td>\n",
       "      <td>-0.004</td>\n",
       "      <td>-0.077</td>\n",
       "      <td>-0.032</td>\n",
       "      <td>-0.078</td>\n",
       "      <td>-0.115</td>\n",
       "      <td>0.015</td>\n",
       "      <td>...</td>\n",
       "      <td>0.014</td>\n",
       "      <td>-0.017</td>\n",
       "      <td>0.277</td>\n",
       "      <td>-0.008</td>\n",
       "      <td>0.018</td>\n",
       "      <td>-0.035</td>\n",
       "      <td>-0.011</td>\n",
       "      <td>-0.018</td>\n",
       "      <td>0.009</td>\n",
       "      <td>-0.001</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 109 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     ain   baby    bad  beautiful   best  better   blue  break    car  change  \\\n",
       "0  0.057  0.256  0.172      0.047  0.074   0.086  0.032  0.081  0.048   0.024   \n",
       "1 -0.016 -0.142 -0.156      0.001  0.053  -0.046 -0.012 -0.010 -0.016  -0.013   \n",
       "2 -0.034 -0.153  0.011     -0.018  0.005   0.003 -0.010 -0.054 -0.030  -0.014   \n",
       "3  0.087  0.064 -0.206     -0.012  0.055   0.083  0.002  0.025 -0.025   0.012   \n",
       "4 -0.008 -0.062 -0.091      0.070 -0.004  -0.077 -0.032 -0.078 -0.115   0.015   \n",
       "\n",
       "   ...  trying   turn  waiting   walk  wanted    won  world  wrong   yeah  \\\n",
       "0  ...   0.024  0.037    0.055  0.023   0.027  0.045  0.047  0.031  0.135   \n",
       "1  ...  -0.014  0.006   -0.033 -0.008   0.015 -0.003 -0.040 -0.014 -0.090   \n",
       "2  ...  -0.016  0.002   -0.075 -0.013  -0.004 -0.015  0.017 -0.016 -0.055   \n",
       "3  ...   0.007  0.023    0.020  0.012  -0.006  0.019 -0.034  0.008  0.017   \n",
       "4  ...   0.014 -0.017    0.277 -0.008   0.018 -0.035 -0.011 -0.018  0.009   \n",
       "\n",
       "   years  \n",
       "0  0.014  \n",
       "1 -0.004  \n",
       "2 -0.002  \n",
       "3  0.010  \n",
       "4 -0.001  \n",
       "\n",
       "[5 rows x 109 columns]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lsa_function(5,doc_matrix2,vectorizer2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "# LSA 2 with Bi-grams"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance: [0.01965149 0.03778659 0.03492571 0.03317169 0.03141722]\n",
      "\n",
      "Topic  0\n",
      "say, time, love, said, night, think, baby, eyes, right, tell, let, come, yeah, look, mind\n",
      "\n",
      "Topic  1\n",
      "night, time, remember, best, things, said, walk, street, place, change, dreams, kiss, life, years, waiting\n",
      "\n",
      "Topic  2\n",
      "home, yeah, gone, thing, said, town, night, eyes, right, long, little, car, won, turn, street\n",
      "\n",
      "Topic  3\n",
      "baby, say, looking, didn, think, did, door, try, waiting, thought, thinking, thing, going, phone, tell\n",
      "\n",
      "Topic  4\n",
      "little, say, eyes, come, hold, walk, years, feel, friends, people, hand, really, face, baby, yeah\n",
      "None\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>ain</th>\n",
       "      <th>baby</th>\n",
       "      <th>bad</th>\n",
       "      <th>beautiful</th>\n",
       "      <th>best</th>\n",
       "      <th>better</th>\n",
       "      <th>blue</th>\n",
       "      <th>break</th>\n",
       "      <th>car</th>\n",
       "      <th>change</th>\n",
       "      <th>...</th>\n",
       "      <th>trying</th>\n",
       "      <th>turn</th>\n",
       "      <th>waiting</th>\n",
       "      <th>walk</th>\n",
       "      <th>wanted</th>\n",
       "      <th>won</th>\n",
       "      <th>world</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yeah</th>\n",
       "      <th>years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.074</td>\n",
       "      <td>0.159</td>\n",
       "      <td>0.083</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.081</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.069</td>\n",
       "      <td>0.075</td>\n",
       "      <td>0.063</td>\n",
       "      <td>0.061</td>\n",
       "      <td>...</td>\n",
       "      <td>0.058</td>\n",
       "      <td>0.069</td>\n",
       "      <td>0.061</td>\n",
       "      <td>0.074</td>\n",
       "      <td>0.065</td>\n",
       "      <td>0.072</td>\n",
       "      <td>0.077</td>\n",
       "      <td>0.069</td>\n",
       "      <td>0.146</td>\n",
       "      <td>0.052</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.017</td>\n",
       "      <td>-0.058</td>\n",
       "      <td>-0.175</td>\n",
       "      <td>-0.034</td>\n",
       "      <td>0.139</td>\n",
       "      <td>-0.062</td>\n",
       "      <td>-0.048</td>\n",
       "      <td>0.056</td>\n",
       "      <td>0.073</td>\n",
       "      <td>0.119</td>\n",
       "      <td>...</td>\n",
       "      <td>0.090</td>\n",
       "      <td>-0.035</td>\n",
       "      <td>0.106</td>\n",
       "      <td>0.127</td>\n",
       "      <td>0.084</td>\n",
       "      <td>-0.118</td>\n",
       "      <td>-0.057</td>\n",
       "      <td>-0.079</td>\n",
       "      <td>-0.111</td>\n",
       "      <td>0.108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.119</td>\n",
       "      <td>-0.083</td>\n",
       "      <td>-0.102</td>\n",
       "      <td>-0.063</td>\n",
       "      <td>-0.073</td>\n",
       "      <td>-0.015</td>\n",
       "      <td>0.032</td>\n",
       "      <td>0.025</td>\n",
       "      <td>0.117</td>\n",
       "      <td>-0.091</td>\n",
       "      <td>...</td>\n",
       "      <td>0.097</td>\n",
       "      <td>0.108</td>\n",
       "      <td>-0.072</td>\n",
       "      <td>0.017</td>\n",
       "      <td>-0.147</td>\n",
       "      <td>0.114</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.092</td>\n",
       "      <td>0.218</td>\n",
       "      <td>0.022</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.021</td>\n",
       "      <td>0.314</td>\n",
       "      <td>-0.081</td>\n",
       "      <td>-0.091</td>\n",
       "      <td>-0.087</td>\n",
       "      <td>0.067</td>\n",
       "      <td>-0.206</td>\n",
       "      <td>0.018</td>\n",
       "      <td>0.004</td>\n",
       "      <td>-0.005</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.049</td>\n",
       "      <td>-0.030</td>\n",
       "      <td>0.133</td>\n",
       "      <td>0.065</td>\n",
       "      <td>0.007</td>\n",
       "      <td>0.038</td>\n",
       "      <td>-0.085</td>\n",
       "      <td>0.044</td>\n",
       "      <td>-0.050</td>\n",
       "      <td>-0.078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.011</td>\n",
       "      <td>0.097</td>\n",
       "      <td>-0.010</td>\n",
       "      <td>0.012</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.015</td>\n",
       "      <td>-0.033</td>\n",
       "      <td>-0.009</td>\n",
       "      <td>0.026</td>\n",
       "      <td>0.045</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.015</td>\n",
       "      <td>-0.170</td>\n",
       "      <td>-0.006</td>\n",
       "      <td>0.147</td>\n",
       "      <td>0.082</td>\n",
       "      <td>-0.096</td>\n",
       "      <td>0.013</td>\n",
       "      <td>0.070</td>\n",
       "      <td>0.096</td>\n",
       "      <td>0.143</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 110 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     ain   baby    bad  beautiful   best  better   blue  break    car  change  \\\n",
       "0  0.074  0.159  0.083      0.051  0.081   0.090  0.069  0.075  0.063   0.061   \n",
       "1  0.017 -0.058 -0.175     -0.034  0.139  -0.062 -0.048  0.056  0.073   0.119   \n",
       "2 -0.119 -0.083 -0.102     -0.063 -0.073  -0.015  0.032  0.025  0.117  -0.091   \n",
       "3  0.021  0.314 -0.081     -0.091 -0.087   0.067 -0.206  0.018  0.004  -0.005   \n",
       "4  0.011  0.097 -0.010      0.012  0.005   0.015 -0.033 -0.009  0.026   0.045   \n",
       "\n",
       "   ...  trying   turn  waiting   walk  wanted    won  world  wrong   yeah  \\\n",
       "0  ...   0.058  0.069    0.061  0.074   0.065  0.072  0.077  0.069  0.146   \n",
       "1  ...   0.090 -0.035    0.106  0.127   0.084 -0.118 -0.057 -0.079 -0.111   \n",
       "2  ...   0.097  0.108   -0.072  0.017  -0.147  0.114  0.047  0.092  0.218   \n",
       "3  ...  -0.049 -0.030    0.133  0.065   0.007  0.038 -0.085  0.044 -0.050   \n",
       "4  ...  -0.015 -0.170   -0.006  0.147   0.082 -0.096  0.013  0.070  0.096   \n",
       "\n",
       "   years  \n",
       "0  0.052  \n",
       "1  0.108  \n",
       "2  0.022  \n",
       "3 -0.078  \n",
       "4  0.143  \n",
       "\n",
       "[5 rows x 110 columns]"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lsa_function(5,doc_matrix_bi2,vectorizer_bi2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# NMF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "def nmf_function(num_components, doc_text_matrix, vectorizer):\n",
    "    nmf = NMF(num_components)\n",
    "    doc_topic = nmf.fit_transform(doc_text_matrix)\n",
    "    \n",
    "    index = []\n",
    "    for i in range(num_components):\n",
    "        index.append(i)\n",
    "    topic_word = pd.DataFrame(nmf.components_.round(3),\n",
    "             index = index,\n",
    "             columns = vectorizer.get_feature_names())\n",
    "    \n",
    "    print(display_topics(nmf, vectorizer.get_feature_names(), 15))\n",
    "    \n",
    "#     return topic_word"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "love, baby, look, time, say, bad, think, good, said, come, yeah, right, tell, feel, did\n",
      "\n",
      "Topic  1\n",
      "stay, time, think, hey, let, say, best, hand, people, said, night, leave, love, little, hold\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "nmf_function(2,doc_matrix2,vectorizer2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "love, baby, time, say, bad, good, think, said, come, yeah, right, tell, did, feel, let\n",
      "\n",
      "Topic  1\n",
      "stay, time, think, hey, let, best, say, hand, people, said, leave, night, little, hold, took\n",
      "\n",
      "Topic  2\n",
      "look, bad, time, home, girl, dreams, honey, world, think, phone, better, thing, hey, come, day\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "nmf_function(3,doc_matrix2,vectorizer2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "love, baby, say, bad, good, think, come, said, feel, yeah, did, right, tell, new, let\n",
      "\n",
      "Topic  1\n",
      "stay, think, hey, say, let, hand, people, best, said, love, little, wanted, beautiful, took, hold\n",
      "\n",
      "Topic  2\n",
      "look, bad, home, girl, dreams, time, honey, world, think, phone, thing, better, hey, come, place\n",
      "\n",
      "Topic  3\n",
      "time, right, life, better, baby, ain, best, break, night, heart, hands, times, yeah, door, eyes\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "nmf_function(4,doc_matrix2,vectorizer2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "love, beautiful, lights, let, home, gone, right, life, end, tell, long, little, hands, knew, high\n",
      "\n",
      "Topic  1\n",
      "stay, hey, say, let, think, hand, people, best, love, beautiful, wanted, hold, took, said, leave\n",
      "\n",
      "Topic  2\n",
      "look, bad, dreams, home, time, honey, world, phone, thing, think, hey, feel, day, place, little\n",
      "\n",
      "Topic  3\n",
      "time, right, life, heart, best, ain, break, times, night, hands, door, let, eyes, better, long\n",
      "\n",
      "Topic  4\n",
      "new, waiting, lights, love, hold, took, heart, night, change, day, wanted, heard, better, let, lost\n",
      "\n",
      "Topic  5\n",
      "girl, say, yeah, forever, better, tell, end, hands, home, said, feeling, door, lost, need, make\n",
      "\n",
      "Topic  6\n",
      "baby, bad, hey, play, break, really, did, life, won, things, ain, think, right, sorry, let\n",
      "\n",
      "Topic  7\n",
      "think, car, said, night, things, little, blue, better, head, dress, place, mind, old, gone, met\n",
      "\n",
      "Topic  8\n",
      "good, light, feel, bad, say, did, world, car, make, remember, love, tell, heart, looking, right\n",
      "\n",
      "Topic  9\n",
      "come, said, right, knew, forget, long, day, rain, things, thought, night, sorry, gone, baby, remember\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "nmf_function(10,doc_matrix2,vectorizer2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Try all of the Above with TF-IDF Vectorizer instead of Count Vectorizer:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "cv_tfidf = TfidfVectorizer(stop_words=stop_words,min_df=0.1,max_df=0.7)\n",
    "x_tfidf = cv_tfidf.fit_transform(spotify.track_lyric).toarray()\n",
    "df_tfidf = pd.DataFrame(x_tfidf,columns=cv_tfidf.get_feature_names())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "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>ain</th>\n",
       "      <th>baby</th>\n",
       "      <th>bad</th>\n",
       "      <th>beautiful</th>\n",
       "      <th>best</th>\n",
       "      <th>better</th>\n",
       "      <th>blue</th>\n",
       "      <th>break</th>\n",
       "      <th>car</th>\n",
       "      <th>change</th>\n",
       "      <th>...</th>\n",
       "      <th>trying</th>\n",
       "      <th>turn</th>\n",
       "      <th>waiting</th>\n",
       "      <th>walk</th>\n",
       "      <th>wanted</th>\n",
       "      <th>won</th>\n",
       "      <th>world</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yeah</th>\n",
       "      <th>years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.264438</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.055182</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.123388</td>\n",
       "      <td>0.403619</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.080724</td>\n",
       "      <td>0.077898</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.339606</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.141475</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.152281</td>\n",
       "      <td>0.497438</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.162098</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.304561</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.095507</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.498249</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.280844</td>\n",
       "      <td>0.28652</td>\n",
       "      <td>0.071603</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.143627</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.287255</td>\n",
       "      <td>0.1522</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</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",
       "      <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>106</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.645764</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.293275</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.095822</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.096601</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.133942</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.200046</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.200046</td>\n",
       "      <td>0.579130</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.415574</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.231696</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.219128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>0.171212</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>111 rows × 109 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          ain      baby       bad  beautiful      best    better      blue  \\\n",
       "0    0.000000  0.000000  0.000000   0.000000  0.000000  0.000000  0.264438   \n",
       "1    0.000000  0.123388  0.403619   0.000000  0.080724  0.077898  0.000000   \n",
       "2    0.000000  0.000000  0.000000   0.339606  0.000000  0.141475  0.000000   \n",
       "3    0.000000  0.000000  0.000000   0.000000  0.000000  0.000000  0.095507   \n",
       "4    0.000000  0.000000  0.000000   0.000000  0.143627  0.000000  0.000000   \n",
       "..        ...       ...       ...        ...       ...       ...       ...   \n",
       "106  0.000000  0.000000  0.645764   0.000000  0.000000  0.000000  0.293275   \n",
       "107  0.000000  0.096601  0.000000   0.000000  0.000000  0.000000  0.133942   \n",
       "108  0.000000  0.000000  0.200046   0.000000  0.200046  0.579130  0.000000   \n",
       "109  0.000000  0.000000  0.000000   0.000000  0.000000  0.000000  0.000000   \n",
       "110  0.171212  0.000000  0.000000   0.000000  0.000000  0.000000  0.000000   \n",
       "\n",
       "        break       car  change  ...    trying      turn  waiting      walk  \\\n",
       "0    0.000000  0.000000     0.0  ...  0.000000  0.055182      0.0  0.000000   \n",
       "1    0.000000  0.000000     0.0  ...  0.000000  0.000000      0.0  0.000000   \n",
       "2    0.152281  0.497438     0.0  ...  0.000000  0.162098      0.0  0.000000   \n",
       "3    0.000000  0.000000     0.0  ...  0.498249  0.000000      0.0  0.000000   \n",
       "4    0.000000  0.000000     0.0  ...  0.000000  0.000000      0.0  0.287255   \n",
       "..        ...       ...     ...  ...       ...       ...      ...       ...   \n",
       "106  0.000000  0.000000     0.0  ...  0.000000  0.000000      0.0  0.000000   \n",
       "107  0.000000  0.000000     0.0  ...  0.000000  0.000000      0.0  0.000000   \n",
       "108  0.000000  0.000000     0.0  ...  0.000000  0.000000      0.0  0.000000   \n",
       "109  0.000000  0.000000     0.0  ...  0.000000  0.000000      0.0  0.000000   \n",
       "110  0.000000  0.000000     0.0  ...  0.000000  0.000000      0.0  0.000000   \n",
       "\n",
       "     wanted       won     world    wrong      yeah     years  \n",
       "0    0.0000  0.000000  0.000000  0.00000  0.000000  0.000000  \n",
       "1    0.0000  0.000000  0.000000  0.00000  0.000000  0.000000  \n",
       "2    0.0000  0.304561  0.000000  0.00000  0.000000  0.000000  \n",
       "3    0.0000  0.000000  0.280844  0.28652  0.071603  0.000000  \n",
       "4    0.1522  0.000000  0.000000  0.00000  0.000000  0.000000  \n",
       "..      ...       ...       ...      ...       ...       ...  \n",
       "106  0.0000  0.000000  0.095822  0.00000  0.000000  0.000000  \n",
       "107  0.0000  0.000000  0.000000  0.00000  0.000000  0.000000  \n",
       "108  0.0000  0.415574  0.000000  0.00000  0.000000  0.231696  \n",
       "109  0.0000  0.000000  0.000000  0.00000  0.000000  0.219128  \n",
       "110  0.0000  0.000000  0.000000  0.00000  0.000000  0.000000  \n",
       "\n",
       "[111 rows x 109 columns]"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_tfidf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Scree Plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Scree plot TF-IDF')"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZMAAAEWCAYAAACjYXoKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXxU1f3/8dc7GyGQDRIgLAEFXBBkMbKorXsLtorWuuDaVW21+2Z3u31/rS21m5WqtUWroNZWqWLVumCrsioim4oou5ARgYQlG5/fH/cGxnGSTJbJJOTzfDzmMXfuPffOZy5DPnPuueccmRnOOedca6SlOgDnnHOdnycT55xzrebJxDnnXKt5MnHOOddqnkycc861micT55xzrebJxLl2IuktSWekOg7nksGTieuUJJ0k6XlJOyVtl/ScpONTHVdbkHSKpI2NbH9UUmX4qJFUHfV6Rrj//qh1lZL+1cCxhkgySRnh67+Gx6sIH8sl/T9J+VH7fEJSXczx/9D2Z8J1JhmpDsC55pKUBzwMfA64D8gCPgBUNfM4GWZW2/YRJpeZTalflvRXYKOZfS9q3SnAZjMb2MK3uNHMvicpGxgF3Ag8J2mCme0Oy7xgZie18PjuEOQ1E9cZHQFgZrPMrM7M9prZ42a2rL6ApM9KWhX+ul4paVy4/i1J35K0DNgtKUPSxLCWs0PSy+Ef4/rj5Ev6s6QtkjZJ+qmk9HhBSbpB0t8l3Ru+74uSRjdQtpuk30jaHD5+E67rATwK9I/61d+/7U5d4sxsn5ktAs4BegOfTEUcrnPwZOI6o9eAOkkzJU2RVBi9UdIFwA3AFUAewR/Dd6KKTAM+AhQAfYFHgJ8CvYCvAw9IKg7LzgRqgWHAWOBDwGcaiW0qcH94rHuAByVlxin3XWAiMAYYDYwHvhf+8p9CULPoGT42N3lGksjMKoAnCGp/zsXlycR1Oma2CzgJMOA2oFzSHEl9wyKfIbhUs8gCa8xsXdQhfmdmG8xsL3AZMNfM5prZfjN7AlgMnBUebwrwZTPbbWbbgJuAixsJb4mZ/d3MaoBfA9kESSPWpcCPzWybmZUDPwIub+Epiad/WNOqf1zYyuNtJkiQ9SbGHD/eZ3RdiLeZuE7JzFYBnwCQdBTwN+A3BLWOQcAbjey+IWp5MHCBpLOj1mUCT4fbMoEtkuq3pcXs3+CxzWx/2JAe7zJVfyA6wa1roFxLxW0zkVQZ9XJEM443ANge9Xq+t5m4aJ5MXKdnZqvDhuirw1UbgKGN7RK1vAG4y8w+G1tIUglBo35RMxrqB0XtnwYMJPhVH2szQbJaEb4ujSqXtKG8zaxn9GtJQ5raR1JP4AzgZ8mJyh0K/DKX63QkHSXpa5IGhq8HEdRI5odFbge+Luk4BYZJGtzA4f4GnC3pw5LSJWWHt9YONLMtwOPAdEl5ktIkDZV0ciPhHSfpY+Gttl8mSEbz45SbBXxPUrGkIuAHYSwAW4He0bfjpkJ4Q8BxwIPAu8BfUhmP69g8mbjOqAKYACyQtJvgj/Vy4GsAZnY/wa/oe8KyD/Le6/0HmNkGgkbz7wDlBDWVb3Dw/8YVBLceryT4g/p3oKSR2B4CLgrLXg58LGw/ifVTgraZZcArwIvhOsxsNUGyWRu2R7T33VzflFRBcFnrTmAJcELUbcHOvY98cizn2oakG4BhZnZZqmNxrr15zcQ551yreTJxzjnXan6ZyznnXKt5zcQ551yrdYl+JkVFRTZkyJBUh+Gcc53KkiVLImZW3HTJLpJMhgwZwuLFi1MdhnPOdSqS1jVdKuCXuZxzzrWaJxPnnHOt5snEOedcq3kycc4512qeTJxzzrWaJxPnnHOt5snEOedcqyU1mUiaLOlVSWskXR9nuyT9Lty+TNK4cH22pIWSXpa0QtKPova5QdImSUvDx1nJ/AxtYeXmXdy/eAM798Qbidw55zq/pHValJQO3AycCWwEFkmaY2Yro4pNAYaHjwnALeFzFXCamVVKygT+J+lRM6ufZOgmM/tVsmJva9Mff5UnV2/ju+nLOf3oPpw3dgCnHNmHrAyvGDrnDg3J7AE/HlhjZmsBJM0mmIQoOplMBe60YLTJ+ZIKJJWEM9zVz1WdGT467YiUWyv2MWpAPmVDCvnXy5t5dPnbFOZkcvbo/pw3dgBjBhUQNce4c851Osn8aTyAYNa6ehvDdQmVCadQXQpsA54wswVR5a4LL4vdIakw3ptLukrSYkmLy8vLW/tZWiVSUc2R/XL54dnH8MK3T+eOT5Rx4rAi7l20gfP++DynT5/H7598nQ3b96Q0Tueca6lkJpN4P7VjaxcNljGzOjMbAwwExksaGW6/BRgKjAG2ANPjvbmZ3WpmZWZWVlyc0DhlSWFmvLO7iuLcbgBkpqdx2lF9+cMl41j0vTO48fxjKc7txvQnXuMDNz7NhTNeYNbC9ezc6+0rzrnOI5mXuTYCg6JeDwQ2N7eMme2Q9AwwGVhuZlvrt0m6DXi4DWNuczv31lBTZxT17Pa+bXnZmVx4/CAuPH4QG9/dw0NLN/PAixv59j9e4YdzVnDm0X05b+wATj6ymMx0b19xznVcyUwmi4Dhkg4DNgEXA5fElJlDcMlqNkHD+04z2yKpGKgJE0l34AzgFwBRbSoA5wHLk/gZWi1SWQVAUc+sRssNLMzh2lOH8flThrJs407++dIm5ry8mUde2UKvHlmcfWwJHxs3kGMH5nv7inOuw0laMjGzWknXAY8B6cAdZrZC0jXh9hnAXOAsYA2wB/hkuHsJMDO8IywNuM/M6msgN0oaQ3A57C3g6mR9hrawrSJIJsVxaibxSGL0oAJGDyrgux85mnmvlvPPlzYxa9EGZr6wjkmH9+bmS8fRq0fjyck559pTUuczMbO5BAkjet2MqGUDro2z3zJgbAPHvLyNw0yqSGU1AEW5iSWTaJnpaZwxoi9njOjLzr01PLBkIz//92rOvfk5br+yjCP65rZ1uM451yJ+IT7JIhX1l7man0yi5XfP5FMnHca9V01kb00dH/vj8zy1emvTOzrnXDvwZJJkkcoqMtJEQffMNjne2NJCHrr2RAb3zuHTMxdz27NrCSp4zjmXOp5MkixSWUXvnlmkpbVdo3n/gu7cf80kJh/Tj5/NXcU3/76Mqtq6Nju+c841lyeTJItUVrf6Elc8OVkZ3HzJOL542jDuX7KRy25fwDvhnWPOOdfePJkkWXlFVVKSCUBamvjqh47k99PGsmzjTs75w3OsfntXUt7LOeca48kkySKVyUsm9c4e3Z/7rp5ETd1+zv/j8/xnpTfMO+falyeTJDIz3qmsPjCUSjKNHlTAnOtO4vDinnz2rsXMmPeGN8w759qNJ5Mk2rW3luq6/U32fm8r/fKzue/qSZw1qoSfP7qar93/sjfMO+faRVI7LXZ15WGDeHvUTOp1z0rnD9PGckSfXG76z2u8FdnNny4va9cYnHNdj9dMkqi8jTosNpckvnTGcG6+ZBwrt+zi3JufY+Vmb5h3ziWPJ5MkOjjIY2pqBR85toT7rz6Buv3Gx2c8z2Mr3k5JHM65Q58nkySKpOAyV6xRA/OZc92JDO/Tk6vvWsLNT6/xhnnnXJvzZJJEkcoq0ttwKJWW6pOXzb1XT+Kc0f355WOv8pV7l7KvxhvmnXNtxxvgkyhSUU3vHm07lEpLZWem89uLx3BE35786vHX+PeKtxk9sICyIYWUDe7FuNJC8nNSm/Scc52XJ5Mkao8Oi80hietOG07ZkF48tuJtlqx7lxnz1lK3/w0AhvfpSdmQQsaVFlI2pBdDeuf4RFzOuYR4Mkmi8sqqFs1jkmwTD+/NxMN7A7CnupaXN+xkybrtLF73Lo8s28KshRsA6N0ji3GDCykbXEjZkEJGDsinW0Z6KkN3znVQnkySKFJRxbA+PVMdRqNysjKYNLQ3k4YGyWX/fmNNeSWL33qXJeveZcm67TwRDs+SlZ7GqIH5lA0uZNzgQo4bXNihal7OudTxZJIkZkaknYZSaUtpaeKIvrkc0TeXSyaUAkF/mRfXB8ll8Vvb+ctzb/GnZ9cCcNzgQqaNL+Ujo0ronuW1Fue6Kk8mSbJrXzCUSqJzv3dkxbnd+PAx/fjwMf0A2FdTx/JNO1nw5nYeWLKRr9//Mj/61wrOGzuAaeNLObokL8URO+famyeTJEl1h8Vkys5Mp2xIL8qG9OLzpwxl4ZvbmbVwPbMXbeDOF9YxZlABl4wv5aOjS8jJ8q+Yc12B/09PklQNpdLeJDHh8N5MOLw3N+yp5h8vbmLWwvV884Fl/PjhlUwd059p40sZOSA/1aE655LIk0mSHKiZ5LbPiMEdQUFOFp866TA+eeIQlqx7l1kLN/D3JRu5e8F6Rg3IZ9r4Us4Z05+e3fxr59yhxv9XJ0kkrJkcCm0mzSXpwGWwH3x0BA8uDWor3/nnK/z0kaC2cvHxpRw7MN/7sTh3iPBkkiSRymrS00RhTtepmcSTn5PJlScM4YpJg1m6YQezFq7nwZc2M2vhBkaU5DFtQilTx/QnL9t73zvXmfnYXEkSqayiVwcZSqUjkMTY0kJu/PhoFn73dH567kgAvv/gcib87EnueuGtlMbnnGsdr5kkSUcbSqUjyc3O5LKJg7l0QimvbNrJr594je8/tIJNO/bxzQ8f6QnYuU7IayZJUl5R1W7T9XZWkjh2YAG3X1HGZRNLmTHvDb5y31Kfati5TshrJkkSqaxmaHHHHkqlo8hIT+MnU0fSv6A7N/77VbbtqmLG5ceRn+Kh+51ziUtqzUTSZEmvSloj6fo42yXpd+H2ZZLGheuzJS2U9LKkFZJ+FLVPL0lPSHo9fC5M5mdoCTOjvLKq0w2lkkqS+Pwpw7jpotEsXredC2e8wOYde1MdlnMuQUlLJpLSgZuBKcAIYJqkETHFpgDDw8dVwC3h+irgNDMbDYwBJkuaGG67HnjSzIYDT4avO5SKqlqqa/d7m0kLnDd2IH/95Hg279jLx/74PKu2+Nz1znUGyayZjAfWmNlaM6sGZgNTY8pMBe60wHygQFJJ+LoyLJMZPixqn5nh8kzg3CR+hhap72PSlTostqUThxVx3zWTALhwxgs8tyaS4oicc01JZjIZAGyIer0xXJdQGUnpkpYC24AnzGxBWKavmW0BCJ/7JCH2VukqQ6kk09Elefzz2hPoX9CdT/xlIf98aWOqQ3LONSKZySTe/Z2WaBkzqzOzMcBAYLykkc16c+kqSYslLS4vL2/Orq0WqawGPJm0Vkl+d+67ZhJlg3vxlXtf5uan12AW+xVyznUEyUwmG4FBUa8HApubW8bMdgDPAJPDVVsllQCEz9vivbmZ3WpmZWZWVlxc3NLP0CL143J5A3zr5XfP5K+fOp6pY/rzy8de5XsPLqe2bn+qw3LOxUhmMlkEDJd0mKQs4GJgTkyZOcAV4V1dE4GdZrZFUrGkAgBJ3YEzgNVR+1wZLl8JPJTEz9Aikcoq0kSXH0qlrXTLSOemC8fwuVOGcveC9VzztyXsqa5NdVjOuShJSyZmVgtcBzwGrALuM7MVkq6RdE1YbC6wFlgD3AZ8PlxfAjwtaRlBUnrCzB4Ot/0cOFPS68CZ4esOJRhKpRvp3pO7zaSliW9NPoqfTD2Gp1ZvY9ptCw7UAJ1zqZfUTotmNpcgYUSvmxG1bMC1cfZbBoxt4JjvAKe3baRtq7yi2nu/J8nlk4bQNy+bL85+ifNveZ6ZnxzPkKIeqQ7LuS7Ph1NJAu+wmFwfOqYf93x2IhX7avnYLc/z0vp3Ux2Sc12eJ5MkiFRUdcl5TNrTuNJCHvjcCeRmZzDttvk8sXJrqkNyrkvzZNLGzCwYMdhrJkl3WFEPHvjcCRzZN5er71rMXfPXpTok57osTyZtrLKqlqra/d5m0k6KenZj1lUTOfXIPnz/weV8+KZn+c1/XuO1rRXeJ8W5duSjBrcx77DY/nKyMvjT5ccxa+F6/rVsC7998nV+85/XOby4B2eNLGHKqH6MKMnzKYKdSyJPJm3Mh1JJjYz0NC6fNITLJw1hW8U+HluxlX8v38It897gD0+vobRXDlNG9WPKyBJG+9zzzrU5TyZtrL7vgyeT1OmTm83lEwdz+cTBbN9dzRMr32buK2/z5/++yZ/mrWVAQXcmj+zHlJH9GFda6DM7OtcGPJm0MR9KpWPp1SOLi44v5aLjS9m5p4b/rNrKo8u3cNcL6/jz/96kb143Jh/Tj8kjSxh/WC/vaOpcCyWcTCT1MLPdyQzmUBCpCIZS6dXDG+A7mvycTM4/biDnHzeQin01PLV6G4++8jb3Lt7AzBfWUdQzizNH9GPyyH6MLS0gL9tnenQuUU0mE0knALcDPYFSSaOBq83s843v2TWVV1bTq0eW/8Lt4HKzM5k6ZgBTxwxgT3Utz7xaztxXtjBn6SZmLVwPwJDeORwzIJ+R/fMZOSCPkf3zKfQfCc7FlUjN5Cbgw4SDNJrZy5I+mNSoOrHyiipvL+lkcrIyOGtUCWeNKmFfTR0L3tzO8k07Wb5pJ8s27uCRZVsOlB1Q0J1j+ucxcsDBBNMnLzuF0TvXMSR0mcvMNsTc/VKXnHA6v0ilJ5POLDsznZOPKObkIw5OW7BjTzUrN+9i+eadLN8UPD+xaiv13ViKc7sxMkwwx4S1mAEF3f2OMdelJJJMNoSXuiwcSv6LBKMAuzgilVUc5gMPHlIKcrI4YVgRJwwrOrCusqqWVVt2hTWY4Hnea+Xst/p9MikbXMhPzx1Fv3yvubhDXyLJ5BrgtwTT6W4EHifOSL8uaigV7/1+yOvZLYPjh/Ti+CG9DqzbW13H6rd3sXzzLlZs2snDy7ZwwZ+e557PTGRQr5wURutc8jWZTMwsAlzaDrF0erur69hXs98vc3VR3bPSGVtayNjSQgAuHl/KlXcs5MI/vcDdn5nA4cU9Uxyhc8nT5NhckmbWz3oYvi6UdEdyw+qcIt773UUZM6iA2VdNpKZuPxf+aT6r396V6pCcS5pEBno8NpyHHQAze5cGJq7q6srre797h0UXOrokj9lXTSIjTVx863yWbdzR9E7OdUKJJJM0SYX1LyT1wnvOx3WwZuJtJu6gYX16cv81k8jNzuCS2xaw6K3tqQ7JuTaXSDKZDjwv6SeSfgI8D9yY3LA6Jx9KxTVkUK8c7rt6En3yunHFnxfyv9cjqQ7JuTbVZDIxszuBjwNbgW3Ax8zsrmQH1hmVV1YjQa8cr5m49yvJ7869V01icO8cPjVzEf/x2SHdISTRybFWA/8AHgIqJZUmL6TOK1JZRa+cLDLSfc4xF19xbjdmXzWRo/vlcs3flvDwss2pDsm5NpHI3VxfIKiVPAE8DDwSPrsYPpSKS0RBThZ/+8wExpYW8MVZL3H/4g2pDsm5VkukIf1LwJFm9k6yg+nsgrnf/RKXa1pudiYzPzWeq+9awjf+vox9NXVcPmlIqsNyrsUSuR6zAdiZ7EAOBZHKKoq9ZuISlJOVwW1XlHHG0X35/kMr+NO8N1IdknMtlkjNZC3wjKRHgKr6lWb266RF1UlFKqr9MpdrluzMdG65bBxfuXcp/+/R1eypruPLZwz3QSJdp5NIMlkfPrLCh4tjd1Ute2vqvMOia7bM9DR+e/FYumem89snX2dvTR3fnnKUJxTXqSQyNteP2iOQzs7nfnetkZ4mfnH+seRkpXPrs2vZU13Lj88Z6fPTu04jkZkWi4FvAscAB8bSNrPTkhhXp1Puvd9dK6WliRvOOYbuWRnMmPcGe6rruPH8Y/1Wc9cpJHKZ627gXuCjBMPRXwmUJzOozsh7v7u2IIlvTT6SHlnpTH/iNapq9nPTRWPIyvCE4jq2RJJJbzP7s6Qvmdk8YJ6keYkcXNJkgrlQ0oHbzeznMdsVbj8L2AN8wsxelDQIuBPoB+wHbjWz34b73AB8loMJ7TtmNjeReJKpvLIawO/mcq0miS+cPpzuWen89JFVbN21j9GDCsjLziSve0b4nEledkbwHC73yMrwy2IuZRJJJjXh8xZJHwE2AwOb2klSOnAzcCbBpFqLJM0xs5VRxaYAw8PHBOCW8LkW+FqYWHKBJZKeiNr3JjP7VQKxt5tIRVUwlEoPv8zl2sZnPnA4Pbtl8Pun1rBqy3p2Vzc+W3aagv4rBxJO1HJudialvbozdcwACv076pIgkWTyU0n5wNeA3wN5wFcS2G88sMbM1gJImg1MBaKTyVTgTjMzYL6kAkklZrYF2AJgZhWSVhHM9LiSDipSWUWhD6Xi2tjF40u5eHwwelFN3X4q99Wya18Nu/bWP9e873VFVJl17+wJy9RSWVXL/z26mo+MKuHSCaUcN7jQ7xhzbSaRu7nqh07ZCZzajGMPIOjwWG8jQa2jqTIDCBMJgKQhBPOnLIgqd52kK4DFBDWYd2PfXNJVwFUApaXJH0osGErFf/G55MlMT6OwR1aLaxavvl3BPQvW8Y8XN/HPlzZxRN+eXDphMOeNG0BedmYbR+u6mgZ/Rkv6Zvj8e0m/i30kcOx4P3msOWUk9QQeAL5sZvXT1N0CDAXGECSd6fHe3MxuNbMyMysrLi5OINzWCeZ+9/YS13Ed2S+XH00dyYLvns4vzh9FdmY6P5yzggk/e5Jv/X2ZT9zlWqWxmsmq8HlxC4+9ERgU9XogQXtLQmUkZRIkkrvN7B/1BczswLjdkm6jgww6GamsZmxpQdMFnUuxnKwMLjq+lIuOL2XZxh3cs2A9Dy3dzL2LNzBqQD6XTijlnDH9ycnyOfBc4hr8tpjZv8JG9JFm9o0WHHsRMFzSYcAm4GLgkpgycwguWc0muAS208y2hHd5/RlYFTtsS1SbCsB5wPIWxNbmvGbiOqNjBxZw7MACvvORo3nopU38bf56rv/HK/zskVWcN24Al0wo5ah+eakO03UCjf70MLM6Sce15MBmVivpOuAxgluD7zCzFZKuCbfPAOYS3Ba8huDW4E+Gu58IXA68ImlpuK7+FuAbJY0huBz2FnB1S+JrS3uqa9lTXefJxHVaedmZXD5pCJdNHMyL69/l7vnrmb1oA3e+sI6ywYVcOrGUKSNLyM5MT3WoroNScCNVIwWk6QS37t4P7K5fH33pqaMrKyuzxYtberWuaevf2cMHf/k0v/z4sVxQNqjpHZzrBN7dXc0DL27k7gXreTOym4KcTD4+biDnjh3AEX1zvSNlFyBpiZmVJVI2kYuivYB3gOjhU4xg5kUHlFfuA/BBHt0hpbBHFp/5wOF8+qTDeOGNd7h7wXr++vxb3P6/N8lMF8P75HJM/zxG9M9jREkeR/fP87vCurBEbg3+ZFNlurryCu/97g5dkjhhWBEnDCtiW8U+5q/dzsrNu1i5ZRdPv7qN+5dsPFB2UK/ujCjJ45j++YwoCRJNSX6292fpAhIZ6DEb+DTvH+jxU0mMq1PxcblcV9EnN5tzRvfnnNH9D6zbtmsfK7bsOpBgVm3exeMrt1J/Bb0gJzNILGFyOaZ/PocX9yDTO/geUhK5zHUXsBr4MPBj4FIO3jbsOJhMfCgV1xX1ycumT142px7Z58C63VW1rH67gpVRSeau+euoqt0PQFZGGqMG5HPisCI+MLyIMYMKPLl0cokkk2FmdoGkqWY2U9I9BHdouVAwlEqm/2dwLtSjWwbHDS7kuMGFB9bV1u3nzchuVm7ZxYrNu1j45nb+8NTr/O7J1+nZLYOJh/fipGFFnDS8mKHFPfzSWCfTnIEed0gaCbwNDElaRJ1QMJSKX+JyrjEZ6WkM75vL8L65TB0zAICde2p4YW2E/74ePP6zahsA/fOzOXFYEScNL+KkYUX09v9fHV4iyeRWSYXA9wk6GfYMl10oUlnt7SXOtUB+TiaTR5YweWQJENxm/9815fzv9QiPrXj7QOP+iJI8PnBEER8YVkzZkELv79IBNZhMJK0kmBhrdjiQ4jzg8PYKrDOJVFYxeqAPpeJca5X2zuHS3oO5dMJg6vYbr2zayf9eL+e/r0e4439v8qd5a+mWkcb4w+oviRVxZN9cH627A2isZjKNYAiUxyVFgFnAvVFDmbhQxC9zOdfm0tPEmEEFjBlUwHWnDWd3VS0L39weXhIr5/89uhoeDcrmZmdQkJNJQfcsCnIyye+eGed1Vrguk/xwXbcMr+G0lcbG5noZeBn4tqSJwEXAAklrgFlmdls7xdih7a2uY3d1HUW5fieXc8nUo1sGpx7Vh1OPCu4ae3vnPp5bE2HDu3vYsaeGnXtr2LGnmh17a9j07l52hK/3NzLIR05Wephcsvjg8CI+84HD/ZJ1CyU0LKiZzSeYvOoh4CbgD4AnEw7eFuw1E+faV7/8bM4/rvFJX/fvNyqra9m5p4Yde2rYsbc6fK5h556Dy1t37eO2/65l5gtvcemEwVz9wcPpk5fd6LHdeyXSafF4gkte5xMMrHgrwThdDthWEXZY9GTiXIeTlqYDUxgP6tV42TfKK7n56TX85bk3+dv8dUwbX8o1Jw+lX74nlUQ01gD/fwSXtt4FZgMnmtnGhsp3Vd773blDw9Dinvz6wjF88bTh3Pz0Gu6av457Fq7n4uMHcc3JQ+lf0D3VIXZojdVMqoApZvZaewXTGfllLucOLUOKevDLC0bzhdOG88dn1nDPgvXMXriBC8oG8rlThjKwMCfVIXZIDd5PZ2Y/8kTStEg4yGNvn//duUNKae8cfn7+sTzzjVO4oGwg9y3ewKm/eoZv/2MZG7bvSXV4HY7fnN1KkcoqCnwoFecOWQMLc/jZeaOY941TmTa+lAeWbOLUXz3DN//+Muve2d30AboI/wvYSj6UinNdQ/+C7vx46kie/eapXDZxMA8t3cxp0+fx1fuWsra8MtXhpVxjDfDjGtvRzF5s+3A6n0hlld/J5VwX0i8/mxvOOYbPnzKUW59dy98WrOPBlzZxzuj+XHfaMIb1yU11iCnRWAP89PA5Gygj6MAo4FhgAXBSckPrHCKVVYzyoVSc63L65GXzvY+O4OqTh3L7f9dy5wvreOjlzVx8/CC+NfkoCnK6VjtqYw3wp5rZqcA6YJyZlZnZccBYYE17BdjRRSqrKfLGd+e6rOLcbnz7rKP537dO5VMnHsZ9izdy+u1UZqQAABaqSURBVPR5/OPFjZg10v3+EJNIm8lRZvZK/QszWw6MSV5Ince+mjoqq2q9zcQ5R++e3fj+R0fw8BdOYnDvHL5638tMu20+a7Z1jfaURJLJKkm3SzpF0smSbsNnWgSCxnfw3u/OuYOOLsnj79ecwP+dN4qVm3cx5bfPMv3xV9lXU5fq0JIqkWTySWAF8CXgy8DKcF2XV17fYdEHeXTORUlLE5dMKOWpr5/C2cf25/dPreFDNz3LvNfKUx1a0jSZTMxsHzADuN7MzjOzm8J1XV7kQM3Ex+5xzr1fUc9u/PqiMdzzmQlkpIkr71jIdfe8yLZdh96f0CaTiaRzgKXAv8PXYyTNSXZgnUGkMuj97jUT51xjThhWxKNf/gBfPfMIHl+5ldOnz2Pm829R19j4+J1MIpe5fgiMB3YAmNlSfA544OC4XL17eJuJc65x3TLS+eLpw3n8yx9kTGkBP5yzgnNvfo5XNu5MdWhtIpFkUmtmh8anbWORyiryu2eSleEDCTjnEjOkqAd3fmo8v582lrd37WPqzf/jhjkrqNhXk+rQWiWRv4LLJV0CpEsaLun3wPNJjqtTCIZS8UtczrnmkcTZo/vzn6+ezGUTBzPzhbc4ffo8Hlm2pdP2TUkkmXwBOIZgSPpZwC6Cu7qaJGmypFclrZF0fZztkvS7cPuy+iFcJA2S9LSkVZJWSPpS1D69JD0h6fXwuTCRWJIhUlnl85g451osv3smP546kgc/fyLFud249p4X+cRfFrH+nc43KrGSlQUlpQOvAWcCG4FFwDQzWxlV5iyCZHUWMAH4rZlNkFQClJjZi5JygSXAuWa2UtKNwHYz+3mYoArN7FuNxVJWVmaLFy9u88946q+e4Zj+efzhkkaHMXPOuSbV1u3nrvnrmP74a1TX7mdgYXe6Z6WTk5VOTlYGOVnpdM9Kp0e4/J513dLpnhm8rl/u0S3YVpiT1eJRzSUtMbOyRMomMm3vEcDXCRrdD5Q3s9Oa2HU8sMbM1obHmQ1MJeinUm8qcKcFGW2+pAJJJWa2BdgSvk+FpFXAgHDfqcAp4f4zgWeARpNJskR8xGDnXBvJSE/jkycexpSRJdz67Fq2Vexjb3Udu6tr2bGnmk076thbXcee6lp2V9dRXbs/oeP+5RPHc+pRfZIcfQLJhGC+9xnA7UBzunAOADZEvd5IUPtoqswAwkQCIGkIwXhgC8JVfcNkg5ltkRT3LEm6CrgKoLS0tBlhJ2ZfTR0VVbV+mcs516b65Wfzg7NHNFmutm4/e2vq2FMdPHZX1R58XVUbPNfUcWS/9hnFOJFkUmtmt7Tg2IqzLvaaWqNlJPUEHgC+bGa7mvPmZnYrcCsEl7mas28i6odS8QZ451wqZKSnkZueRm52ZqpDARJrgP+XpM9LKgkbv3tJ6pXAfhuBQVGvBwKbEy0jKZMgkdxtZv+IKrM1bFMhfN6WQCxtrr6PiddMnHMusWRyJfANgtuBl4SPRFqzFwHDJR0mKQu4GIjtOT8HuCK8q2sisDO8dCXgz8AqM/t1nH2ujIrtoQRiaXMHer97m4lzzjV9mcvMDmvJgc2sVtJ1wGNAOnCHma2QdE24fQYwl+BOrjXAHg4OIHkicDnwiqSl4brvmNlc4OfAfZI+DawHLmhJfK1VXzPxZOKcc41P23uamT0l6WPxtsdceoor/OM/N2bdjKhlA66Ns9//iN+egpm9A5ze1HsnW/0gj729zcQ55xqtmZwMPAWcHWebAU0mk0NZpLKKvOwMumWkpzoU55xLuQaTiZn9MHz2uUviKK+sosgb351zDkjs1mAkfYRgSJUDE3eY2Y+TFVRnEKmo9hkWnXMulMh8JjOAiwiGPRFBg/fgJMfV4UW8ZuKccwckcmvwCWZ2BfCumf0ImMR7+4Z0SeWVVV4zcc65UCLJZG/4vEdSf6AGaNHtwoeKfTV1VOyr9d7vzjkXSqTN5GFJBcAvgRcJ7uS6PalRdXDex8Q5594rkU6LPwkXH5D0MJDd1WderO/97kOpOOdcoLFOi3E7K4bbEuq0eKiKVHjNxDnnojVWM4nXWbFel+60eOAyl9dMnHMOaLzTondWbEB9MundwxvgnXMOEutn0jucp/1FSUsk/VZS7/YIrqOKVFaTm51BdqYPpeKcc5DYrcGzgXLgfODj4fK9yQyqoyuv8D4mzjkXLZFbg3tF3dEF8FNJ5yYroM7Ax+Vyzrn3SqRm8rSkiyWlhY8LgUeSHVhHFvHe78459x6JJJOrgXuAqvAxG/iqpApJzZqX/VARqajy3u/OORclkU6Lue0RSGdRVVvHrn213sfEOeeiJHI316djXqdL+mHyQurYDsz97m0mzjl3QCKXuU6XNFdSiaRRwHygy9ZW6nu/e5uJc84dlMhlrkskXQS8AuwBppnZc0mPrIPy3u/OOfd+iVzmGg58CXgAeAu4XFJOkuPqsA6OGOwN8M45Vy+Ry1z/Ar5vZlcDJwOvA4uSGlUHdqDNxC9zOefcAYl0WhxvZrsAzMyA6ZLmJDesjqu8oorcbj6UinPORWuwZiLpmwBmtkvSBTGbu+wgkOWVVT6PiXPOxWjsMtfFUcvfjtk2OQmxdApBh0VPJs45F62xZKIGluO97jIilVUU5Xrju3PORWssmVgDy/FedxmRymqvmTjnXIzGGuBHh2NvCegeNQ6XgOykR9YBVdXWsXNvjScT55yL0WDNxMzSzSzPzHLNLCNcrn+dmcjBJU2W9KqkNZKuj7Nd4cRbayQtkzQuatsdkrZJWh6zzw2SNklaGj7Oas4Hbo13/LZg55yLK5F+Ji0iKR24GZgCjACmSRoRU2wKMDx8XAXcErXtrzTc0H+TmY0JH3PbNPBG1HdY9Lu5nHPuvZKWTIDxwBozW2tm1QRD10+NKTMVuNMC84ECSSUAZvYssD2J8TWb9353zrn4kplMBgAbol5vDNc1t0w814WXxe6QVNi6MBMXqfDLXM45F08yk0m824dj7wJLpEysW4ChwBhgCzA97ptLV0laLGlxeXl5U7EmpNwvcznnXFzJTCYbgUFRrwcCm1tQ5j3MbKuZ1ZnZfuA2gstp8crdamZlZlZWXFzc7ODjiVRW0dOHUnHOufdJZjJZBAyXdJikLIIe9bFjes0Brgjv6poI7DSzLY0dtL5NJXQesLyhsm2tvMKHUnHOuXgSGeixRcysVtJ1wGNAOnCHma2QdE24fQYwFzgLWEMwV8qBMb8kzQJOAYokbQR+aGZ/Bm6UNIbgcthbBHPUt4tIpc/97pxz8SQtmQCEt+3OjVk3I2rZgGsb2HdaA+svb8sYmyNSWc3wPj1T9fbOOddhJfMy1yEnqJn4ZS7nnIvlySRB1bX72bHHh1Jxzrl4PJkk6J3dfluwc841xJNJgg52WPQGeOeci+XJJEEHhlLxmolzzr2PJ5MEHej97m0mzjn3Pp5MEnRwkEdPJs45F8uTSYIiFdX0yEqne5YPpeKcc7E8mSSovNKHUnHOuYZ4MklQpMI7LDrnXEM8mSTIe78751zDPJkkKFJZRVGu9zFxzrl4PJkkoKZuP+/6UCrOOdcgTyYJeKcy6P3uDfDOORefJ5MEeB8T55xrnCeTBJR7MnHOuUZ5MklApMKHUnHOucZ4MklAJGwz8bu5nHMuPk8mCSivqCInK52crKTOcuycc52WJ5MERHwoFeeca5QnkwR473fnnGucJ5MEBMnE20ucc64hnkwSEKms9pqJc841wpNJE4KhVDyZOOdcYzyZNGH77mrMfCgV55xrjCeTJpRXeO9355xriieTJtSPy1XsHRadc65BnkyacKD3u9dMnHOuQZ5MmuAjBjvnXNOSmkwkTZb0qqQ1kq6Ps12SfhduXyZpXNS2OyRtk7Q8Zp9ekp6Q9Hr4XJjMz1A/lEqPbj6UinPONSRpyURSOnAzMAUYAUyTNCKm2BRgePi4CrglattfgclxDn098KSZDQeeDF8njfd+d865piWzZjIeWGNma82sGpgNTI0pMxW40wLzgQJJJQBm9iywPc5xpwIzw+WZwLlJiT7kvd+dc65pyUwmA4ANUa83huuaWyZWXzPbAhA+94lXSNJVkhZLWlxeXt6swKNFKrzDonPONSWZyURx1lkLyrSImd1qZmVmVlZcXNzi40QqqyjyDovOOdeoZCaTjcCgqNcDgc0tKBNra/2lsPB5WyvjbFBt3X62+1AqzjnXpGQmk0XAcEmHScoCLgbmxJSZA1wR3tU1EdhZfwmrEXOAK8PlK4GH2jLoaD6UinPOJSZpycTMaoHrgMeAVcB9ZrZC0jWSrgmLzQXWAmuA24DP1+8vaRbwAnCkpI2SPh1u+jlwpqTXgTPD10lRXt/73RvgnXOuUUntPGFmcwkSRvS6GVHLBlzbwL7TGlj/DnB6G4bZIO/97pxzifEe8I2I+CCPzjmXEE8mjTgwlIq3mTjnXKM8mTSivKKK7pnp9MhKT3UozjnXoXkyacTwvj05Z3R/pHjdYZxzztXz0QsbcdHxpVx0fGmqw3DOuQ7PaybOOedazZOJc865VvNk4pxzrtU8mTjnnGs1TybOOedazZOJc865VvNk4pxzrtU8mTjnnGs1BQP3HtoklQPrWrh7ERBpw3Dag8ecfJ0tXvCY20tni7mxeAebWUJT1XaJZNIakhabWVmq42gOjzn5Olu84DG3l84Wc1vF65e5nHPOtZonE+ecc63myaRpt6Y6gBbwmJOvs8ULHnN76Wwxt0m83mbinHOu1bxm4pxzrtU8mTjnnGs1TyYhSZMlvSppjaTr42yXpN+F25dJGpeKOKPiGSTpaUmrJK2Q9KU4ZU6RtFPS0vDxg1TEGhXPW5JeCWNZHGd7RzvHR0adu6WSdkn6ckyZlJ9jSXdI2iZpedS6XpKekPR6+FzYwL6Nfu/bOeZfSlod/tv/U1JBA/s2+j1q55hvkLQp6t//rAb2bffz3EC890bF+pakpQ3s2/xzbGZd/gGkA28AhwNZwMvAiJgyZwGPAgImAgtSHHMJMC5czgVeixPzKcDDqT6/UfG8BRQ1sr1DneM435G3CTpxdahzDHwQGAcsj1p3I3B9uHw98IsGPlOj3/t2jvlDQEa4/It4MSfyPWrnmG8Avp7Ad6fdz3O8eGO2Twd+0Fbn2GsmgfHAGjNba2bVwGxgakyZqcCdFpgPFEgqae9A65nZFjN7MVyuAFYBA1IVTxvpUOc4xunAG2bW0pEUksbMngW2x6yeCswMl2cC58bZNZHvfVLEi9nMHjez2vDlfGBge8SSqAbOcyJScp4bi1eSgAuBWW31fp5MAgOADVGvN/L+P8yJlEkJSUOAscCCOJsnSXpZ0qOSjmnXwN7PgMclLZF0VZztHfYcAxfT8H+8jnSO6/U1sy0Q/PAA+sQp05HP96cIaqnxNPU9am/XhZfm7mjgcmJHPM8fALaa2esNbG/2OfZkElCcdbH3TCdSpt1J6gk8AHzZzHbFbH6R4LLMaOD3wIPtHV+ME81sHDAFuFbSB2O2d9RznAWcA9wfZ3NHO8fN0VHP93eBWuDuBoo09T1qT7cAQ4ExwBaCS0exOuJ5nkbjtZJmn2NPJoGNwKCo1wOBzS0o064kZRIkkrvN7B+x281sl5lVhstzgUxJRe0cZnQ8m8PnbcA/Car/0TrcOQ5NAV40s62xGzraOY6ytf4SYfi8LU6ZDne+JV0JfBS41MKL97ES+B61GzPbamZ1ZrYfuK2BWDrUeZaUAXwMuLehMi05x55MAouA4ZIOC3+FXgzMiSkzB7givONoIrCz/jJCKoTXPP8MrDKzXzdQpl9YDknjCf6932m/KN8TSw9JufXLBI2ty2OKdahzHKXBX3Ed6RzHmANcGS5fCTwUp0wi3/t2I2ky8C3gHDPb00CZRL5H7SamTe+8BmLpUOcZOANYbWYb421s8TlO9h0FneVBcCfRawR3XXw3XHcNcE24LODmcPsrQFmK4z2JoKq8DFgaPs6Kifk6YAXB3SPzgRNSGO/hYRwvhzF1+HMcxpRDkBzyo9Z1qHNMkOi2ADUEv4I/DfQGngReD597hWX7A3Oj9n3f9z6FMa8haFuo/z7PiI25oe9RCmO+K/yuLiNIECUd5TzHizdc/9f6729U2VafYx9OxTnnXKv5ZS7nnHOt5snEOedcq3kycc4512qeTJxzzrWaJxPnnHOt5snEdUiSTNL0qNdfl3RDGx37r5I+3hbHauJ9LlAwqvPTcbYdIWluOIrsKkn3Seqb7JiSSdK5kkakOg6XGp5MXEdVBXysg/QmP0BSejOKfxr4vJmdGnOMbOAR4BYzG2ZmRxMMy1HcdpGmxLmAJ5MuypOJ66hqCeam/krshtiahaTK8PkUSfPCX/mvSfq5pEslLQznZhgadZgzJP03LPfRcP90BXNqLAoH7rs66rhPS7qHoINabDzTwuMvl/SLcN0PCDqWzpD0y5hdLgFeMLN/1a8ws6fNbLmkbEl/CY/3kqRTw+N9QtKDkv4l6U1J10n6alhmvqReYblnJP1G0vNhPOPD9b3C/ZeF5Y8N19+gYIDCZyStlfTFqM91WXjulkr6U30ilVQp6WcKBrecL6mvpBMIxi/7ZVh+qKQvSloZvufsRP7RXSfWXr1H/eGP5jyASiCPYF6FfODrwA3htr8CH48uGz6fAuwgmOulG7AJ+FG47UvAb6L2/zfBj6nhBL2Ds4GrgO+FZboBi4HDwuPuBg6LE2d/YD1BrSIDeAo4N9z2DHF68QO/Br7UwOf+GvCXcPmo8NjZwCcIeojnhu+1k4O98G8iGOiz/j1vC5c/SDiXBcEglD8Ml08DlobLNwDPh5+3iKC3fyZwNPAvIDMs90fginDZgLPD5Rujzlnsv8tmoFu4XJDq75Q/kvvwmonrsCwYBflO4ItNlY2yyIK5XqoIhq54PFz/CjAkqtx9ZrbfgiG41xL84f4QwdhgSwmG8+9NkGwAFprZm3He73jgGTMrt2AujrsJ/oi31EkEQ3RgZquBdcAR4banzazCzMoJkkl9zSb2s80K938WyFMwY2H0cZ8CekvKD8s/YmZVZhYhGBCyL8H8LccBi8LzcTrBMBsA1cDD4fKSmPeOtgy4W9JlBDVNdwjLSHUAzjXhNwTDvP8lal0t4SXacJDFrKhtVVHL+6Ne7+e93/fYcYSMYGywL5jZY9EbJJ1CUDOJJ97w4k1ZAZzcguO19rPFqi8Xfdy68FgCZprZt+PsV2NmFlM+no8QJNZzgO9LOsYOTn7lDjFeM3EdmpltB+4jaMyu9xbBr2YIZqzLbMGhL5CUFrajHA68CjwGfE7B0P71d1z1aOI4C4CTJRWFbQrTgHlN7HMPcIKkj9SvUDBH+CjgWeDS+vcHSsPYmuOicP+TCEZe3hlz3FOAiL1//ptoTwIfl9Qn3KeXpMFNvG8FwWU4JKUBg8zsaeCbQAHQs5mfw3UiXjNxncF0gtF5690GPCRpIcEfvYZqDY15leCPfl+Ctod9km4nuGTzYljjKSf+dLcHmNkWSd8Gnib4NT/XzOIN9x69z96w0f83kn5DMKrrMoJ2nT8SNNq/QlAD+4SZVQXhJOxdSc8TtDl9Klx3A/AXScuAPRwcnr6hGFdK+h7BbHtpYYzXElx2a8hs4LawEf9i4M/hpTQBN5nZjuZ8CNe5+KjBzh1CJD0DfN3MFqc6Fte1+GUu55xzreY1E+ecc63mNRPnnHOt5snEOedcq3kycc4512qeTJxzzrWaJxPnnHOt9v8Bak/CROWr7uIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "lsascree2 = TruncatedSVD(18)\n",
    "lsascree2.fit(df_tfidf)\n",
    "lsa_features2 = lsascree2.transform(df_tfidf)\n",
    "plt.plot(lsascree2.explained_variance_ratio_)\n",
    "plt.xlabel('Number of Components')\n",
    "plt.ylabel('Explained Variance')\n",
    "plt.title('Scree plot TF-IDF')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LSA TF-IDF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance: [0.00879464 0.03491028]\n",
      "\n",
      "Topic  0\n",
      "love, baby, time, yeah, say, said, think, tell, home, night, right, come, bad, look, make\n",
      "\n",
      "Topic  1\n",
      "baby, love, bad, smile, ain, sorry, good, time, really, did, come, hey, beautiful, right, high\n",
      "None\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>ain</th>\n",
       "      <th>baby</th>\n",
       "      <th>bad</th>\n",
       "      <th>beautiful</th>\n",
       "      <th>best</th>\n",
       "      <th>better</th>\n",
       "      <th>blue</th>\n",
       "      <th>break</th>\n",
       "      <th>car</th>\n",
       "      <th>change</th>\n",
       "      <th>...</th>\n",
       "      <th>trying</th>\n",
       "      <th>turn</th>\n",
       "      <th>waiting</th>\n",
       "      <th>walk</th>\n",
       "      <th>wanted</th>\n",
       "      <th>won</th>\n",
       "      <th>world</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yeah</th>\n",
       "      <th>years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.085</td>\n",
       "      <td>0.241</td>\n",
       "      <td>0.131</td>\n",
       "      <td>0.072</td>\n",
       "      <td>0.100</td>\n",
       "      <td>0.117</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.076</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.042</td>\n",
       "      <td>...</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.089</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.194</td>\n",
       "      <td>0.044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.130</td>\n",
       "      <td>0.568</td>\n",
       "      <td>0.153</td>\n",
       "      <td>0.087</td>\n",
       "      <td>-0.105</td>\n",
       "      <td>-0.039</td>\n",
       "      <td>-0.061</td>\n",
       "      <td>0.051</td>\n",
       "      <td>-0.059</td>\n",
       "      <td>0.010</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.078</td>\n",
       "      <td>-0.021</td>\n",
       "      <td>0.017</td>\n",
       "      <td>-0.034</td>\n",
       "      <td>-0.013</td>\n",
       "      <td>0.070</td>\n",
       "      <td>-0.025</td>\n",
       "      <td>-0.035</td>\n",
       "      <td>-0.280</td>\n",
       "      <td>-0.042</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 109 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     ain   baby    bad  beautiful   best  better   blue  break    car  change  \\\n",
       "0  0.085  0.241  0.131      0.072  0.100   0.117  0.057  0.076  0.060   0.042   \n",
       "1  0.130  0.568  0.153      0.087 -0.105  -0.039 -0.061  0.051 -0.059   0.010   \n",
       "\n",
       "   ...  trying   turn  waiting   walk  wanted    won  world  wrong   yeah  \\\n",
       "0  ...   0.080  0.048    0.080  0.060   0.043  0.089  0.051  0.060  0.194   \n",
       "1  ...  -0.078 -0.021    0.017 -0.034  -0.013  0.070 -0.025 -0.035 -0.280   \n",
       "\n",
       "   years  \n",
       "0  0.044  \n",
       "1 -0.042  \n",
       "\n",
       "[2 rows x 109 columns]"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lsa_function(2,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance: [0.00879339 0.03476453 0.03388777]\n",
      "\n",
      "Topic  0\n",
      "love, baby, time, yeah, say, said, think, tell, home, night, right, come, bad, look, make\n",
      "\n",
      "Topic  1\n",
      "baby, come, bad, did, sorry, smile, good, coming, hey, ain, new, really, right, play, high\n",
      "\n",
      "Topic  2\n",
      "love, good, beautiful, right, end, bad, light, miss, heart, coming, knew, long, loved, feel, come\n",
      "None\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>ain</th>\n",
       "      <th>baby</th>\n",
       "      <th>bad</th>\n",
       "      <th>beautiful</th>\n",
       "      <th>best</th>\n",
       "      <th>better</th>\n",
       "      <th>blue</th>\n",
       "      <th>break</th>\n",
       "      <th>car</th>\n",
       "      <th>change</th>\n",
       "      <th>...</th>\n",
       "      <th>trying</th>\n",
       "      <th>turn</th>\n",
       "      <th>waiting</th>\n",
       "      <th>walk</th>\n",
       "      <th>wanted</th>\n",
       "      <th>won</th>\n",
       "      <th>world</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yeah</th>\n",
       "      <th>years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.085</td>\n",
       "      <td>0.241</td>\n",
       "      <td>0.131</td>\n",
       "      <td>0.072</td>\n",
       "      <td>0.100</td>\n",
       "      <td>0.117</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.076</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.042</td>\n",
       "      <td>...</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.089</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.194</td>\n",
       "      <td>0.044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.109</td>\n",
       "      <td>0.547</td>\n",
       "      <td>0.168</td>\n",
       "      <td>0.037</td>\n",
       "      <td>-0.147</td>\n",
       "      <td>-0.090</td>\n",
       "      <td>-0.067</td>\n",
       "      <td>0.055</td>\n",
       "      <td>-0.070</td>\n",
       "      <td>0.030</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.079</td>\n",
       "      <td>-0.026</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.039</td>\n",
       "      <td>-0.009</td>\n",
       "      <td>-0.003</td>\n",
       "      <td>-0.211</td>\n",
       "      <td>-0.032</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.112</td>\n",
       "      <td>-0.319</td>\n",
       "      <td>0.109</td>\n",
       "      <td>0.204</td>\n",
       "      <td>-0.150</td>\n",
       "      <td>-0.320</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.019</td>\n",
       "      <td>0.023</td>\n",
       "      <td>0.006</td>\n",
       "      <td>...</td>\n",
       "      <td>0.043</td>\n",
       "      <td>-0.001</td>\n",
       "      <td>-0.041</td>\n",
       "      <td>-0.029</td>\n",
       "      <td>-0.027</td>\n",
       "      <td>-0.162</td>\n",
       "      <td>0.045</td>\n",
       "      <td>-0.014</td>\n",
       "      <td>0.045</td>\n",
       "      <td>-0.011</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 109 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     ain   baby    bad  beautiful   best  better   blue  break    car  change  \\\n",
       "0  0.085  0.241  0.131      0.072  0.100   0.117  0.057  0.076  0.060   0.042   \n",
       "1  0.109  0.547  0.168      0.037 -0.147  -0.090 -0.067  0.055 -0.070   0.030   \n",
       "2 -0.112 -0.319  0.109      0.204 -0.150  -0.320  0.048  0.019  0.023   0.006   \n",
       "\n",
       "   ...  trying   turn  waiting   walk  wanted    won  world  wrong   yeah  \\\n",
       "0  ...   0.080  0.048    0.080  0.060   0.043  0.089  0.051  0.060  0.194   \n",
       "1  ...  -0.079 -0.026    0.002  0.005   0.008  0.039 -0.009 -0.003 -0.211   \n",
       "2  ...   0.043 -0.001   -0.041 -0.029  -0.027 -0.162  0.045 -0.014  0.045   \n",
       "\n",
       "   years  \n",
       "0  0.044  \n",
       "1 -0.032  \n",
       "2 -0.011  \n",
       "\n",
       "[3 rows x 109 columns]"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lsa_function(3,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance: [0.00879384 0.03523428 0.0342991  0.03213068]\n",
      "\n",
      "Topic  0\n",
      "love, baby, time, yeah, say, said, think, tell, home, night, right, come, bad, look, make\n",
      "\n",
      "Topic  1\n",
      "baby, love, bad, good, sorry, smile, come, ain, hey, beautiful, coming, time, did, really, high\n",
      "\n",
      "Topic  2\n",
      "love, beautiful, good, right, end, come, heart, miss, knew, light, loved, lights, long, saw, feel\n",
      "\n",
      "Topic  3\n",
      "said, gone, did, think, girl, forever, thought, wrong, come, feel, home, rain, phone, door, kiss\n",
      "None\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>ain</th>\n",
       "      <th>baby</th>\n",
       "      <th>bad</th>\n",
       "      <th>beautiful</th>\n",
       "      <th>best</th>\n",
       "      <th>better</th>\n",
       "      <th>blue</th>\n",
       "      <th>break</th>\n",
       "      <th>car</th>\n",
       "      <th>change</th>\n",
       "      <th>...</th>\n",
       "      <th>trying</th>\n",
       "      <th>turn</th>\n",
       "      <th>waiting</th>\n",
       "      <th>walk</th>\n",
       "      <th>wanted</th>\n",
       "      <th>won</th>\n",
       "      <th>world</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yeah</th>\n",
       "      <th>years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.085</td>\n",
       "      <td>0.241</td>\n",
       "      <td>0.131</td>\n",
       "      <td>0.072</td>\n",
       "      <td>0.100</td>\n",
       "      <td>0.117</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.076</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.042</td>\n",
       "      <td>...</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.089</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.194</td>\n",
       "      <td>0.044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.111</td>\n",
       "      <td>0.502</td>\n",
       "      <td>0.164</td>\n",
       "      <td>0.098</td>\n",
       "      <td>-0.141</td>\n",
       "      <td>-0.095</td>\n",
       "      <td>-0.057</td>\n",
       "      <td>0.072</td>\n",
       "      <td>-0.064</td>\n",
       "      <td>0.015</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.103</td>\n",
       "      <td>-0.020</td>\n",
       "      <td>0.007</td>\n",
       "      <td>-0.028</td>\n",
       "      <td>-0.011</td>\n",
       "      <td>0.038</td>\n",
       "      <td>-0.023</td>\n",
       "      <td>-0.043</td>\n",
       "      <td>-0.271</td>\n",
       "      <td>-0.044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.108</td>\n",
       "      <td>-0.412</td>\n",
       "      <td>0.025</td>\n",
       "      <td>0.252</td>\n",
       "      <td>-0.093</td>\n",
       "      <td>-0.254</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.012</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.016</td>\n",
       "      <td>...</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.008</td>\n",
       "      <td>-0.020</td>\n",
       "      <td>-0.017</td>\n",
       "      <td>-0.158</td>\n",
       "      <td>0.044</td>\n",
       "      <td>-0.029</td>\n",
       "      <td>0.057</td>\n",
       "      <td>-0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.106</td>\n",
       "      <td>-0.032</td>\n",
       "      <td>-0.080</td>\n",
       "      <td>0.047</td>\n",
       "      <td>-0.213</td>\n",
       "      <td>-0.185</td>\n",
       "      <td>-0.021</td>\n",
       "      <td>0.070</td>\n",
       "      <td>0.047</td>\n",
       "      <td>-0.011</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.152</td>\n",
       "      <td>0.044</td>\n",
       "      <td>-0.055</td>\n",
       "      <td>0.012</td>\n",
       "      <td>-0.063</td>\n",
       "      <td>-0.060</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.148</td>\n",
       "      <td>-0.397</td>\n",
       "      <td>-0.012</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4 rows × 109 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     ain   baby    bad  beautiful   best  better   blue  break    car  change  \\\n",
       "0  0.085  0.241  0.131      0.072  0.100   0.117  0.057  0.076  0.060   0.042   \n",
       "1  0.111  0.502  0.164      0.098 -0.141  -0.095 -0.057  0.072 -0.064   0.015   \n",
       "2 -0.108 -0.412  0.025      0.252 -0.093  -0.254  0.047  0.012  0.045   0.016   \n",
       "3 -0.106 -0.032 -0.080      0.047 -0.213  -0.185 -0.021  0.070  0.047  -0.011   \n",
       "\n",
       "   ...  trying   turn  waiting   walk  wanted    won  world  wrong   yeah  \\\n",
       "0  ...   0.080  0.048    0.080  0.060   0.043  0.089  0.051  0.060  0.194   \n",
       "1  ...  -0.103 -0.020    0.007 -0.028  -0.011  0.038 -0.023 -0.043 -0.271   \n",
       "2  ...   0.051  0.002    0.008 -0.020  -0.017 -0.158  0.044 -0.029  0.057   \n",
       "3  ...  -0.152  0.044   -0.055  0.012  -0.063 -0.060  0.000  0.148 -0.397   \n",
       "\n",
       "   years  \n",
       "0  0.044  \n",
       "1 -0.044  \n",
       "2 -0.000  \n",
       "3 -0.012  \n",
       "\n",
       "[4 rows x 109 columns]"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lsa_function(4,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Explained Variance: [0.00879408 0.03524024 0.03432569 0.03231191 0.03155085]\n",
      "\n",
      "Topic  0\n",
      "love, baby, time, yeah, say, said, think, tell, home, night, right, come, bad, look, make\n",
      "\n",
      "Topic  1\n",
      "baby, love, bad, good, come, sorry, smile, ain, hey, right, coming, really, did, high, time\n",
      "\n",
      "Topic  2\n",
      "love, beautiful, good, right, end, come, heart, knew, long, miss, light, lights, saw, loved, home\n",
      "\n",
      "Topic  3\n",
      "said, gone, did, come, think, forever, girl, thought, feel, wrong, rain, phone, home, door, kiss\n",
      "\n",
      "Topic  4\n",
      "stay, beautiful, think, love, little, better, won, light, car, try, make, mind, smile, remember, turn\n",
      "None\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>ain</th>\n",
       "      <th>baby</th>\n",
       "      <th>bad</th>\n",
       "      <th>beautiful</th>\n",
       "      <th>best</th>\n",
       "      <th>better</th>\n",
       "      <th>blue</th>\n",
       "      <th>break</th>\n",
       "      <th>car</th>\n",
       "      <th>change</th>\n",
       "      <th>...</th>\n",
       "      <th>trying</th>\n",
       "      <th>turn</th>\n",
       "      <th>waiting</th>\n",
       "      <th>walk</th>\n",
       "      <th>wanted</th>\n",
       "      <th>won</th>\n",
       "      <th>world</th>\n",
       "      <th>wrong</th>\n",
       "      <th>yeah</th>\n",
       "      <th>years</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.085</td>\n",
       "      <td>0.241</td>\n",
       "      <td>0.131</td>\n",
       "      <td>0.072</td>\n",
       "      <td>0.100</td>\n",
       "      <td>0.117</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.076</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.042</td>\n",
       "      <td>...</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.089</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.194</td>\n",
       "      <td>0.044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.115</td>\n",
       "      <td>0.529</td>\n",
       "      <td>0.160</td>\n",
       "      <td>0.079</td>\n",
       "      <td>-0.143</td>\n",
       "      <td>-0.096</td>\n",
       "      <td>-0.057</td>\n",
       "      <td>0.061</td>\n",
       "      <td>-0.064</td>\n",
       "      <td>0.013</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.096</td>\n",
       "      <td>-0.023</td>\n",
       "      <td>0.017</td>\n",
       "      <td>-0.025</td>\n",
       "      <td>-0.011</td>\n",
       "      <td>0.041</td>\n",
       "      <td>-0.025</td>\n",
       "      <td>-0.043</td>\n",
       "      <td>-0.258</td>\n",
       "      <td>-0.042</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.094</td>\n",
       "      <td>-0.385</td>\n",
       "      <td>0.010</td>\n",
       "      <td>0.250</td>\n",
       "      <td>-0.098</td>\n",
       "      <td>-0.271</td>\n",
       "      <td>0.041</td>\n",
       "      <td>0.007</td>\n",
       "      <td>0.026</td>\n",
       "      <td>0.010</td>\n",
       "      <td>...</td>\n",
       "      <td>0.033</td>\n",
       "      <td>0.002</td>\n",
       "      <td>0.015</td>\n",
       "      <td>-0.029</td>\n",
       "      <td>-0.023</td>\n",
       "      <td>-0.162</td>\n",
       "      <td>0.039</td>\n",
       "      <td>-0.036</td>\n",
       "      <td>0.027</td>\n",
       "      <td>-0.005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.104</td>\n",
       "      <td>-0.058</td>\n",
       "      <td>-0.071</td>\n",
       "      <td>0.031</td>\n",
       "      <td>-0.212</td>\n",
       "      <td>-0.158</td>\n",
       "      <td>-0.024</td>\n",
       "      <td>0.051</td>\n",
       "      <td>0.010</td>\n",
       "      <td>-0.000</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.154</td>\n",
       "      <td>0.035</td>\n",
       "      <td>-0.040</td>\n",
       "      <td>-0.004</td>\n",
       "      <td>-0.058</td>\n",
       "      <td>-0.082</td>\n",
       "      <td>0.006</td>\n",
       "      <td>0.143</td>\n",
       "      <td>-0.400</td>\n",
       "      <td>-0.015</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.027</td>\n",
       "      <td>-0.082</td>\n",
       "      <td>-0.170</td>\n",
       "      <td>0.329</td>\n",
       "      <td>0.026</td>\n",
       "      <td>0.105</td>\n",
       "      <td>-0.033</td>\n",
       "      <td>0.056</td>\n",
       "      <td>0.087</td>\n",
       "      <td>-0.023</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.129</td>\n",
       "      <td>0.059</td>\n",
       "      <td>-0.001</td>\n",
       "      <td>-0.004</td>\n",
       "      <td>-0.027</td>\n",
       "      <td>0.101</td>\n",
       "      <td>-0.065</td>\n",
       "      <td>-0.041</td>\n",
       "      <td>-0.275</td>\n",
       "      <td>0.004</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 109 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     ain   baby    bad  beautiful   best  better   blue  break    car  change  \\\n",
       "0  0.085  0.241  0.131      0.072  0.100   0.117  0.057  0.076  0.060   0.042   \n",
       "1  0.115  0.529  0.160      0.079 -0.143  -0.096 -0.057  0.061 -0.064   0.013   \n",
       "2 -0.094 -0.385  0.010      0.250 -0.098  -0.271  0.041  0.007  0.026   0.010   \n",
       "3 -0.104 -0.058 -0.071      0.031 -0.212  -0.158 -0.024  0.051  0.010  -0.000   \n",
       "4 -0.027 -0.082 -0.170      0.329  0.026   0.105 -0.033  0.056  0.087  -0.023   \n",
       "\n",
       "   ...  trying   turn  waiting   walk  wanted    won  world  wrong   yeah  \\\n",
       "0  ...   0.080  0.048    0.080  0.060   0.043  0.089  0.051  0.060  0.194   \n",
       "1  ...  -0.096 -0.023    0.017 -0.025  -0.011  0.041 -0.025 -0.043 -0.258   \n",
       "2  ...   0.033  0.002    0.015 -0.029  -0.023 -0.162  0.039 -0.036  0.027   \n",
       "3  ...  -0.154  0.035   -0.040 -0.004  -0.058 -0.082  0.006  0.143 -0.400   \n",
       "4  ...  -0.129  0.059   -0.001 -0.004  -0.027  0.101 -0.065 -0.041 -0.275   \n",
       "\n",
       "   years  \n",
       "0  0.044  \n",
       "1 -0.042  \n",
       "2 -0.005  \n",
       "3 -0.015  \n",
       "4  0.004  \n",
       "\n",
       "[5 rows x 109 columns]"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lsa_function(5,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LSA TF-IDF with Bi-grams"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "lsa2bi = TruncatedSVD(5)\n",
    "doc_topic2bi = lsa2bi.fit_transform(doc_matrix_bi2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "topic_word2bi = pd.DataFrame(lsa2bi.components_.round(3),\n",
    "             index = [\"component_1\",\"component_2\",\"component_3\",\"component_4\",\"component_5\"],\n",
    "             columns = vectorizer_bi2.get_feature_names())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "say, time, love, said, night, think, baby, eyes, right, tell\n",
      "\n",
      "Topic  1\n",
      "night, time, remember, best, things, said, place, street, walk, change\n",
      "\n",
      "Topic  2\n",
      "home, yeah, thing, gone, said, eyes, little, right, town, night\n",
      "\n",
      "Topic  3\n",
      "baby, say, didn, looking, think, thinking, try, door, did, phone\n",
      "\n",
      "Topic  4\n",
      "feel, walk, little, eyes, friends, beautiful, change, rain, going, wanted\n"
     ]
    }
   ],
   "source": [
    "display_topics(lsa2bi, vectorizer_bi2.get_feature_names(), 10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# NMF TF-IDF"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "yeah, think, said, home, say, girl, better, night, tell, best, look, remember, dress, stay, time\n",
      "\n",
      "Topic  1\n",
      "love, baby, bad, come, good, time, right, smile, beautiful, let, sorry, coming, ain, new, life\n",
      "None\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lindsayread/opt/anaconda3/envs/metis/lib/python3.8/site-packages/sklearn/decomposition/_nmf.py:1076: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.\n",
      "  warnings.warn(\"Maximum number of iterations %d reached. Increase it to\"\n"
     ]
    }
   ],
   "source": [
    "nmf_function(2,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "yeah, think, said, home, say, girl, night, tell, best, look, dress, remember, better, stay, time\n",
      "\n",
      "Topic  1\n",
      "baby, sorry, bad, time, ain, won, smile, did, things, come, really, let, hey, better, night\n",
      "\n",
      "Topic  2\n",
      "love, good, beautiful, right, come, bad, heart, knew, light, saw, time, coming, lights, say, long\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "nmf_function(3,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "said, think, girl, home, gone, look, did, forever, come, say, feel, wrong, thought, didn, long\n",
      "\n",
      "Topic  1\n",
      "baby, time, sorry, bad, ain, won, smile, better, things, let, really, hey, tell, life, come\n",
      "\n",
      "Topic  2\n",
      "love, good, beautiful, right, heart, light, saw, knew, time, bad, coming, lights, let, break, say\n",
      "\n",
      "Topic  3\n",
      "yeah, best, night, stay, trying, tell, remember, time, right, say, better, day, head, hands, hold\n",
      "None\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lindsayread/opt/anaconda3/envs/metis/lib/python3.8/site-packages/sklearn/decomposition/_nmf.py:1076: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.\n",
      "  warnings.warn(\"Maximum number of iterations %d reached. Increase it to\"\n"
     ]
    }
   ],
   "source": [
    "nmf_function(4,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "said, girl, home, think, gone, come, did, look, say, forever, feel, wrong, thought, didn, long\n",
      "\n",
      "Topic  1\n",
      "baby, time, sorry, ain, bad, won, smile, better, things, let, tell, really, night, life, hey\n",
      "\n",
      "Topic  2\n",
      "love, good, beautiful, right, heart, saw, bad, light, coming, time, knew, lights, let, break, say\n",
      "\n",
      "Topic  3\n",
      "yeah, best, trying, night, tell, remember, time, head, hands, right, new, feeling, say, end, day\n",
      "\n",
      "Topic  4\n",
      "stay, think, little, beautiful, won, home, right, hand, light, hold, getting, hey, break, make, let\n",
      "None\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lindsayread/opt/anaconda3/envs/metis/lib/python3.8/site-packages/sklearn/decomposition/_nmf.py:1076: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.\n",
      "  warnings.warn(\"Maximum number of iterations %d reached. Increase it to\"\n"
     ]
    }
   ],
   "source": [
    "nmf_function(5,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "said, think, gone, did, home, wrong, girl, thought, forever, dress, feel, didn, little, say, rain\n",
      "\n",
      "Topic  1\n",
      "baby, bad, smile, won, ain, really, time, let, hey, sorry, high, make, coming, life, things\n",
      "\n",
      "Topic  2\n",
      "love, good, beautiful, heart, right, saw, lights, light, knew, coming, let, bad, eyes, break, time\n",
      "\n",
      "Topic  3\n",
      "yeah, trying, girl, home, end, head, look, bad, hands, best, feeling, right, blue, heard, new\n",
      "\n",
      "Topic  4\n",
      "stay, little, beautiful, think, right, won, hand, hold, home, getting, hey, time, light, let, break\n",
      "\n",
      "Topic  5\n",
      "time, come, tell, day, say, remember, night, long, looking, better, need, best, dreams, sorry, things\n",
      "None\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lindsayread/opt/anaconda3/envs/metis/lib/python3.8/site-packages/sklearn/decomposition/_nmf.py:1076: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.\n",
      "  warnings.warn(\"Maximum number of iterations %d reached. Increase it to\"\n"
     ]
    }
   ],
   "source": [
    "nmf_function(6,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "girl, home, look, forever, long, feel, drive, say, place, yeah, lights, trying, phone, world, lost\n",
      "\n",
      "Topic  1\n",
      "baby, bad, smile, won, ain, really, hey, let, coming, make, high, time, eyes, better, sorry\n",
      "\n",
      "Topic  2\n",
      "love, good, beautiful, heart, right, saw, light, lights, knew, coming, let, bad, break, eyes, time\n",
      "\n",
      "Topic  3\n",
      "yeah, trying, head, best, end, hands, new, right, dress, feeling, street, waiting, bad, night, blue\n",
      "\n",
      "Topic  4\n",
      "stay, little, beautiful, think, won, right, hand, home, hold, getting, hey, time, light, let, make\n",
      "\n",
      "Topic  5\n",
      "time, come, tell, day, say, night, looking, long, need, better, sorry, remember, dreams, best, things\n",
      "\n",
      "Topic  6\n",
      "said, think, gone, did, thought, wrong, dress, things, little, head, mind, didn, rain, say, remember\n",
      "None\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/lindsayread/opt/anaconda3/envs/metis/lib/python3.8/site-packages/sklearn/decomposition/_nmf.py:1076: ConvergenceWarning: Maximum number of iterations 200 reached. Increase it to improve convergence.\n",
      "  warnings.warn(\"Maximum number of iterations %d reached. Increase it to\"\n"
     ]
    }
   ],
   "source": [
    "nmf_function(7,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "girl, home, look, forever, long, say, drive, feel, yeah, place, lights, trying, phone, world, lost\n",
      "\n",
      "Topic  1\n",
      "baby, bad, smile, won, ain, really, hey, coming, high, let, make, life, eyes, hand, sorry\n",
      "\n",
      "Topic  2\n",
      "love, good, beautiful, heart, saw, right, light, lights, knew, time, let, coming, hands, eyes, break\n",
      "\n",
      "Topic  3\n",
      "yeah, trying, head, end, best, hands, new, right, feeling, street, waiting, dress, bad, blue, night\n",
      "\n",
      "Topic  4\n",
      "stay, little, beautiful, think, right, won, hand, home, hold, hey, getting, light, let, break, make\n",
      "\n",
      "Topic  5\n",
      "time, better, tell, night, looking, say, best, sorry, remember, dreams, try, ain, didn, long, day\n",
      "\n",
      "Topic  6\n",
      "said, think, gone, did, thought, wrong, things, dress, little, head, remember, mind, say, didn, forever\n",
      "\n",
      "Topic  7\n",
      "come, day, right, miss, feeling, new, kiss, need, feel, forget, play, long, fall, rain, getting\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "nmf_function(8,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "girl, home, look, forever, long, say, yeah, drive, place, feel, trying, phone, lights, world, fine\n",
      "\n",
      "Topic  1\n",
      "baby, bad, smile, won, ain, really, high, hey, let, life, hand, coming, eyes, new, things\n",
      "\n",
      "Topic  2\n",
      "love, beautiful, lights, knew, time, life, saw, heart, long, high, right, eyes, little, smile, let\n",
      "\n",
      "Topic  3\n",
      "yeah, trying, head, end, best, hands, new, right, feeling, street, waiting, dress, night, blue, bad\n",
      "\n",
      "Topic  4\n",
      "stay, little, think, beautiful, right, hand, won, home, hold, hey, getting, light, let, break, make\n",
      "\n",
      "Topic  5\n",
      "better, time, tell, night, sorry, say, looking, best, try, didn, dreams, ain, let, run, long\n",
      "\n",
      "Topic  6\n",
      "said, think, gone, did, thought, wrong, things, dress, little, head, didn, mind, forever, say, rain\n",
      "\n",
      "Topic  7\n",
      "come, day, right, feeling, miss, kiss, new, long, forget, feel, need, rain, fall, change, fight\n",
      "\n",
      "Topic  8\n",
      "good, remember, bad, coming, looking, break, play, light, say, make, right, room, heart, hey, getting\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "nmf_function(9,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "girl, home, look, forever, long, yeah, say, place, drive, feel, trying, phone, lights, world, fine\n",
      "\n",
      "Topic  1\n",
      "baby, bad, smile, won, ain, really, high, hey, life, hand, new, things, eyes, let, make\n",
      "\n",
      "Topic  2\n",
      "love, beautiful, lights, knew, life, saw, eyes, high, heart, little, right, smile, didn, end, home\n",
      "\n",
      "Topic  3\n",
      "yeah, trying, head, end, best, hands, new, right, feeling, street, waiting, dress, night, bad, blue\n",
      "\n",
      "Topic  4\n",
      "stay, little, think, beautiful, right, hand, hold, home, hey, won, getting, let, time, light, break\n",
      "\n",
      "Topic  5\n",
      "time, tell, night, sorry, long, dreams, say, let, need, remember, run, hard, looking, ain, hold\n",
      "\n",
      "Topic  6\n",
      "said, gone, think, did, thought, wrong, things, dress, little, head, didn, forever, say, mind, phone\n",
      "\n",
      "Topic  7\n",
      "come, day, right, feeling, miss, kiss, new, forget, feel, rain, long, need, fall, change, things\n",
      "\n",
      "Topic  8\n",
      "good, bad, remember, coming, looking, break, play, light, say, make, right, room, heart, hey, getting\n",
      "\n",
      "Topic  9\n",
      "better, best, won, make, didn, people, think, try, looking, eyes, mind, day, dress, lost, ain\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "nmf_function(10,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "girl, home, look, forever, long, yeah, drive, place, phone, trying, lights, say, fine, world, heart\n",
      "\n",
      "Topic  1\n",
      "baby, bad, smile, won, high, hey, really, ain, let, hand, new, coming, things, eyes, run\n",
      "\n",
      "Topic  2\n",
      "love, beautiful, lights, knew, life, saw, high, eyes, little, smile, heart, right, home, didn, end\n",
      "\n",
      "Topic  3\n",
      "yeah, trying, head, end, best, hands, new, right, street, feeling, waiting, dress, bad, blue, night\n",
      "\n",
      "Topic  4\n",
      "stay, little, beautiful, think, right, hand, home, hey, hold, won, getting, let, light, night, say\n",
      "\n",
      "Topic  5\n",
      "time, sorry, ain, long, night, life, change, town, turn, hold, times, drive, really, let, dreams\n",
      "\n",
      "Topic  6\n",
      "said, think, gone, did, thought, wrong, things, little, dress, head, mind, rain, didn, thinking, phone\n",
      "\n",
      "Topic  7\n",
      "come, day, right, miss, feeling, new, kiss, forget, long, rain, feel, fall, things, need, change\n",
      "\n",
      "Topic  8\n",
      "good, remember, bad, coming, looking, break, play, light, right, make, say, heart, room, getting, hey\n",
      "\n",
      "Topic  9\n",
      "better, best, won, make, didn, people, try, think, looking, eyes, mind, day, dress, smile, lost\n",
      "\n",
      "Topic  10\n",
      "tell, say, need, forever, let, night, run, dreams, feel, said, face, waiting, room, took, knew\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "nmf_function(11,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Topic  0\n",
      "girl, home, look, long, forever, yeah, drive, place, phone, lights, fine, world, trying, heart, thing\n",
      "\n",
      "Topic  1\n",
      "baby, smile, bad, won, high, things, ain, new, hey, hand, really, life, let, run, lights\n",
      "\n",
      "Topic  2\n",
      "love, beautiful, lights, life, knew, high, saw, little, eyes, smile, right, didn, long, home, heart\n",
      "\n",
      "Topic  3\n",
      "yeah, trying, head, end, right, new, hands, feeling, best, street, bad, waiting, heard, blue, old\n",
      "\n",
      "Topic  4\n",
      "stay, little, think, beautiful, right, hand, hey, home, won, hold, let, getting, light, say, break\n",
      "\n",
      "Topic  5\n",
      "remember, night, dreams, best, hold, forget, long, things, thing, place, lost, looking, hard, said, getting\n",
      "\n",
      "Topic  6\n",
      "said, think, gone, did, thought, wrong, things, head, little, dress, mind, rain, didn, phone, thinking\n",
      "\n",
      "Topic  7\n",
      "come, day, right, miss, feeling, new, kiss, rain, forget, long, feel, knew, need, fall, change\n",
      "\n",
      "Topic  8\n",
      "good, bad, coming, play, break, light, make, right, heart, looking, say, hey, getting, room, did\n",
      "\n",
      "Topic  9\n",
      "better, best, won, make, didn, people, think, try, eyes, looking, mind, day, smile, dress, lost\n",
      "\n",
      "Topic  10\n",
      "tell, say, need, feel, forever, run, let, waiting, face, room, knew, said, took, hey, door\n",
      "\n",
      "Topic  11\n",
      "time, sorry, ain, let, really, change, turn, drive, life, town, night, break, mind, people, door\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "nmf_function(12,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "NMF with TF-IDF vectorizer and 9, 10, or 11 topics seems to produce the best topics."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "def nmf_HMatrix(num_components, doc_text_matrix, vectorizer):\n",
    "    nmf = NMF(num_components)\n",
    "    doc_topic = nmf.fit_transform(doc_text_matrix)\n",
    "    \n",
    "    idx = []\n",
    "    for i in range(num_components):\n",
    "        idx.append(i) \n",
    "    H = pd.DataFrame(doc_topic.round(3),\n",
    "                    index = song_titles,\n",
    "                    columns = idx)\n",
    "    return H"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "h9 = nmf_HMatrix(9,df_tfidf,cv_tfidf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Tim McGraw</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.071</td>\n",
       "      <td>0.028</td>\n",
       "      <td>0.353</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Picture To Burn</th>\n",
       "      <td>0.010</td>\n",
       "      <td>0.214</td>\n",
       "      <td>0.066</td>\n",
       "      <td>0.034</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.199</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.010</td>\n",
       "      <td>0.028</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Teardrops On My Guitar</th>\n",
       "      <td>0.106</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.254</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.086</td>\n",
       "      <td>0.132</td>\n",
       "      <td>0.013</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.028</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A Place In This World</th>\n",
       "      <td>0.265</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.109</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Cold as You</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.023</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.004</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.072</td>\n",
       "      <td>0.066</td>\n",
       "      <td>0.201</td>\n",
       "      <td>0.056</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Miss Americana &amp; the Heartbreak Prince</th>\n",
       "      <td>0.139</td>\n",
       "      <td>0.148</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.038</td>\n",
       "      <td>0.059</td>\n",
       "      <td>0.154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Paper Rings</th>\n",
       "      <td>0.026</td>\n",
       "      <td>0.107</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.009</td>\n",
       "      <td>0.112</td>\n",
       "      <td>0.003</td>\n",
       "      <td>0.113</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Soon You'll Get Better (feat. Dixie Chicks)</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.032</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.024</td>\n",
       "      <td>0.487</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Archer</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.078</td>\n",
       "      <td>0.358</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.138</td>\n",
       "      <td>0.049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>You Need to Calm Down</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.012</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.023</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.151</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.072</td>\n",
       "      <td>0.048</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>111 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 0      1      2      3  \\\n",
       "Tim McGraw                                   0.000  0.000  0.000  0.057   \n",
       "Picture To Burn                              0.010  0.214  0.066  0.034   \n",
       "Teardrops On My Guitar                       0.106  0.048  0.254  0.000   \n",
       "A Place In This World                        0.265  0.000  0.000  0.109   \n",
       "Cold as You                                  0.000  0.023  0.000  0.004   \n",
       "...                                            ...    ...    ...    ...   \n",
       "Miss Americana & the Heartbreak Prince       0.139  0.148  0.008  0.048   \n",
       "Paper Rings                                  0.026  0.107  0.054  0.045   \n",
       "Soon You'll Get Better (feat. Dixie Chicks)  0.000  0.032  0.000  0.001   \n",
       "The Archer                                   0.000  0.000  0.000  0.078   \n",
       "You Need to Calm Down                        0.000  0.012  0.000  0.023   \n",
       "\n",
       "                                                 4      5      6      7      8  \n",
       "Tim McGraw                                   0.071  0.028  0.353  0.000  0.000  \n",
       "Picture To Burn                              0.000  0.199  0.000  0.010  0.028  \n",
       "Teardrops On My Guitar                       0.086  0.132  0.013  0.000  0.028  \n",
       "A Place In This World                        0.000  0.000  0.000  0.000  0.000  \n",
       "Cold as You                                  0.005  0.072  0.066  0.201  0.056  \n",
       "...                                            ...    ...    ...    ...    ...  \n",
       "Miss Americana & the Heartbreak Prince       0.000  0.000  0.038  0.059  0.154  \n",
       "Paper Rings                                  0.009  0.112  0.003  0.113  0.000  \n",
       "Soon You'll Get Better (feat. Dixie Chicks)  0.024  0.487  0.000  0.000  0.000  \n",
       "The Archer                                   0.358  0.000  0.000  0.138  0.049  \n",
       "You Need to Calm Down                        0.008  0.151  0.000  0.072  0.048  \n",
       "\n",
       "[111 rows x 9 columns]"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "h9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Tim McGraw</th>\n",
       "      <td>0.007</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.057</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.364</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.048</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Picture To Burn</th>\n",
       "      <td>0.010</td>\n",
       "      <td>0.201</td>\n",
       "      <td>0.042</td>\n",
       "      <td>0.032</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.037</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.035</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.040</td>\n",
       "      <td>0.287</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Teardrops On My Guitar</th>\n",
       "      <td>0.098</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.258</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.062</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.016</td>\n",
       "      <td>0.039</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.192</td>\n",
       "      <td>0.077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>A Place In This World</th>\n",
       "      <td>0.233</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.097</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.046</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Cold as You</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.020</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.005</td>\n",
       "      <td>0.082</td>\n",
       "      <td>0.060</td>\n",
       "      <td>0.043</td>\n",
       "      <td>0.197</td>\n",
       "      <td>0.073</td>\n",
       "      <td>0.013</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Miss Americana &amp; the Heartbreak Prince</th>\n",
       "      <td>0.131</td>\n",
       "      <td>0.144</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.047</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.156</td>\n",
       "      <td>0.058</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.062</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Paper Rings</th>\n",
       "      <td>0.028</td>\n",
       "      <td>0.105</td>\n",
       "      <td>0.049</td>\n",
       "      <td>0.042</td>\n",
       "      <td>0.009</td>\n",
       "      <td>0.086</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.105</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Soon You'll Get Better (feat. Dixie Chicks)</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.035</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.009</td>\n",
       "      <td>0.086</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.006</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.659</td>\n",
       "      <td>0.044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Archer</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.084</td>\n",
       "      <td>0.370</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.064</td>\n",
       "      <td>0.142</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.004</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>You Need to Calm Down</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.009</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.009</td>\n",
       "      <td>0.014</td>\n",
       "      <td>0.273</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.026</td>\n",
       "      <td>0.015</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>111 rows × 11 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                0      1      2      3   \\\n",
       "Tim McGraw                                   0.007  0.000  0.000  0.057   \n",
       "Picture To Burn                              0.010  0.201  0.042  0.032   \n",
       "Teardrops On My Guitar                       0.098  0.043  0.258  0.000   \n",
       "A Place In This World                        0.233  0.000  0.000  0.097   \n",
       "Cold as You                                  0.000  0.020  0.000  0.001   \n",
       "...                                            ...    ...    ...    ...   \n",
       "Miss Americana & the Heartbreak Prince       0.131  0.144  0.000  0.047   \n",
       "Paper Rings                                  0.028  0.105  0.049  0.042   \n",
       "Soon You'll Get Better (feat. Dixie Chicks)  0.000  0.035  0.000  0.000   \n",
       "The Archer                                   0.000  0.000  0.000  0.084   \n",
       "You Need to Calm Down                        0.000  0.009  0.000  0.009   \n",
       "\n",
       "                                                4      5      6      7   \\\n",
       "Tim McGraw                                   0.055  0.000  0.364  0.000   \n",
       "Picture To Burn                              0.000  0.037  0.000  0.035   \n",
       "Teardrops On My Guitar                       0.062  0.008  0.016  0.039   \n",
       "A Place In This World                        0.000  0.046  0.000  0.000   \n",
       "Cold as You                                  0.005  0.082  0.060  0.043   \n",
       "...                                            ...    ...    ...    ...   \n",
       "Miss Americana & the Heartbreak Prince       0.000  0.000  0.045  0.156   \n",
       "Paper Rings                                  0.009  0.086  0.000  0.000   \n",
       "Soon You'll Get Better (feat. Dixie Chicks)  0.009  0.086  0.000  0.006   \n",
       "The Archer                                   0.370  0.000  0.000  0.064   \n",
       "You Need to Calm Down                        0.014  0.273  0.000  0.000   \n",
       "\n",
       "                                                8      9      10  \n",
       "Tim McGraw                                   0.000  0.043  0.048  \n",
       "Picture To Burn                              0.008  0.040  0.287  \n",
       "Teardrops On My Guitar                       0.000  0.192  0.077  \n",
       "A Place In This World                        0.000  0.000  0.000  \n",
       "Cold as You                                  0.197  0.073  0.013  \n",
       "...                                            ...    ...    ...  \n",
       "Miss Americana & the Heartbreak Prince       0.058  0.000  0.062  \n",
       "Paper Rings                                  0.105  0.000  0.122  \n",
       "Soon You'll Get Better (feat. Dixie Chicks)  0.000  0.659  0.044  \n",
       "The Archer                                   0.142  0.000  0.004  \n",
       "You Need to Calm Down                        0.045  0.026  0.015  \n",
       "\n",
       "[111 rows x 11 columns]"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "h11 = nmf_HMatrix(11,df_tfidf,cv_tfidf)\n",
    "h11"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## K-Means Clustering on NMF (with TF-IDF):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "kmeans9 = KMeans(n_clusters=5,random_state=555)\n",
    "clustering_ori9 = kmeans9.fit_predict(h9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Miss Americana &amp; the Heartbreak Prince</th>\n",
       "      <td>0.139</td>\n",
       "      <td>0.148</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.048</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.038</td>\n",
       "      <td>0.059</td>\n",
       "      <td>0.154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Paper Rings</th>\n",
       "      <td>0.026</td>\n",
       "      <td>0.107</td>\n",
       "      <td>0.054</td>\n",
       "      <td>0.045</td>\n",
       "      <td>0.009</td>\n",
       "      <td>0.112</td>\n",
       "      <td>0.003</td>\n",
       "      <td>0.113</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Soon You'll Get Better (feat. Dixie Chicks)</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.032</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.001</td>\n",
       "      <td>0.024</td>\n",
       "      <td>0.487</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Archer</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.078</td>\n",
       "      <td>0.358</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.138</td>\n",
       "      <td>0.049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>You Need to Calm Down</th>\n",
       "      <td>0.000</td>\n",
       "      <td>0.012</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.023</td>\n",
       "      <td>0.008</td>\n",
       "      <td>0.151</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.072</td>\n",
       "      <td>0.048</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 0      1      2      3  \\\n",
       "Miss Americana & the Heartbreak Prince       0.139  0.148  0.008  0.048   \n",
       "Paper Rings                                  0.026  0.107  0.054  0.045   \n",
       "Soon You'll Get Better (feat. Dixie Chicks)  0.000  0.032  0.000  0.001   \n",
       "The Archer                                   0.000  0.000  0.000  0.078   \n",
       "You Need to Calm Down                        0.000  0.012  0.000  0.023   \n",
       "\n",
       "                                                 4      5      6      7      8  \n",
       "Miss Americana & the Heartbreak Prince       0.000  0.000  0.038  0.059  0.154  \n",
       "Paper Rings                                  0.009  0.112  0.003  0.113  0.000  \n",
       "Soon You'll Get Better (feat. Dixie Chicks)  0.024  0.487  0.000  0.000  0.000  \n",
       "The Archer                                   0.358  0.000  0.000  0.138  0.049  \n",
       "You Need to Calm Down                        0.008  0.151  0.000  0.072  0.048  "
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "h9.tail(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.01514286, 0.04971429, 0.03085714, 0.3645    , 0.02928571,\n",
       "        0.05471429, 0.03785714, 0.01528571, 0.01785714],\n",
       "       [0.02329167, 0.04733333, 0.03933333, 0.033625  , 0.01133333,\n",
       "        0.26366667, 0.03154167, 0.04620833, 0.018625  ],\n",
       "       [0.03475   , 0.0471875 , 0.284     , 0.0330625 , 0.177125  ,\n",
       "        0.040125  , 0.0391875 , 0.016875  , 0.021375  ],\n",
       "       [0.06438298, 0.02614894, 0.04942553, 0.03370213, 0.01719149,\n",
       "        0.02810638, 0.11795745, 0.07742553, 0.09861702],\n",
       "       [0.0243    , 0.4189    , 0.0528    , 0.0201    , 0.0267    ,\n",
       "        0.0623    , 0.0249    , 0.0469    , 0.0333    ]])"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "kmeans9.cluster_centers_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Clusters with Relation to Topics (based on cluster_centers_)\n",
    "\n",
    "Cluster 0: Mostly about Topic 3\n",
    "\n",
    "Cluster 1: Mostly about Topic 5\n",
    "\n",
    "Cluster 2: Mostly about Topics 2 and 4\n",
    "\n",
    "Cluster 3: Mostly about Topics 6, 8, 7, and 0\n",
    "\n",
    "Cluster 4: Mostly about Topic 1. Some Topic 5 and 2."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualization with DBSCAN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "dbscan9 = DBSCAN(eps=0.2,min_samples=3).fit(h9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0,  0,  0,  0,  0,  0,  0, -1,  0,  0,  0,  0,  0,  1,  0,  0, -1,\n",
       "        0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0, -1,  0,  0,\n",
       "        2,  0,  0,  0,  0,  0,  0,  0, -1,  0,  1,  0,  0,  0,  0,  0,  0,\n",
       "        0,  2,  0,  1,  0,  0,  0,  0,  0,  0, -1,  0,  0,  0, -1,  1,  2,\n",
       "        0, -1,  0,  0,  0,  0,  0,  0,  0,  0, -1,  0,  0,  1,  0,  0,  0,\n",
       "       -1,  0,  0,  0,  0,  0,  0, -1,  0,  0,  0,  0,  1,  0, -1,  0, -1,\n",
       "       -1, -1,  0, -1,  0,  0,  0, -1,  0])"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dbscan9.labels_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = StandardScaler().fit_transform(h9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "db = DBSCAN(eps=0.2, min_samples=4).fit(X)\n",
    "\n",
    "# Let's find the observations DBSCAN called \"core\"\n",
    "core_samples_mask = np.zeros_like(db.labels_, dtype=bool)\n",
    "core_samples_mask[db.core_sample_indices_] = True\n",
    "labels = db.labels_\n",
    "\n",
    "# Number of clusters in labels, ignoring noise if present.\n",
    "n_clusters_ = len(set(labels)) - (1 if -1 in labels else 0)\n",
    "\n",
    "unique_labels = set(labels)\n",
    "colors = plt.cm.rainbow(np.linspace(0, 1, len(unique_labels)))\n",
    "plt.figure(dpi=200)\n",
    "show_core = True\n",
    "show_non_core = True\n",
    "for k, col in zip(unique_labels, colors):\n",
    "    if k == -1:\n",
    "        # Black used for noise.\n",
    "        col = 'k'\n",
    "\n",
    "    class_member_mask = (labels == k)\n",
    "    if show_core:\n",
    "        xy = X[class_member_mask & core_samples_mask]\n",
    "        x, y = xy[:,0], xy[:,1]\n",
    "        plt.scatter(x, y, c=col, edgecolors='k',  s=20, linewidths=1.1) # add black border for core points\n",
    "    \n",
    "    if show_non_core:\n",
    "        xy = X[class_member_mask & ~core_samples_mask]\n",
    "        x, y = xy[:,0], xy[:,1]\n",
    "        plt.scatter(x, y, c=col, s=20, linewidths=1.1)\n",
    "\n",
    "plt.title('Estimated number of clusters: %d' % n_clusters_);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualization with TSNE"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "Xtsne9 = TSNE(n_components=2,random_state=467).fit_transform(h9)\n",
    "dftsne9 = pd.DataFrame(Xtsne9)\n",
    "dftsne9['cluster'] = clustering_ori9\n",
    "dftsne9.columns = ['x1','x2','cluster']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "dftsne2d9 = dftsne9\n",
    "dftsne2d9['cluster'] = clustering_ori9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "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>x1</th>\n",
       "      <th>x2</th>\n",
       "      <th>cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-5.565605</td>\n",
       "      <td>-2.144294</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.037211</td>\n",
       "      <td>-2.131413</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.306935</td>\n",
       "      <td>-2.582134</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.567557</td>\n",
       "      <td>0.992481</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.459243</td>\n",
       "      <td>-1.283168</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         x1        x2  cluster\n",
       "0 -5.565605 -2.144294        3\n",
       "1  4.037211 -2.131413        1\n",
       "2 -0.306935 -2.582134        2\n",
       "3  0.567557  0.992481        3\n",
       "4 -0.459243 -1.283168        3"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dftsne2d9.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "tableau_df = pd.concat([spotify,dftsne2d9],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "scrolled": false
   },
   "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>track_album</th>\n",
       "      <th>track_artist</th>\n",
       "      <th>track_title</th>\n",
       "      <th>danceability</th>\n",
       "      <th>energy</th>\n",
       "      <th>key</th>\n",
       "      <th>loudness</th>\n",
       "      <th>mode</th>\n",
       "      <th>speechiness</th>\n",
       "      <th>acousticness</th>\n",
       "      <th>...</th>\n",
       "      <th>time_signature</th>\n",
       "      <th>track_uri</th>\n",
       "      <th>track_id</th>\n",
       "      <th>track_lyric</th>\n",
       "      <th>genres</th>\n",
       "      <th>year_released</th>\n",
       "      <th>world_sales_USD</th>\n",
       "      <th>x1</th>\n",
       "      <th>x2</th>\n",
       "      <th>cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Tim McGraw</td>\n",
       "      <td>0.580</td>\n",
       "      <td>0.491</td>\n",
       "      <td>0</td>\n",
       "      <td>-6.462</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0251</td>\n",
       "      <td>0.575</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:0Om9WAB5RS09L80DyOfTNa</td>\n",
       "      <td>0Om9WAB5RS09L80DyOfTNa</td>\n",
       "      <td>he said the way my blue eyes shined put those ...</td>\n",
       "      <td>Country</td>\n",
       "      <td>2006</td>\n",
       "      <td>7000000</td>\n",
       "      <td>-5.565605</td>\n",
       "      <td>-2.144294</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Picture To Burn</td>\n",
       "      <td>0.658</td>\n",
       "      <td>0.877</td>\n",
       "      <td>7</td>\n",
       "      <td>-2.098</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0323</td>\n",
       "      <td>0.173</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:32mVHdy0bi1XKgr0ajsBlG</td>\n",
       "      <td>32mVHdy0bi1XKgr0ajsBlG</td>\n",
       "      <td>state the obvious  i didn t get my perfect fan...</td>\n",
       "      <td>Country Rock</td>\n",
       "      <td>2006</td>\n",
       "      <td>7000000</td>\n",
       "      <td>4.037211</td>\n",
       "      <td>-2.131413</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Teardrops On My Guitar</td>\n",
       "      <td>0.621</td>\n",
       "      <td>0.417</td>\n",
       "      <td>10</td>\n",
       "      <td>-6.941</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0231</td>\n",
       "      <td>0.288</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:7zMcNqs55Mxer82bvZFkpg</td>\n",
       "      <td>7zMcNqs55Mxer82bvZFkpg</td>\n",
       "      <td>drew looks at me i fake a smile so he won t se...</td>\n",
       "      <td>Country Pop</td>\n",
       "      <td>2006</td>\n",
       "      <td>7000000</td>\n",
       "      <td>-0.306935</td>\n",
       "      <td>-2.582134</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>A Place In This World</td>\n",
       "      <td>0.576</td>\n",
       "      <td>0.777</td>\n",
       "      <td>9</td>\n",
       "      <td>-2.881</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0324</td>\n",
       "      <td>0.051</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:73OX8GdpOeGzKC6OvGSbsv</td>\n",
       "      <td>73OX8GdpOeGzKC6OvGSbsv</td>\n",
       "      <td>i don t know what i want  so don t ask me caus...</td>\n",
       "      <td>Country Pop</td>\n",
       "      <td>2006</td>\n",
       "      <td>7000000</td>\n",
       "      <td>0.567557</td>\n",
       "      <td>0.992481</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Cold as You</td>\n",
       "      <td>0.418</td>\n",
       "      <td>0.482</td>\n",
       "      <td>5</td>\n",
       "      <td>-5.769</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0266</td>\n",
       "      <td>0.217</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:7an1exwMnfYRcdVQm0yDev</td>\n",
       "      <td>7an1exwMnfYRcdVQm0yDev</td>\n",
       "      <td>you have a way of coming easily to me and when...</td>\n",
       "      <td>Country</td>\n",
       "      <td>2006</td>\n",
       "      <td>7000000</td>\n",
       "      <td>-0.459243</td>\n",
       "      <td>-1.283168</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    track_album  track_artist             track_title  danceability  energy  \\\n",
       "0  Taylor Swift  Taylor Swift              Tim McGraw         0.580   0.491   \n",
       "1  Taylor Swift  Taylor Swift         Picture To Burn         0.658   0.877   \n",
       "2  Taylor Swift  Taylor Swift  Teardrops On My Guitar         0.621   0.417   \n",
       "3  Taylor Swift  Taylor Swift   A Place In This World         0.576   0.777   \n",
       "4  Taylor Swift  Taylor Swift             Cold as You         0.418   0.482   \n",
       "\n",
       "   key  loudness  mode  speechiness  acousticness  ...  time_signature  \\\n",
       "0    0    -6.462     1       0.0251         0.575  ...               4   \n",
       "1    7    -2.098     1       0.0323         0.173  ...               4   \n",
       "2   10    -6.941     1       0.0231         0.288  ...               4   \n",
       "3    9    -2.881     1       0.0324         0.051  ...               4   \n",
       "4    5    -5.769     1       0.0266         0.217  ...               4   \n",
       "\n",
       "                              track_uri                track_id  \\\n",
       "0  spotify:track:0Om9WAB5RS09L80DyOfTNa  0Om9WAB5RS09L80DyOfTNa   \n",
       "1  spotify:track:32mVHdy0bi1XKgr0ajsBlG  32mVHdy0bi1XKgr0ajsBlG   \n",
       "2  spotify:track:7zMcNqs55Mxer82bvZFkpg  7zMcNqs55Mxer82bvZFkpg   \n",
       "3  spotify:track:73OX8GdpOeGzKC6OvGSbsv  73OX8GdpOeGzKC6OvGSbsv   \n",
       "4  spotify:track:7an1exwMnfYRcdVQm0yDev  7an1exwMnfYRcdVQm0yDev   \n",
       "\n",
       "                                         track_lyric        genres  \\\n",
       "0  he said the way my blue eyes shined put those ...       Country   \n",
       "1  state the obvious  i didn t get my perfect fan...  Country Rock   \n",
       "2  drew looks at me i fake a smile so he won t se...   Country Pop   \n",
       "3  i don t know what i want  so don t ask me caus...   Country Pop   \n",
       "4  you have a way of coming easily to me and when...       Country   \n",
       "\n",
       "   year_released world_sales_USD        x1        x2 cluster  \n",
       "0           2006         7000000 -5.565605 -2.144294       3  \n",
       "1           2006         7000000  4.037211 -2.131413       1  \n",
       "2           2006         7000000 -0.306935 -2.582134       2  \n",
       "3           2006         7000000  0.567557  0.992481       3  \n",
       "4           2006         7000000 -0.459243 -1.283168       3  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tableau_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "tableau_df.to_csv(r'full_df2.csv',index=False, header=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "kmeans11 = KMeans(n_clusters=5,random_state=170)\n",
    "clustering_ori11 = kmeans11.fit_predict(h11)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "Xtsne = TSNE(n_components=2,random_state=879).fit_transform(h11)\n",
    "dftsne = pd.DataFrame(Xtsne)\n",
    "dftsne['cluster'] = clustering_ori11\n",
    "dftsne.columns = ['x1','x2','cluster']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "dftsne2d11 = dftsne\n",
    "dftsne2d11['cluster'] = clustering_ori11"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "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>x1</th>\n",
       "      <th>x2</th>\n",
       "      <th>cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.893362</td>\n",
       "      <td>-5.143504</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-2.943644</td>\n",
       "      <td>-0.703904</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.523659</td>\n",
       "      <td>1.260154</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2.376999</td>\n",
       "      <td>-4.124897</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.987731</td>\n",
       "      <td>-1.413858</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>3.221255</td>\n",
       "      <td>-1.897877</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>0.254182</td>\n",
       "      <td>-2.292215</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>-2.188454</td>\n",
       "      <td>-8.961349</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>0.636082</td>\n",
       "      <td>3.748459</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>-1.046318</td>\n",
       "      <td>-3.882412</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>111 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           x1        x2  cluster\n",
       "0    5.893362 -5.143504        0\n",
       "1   -2.943644 -0.703904        4\n",
       "2    3.523659  1.260154        0\n",
       "3    2.376999 -4.124897        0\n",
       "4    0.987731 -1.413858        0\n",
       "..        ...       ...      ...\n",
       "106  3.221255 -1.897877        0\n",
       "107  0.254182 -2.292215        0\n",
       "108 -2.188454 -8.961349        1\n",
       "109  0.636082  3.748459        2\n",
       "110 -1.046318 -3.882412        0\n",
       "\n",
       "[111 rows x 3 columns]"
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dftsne2d11"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(1, 2, figsize=(12,6))\n",
    "sns.scatterplot(data=dftsne2d9,x='x1',y='x2',hue='cluster',legend=\"full\",alpha=0.7,ax=ax[0],palette='Paired')\n",
    "ax[0].set_title('Visualized on TSNE 5D>2D')\n",
    "ax[0].legend(bbox_to_anchor=(-0.3,1),loc=2)\n",
    "sns.scatterplot(data=dftsne2d11,x='x1',y='x2',hue='cluster',legend=\"full\",alpha=0.7,ax=ax[1],palette='Paired')\n",
    "ax[1].set_title('Visualized on TSNE 9 topics')\n",
    "ax[1].legend(bbox_to_anchor=(0.9,1),loc=2)\n",
    "fig.suptitle('Comparing clustering result when visualized using TSNE 5D>2D vs. TSNE 34D>2D')\n",
    "display(fig)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Mean World Wide Sales Per Cluster:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_df = tableau_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "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>track_album</th>\n",
       "      <th>track_artist</th>\n",
       "      <th>track_title</th>\n",
       "      <th>danceability</th>\n",
       "      <th>energy</th>\n",
       "      <th>key</th>\n",
       "      <th>loudness</th>\n",
       "      <th>mode</th>\n",
       "      <th>speechiness</th>\n",
       "      <th>acousticness</th>\n",
       "      <th>...</th>\n",
       "      <th>time_signature</th>\n",
       "      <th>track_uri</th>\n",
       "      <th>track_id</th>\n",
       "      <th>track_lyric</th>\n",
       "      <th>genres</th>\n",
       "      <th>year_released</th>\n",
       "      <th>world_sales_USD</th>\n",
       "      <th>x1</th>\n",
       "      <th>x2</th>\n",
       "      <th>cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Tim McGraw</td>\n",
       "      <td>0.580</td>\n",
       "      <td>0.491</td>\n",
       "      <td>0</td>\n",
       "      <td>-6.462</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0251</td>\n",
       "      <td>0.575</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:0Om9WAB5RS09L80DyOfTNa</td>\n",
       "      <td>0Om9WAB5RS09L80DyOfTNa</td>\n",
       "      <td>he said the way my blue eyes shined put those ...</td>\n",
       "      <td>Country</td>\n",
       "      <td>2006</td>\n",
       "      <td>7000000</td>\n",
       "      <td>-5.565605</td>\n",
       "      <td>-2.144294</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Picture To Burn</td>\n",
       "      <td>0.658</td>\n",
       "      <td>0.877</td>\n",
       "      <td>7</td>\n",
       "      <td>-2.098</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0323</td>\n",
       "      <td>0.173</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:32mVHdy0bi1XKgr0ajsBlG</td>\n",
       "      <td>32mVHdy0bi1XKgr0ajsBlG</td>\n",
       "      <td>state the obvious  i didn t get my perfect fan...</td>\n",
       "      <td>Country Rock</td>\n",
       "      <td>2006</td>\n",
       "      <td>7000000</td>\n",
       "      <td>4.037211</td>\n",
       "      <td>-2.131413</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Teardrops On My Guitar</td>\n",
       "      <td>0.621</td>\n",
       "      <td>0.417</td>\n",
       "      <td>10</td>\n",
       "      <td>-6.941</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0231</td>\n",
       "      <td>0.288</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:7zMcNqs55Mxer82bvZFkpg</td>\n",
       "      <td>7zMcNqs55Mxer82bvZFkpg</td>\n",
       "      <td>drew looks at me i fake a smile so he won t se...</td>\n",
       "      <td>Country Pop</td>\n",
       "      <td>2006</td>\n",
       "      <td>7000000</td>\n",
       "      <td>-0.306935</td>\n",
       "      <td>-2.582134</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>A Place In This World</td>\n",
       "      <td>0.576</td>\n",
       "      <td>0.777</td>\n",
       "      <td>9</td>\n",
       "      <td>-2.881</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0324</td>\n",
       "      <td>0.051</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:73OX8GdpOeGzKC6OvGSbsv</td>\n",
       "      <td>73OX8GdpOeGzKC6OvGSbsv</td>\n",
       "      <td>i don t know what i want  so don t ask me caus...</td>\n",
       "      <td>Country Pop</td>\n",
       "      <td>2006</td>\n",
       "      <td>7000000</td>\n",
       "      <td>0.567557</td>\n",
       "      <td>0.992481</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Taylor Swift</td>\n",
       "      <td>Cold as You</td>\n",
       "      <td>0.418</td>\n",
       "      <td>0.482</td>\n",
       "      <td>5</td>\n",
       "      <td>-5.769</td>\n",
       "      <td>1</td>\n",
       "      <td>0.0266</td>\n",
       "      <td>0.217</td>\n",
       "      <td>...</td>\n",
       "      <td>4</td>\n",
       "      <td>spotify:track:7an1exwMnfYRcdVQm0yDev</td>\n",
       "      <td>7an1exwMnfYRcdVQm0yDev</td>\n",
       "      <td>you have a way of coming easily to me and when...</td>\n",
       "      <td>Country</td>\n",
       "      <td>2006</td>\n",
       "      <td>7000000</td>\n",
       "      <td>-0.459243</td>\n",
       "      <td>-1.283168</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 25 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    track_album  track_artist             track_title  danceability  energy  \\\n",
       "0  Taylor Swift  Taylor Swift              Tim McGraw         0.580   0.491   \n",
       "1  Taylor Swift  Taylor Swift         Picture To Burn         0.658   0.877   \n",
       "2  Taylor Swift  Taylor Swift  Teardrops On My Guitar         0.621   0.417   \n",
       "3  Taylor Swift  Taylor Swift   A Place In This World         0.576   0.777   \n",
       "4  Taylor Swift  Taylor Swift             Cold as You         0.418   0.482   \n",
       "\n",
       "   key  loudness  mode  speechiness  acousticness  ...  time_signature  \\\n",
       "0    0    -6.462     1       0.0251         0.575  ...               4   \n",
       "1    7    -2.098     1       0.0323         0.173  ...               4   \n",
       "2   10    -6.941     1       0.0231         0.288  ...               4   \n",
       "3    9    -2.881     1       0.0324         0.051  ...               4   \n",
       "4    5    -5.769     1       0.0266         0.217  ...               4   \n",
       "\n",
       "                              track_uri                track_id  \\\n",
       "0  spotify:track:0Om9WAB5RS09L80DyOfTNa  0Om9WAB5RS09L80DyOfTNa   \n",
       "1  spotify:track:32mVHdy0bi1XKgr0ajsBlG  32mVHdy0bi1XKgr0ajsBlG   \n",
       "2  spotify:track:7zMcNqs55Mxer82bvZFkpg  7zMcNqs55Mxer82bvZFkpg   \n",
       "3  spotify:track:73OX8GdpOeGzKC6OvGSbsv  73OX8GdpOeGzKC6OvGSbsv   \n",
       "4  spotify:track:7an1exwMnfYRcdVQm0yDev  7an1exwMnfYRcdVQm0yDev   \n",
       "\n",
       "                                         track_lyric        genres  \\\n",
       "0  he said the way my blue eyes shined put those ...       Country   \n",
       "1  state the obvious  i didn t get my perfect fan...  Country Rock   \n",
       "2  drew looks at me i fake a smile so he won t se...   Country Pop   \n",
       "3  i don t know what i want  so don t ask me caus...   Country Pop   \n",
       "4  you have a way of coming easily to me and when...       Country   \n",
       "\n",
       "   year_released world_sales_USD        x1        x2 cluster  \n",
       "0           2006         7000000 -5.565605 -2.144294       3  \n",
       "1           2006         7000000  4.037211 -2.131413       1  \n",
       "2           2006         7000000 -0.306935 -2.582134       2  \n",
       "3           2006         7000000  0.567557  0.992481       3  \n",
       "4           2006         7000000 -0.459243 -1.283168       3  \n",
       "\n",
       "[5 rows x 25 columns]"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "final_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_cluster0 = final_df[final_df['cluster']==0]\n",
    "final_cluster1 = final_df[final_df['cluster']==1]\n",
    "final_cluster2 = final_df[final_df['cluster']==2]\n",
    "final_cluster3 = final_df[final_df['cluster']==3]\n",
    "final_cluster4 = final_df[final_df['cluster']==4]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5535714.285714285\n",
      "7916666.666666667\n",
      "7218750.0\n",
      "6861702.127659574\n",
      "6700000.0\n"
     ]
    }
   ],
   "source": [
    "print(final_cluster0['world_sales_USD'].mean())\n",
    "print(final_cluster1['world_sales_USD'].mean())\n",
    "print(final_cluster2['world_sales_USD'].mean())\n",
    "print(final_cluster3['world_sales_USD'].mean())\n",
    "print(final_cluster4['world_sales_USD'].mean())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "KM9 = KMeans(n_clusters=5,random_state=199)\n",
    "clustering_oriKM9 = kmeans9.fit_predict(h9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [],
   "source": [
    "Xtsne9 = TSNE(n_components=2,random_state=24).fit_transform(h9)\n",
    "tsne9 = pd.DataFrame(Xtsne9)\n",
    "tsne9['cluster'] = clustering_oriKM9\n",
    "tsne9.columns = ['x1','x2','cluster']\n",
    "\n",
    "tsne9['cluster'] = clustering_oriKM9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "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>x1</th>\n",
       "      <th>x2</th>\n",
       "      <th>cluster</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5.211853</td>\n",
       "      <td>7.981372</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-3.652538</td>\n",
       "      <td>1.448218</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-4.464827</td>\n",
       "      <td>4.929675</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.671046</td>\n",
       "      <td>8.984304</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.130948</td>\n",
       "      <td>5.479666</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>-0.080085</td>\n",
       "      <td>7.039864</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>-0.749628</td>\n",
       "      <td>4.254309</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>-0.974250</td>\n",
       "      <td>-0.404637</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>2.787287</td>\n",
       "      <td>-1.435306</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>-0.240991</td>\n",
       "      <td>3.837819</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>111 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "           x1        x2  cluster\n",
       "0    5.211853  7.981372        3\n",
       "1   -3.652538  1.448218        1\n",
       "2   -4.464827  4.929675        2\n",
       "3    0.671046  8.984304        3\n",
       "4   -1.130948  5.479666        3\n",
       "..        ...       ...      ...\n",
       "106 -0.080085  7.039864        3\n",
       "107 -0.749628  4.254309        1\n",
       "108 -0.974250 -0.404637        1\n",
       "109  2.787287 -1.435306        2\n",
       "110 -0.240991  3.837819        1\n",
       "\n",
       "[111 rows x 3 columns]"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tsne9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14\n",
      "24\n",
      "16\n",
      "47\n",
      "10\n"
     ]
    }
   ],
   "source": [
    "print(len(tsne9[tsne9['cluster']==0]))\n",
    "print(len(tsne9[tsne9['cluster']==1]))\n",
    "print(len(tsne9[tsne9['cluster']==2]))\n",
    "print(len(tsne9[tsne9['cluster']==3]))\n",
    "print(len(tsne9[tsne9['cluster']==4]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Exploring Similar Songs with Cosine Similarity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>101</th>\n",
       "      <th>102</th>\n",
       "      <th>103</th>\n",
       "      <th>104</th>\n",
       "      <th>105</th>\n",
       "      <th>106</th>\n",
       "      <th>107</th>\n",
       "      <th>108</th>\n",
       "      <th>109</th>\n",
       "      <th>110</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.067758</td>\n",
       "      <td>0.122150</td>\n",
       "      <td>0.059303</td>\n",
       "      <td>0.305801</td>\n",
       "      <td>0.079529</td>\n",
       "      <td>0.007457</td>\n",
       "      <td>0.133400</td>\n",
       "      <td>0.906208</td>\n",
       "      <td>0.808839</td>\n",
       "      <td>...</td>\n",
       "      <td>0.171491</td>\n",
       "      <td>0.220100</td>\n",
       "      <td>0.080442</td>\n",
       "      <td>0.237896</td>\n",
       "      <td>0.046670</td>\n",
       "      <td>0.164348</td>\n",
       "      <td>0.098197</td>\n",
       "      <td>0.086180</td>\n",
       "      <td>0.207002</td>\n",
       "      <td>0.094826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.067758</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.564585</td>\n",
       "      <td>0.073174</td>\n",
       "      <td>0.327065</td>\n",
       "      <td>0.744168</td>\n",
       "      <td>0.763424</td>\n",
       "      <td>0.158216</td>\n",
       "      <td>0.242070</td>\n",
       "      <td>0.129719</td>\n",
       "      <td>...</td>\n",
       "      <td>0.793062</td>\n",
       "      <td>0.100091</td>\n",
       "      <td>0.246692</td>\n",
       "      <td>0.253967</td>\n",
       "      <td>0.705824</td>\n",
       "      <td>0.492484</td>\n",
       "      <td>0.827106</td>\n",
       "      <td>0.700721</td>\n",
       "      <td>0.045179</td>\n",
       "      <td>0.664158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.122150</td>\n",
       "      <td>0.564585</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.304232</td>\n",
       "      <td>0.180402</td>\n",
       "      <td>0.431211</td>\n",
       "      <td>0.257418</td>\n",
       "      <td>0.756174</td>\n",
       "      <td>0.107208</td>\n",
       "      <td>0.457543</td>\n",
       "      <td>...</td>\n",
       "      <td>0.392652</td>\n",
       "      <td>0.252233</td>\n",
       "      <td>0.677175</td>\n",
       "      <td>0.658163</td>\n",
       "      <td>0.170345</td>\n",
       "      <td>0.331122</td>\n",
       "      <td>0.560281</td>\n",
       "      <td>0.431140</td>\n",
       "      <td>0.252943</td>\n",
       "      <td>0.397129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.059303</td>\n",
       "      <td>0.073174</td>\n",
       "      <td>0.304232</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.006569</td>\n",
       "      <td>0.009653</td>\n",
       "      <td>0.008976</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.052445</td>\n",
       "      <td>0.171906</td>\n",
       "      <td>...</td>\n",
       "      <td>0.212248</td>\n",
       "      <td>0.631752</td>\n",
       "      <td>0.196284</td>\n",
       "      <td>0.870973</td>\n",
       "      <td>0.055882</td>\n",
       "      <td>0.546240</td>\n",
       "      <td>0.199744</td>\n",
       "      <td>0.000778</td>\n",
       "      <td>0.075197</td>\n",
       "      <td>0.049671</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.305801</td>\n",
       "      <td>0.327065</td>\n",
       "      <td>0.180402</td>\n",
       "      <td>0.006569</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.344868</td>\n",
       "      <td>0.130020</td>\n",
       "      <td>0.014830</td>\n",
       "      <td>0.301250</td>\n",
       "      <td>0.445539</td>\n",
       "      <td>...</td>\n",
       "      <td>0.200605</td>\n",
       "      <td>0.024420</td>\n",
       "      <td>0.015208</td>\n",
       "      <td>0.136646</td>\n",
       "      <td>0.112778</td>\n",
       "      <td>0.427096</td>\n",
       "      <td>0.705210</td>\n",
       "      <td>0.317411</td>\n",
       "      <td>0.356531</td>\n",
       "      <td>0.697098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.079529</td>\n",
       "      <td>0.744168</td>\n",
       "      <td>0.431211</td>\n",
       "      <td>0.009653</td>\n",
       "      <td>0.344868</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.172851</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.040293</td>\n",
       "      <td>0.002790</td>\n",
       "      <td>...</td>\n",
       "      <td>0.363969</td>\n",
       "      <td>0.019962</td>\n",
       "      <td>0.015884</td>\n",
       "      <td>0.110198</td>\n",
       "      <td>0.118066</td>\n",
       "      <td>0.131135</td>\n",
       "      <td>0.603353</td>\n",
       "      <td>0.991333</td>\n",
       "      <td>0.018322</td>\n",
       "      <td>0.886560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.007457</td>\n",
       "      <td>0.763424</td>\n",
       "      <td>0.257418</td>\n",
       "      <td>0.008976</td>\n",
       "      <td>0.130020</td>\n",
       "      <td>0.172851</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.077941</td>\n",
       "      <td>0.337200</td>\n",
       "      <td>0.072715</td>\n",
       "      <td>...</td>\n",
       "      <td>0.851136</td>\n",
       "      <td>0.018562</td>\n",
       "      <td>0.105515</td>\n",
       "      <td>0.139194</td>\n",
       "      <td>0.988062</td>\n",
       "      <td>0.586795</td>\n",
       "      <td>0.575807</td>\n",
       "      <td>0.114120</td>\n",
       "      <td>0.012607</td>\n",
       "      <td>0.129552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.133400</td>\n",
       "      <td>0.158216</td>\n",
       "      <td>0.756174</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.014830</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.077941</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.331034</td>\n",
       "      <td>...</td>\n",
       "      <td>0.278286</td>\n",
       "      <td>0.344796</td>\n",
       "      <td>0.622188</td>\n",
       "      <td>0.379788</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.021631</td>\n",
       "      <td>0.220419</td>\n",
       "      <td>0.033741</td>\n",
       "      <td>0.623281</td>\n",
       "      <td>0.031201</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.906208</td>\n",
       "      <td>0.242070</td>\n",
       "      <td>0.107208</td>\n",
       "      <td>0.052445</td>\n",
       "      <td>0.301250</td>\n",
       "      <td>0.040293</td>\n",
       "      <td>0.337200</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.804909</td>\n",
       "      <td>...</td>\n",
       "      <td>0.333650</td>\n",
       "      <td>0.020389</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.224635</td>\n",
       "      <td>0.388030</td>\n",
       "      <td>0.349404</td>\n",
       "      <td>0.197482</td>\n",
       "      <td>0.022282</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.023180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.808839</td>\n",
       "      <td>0.129719</td>\n",
       "      <td>0.457543</td>\n",
       "      <td>0.171906</td>\n",
       "      <td>0.445539</td>\n",
       "      <td>0.002790</td>\n",
       "      <td>0.072715</td>\n",
       "      <td>0.331034</td>\n",
       "      <td>0.804909</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.082751</td>\n",
       "      <td>0.066832</td>\n",
       "      <td>0.390029</td>\n",
       "      <td>0.442532</td>\n",
       "      <td>0.075221</td>\n",
       "      <td>0.292844</td>\n",
       "      <td>0.296447</td>\n",
       "      <td>0.001543</td>\n",
       "      <td>0.082407</td>\n",
       "      <td>0.097919</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.750550</td>\n",
       "      <td>0.143976</td>\n",
       "      <td>0.236148</td>\n",
       "      <td>0.367996</td>\n",
       "      <td>0.590149</td>\n",
       "      <td>0.051421</td>\n",
       "      <td>0.137904</td>\n",
       "      <td>0.024096</td>\n",
       "      <td>0.791110</td>\n",
       "      <td>0.816831</td>\n",
       "      <td>...</td>\n",
       "      <td>0.205990</td>\n",
       "      <td>0.143065</td>\n",
       "      <td>0.037602</td>\n",
       "      <td>0.553043</td>\n",
       "      <td>0.175495</td>\n",
       "      <td>0.651003</td>\n",
       "      <td>0.301230</td>\n",
       "      <td>0.007480</td>\n",
       "      <td>0.153451</td>\n",
       "      <td>0.232178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.146410</td>\n",
       "      <td>0.576124</td>\n",
       "      <td>0.428751</td>\n",
       "      <td>0.518775</td>\n",
       "      <td>0.248299</td>\n",
       "      <td>0.771262</td>\n",
       "      <td>0.050735</td>\n",
       "      <td>0.009250</td>\n",
       "      <td>0.019090</td>\n",
       "      <td>0.062574</td>\n",
       "      <td>...</td>\n",
       "      <td>0.362971</td>\n",
       "      <td>0.556788</td>\n",
       "      <td>0.281396</td>\n",
       "      <td>0.479229</td>\n",
       "      <td>0.020341</td>\n",
       "      <td>0.271497</td>\n",
       "      <td>0.579275</td>\n",
       "      <td>0.766724</td>\n",
       "      <td>0.120020</td>\n",
       "      <td>0.729791</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.163095</td>\n",
       "      <td>0.394946</td>\n",
       "      <td>0.839327</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.139152</td>\n",
       "      <td>0.044865</td>\n",
       "      <td>0.359090</td>\n",
       "      <td>0.814874</td>\n",
       "      <td>0.205727</td>\n",
       "      <td>0.544440</td>\n",
       "      <td>...</td>\n",
       "      <td>0.310127</td>\n",
       "      <td>0.104479</td>\n",
       "      <td>0.795201</td>\n",
       "      <td>0.420751</td>\n",
       "      <td>0.259501</td>\n",
       "      <td>0.274512</td>\n",
       "      <td>0.410353</td>\n",
       "      <td>0.026866</td>\n",
       "      <td>0.222060</td>\n",
       "      <td>0.083124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.152249</td>\n",
       "      <td>0.314622</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.231110</td>\n",
       "      <td>0.102415</td>\n",
       "      <td>0.086869</td>\n",
       "      <td>0.213471</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.133818</td>\n",
       "      <td>...</td>\n",
       "      <td>0.051220</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.276409</td>\n",
       "      <td>0.303933</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.556457</td>\n",
       "      <td>0.076975</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.118704</td>\n",
       "      <td>0.260484</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.257667</td>\n",
       "      <td>0.355705</td>\n",
       "      <td>0.187964</td>\n",
       "      <td>0.353658</td>\n",
       "      <td>0.135261</td>\n",
       "      <td>0.323783</td>\n",
       "      <td>0.118209</td>\n",
       "      <td>0.136312</td>\n",
       "      <td>0.078292</td>\n",
       "      <td>0.046771</td>\n",
       "      <td>...</td>\n",
       "      <td>0.416806</td>\n",
       "      <td>0.830942</td>\n",
       "      <td>0.479723</td>\n",
       "      <td>0.314611</td>\n",
       "      <td>0.084831</td>\n",
       "      <td>0.219032</td>\n",
       "      <td>0.415230</td>\n",
       "      <td>0.310357</td>\n",
       "      <td>0.363814</td>\n",
       "      <td>0.388318</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>15 rows × 111 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         0         1         2         3         4         5         6    \\\n",
       "0   1.000000  0.067758  0.122150  0.059303  0.305801  0.079529  0.007457   \n",
       "1   0.067758  1.000000  0.564585  0.073174  0.327065  0.744168  0.763424   \n",
       "2   0.122150  0.564585  1.000000  0.304232  0.180402  0.431211  0.257418   \n",
       "3   0.059303  0.073174  0.304232  1.000000  0.006569  0.009653  0.008976   \n",
       "4   0.305801  0.327065  0.180402  0.006569  1.000000  0.344868  0.130020   \n",
       "5   0.079529  0.744168  0.431211  0.009653  0.344868  1.000000  0.172851   \n",
       "6   0.007457  0.763424  0.257418  0.008976  0.130020  0.172851  1.000000   \n",
       "7   0.133400  0.158216  0.756174  0.000000  0.014830  0.000000  0.077941   \n",
       "8   0.906208  0.242070  0.107208  0.052445  0.301250  0.040293  0.337200   \n",
       "9   0.808839  0.129719  0.457543  0.171906  0.445539  0.002790  0.072715   \n",
       "10  0.750550  0.143976  0.236148  0.367996  0.590149  0.051421  0.137904   \n",
       "11  0.146410  0.576124  0.428751  0.518775  0.248299  0.771262  0.050735   \n",
       "12  0.163095  0.394946  0.839327  0.000000  0.139152  0.044865  0.359090   \n",
       "13  0.000000  0.152249  0.314622  0.000000  0.231110  0.102415  0.086869   \n",
       "14  0.257667  0.355705  0.187964  0.353658  0.135261  0.323783  0.118209   \n",
       "\n",
       "         7         8         9    ...       101       102       103       104  \\\n",
       "0   0.133400  0.906208  0.808839  ...  0.171491  0.220100  0.080442  0.237896   \n",
       "1   0.158216  0.242070  0.129719  ...  0.793062  0.100091  0.246692  0.253967   \n",
       "2   0.756174  0.107208  0.457543  ...  0.392652  0.252233  0.677175  0.658163   \n",
       "3   0.000000  0.052445  0.171906  ...  0.212248  0.631752  0.196284  0.870973   \n",
       "4   0.014830  0.301250  0.445539  ...  0.200605  0.024420  0.015208  0.136646   \n",
       "5   0.000000  0.040293  0.002790  ...  0.363969  0.019962  0.015884  0.110198   \n",
       "6   0.077941  0.337200  0.072715  ...  0.851136  0.018562  0.105515  0.139194   \n",
       "7   1.000000  0.000000  0.331034  ...  0.278286  0.344796  0.622188  0.379788   \n",
       "8   0.000000  1.000000  0.804909  ...  0.333650  0.020389  0.000000  0.224635   \n",
       "9   0.331034  0.804909  1.000000  ...  0.082751  0.066832  0.390029  0.442532   \n",
       "10  0.024096  0.791110  0.816831  ...  0.205990  0.143065  0.037602  0.553043   \n",
       "11  0.009250  0.019090  0.062574  ...  0.362971  0.556788  0.281396  0.479229   \n",
       "12  0.814874  0.205727  0.544440  ...  0.310127  0.104479  0.795201  0.420751   \n",
       "13  0.213471  0.000000  0.133818  ...  0.051220  0.000000  0.276409  0.303933   \n",
       "14  0.136312  0.078292  0.046771  ...  0.416806  0.830942  0.479723  0.314611   \n",
       "\n",
       "         105       106       107       108       109       110  \n",
       "0   0.046670  0.164348  0.098197  0.086180  0.207002  0.094826  \n",
       "1   0.705824  0.492484  0.827106  0.700721  0.045179  0.664158  \n",
       "2   0.170345  0.331122  0.560281  0.431140  0.252943  0.397129  \n",
       "3   0.055882  0.546240  0.199744  0.000778  0.075197  0.049671  \n",
       "4   0.112778  0.427096  0.705210  0.317411  0.356531  0.697098  \n",
       "5   0.118066  0.131135  0.603353  0.991333  0.018322  0.886560  \n",
       "6   0.988062  0.586795  0.575807  0.114120  0.012607  0.129552  \n",
       "7   0.000000  0.021631  0.220419  0.033741  0.623281  0.031201  \n",
       "8   0.388030  0.349404  0.197482  0.022282  0.000000  0.023180  \n",
       "9   0.075221  0.292844  0.296447  0.001543  0.082407  0.097919  \n",
       "10  0.175495  0.651003  0.301230  0.007480  0.153451  0.232178  \n",
       "11  0.020341  0.271497  0.579275  0.766724  0.120020  0.729791  \n",
       "12  0.259501  0.274512  0.410353  0.026866  0.222060  0.083124  \n",
       "13  0.000000  0.556457  0.076975  0.000000  0.118704  0.260484  \n",
       "14  0.084831  0.219032  0.415230  0.310357  0.363814  0.388318  \n",
       "\n",
       "[15 rows x 111 columns]"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc_similarity_matrix9 = pd.DataFrame(cosine_similarity(h9))\n",
    "doc_similarity_matrix9.head(15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "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>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>101</th>\n",
       "      <th>102</th>\n",
       "      <th>103</th>\n",
       "      <th>104</th>\n",
       "      <th>105</th>\n",
       "      <th>106</th>\n",
       "      <th>107</th>\n",
       "      <th>108</th>\n",
       "      <th>109</th>\n",
       "      <th>110</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>96</th>\n",
       "      <td>0.153583</td>\n",
       "      <td>0.235799</td>\n",
       "      <td>0.048011</td>\n",
       "      <td>0.387427</td>\n",
       "      <td>0.042330</td>\n",
       "      <td>0.048569</td>\n",
       "      <td>0.191967</td>\n",
       "      <td>0.014384</td>\n",
       "      <td>0.057248</td>\n",
       "      <td>0.015475</td>\n",
       "      <td>...</td>\n",
       "      <td>0.343644</td>\n",
       "      <td>0.779901</td>\n",
       "      <td>0.526238</td>\n",
       "      <td>0.292274</td>\n",
       "      <td>0.166300</td>\n",
       "      <td>0.295955</td>\n",
       "      <td>0.308211</td>\n",
       "      <td>0.012885</td>\n",
       "      <td>0.199283</td>\n",
       "      <td>0.149904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>97</th>\n",
       "      <td>0.026822</td>\n",
       "      <td>0.232006</td>\n",
       "      <td>0.112844</td>\n",
       "      <td>0.072935</td>\n",
       "      <td>0.254673</td>\n",
       "      <td>0.128860</td>\n",
       "      <td>0.232727</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.059688</td>\n",
       "      <td>0.005607</td>\n",
       "      <td>...</td>\n",
       "      <td>0.234048</td>\n",
       "      <td>0.138254</td>\n",
       "      <td>0.114028</td>\n",
       "      <td>0.274112</td>\n",
       "      <td>0.174041</td>\n",
       "      <td>0.686319</td>\n",
       "      <td>0.128973</td>\n",
       "      <td>0.011752</td>\n",
       "      <td>0.154412</td>\n",
       "      <td>0.298535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>98</th>\n",
       "      <td>0.140557</td>\n",
       "      <td>0.589646</td>\n",
       "      <td>0.876019</td>\n",
       "      <td>0.255198</td>\n",
       "      <td>0.124838</td>\n",
       "      <td>0.188862</td>\n",
       "      <td>0.528058</td>\n",
       "      <td>0.606544</td>\n",
       "      <td>0.290146</td>\n",
       "      <td>0.554501</td>\n",
       "      <td>...</td>\n",
       "      <td>0.446804</td>\n",
       "      <td>0.099213</td>\n",
       "      <td>0.714639</td>\n",
       "      <td>0.576250</td>\n",
       "      <td>0.457450</td>\n",
       "      <td>0.426441</td>\n",
       "      <td>0.557248</td>\n",
       "      <td>0.167169</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.148930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>99</th>\n",
       "      <td>0.454444</td>\n",
       "      <td>0.265143</td>\n",
       "      <td>0.633106</td>\n",
       "      <td>0.009034</td>\n",
       "      <td>0.301937</td>\n",
       "      <td>0.159143</td>\n",
       "      <td>0.110315</td>\n",
       "      <td>0.483200</td>\n",
       "      <td>0.430941</td>\n",
       "      <td>0.687540</td>\n",
       "      <td>...</td>\n",
       "      <td>0.080732</td>\n",
       "      <td>0.018681</td>\n",
       "      <td>0.596685</td>\n",
       "      <td>0.432949</td>\n",
       "      <td>0.022194</td>\n",
       "      <td>0.421569</td>\n",
       "      <td>0.239164</td>\n",
       "      <td>0.097350</td>\n",
       "      <td>0.076782</td>\n",
       "      <td>0.244984</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>100</th>\n",
       "      <td>0.924889</td>\n",
       "      <td>0.068724</td>\n",
       "      <td>0.096715</td>\n",
       "      <td>0.175867</td>\n",
       "      <td>0.259688</td>\n",
       "      <td>0.011732</td>\n",
       "      <td>0.019220</td>\n",
       "      <td>0.056307</td>\n",
       "      <td>0.829105</td>\n",
       "      <td>0.775317</td>\n",
       "      <td>...</td>\n",
       "      <td>0.132003</td>\n",
       "      <td>0.363685</td>\n",
       "      <td>0.304898</td>\n",
       "      <td>0.284921</td>\n",
       "      <td>0.042699</td>\n",
       "      <td>0.209766</td>\n",
       "      <td>0.134116</td>\n",
       "      <td>0.000946</td>\n",
       "      <td>0.091391</td>\n",
       "      <td>0.060368</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>101</th>\n",
       "      <td>0.171491</td>\n",
       "      <td>0.793062</td>\n",
       "      <td>0.392652</td>\n",
       "      <td>0.212248</td>\n",
       "      <td>0.200605</td>\n",
       "      <td>0.363969</td>\n",
       "      <td>0.851136</td>\n",
       "      <td>0.278286</td>\n",
       "      <td>0.333650</td>\n",
       "      <td>0.082751</td>\n",
       "      <td>...</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.415900</td>\n",
       "      <td>0.106462</td>\n",
       "      <td>0.360801</td>\n",
       "      <td>0.846419</td>\n",
       "      <td>0.609998</td>\n",
       "      <td>0.656081</td>\n",
       "      <td>0.331535</td>\n",
       "      <td>0.414446</td>\n",
       "      <td>0.337173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>0.220100</td>\n",
       "      <td>0.100091</td>\n",
       "      <td>0.252233</td>\n",
       "      <td>0.631752</td>\n",
       "      <td>0.024420</td>\n",
       "      <td>0.019962</td>\n",
       "      <td>0.018562</td>\n",
       "      <td>0.344796</td>\n",
       "      <td>0.020389</td>\n",
       "      <td>0.066832</td>\n",
       "      <td>...</td>\n",
       "      <td>0.415900</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.405907</td>\n",
       "      <td>0.602888</td>\n",
       "      <td>0.021725</td>\n",
       "      <td>0.326441</td>\n",
       "      <td>0.239055</td>\n",
       "      <td>0.026262</td>\n",
       "      <td>0.610885</td>\n",
       "      <td>0.125513</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>103</th>\n",
       "      <td>0.080442</td>\n",
       "      <td>0.246692</td>\n",
       "      <td>0.677175</td>\n",
       "      <td>0.196284</td>\n",
       "      <td>0.015208</td>\n",
       "      <td>0.015884</td>\n",
       "      <td>0.105515</td>\n",
       "      <td>0.622188</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.390029</td>\n",
       "      <td>...</td>\n",
       "      <td>0.106462</td>\n",
       "      <td>0.405907</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.423498</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.132563</td>\n",
       "      <td>0.337026</td>\n",
       "      <td>0.001056</td>\n",
       "      <td>0.105236</td>\n",
       "      <td>0.074474</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>104</th>\n",
       "      <td>0.237896</td>\n",
       "      <td>0.253967</td>\n",
       "      <td>0.658163</td>\n",
       "      <td>0.870973</td>\n",
       "      <td>0.136646</td>\n",
       "      <td>0.110198</td>\n",
       "      <td>0.139194</td>\n",
       "      <td>0.379788</td>\n",
       "      <td>0.224635</td>\n",
       "      <td>0.442532</td>\n",
       "      <td>...</td>\n",
       "      <td>0.360801</td>\n",
       "      <td>0.602888</td>\n",
       "      <td>0.423498</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.140059</td>\n",
       "      <td>0.679396</td>\n",
       "      <td>0.341614</td>\n",
       "      <td>0.087313</td>\n",
       "      <td>0.259005</td>\n",
       "      <td>0.168469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>0.046670</td>\n",
       "      <td>0.705824</td>\n",
       "      <td>0.170345</td>\n",
       "      <td>0.055882</td>\n",
       "      <td>0.112778</td>\n",
       "      <td>0.118066</td>\n",
       "      <td>0.988062</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.388030</td>\n",
       "      <td>0.075221</td>\n",
       "      <td>...</td>\n",
       "      <td>0.846419</td>\n",
       "      <td>0.021725</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.140059</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.587103</td>\n",
       "      <td>0.525986</td>\n",
       "      <td>0.065291</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.067923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>106</th>\n",
       "      <td>0.164348</td>\n",
       "      <td>0.492484</td>\n",
       "      <td>0.331122</td>\n",
       "      <td>0.546240</td>\n",
       "      <td>0.427096</td>\n",
       "      <td>0.131135</td>\n",
       "      <td>0.586795</td>\n",
       "      <td>0.021631</td>\n",
       "      <td>0.349404</td>\n",
       "      <td>0.292844</td>\n",
       "      <td>...</td>\n",
       "      <td>0.609998</td>\n",
       "      <td>0.326441</td>\n",
       "      <td>0.132563</td>\n",
       "      <td>0.679396</td>\n",
       "      <td>0.587103</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.520394</td>\n",
       "      <td>0.036428</td>\n",
       "      <td>0.183237</td>\n",
       "      <td>0.306713</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>107</th>\n",
       "      <td>0.098197</td>\n",
       "      <td>0.827106</td>\n",
       "      <td>0.560281</td>\n",
       "      <td>0.199744</td>\n",
       "      <td>0.705210</td>\n",
       "      <td>0.603353</td>\n",
       "      <td>0.575807</td>\n",
       "      <td>0.220419</td>\n",
       "      <td>0.197482</td>\n",
       "      <td>0.296447</td>\n",
       "      <td>...</td>\n",
       "      <td>0.656081</td>\n",
       "      <td>0.239055</td>\n",
       "      <td>0.337026</td>\n",
       "      <td>0.341614</td>\n",
       "      <td>0.525986</td>\n",
       "      <td>0.520394</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.578247</td>\n",
       "      <td>0.274562</td>\n",
       "      <td>0.755909</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>108</th>\n",
       "      <td>0.086180</td>\n",
       "      <td>0.700721</td>\n",
       "      <td>0.431140</td>\n",
       "      <td>0.000778</td>\n",
       "      <td>0.317411</td>\n",
       "      <td>0.991333</td>\n",
       "      <td>0.114120</td>\n",
       "      <td>0.033741</td>\n",
       "      <td>0.022282</td>\n",
       "      <td>0.001543</td>\n",
       "      <td>...</td>\n",
       "      <td>0.331535</td>\n",
       "      <td>0.026262</td>\n",
       "      <td>0.001056</td>\n",
       "      <td>0.087313</td>\n",
       "      <td>0.065291</td>\n",
       "      <td>0.036428</td>\n",
       "      <td>0.578247</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.044967</td>\n",
       "      <td>0.861343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>0.207002</td>\n",
       "      <td>0.045179</td>\n",
       "      <td>0.252943</td>\n",
       "      <td>0.075197</td>\n",
       "      <td>0.356531</td>\n",
       "      <td>0.018322</td>\n",
       "      <td>0.012607</td>\n",
       "      <td>0.623281</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.082407</td>\n",
       "      <td>...</td>\n",
       "      <td>0.414446</td>\n",
       "      <td>0.610885</td>\n",
       "      <td>0.105236</td>\n",
       "      <td>0.259005</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.183237</td>\n",
       "      <td>0.274562</td>\n",
       "      <td>0.044967</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.243820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>110</th>\n",
       "      <td>0.094826</td>\n",
       "      <td>0.664158</td>\n",
       "      <td>0.397129</td>\n",
       "      <td>0.049671</td>\n",
       "      <td>0.697098</td>\n",
       "      <td>0.886560</td>\n",
       "      <td>0.129552</td>\n",
       "      <td>0.031201</td>\n",
       "      <td>0.023180</td>\n",
       "      <td>0.097919</td>\n",
       "      <td>...</td>\n",
       "      <td>0.337173</td>\n",
       "      <td>0.125513</td>\n",
       "      <td>0.074474</td>\n",
       "      <td>0.168469</td>\n",
       "      <td>0.067923</td>\n",
       "      <td>0.306713</td>\n",
       "      <td>0.755909</td>\n",
       "      <td>0.861343</td>\n",
       "      <td>0.243820</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>15 rows × 111 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1         2         3         4         5         6    \\\n",
       "96   0.153583  0.235799  0.048011  0.387427  0.042330  0.048569  0.191967   \n",
       "97   0.026822  0.232006  0.112844  0.072935  0.254673  0.128860  0.232727   \n",
       "98   0.140557  0.589646  0.876019  0.255198  0.124838  0.188862  0.528058   \n",
       "99   0.454444  0.265143  0.633106  0.009034  0.301937  0.159143  0.110315   \n",
       "100  0.924889  0.068724  0.096715  0.175867  0.259688  0.011732  0.019220   \n",
       "101  0.171491  0.793062  0.392652  0.212248  0.200605  0.363969  0.851136   \n",
       "102  0.220100  0.100091  0.252233  0.631752  0.024420  0.019962  0.018562   \n",
       "103  0.080442  0.246692  0.677175  0.196284  0.015208  0.015884  0.105515   \n",
       "104  0.237896  0.253967  0.658163  0.870973  0.136646  0.110198  0.139194   \n",
       "105  0.046670  0.705824  0.170345  0.055882  0.112778  0.118066  0.988062   \n",
       "106  0.164348  0.492484  0.331122  0.546240  0.427096  0.131135  0.586795   \n",
       "107  0.098197  0.827106  0.560281  0.199744  0.705210  0.603353  0.575807   \n",
       "108  0.086180  0.700721  0.431140  0.000778  0.317411  0.991333  0.114120   \n",
       "109  0.207002  0.045179  0.252943  0.075197  0.356531  0.018322  0.012607   \n",
       "110  0.094826  0.664158  0.397129  0.049671  0.697098  0.886560  0.129552   \n",
       "\n",
       "          7         8         9    ...       101       102       103  \\\n",
       "96   0.014384  0.057248  0.015475  ...  0.343644  0.779901  0.526238   \n",
       "97   0.000000  0.059688  0.005607  ...  0.234048  0.138254  0.114028   \n",
       "98   0.606544  0.290146  0.554501  ...  0.446804  0.099213  0.714639   \n",
       "99   0.483200  0.430941  0.687540  ...  0.080732  0.018681  0.596685   \n",
       "100  0.056307  0.829105  0.775317  ...  0.132003  0.363685  0.304898   \n",
       "101  0.278286  0.333650  0.082751  ...  1.000000  0.415900  0.106462   \n",
       "102  0.344796  0.020389  0.066832  ...  0.415900  1.000000  0.405907   \n",
       "103  0.622188  0.000000  0.390029  ...  0.106462  0.405907  1.000000   \n",
       "104  0.379788  0.224635  0.442532  ...  0.360801  0.602888  0.423498   \n",
       "105  0.000000  0.388030  0.075221  ...  0.846419  0.021725  0.000000   \n",
       "106  0.021631  0.349404  0.292844  ...  0.609998  0.326441  0.132563   \n",
       "107  0.220419  0.197482  0.296447  ...  0.656081  0.239055  0.337026   \n",
       "108  0.033741  0.022282  0.001543  ...  0.331535  0.026262  0.001056   \n",
       "109  0.623281  0.000000  0.082407  ...  0.414446  0.610885  0.105236   \n",
       "110  0.031201  0.023180  0.097919  ...  0.337173  0.125513  0.074474   \n",
       "\n",
       "          104       105       106       107       108       109       110  \n",
       "96   0.292274  0.166300  0.295955  0.308211  0.012885  0.199283  0.149904  \n",
       "97   0.274112  0.174041  0.686319  0.128973  0.011752  0.154412  0.298535  \n",
       "98   0.576250  0.457450  0.426441  0.557248  0.167169  0.000000  0.148930  \n",
       "99   0.432949  0.022194  0.421569  0.239164  0.097350  0.076782  0.244984  \n",
       "100  0.284921  0.042699  0.209766  0.134116  0.000946  0.091391  0.060368  \n",
       "101  0.360801  0.846419  0.609998  0.656081  0.331535  0.414446  0.337173  \n",
       "102  0.602888  0.021725  0.326441  0.239055  0.026262  0.610885  0.125513  \n",
       "103  0.423498  0.000000  0.132563  0.337026  0.001056  0.105236  0.074474  \n",
       "104  1.000000  0.140059  0.679396  0.341614  0.087313  0.259005  0.168469  \n",
       "105  0.140059  1.000000  0.587103  0.525986  0.065291  0.000000  0.067923  \n",
       "106  0.679396  0.587103  1.000000  0.520394  0.036428  0.183237  0.306713  \n",
       "107  0.341614  0.525986  0.520394  1.000000  0.578247  0.274562  0.755909  \n",
       "108  0.087313  0.065291  0.036428  0.578247  1.000000  0.044967  0.861343  \n",
       "109  0.259005  0.000000  0.183237  0.274562  0.044967  1.000000  0.243820  \n",
       "110  0.168469  0.067923  0.306713  0.755909  0.861343  0.243820  1.000000  \n",
       "\n",
       "[15 rows x 111 columns]"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc_similarity_matrix9.tail(15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.068"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc_similarity_matrix9[0][1].round(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      1.000000\n",
       "75     0.978577\n",
       "38     0.973273\n",
       "91     0.936580\n",
       "100    0.924889\n",
       "Name: 0, dtype: float64"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc_similarity_matrix9[0].nlargest(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5      1.000000\n",
       "50     0.993692\n",
       "108    0.991333\n",
       "57     0.981160\n",
       "53     0.975672\n",
       "Name: 5, dtype: float64"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc_similarity_matrix9[5].nlargest(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('The Outside', 'I Almost Do')"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify.track_title[5],spotify.track_title[50]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1     1.000000\n",
       "79    0.985590\n",
       "22    0.948287\n",
       "18    0.897878\n",
       "19    0.886011\n",
       "Name: 1, dtype: float64"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc_similarity_matrix9[1].nlargest(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('Picture To Burn', '...Ready for It?', 'Youre Not Sorry')"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify.track_title[1],spotify.track_title[79],spotify.track_title[22]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'state the obvious  i didn t get my perfect fantasy i realize you love yourself more than you could ever love me so go and tell your friends that i m obsessive and crazy that s fine  i ll tell mine that you re gay by the way i hate that stupid old pickup truck you never let me drive you re a redneck heartbreak who s really bad at lying so watch me strike a match on all my wasted time as far as i m concerned you re just another picture to burn there s no time for tears i m just sitting here planning my revenge there s nothing stopping me from going out with all of your best friends and if you come around saying sorry to me my daddy s gonna show you how sorry you ll be i hate that stupid old pickup truck you never let me drive you re a redneck heartbreak who s really bad at lying so watch me strike a match on all my wasted time as far as i m concerned you re just another picture to burn and if you re missing me you d better keep it to yourself  cause coming back around here would be bad for your health cause i hate that stupid old pickup truck you never let me drive you re a redneck heartbreak who s really bad at lying so watch me strike a match on all my wasted time in case you haven t heard i really really hate that i hate that stupid old pickup truck you never let me drive you re a redneck heartbreak who s really bad at lying so watch me strike a match on all my wasted time as far as i m concerned you re just another picture to burn burn  burn  burn  baby  burn you re just another picture to burn baby  burn'"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify.track_lyric[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'knew he was a killer first time that i saw him wondered how many girls he had loved and left haunted but if he s a ghost  then i can be a phantom holdin  him for ransom  some some boys are tryin  too hard  he don t try at all though younger than my exes  but he act like such a man  so i see nothing better  i keep him forever like a vendetta ta i  i  i see how this is gon  go touch me and you ll never be alone i island breeze and lights down low no one has to know in the middle of the night  in my dreams you should see the things we do  baby in the middle of the night  in my dreams i know i m gonna be with you so i take my time are you ready for it  knew i was a robber first time that he saw me stealing hearts and running off and never sayin  sorry but if i m a thief  then he can join the heist  and we ll move to an island  and and he can be my jailer  burton to this taylor every love i ve known in comparison is a failure i forget their names now  i m so very tame now never be the same now  now i  i  i see how this is gon  go touch me and you ll never be alone i island breeze and lights down low no one has to know  no one has to know  in the middle of the night  in my dreams you should see the things we do  baby in the middle of the night in my dreams i know i m gonna be with you so i take my time are you ready for it  oh  are you ready for it  baby  let the games begin let the games begin let the games begin baby  let the games begin let the games begin let the games begin i  i  i see how this is gon  go touch me and you ll never be alone i island breeze and lights down low no one has to know in the middle of the night  in my dreams you should see the things we do  baby in the middle of the night  in my dreams i know i m gonna be with you so i take my time in the middle of the night baby  let the games begin let the games begin let the games begin are you ready for it  baby  let the games begin let the games begin let the games begin are you ready for it '"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify.track_lyric[79]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'all this time i was wasting hoping you would come around i ve been giving out chances every time and all you do is let me down and it s taken me this long baby but i ve figured you out and you re thinking we ll be fine again but not this time around you don t have to call anymore i won t pick up the phone this is the last straw don t wanna hurt anymore and you can tell me that you re sorry but i don t believe you baby like i did before you re not sorry  no no no no  you re looking so innocent i might believe you if i didn t know could ve loved you all my life if you hadn t left me waiting in the cold and you ve got your share of secrets and i m tired of being last to know and now you re asking me to listen  cause it s worked each time before but you don t have to call anymore i won t pick up the phone this is the last straw don t wanna hurt anymore and you can tell me that you re sorry but i don t believe you baby like i did before you re not sorry  no no no no  you re not sorry  no no no no  you had me crawling for you honey and it never would have gone away you used to shine so bright but i watched all of it fade so you don t have to call anymore i won t pick up the phone this is the last straw there s nothing left to beg for and you can tell me that you re sorry but i don t believe you baby like i did before you re not sorry  no no no no  you re not sorry  no no no no   no  no  no    '"
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify.track_lyric[22]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2     1.000000\n",
       "35    0.931094\n",
       "64    0.901478\n",
       "98    0.876019\n",
       "41    0.866561\n",
       "Name: 2, dtype: float64"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc_similarity_matrix9[2].nlargest(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('Teardrops On My Guitar', 'Enchanted')"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify.track_title[2],spotify.track_title[35]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "61    0.002785\n",
       "42    0.013266\n",
       "32    0.022145\n",
       "96    0.048011\n",
       "75    0.061345\n",
       "Name: 2, dtype: float64"
      ]
     },
     "execution_count": 110,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "doc_similarity_matrix9[2].nsmallest(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('Come Back... Be Here', 'If This Was a Movie')"
      ]
     },
     "execution_count": 111,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "spotify.track_title[61],spotify.track_title[42]"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:metis] *",
   "language": "python",
   "name": "conda-env-metis-py"
  },
  "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": 4
}