CS_music_recommender_cleaning+eda_part1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Part 1: Cleaning and EDA for SVD Recommender System \n",
    "\n",
    "## Cleaning\n",
    "\n",
    "#### Cleaning for song df\n",
    "- Remove duplicates\n",
    "- Handle Null values\n",
    "- Impute for null values where possible\n",
    "- Make decision on other null values that can't be imputed\n",
    "\n",
    "#### Cleaning for triplets df\n",
    "- Check for duplicates\n",
    "- Check for nulls\n",
    "\n",
    "## EDA\n",
    "1. Songs with most listens (top 10) + visualisation\n",
    "2. Highest average number of listens (top 10)\n",
    "3. Songs with highest different number of listeners (unique listeners - top 10)\n",
    "4. Artists with most listens + visualisation (top 10)\n",
    "5. Artists with highest average number of plays (top 10)\n",
    "6. Most polarising songs (largest std in listen count)\n",
    "7. Analysis by year\n",
    "\n",
    "####  Item-item relationships - Cosine Similarity\n",
    "- Create csr matrix\n",
    "- Pass into cosine similarity \n",
    "- Form data frame\n",
    "- Filter for top 10 songs song_ids\n",
    "- Convert song_ids to title + artist\n",
    "- Filter for threshold level of similarity\n",
    "- Visualise similarites in Heatmap\n",
    "\n",
    "#### User-item relationships - Cosine Similarity\n",
    "- Review listen count statstics\n",
    "- Filter for listen counts under median\n",
    "- Filter matrix\n",
    "- Sample matrix to handle user quantity\n",
    "- Get cosine similarity for sample\n",
    "- Sample 10 columns and analyse similairty to all other users\n",
    "- Show on heatmap\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from scipy.sparse import csr_matrix\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "pd.set_option('display.float_format', '{:.2f}'.format)\n",
    "from scipy.stats import zscore"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "song_df = pd.read_csv(r'C:\\Users\\Ollie\\Downloads\\song_data.csv')\n",
    "triplets_df = pd.read_csv(r'C:/Users/Ollie/Downloads/triplets_file.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# View Song Data - this is all the songs names, artist names, release year and song id"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>song_id</th>\n",
       "      <th>title</th>\n",
       "      <th>release</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>SOQMMHC12AB0180CB8</td>\n",
       "      <td>Silent Night</td>\n",
       "      <td>Monster Ballads X-Mas</td>\n",
       "      <td>Faster Pussy cat</td>\n",
       "      <td>2003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>SOVFVAK12A8C1350D9</td>\n",
       "      <td>Tanssi vaan</td>\n",
       "      <td>Karkuteillä</td>\n",
       "      <td>Karkkiautomaatti</td>\n",
       "      <td>1995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>SOGTUKN12AB017F4F1</td>\n",
       "      <td>No One Could Ever</td>\n",
       "      <td>Butter</td>\n",
       "      <td>Hudson Mohawke</td>\n",
       "      <td>2006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SOBNYVR12A8C13558C</td>\n",
       "      <td>Si Vos Querés</td>\n",
       "      <td>De Culo</td>\n",
       "      <td>Yerba Brava</td>\n",
       "      <td>2003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>SOHSBXH12A8C13B0DF</td>\n",
       "      <td>Tangle Of Aspens</td>\n",
       "      <td>Rene Ablaze Presents Winter Sessions</td>\n",
       "      <td>Der Mystic</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999995</th>\n",
       "      <td>SOTXAME12AB018F136</td>\n",
       "      <td>O Samba Da Vida</td>\n",
       "      <td>Pacha V.I.P.</td>\n",
       "      <td>Kiko Navarro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999996</th>\n",
       "      <td>SOXQYIQ12A8C137FBB</td>\n",
       "      <td>Jago Chhadeo</td>\n",
       "      <td>Naale Baba Lassi Pee Gya</td>\n",
       "      <td>Kuldeep Manak</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999997</th>\n",
       "      <td>SOHODZI12A8C137BB3</td>\n",
       "      <td>Novemba</td>\n",
       "      <td>Dub_Connected: electronic music</td>\n",
       "      <td>Gabriel Le Mar</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999998</th>\n",
       "      <td>SOLXGOR12A81C21EB7</td>\n",
       "      <td>Faraday</td>\n",
       "      <td>The Trance Collection Vol. 2</td>\n",
       "      <td>Elude</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999999</th>\n",
       "      <td>SOWXJXQ12AB0189F43</td>\n",
       "      <td>Fernweh feat. Sektion Kuchikäschtli</td>\n",
       "      <td>So Oder So</td>\n",
       "      <td>Texta</td>\n",
       "      <td>2004</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1000000 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   song_id                                title  \\\n",
       "0       SOQMMHC12AB0180CB8                         Silent Night   \n",
       "1       SOVFVAK12A8C1350D9                          Tanssi vaan   \n",
       "2       SOGTUKN12AB017F4F1                    No One Could Ever   \n",
       "3       SOBNYVR12A8C13558C                        Si Vos Querés   \n",
       "4       SOHSBXH12A8C13B0DF                     Tangle Of Aspens   \n",
       "...                    ...                                  ...   \n",
       "999995  SOTXAME12AB018F136                      O Samba Da Vida   \n",
       "999996  SOXQYIQ12A8C137FBB                         Jago Chhadeo   \n",
       "999997  SOHODZI12A8C137BB3                              Novemba   \n",
       "999998  SOLXGOR12A81C21EB7                              Faraday   \n",
       "999999  SOWXJXQ12AB0189F43  Fernweh feat. Sektion Kuchikäschtli   \n",
       "\n",
       "                                     release       artist_name  year  \n",
       "0                      Monster Ballads X-Mas  Faster Pussy cat  2003  \n",
       "1                                Karkuteillä  Karkkiautomaatti  1995  \n",
       "2                                     Butter    Hudson Mohawke  2006  \n",
       "3                                    De Culo       Yerba Brava  2003  \n",
       "4       Rene Ablaze Presents Winter Sessions        Der Mystic     0  \n",
       "...                                      ...               ...   ...  \n",
       "999995                          Pacha V.I.P.      Kiko Navarro     0  \n",
       "999996              Naale Baba Lassi Pee Gya     Kuldeep Manak     0  \n",
       "999997       Dub_Connected: electronic music    Gabriel Le Mar     0  \n",
       "999998          The Trance Collection Vol. 2             Elude     0  \n",
       "999999                            So Oder So             Texta  2004  \n",
       "\n",
       "[1000000 rows x 5 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "song_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "999056"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "song_df['song_id'].nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# View other dataframe of users, song_id's and listen counts for each"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>user_id</th>\n",
       "      <th>song_id</th>\n",
       "      <th>listen_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOAKIMP12A8C130995</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBBMDR12A8C13253B</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBXHDL12A81C204C0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBYHAJ12A6701BF1D</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SODACBL12A8C13C273</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999995</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJEYPO12AAA8C6B0E</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999996</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJJYDE12AF729FC16</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999997</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJKQSF12A6D4F5EE9</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999998</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJUXGA12AC961885C</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999999</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJYOLS12A8C13C06F</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2000000 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          user_id             song_id  \\\n",
       "0        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOAKIMP12A8C130995   \n",
       "1        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBBMDR12A8C13253B   \n",
       "2        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBXHDL12A81C204C0   \n",
       "3        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBYHAJ12A6701BF1D   \n",
       "4        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SODACBL12A8C13C273   \n",
       "...                                           ...                 ...   \n",
       "1999995  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJEYPO12AAA8C6B0E   \n",
       "1999996  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJJYDE12AF729FC16   \n",
       "1999997  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJKQSF12A6D4F5EE9   \n",
       "1999998  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJUXGA12AC961885C   \n",
       "1999999  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJYOLS12A8C13C06F   \n",
       "\n",
       "         listen_count  \n",
       "0                   1  \n",
       "1                   2  \n",
       "2                   1  \n",
       "3                   1  \n",
       "4                   1  \n",
       "...               ...  \n",
       "1999995             2  \n",
       "1999996             4  \n",
       "1999997             3  \n",
       "1999998             1  \n",
       "1999999             1  \n",
       "\n",
       "[2000000 rows x 3 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "triplets_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# -- Handle duplicates and null values in either dataframe --"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "999056"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Song_df first, nunique() on 'song_id' being lower than total rows means there must be duplicates\n",
    "song_df['song_id'].nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>song_id</th>\n",
       "      <th>title</th>\n",
       "      <th>release</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>907282</th>\n",
       "      <td>SOYWFPT12AB018742C</td>\n",
       "      <td>Landing In London</td>\n",
       "      <td>Landing In London</td>\n",
       "      <td>3 Doors Down</td>\n",
       "      <td>2005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>944290</th>\n",
       "      <td>SOUZNSP12A6701FBD5</td>\n",
       "      <td>It's Not Me</td>\n",
       "      <td>Seventeen Days</td>\n",
       "      <td>3 Doors Down</td>\n",
       "      <td>2003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>466874</th>\n",
       "      <td>SOUEHCD12AB0188F80</td>\n",
       "      <td>Here Without You</td>\n",
       "      <td>Acoustic EP</td>\n",
       "      <td>3 Doors Down</td>\n",
       "      <td>2002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>886850</th>\n",
       "      <td>SOGHVAN12A81C228BA</td>\n",
       "      <td>Man Down (Censored)</td>\n",
       "      <td>Curtis</td>\n",
       "      <td>50 Cent</td>\n",
       "      <td>2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>564397</th>\n",
       "      <td>SOCQKVJ12AB017B52C</td>\n",
       "      <td>God Gave Me Style</td>\n",
       "      <td>The Massacre</td>\n",
       "      <td>50 Cent</td>\n",
       "      <td>2005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>637540</th>\n",
       "      <td>SOEYSCL12A8C131F19</td>\n",
       "      <td>Space Rock</td>\n",
       "      <td>Maladroit</td>\n",
       "      <td>Weezer</td>\n",
       "      <td>2002</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>662376</th>\n",
       "      <td>SOZHPEI12A8C138502</td>\n",
       "      <td>Dreamin'</td>\n",
       "      <td>Weezer (Red Album)</td>\n",
       "      <td>Weezer</td>\n",
       "      <td>2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>399501</th>\n",
       "      <td>SOCRCNK12A8C133AA7</td>\n",
       "      <td>Your Woman</td>\n",
       "      <td>Women In Technology</td>\n",
       "      <td>White Town</td>\n",
       "      <td>1996</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>233017</th>\n",
       "      <td>SOHBZKL12A6D4F9354</td>\n",
       "      <td>Down On My Head</td>\n",
       "      <td>Lights And Sounds</td>\n",
       "      <td>Yellowcard</td>\n",
       "      <td>2006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>675804</th>\n",
       "      <td>SOQXGUH12A8C13241E</td>\n",
       "      <td>Oh_ My Love (Album Version)</td>\n",
       "      <td>Make Some Noise: The Amnesty International Cam...</td>\n",
       "      <td>Yellowcard</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>944 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   song_id                        title  \\\n",
       "907282  SOYWFPT12AB018742C            Landing In London   \n",
       "944290  SOUZNSP12A6701FBD5                  It's Not Me   \n",
       "466874  SOUEHCD12AB0188F80             Here Without You   \n",
       "886850  SOGHVAN12A81C228BA          Man Down (Censored)   \n",
       "564397  SOCQKVJ12AB017B52C            God Gave Me Style   \n",
       "...                    ...                          ...   \n",
       "637540  SOEYSCL12A8C131F19                   Space Rock   \n",
       "662376  SOZHPEI12A8C138502                     Dreamin'   \n",
       "399501  SOCRCNK12A8C133AA7                   Your Woman   \n",
       "233017  SOHBZKL12A6D4F9354              Down On My Head   \n",
       "675804  SOQXGUH12A8C13241E  Oh_ My Love (Album Version)   \n",
       "\n",
       "                                                  release   artist_name  year  \n",
       "907282                                  Landing In London  3 Doors Down  2005  \n",
       "944290                                     Seventeen Days  3 Doors Down  2003  \n",
       "466874                                        Acoustic EP  3 Doors Down  2002  \n",
       "886850                                             Curtis       50 Cent  2007  \n",
       "564397                                       The Massacre       50 Cent  2005  \n",
       "...                                                   ...           ...   ...  \n",
       "637540                                          Maladroit        Weezer  2002  \n",
       "662376                                 Weezer (Red Album)        Weezer  2008  \n",
       "399501                                Women In Technology    White Town  1996  \n",
       "233017                                  Lights And Sounds    Yellowcard  2006  \n",
       "675804  Make Some Noise: The Amnesty International Cam...    Yellowcard     0  \n",
       "\n",
       "[944 rows x 5 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Duplicates where song_id and artist_name are equal\n",
    "duplicates = song_df[song_df.duplicated(subset=['song_id'])]\n",
    "duplicates.sort_values(['artist_name'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>song_id</th>\n",
       "      <th>title</th>\n",
       "      <th>release</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>827195</th>\n",
       "      <td>SOKBIIQ12AB017FEFC</td>\n",
       "      <td>Did It Again</td>\n",
       "      <td>Superventas 2010</td>\n",
       "      <td>Shakira</td>\n",
       "      <td>2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>912127</th>\n",
       "      <td>SOKBIIQ12AB017FEFC</td>\n",
       "      <td>Did It Again (featuring Kid Cudi)</td>\n",
       "      <td>She Wolf</td>\n",
       "      <td>Shakira</td>\n",
       "      <td>2009</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   song_id                              title  \\\n",
       "827195  SOKBIIQ12AB017FEFC                       Did It Again   \n",
       "912127  SOKBIIQ12AB017FEFC  Did It Again (featuring Kid Cudi)   \n",
       "\n",
       "                 release artist_name  year  \n",
       "827195  Superventas 2010     Shakira  2009  \n",
       "912127          She Wolf     Shakira  2009  "
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We can see from the below that there are instances where song_id and artist name are the same\n",
    "# But the song is actually a different version with another artist\n",
    "# Ideally we distinguish between these as they could reflect preferences however our user-item df only has song_id\n",
    "# This means there is no way to know which version they listened to, and thus derive a more informed recommendation \n",
    "# Therefore we will need to simply drop any duplicates on song_id, so that all song_id's are unique.\n",
    "# This however should still be fairly accurate as a recommendation. \n",
    "duplicates[duplicates['song_id']=='SOKBIIQ12AB017FEFC']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Drop where song_id is duplicated \n",
    "song_df = song_df.drop_duplicates(subset=['song_id'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>song_id</th>\n",
       "      <th>title</th>\n",
       "      <th>release</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>SOQMMHC12AB0180CB8</td>\n",
       "      <td>Silent Night</td>\n",
       "      <td>Monster Ballads X-Mas</td>\n",
       "      <td>Faster Pussy cat</td>\n",
       "      <td>2003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>SOVFVAK12A8C1350D9</td>\n",
       "      <td>Tanssi vaan</td>\n",
       "      <td>Karkuteillä</td>\n",
       "      <td>Karkkiautomaatti</td>\n",
       "      <td>1995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>SOGTUKN12AB017F4F1</td>\n",
       "      <td>No One Could Ever</td>\n",
       "      <td>Butter</td>\n",
       "      <td>Hudson Mohawke</td>\n",
       "      <td>2006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SOBNYVR12A8C13558C</td>\n",
       "      <td>Si Vos Querés</td>\n",
       "      <td>De Culo</td>\n",
       "      <td>Yerba Brava</td>\n",
       "      <td>2003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>SOHSBXH12A8C13B0DF</td>\n",
       "      <td>Tangle Of Aspens</td>\n",
       "      <td>Rene Ablaze Presents Winter Sessions</td>\n",
       "      <td>Der Mystic</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999995</th>\n",
       "      <td>SOTXAME12AB018F136</td>\n",
       "      <td>O Samba Da Vida</td>\n",
       "      <td>Pacha V.I.P.</td>\n",
       "      <td>Kiko Navarro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999996</th>\n",
       "      <td>SOXQYIQ12A8C137FBB</td>\n",
       "      <td>Jago Chhadeo</td>\n",
       "      <td>Naale Baba Lassi Pee Gya</td>\n",
       "      <td>Kuldeep Manak</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999997</th>\n",
       "      <td>SOHODZI12A8C137BB3</td>\n",
       "      <td>Novemba</td>\n",
       "      <td>Dub_Connected: electronic music</td>\n",
       "      <td>Gabriel Le Mar</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999998</th>\n",
       "      <td>SOLXGOR12A81C21EB7</td>\n",
       "      <td>Faraday</td>\n",
       "      <td>The Trance Collection Vol. 2</td>\n",
       "      <td>Elude</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>999999</th>\n",
       "      <td>SOWXJXQ12AB0189F43</td>\n",
       "      <td>Fernweh feat. Sektion Kuchikäschtli</td>\n",
       "      <td>So Oder So</td>\n",
       "      <td>Texta</td>\n",
       "      <td>2004</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>999056 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                   song_id                                title  \\\n",
       "0       SOQMMHC12AB0180CB8                         Silent Night   \n",
       "1       SOVFVAK12A8C1350D9                          Tanssi vaan   \n",
       "2       SOGTUKN12AB017F4F1                    No One Could Ever   \n",
       "3       SOBNYVR12A8C13558C                        Si Vos Querés   \n",
       "4       SOHSBXH12A8C13B0DF                     Tangle Of Aspens   \n",
       "...                    ...                                  ...   \n",
       "999995  SOTXAME12AB018F136                      O Samba Da Vida   \n",
       "999996  SOXQYIQ12A8C137FBB                         Jago Chhadeo   \n",
       "999997  SOHODZI12A8C137BB3                              Novemba   \n",
       "999998  SOLXGOR12A81C21EB7                              Faraday   \n",
       "999999  SOWXJXQ12AB0189F43  Fernweh feat. Sektion Kuchikäschtli   \n",
       "\n",
       "                                     release       artist_name  year  \n",
       "0                      Monster Ballads X-Mas  Faster Pussy cat  2003  \n",
       "1                                Karkuteillä  Karkkiautomaatti  1995  \n",
       "2                                     Butter    Hudson Mohawke  2006  \n",
       "3                                    De Culo       Yerba Brava  2003  \n",
       "4       Rene Ablaze Presents Winter Sessions        Der Mystic     0  \n",
       "...                                      ...               ...   ...  \n",
       "999995                          Pacha V.I.P.      Kiko Navarro     0  \n",
       "999996              Naale Baba Lassi Pee Gya     Kuldeep Manak     0  \n",
       "999997       Dub_Connected: electronic music    Gabriel Le Mar     0  \n",
       "999998          The Trance Collection Vol. 2             Elude     0  \n",
       "999999                            So Oder So             Texta  2004  \n",
       "\n",
       "[999056 rows x 5 columns]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "song_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# -- Handle song_df null values -- "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "song_id         0\n",
       "title          15\n",
       "release         5\n",
       "artist_name     0\n",
       "year            0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check null values, a few have been found\n",
    "song_df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>song_id</th>\n",
       "      <th>title</th>\n",
       "      <th>release</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>64262</th>\n",
       "      <td>SOVICLT12A58A7C4D0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Of Flies And Men</td>\n",
       "      <td>Milton</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>114672</th>\n",
       "      <td>SOAGRAA12AB018D567</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138011</th>\n",
       "      <td>SOOBGAQ12A6D4F6762</td>\n",
       "      <td>Solitude</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197822</th>\n",
       "      <td>SOGFJER12A6D4F675F</td>\n",
       "      <td>All The Beautiful People</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>223019</th>\n",
       "      <td>SOAFJBH12A6D4F6765</td>\n",
       "      <td>Down The Aisles</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>229353</th>\n",
       "      <td>SOOAFJX12AB018A028</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>264639</th>\n",
       "      <td>SOLDTFD12AB018AFE6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>458083</th>\n",
       "      <td>SOAUTVB12AB018AFF0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>557125</th>\n",
       "      <td>SOMPGYE12AB018AFE0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>578241</th>\n",
       "      <td>SOVHAZH12AB018D5B3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>652287</th>\n",
       "      <td>SOECGBQ12AB018D582</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>693232</th>\n",
       "      <td>SOCKQGJ12A6D4F6768</td>\n",
       "      <td>Provincial</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>697789</th>\n",
       "      <td>SOTJWHR12AB018D5A0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>697813</th>\n",
       "      <td>SOBDDHR12AB01888A6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Grand Glam Outtakes</td>\n",
       "      <td>Remute</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>725270</th>\n",
       "      <td>SOFDILP12AC960A4EF</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Recupera tus Clásicos - Fobia</td>\n",
       "      <td>Fobia</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>795481</th>\n",
       "      <td>SOSFNKO12AB018D5BB</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>841107</th>\n",
       "      <td>SOSQKYC12A6D4F6763</td>\n",
       "      <td>Middlemen</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>853170</th>\n",
       "      <td>SOASSAS12AB018AFCF</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>857992</th>\n",
       "      <td>SOZDBDL12AB018AFFF</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>925868</th>\n",
       "      <td>SOQUGMS12AB018B01D</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   song_id                     title  \\\n",
       "64262   SOVICLT12A58A7C4D0                       NaN   \n",
       "114672  SOAGRAA12AB018D567                       NaN   \n",
       "138011  SOOBGAQ12A6D4F6762                  Solitude   \n",
       "197822  SOGFJER12A6D4F675F  All The Beautiful People   \n",
       "223019  SOAFJBH12A6D4F6765           Down The Aisles   \n",
       "229353  SOOAFJX12AB018A028                       NaN   \n",
       "264639  SOLDTFD12AB018AFE6                       NaN   \n",
       "458083  SOAUTVB12AB018AFF0                       NaN   \n",
       "557125  SOMPGYE12AB018AFE0                       NaN   \n",
       "578241  SOVHAZH12AB018D5B3                       NaN   \n",
       "652287  SOECGBQ12AB018D582                       NaN   \n",
       "693232  SOCKQGJ12A6D4F6768                Provincial   \n",
       "697789  SOTJWHR12AB018D5A0                       NaN   \n",
       "697813  SOBDDHR12AB01888A6                       NaN   \n",
       "725270  SOFDILP12AC960A4EF                       NaN   \n",
       "795481  SOSFNKO12AB018D5BB                       NaN   \n",
       "841107  SOSQKYC12A6D4F6763                 Middlemen   \n",
       "853170  SOASSAS12AB018AFCF                       NaN   \n",
       "857992  SOZDBDL12AB018AFFF                       NaN   \n",
       "925868  SOQUGMS12AB018B01D                       NaN   \n",
       "\n",
       "                              release     artist_name  year  \n",
       "64262                Of Flies And Men          Milton     0  \n",
       "114672                   Puce de luxe  Sébastien Roch     0  \n",
       "138011                            NaN           Metro     0  \n",
       "197822                            NaN           Metro     0  \n",
       "223019                            NaN           Metro     0  \n",
       "229353                   Puce de luxe  Sébastien Roch     0  \n",
       "264639                   Puce de luxe  Sébastien Roch     0  \n",
       "458083                   Puce de luxe  Sébastien Roch     0  \n",
       "557125                   Puce de luxe  Sébastien Roch     0  \n",
       "578241                   Puce de luxe  Sébastien Roch     0  \n",
       "652287                   Puce de luxe  Sébastien Roch     0  \n",
       "693232                            NaN           Metro     0  \n",
       "697789                   Puce de luxe  Sébastien Roch     0  \n",
       "697813            Grand Glam Outtakes          Remute     0  \n",
       "725270  Recupera tus Clásicos - Fobia           Fobia     0  \n",
       "795481                   Puce de luxe  Sébastien Roch     0  \n",
       "841107                            NaN           Metro     0  \n",
       "853170                   Puce de luxe  Sébastien Roch     0  \n",
       "857992                   Puce de luxe  Sébastien Roch     0  \n",
       "925868                   Puce de luxe  Sébastien Roch     0  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Identify these rows\n",
    "rows_with_na = song_df[song_df.isna().any(axis=1)]\n",
    "rows_with_na"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Interestingly almost all NaNs (12/15) on title are from the same artist and the same release, Sebastien Roch - Puce de Luxe\n",
    "# As we have unique song ids for this we can fill these with Track 1, Track 2,... \n",
    "# The user can then deduce themselves which song on the album release that it is"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>song_id</th>\n",
       "      <th>title</th>\n",
       "      <th>release</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>114672</th>\n",
       "      <td>SOAGRAA12AB018D567</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>229353</th>\n",
       "      <td>SOOAFJX12AB018A028</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>264639</th>\n",
       "      <td>SOLDTFD12AB018AFE6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>458083</th>\n",
       "      <td>SOAUTVB12AB018AFF0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>557125</th>\n",
       "      <td>SOMPGYE12AB018AFE0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>578241</th>\n",
       "      <td>SOVHAZH12AB018D5B3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>652287</th>\n",
       "      <td>SOECGBQ12AB018D582</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>697789</th>\n",
       "      <td>SOTJWHR12AB018D5A0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>795481</th>\n",
       "      <td>SOSFNKO12AB018D5BB</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>853170</th>\n",
       "      <td>SOASSAS12AB018AFCF</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>857992</th>\n",
       "      <td>SOZDBDL12AB018AFFF</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>925868</th>\n",
       "      <td>SOQUGMS12AB018B01D</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Puce de luxe</td>\n",
       "      <td>Sébastien Roch</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   song_id title       release     artist_name  year\n",
       "114672  SOAGRAA12AB018D567   NaN  Puce de luxe  Sébastien Roch     0\n",
       "229353  SOOAFJX12AB018A028   NaN  Puce de luxe  Sébastien Roch     0\n",
       "264639  SOLDTFD12AB018AFE6   NaN  Puce de luxe  Sébastien Roch     0\n",
       "458083  SOAUTVB12AB018AFF0   NaN  Puce de luxe  Sébastien Roch     0\n",
       "557125  SOMPGYE12AB018AFE0   NaN  Puce de luxe  Sébastien Roch     0\n",
       "578241  SOVHAZH12AB018D5B3   NaN  Puce de luxe  Sébastien Roch     0\n",
       "652287  SOECGBQ12AB018D582   NaN  Puce de luxe  Sébastien Roch     0\n",
       "697789  SOTJWHR12AB018D5A0   NaN  Puce de luxe  Sébastien Roch     0\n",
       "795481  SOSFNKO12AB018D5BB   NaN  Puce de luxe  Sébastien Roch     0\n",
       "853170  SOASSAS12AB018AFCF   NaN  Puce de luxe  Sébastien Roch     0\n",
       "857992  SOZDBDL12AB018AFFF   NaN  Puce de luxe  Sébastien Roch     0\n",
       "925868  SOQUGMS12AB018B01D   NaN  Puce de luxe  Sébastien Roch     0"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Grab NaN rows from this release/artist, calling just the artist and release is enough to find all 12 of their NaNs\n",
    "imputable_nan = song_df[(song_df['artist_name']== 'Sébastien Roch') & (song_df['release']=='Puce de luxe')]\n",
    "imputable_nan"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Ollie\\anaconda3\\lib\\site-packages\\pandas\\core\\indexing.py:1773: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  self._setitem_single_column(ilocs[0], value, pi)\n"
     ]
    }
   ],
   "source": [
    "# Fill values with for loop for all twelve NaNs\n",
    "song_df.loc[imputable_nan.index, 'title'] = [f'Title {i+1}' for i in range(len(imputable_nan))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "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>song_id</th>\n",
       "      <th>title</th>\n",
       "      <th>release</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>64262</th>\n",
       "      <td>SOVICLT12A58A7C4D0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Of Flies And Men</td>\n",
       "      <td>Milton</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138011</th>\n",
       "      <td>SOOBGAQ12A6D4F6762</td>\n",
       "      <td>Solitude</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197822</th>\n",
       "      <td>SOGFJER12A6D4F675F</td>\n",
       "      <td>All The Beautiful People</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>223019</th>\n",
       "      <td>SOAFJBH12A6D4F6765</td>\n",
       "      <td>Down The Aisles</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>693232</th>\n",
       "      <td>SOCKQGJ12A6D4F6768</td>\n",
       "      <td>Provincial</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>697813</th>\n",
       "      <td>SOBDDHR12AB01888A6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Grand Glam Outtakes</td>\n",
       "      <td>Remute</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>725270</th>\n",
       "      <td>SOFDILP12AC960A4EF</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Recupera tus Clásicos - Fobia</td>\n",
       "      <td>Fobia</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>841107</th>\n",
       "      <td>SOSQKYC12A6D4F6763</td>\n",
       "      <td>Middlemen</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Metro</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   song_id                     title  \\\n",
       "64262   SOVICLT12A58A7C4D0                       NaN   \n",
       "138011  SOOBGAQ12A6D4F6762                  Solitude   \n",
       "197822  SOGFJER12A6D4F675F  All The Beautiful People   \n",
       "223019  SOAFJBH12A6D4F6765           Down The Aisles   \n",
       "693232  SOCKQGJ12A6D4F6768                Provincial   \n",
       "697813  SOBDDHR12AB01888A6                       NaN   \n",
       "725270  SOFDILP12AC960A4EF                       NaN   \n",
       "841107  SOSQKYC12A6D4F6763                 Middlemen   \n",
       "\n",
       "                              release artist_name  year  \n",
       "64262                Of Flies And Men      Milton     0  \n",
       "138011                            NaN       Metro     0  \n",
       "197822                            NaN       Metro     0  \n",
       "223019                            NaN       Metro     0  \n",
       "693232                            NaN       Metro     0  \n",
       "697813            Grand Glam Outtakes      Remute     0  \n",
       "725270  Recupera tus Clásicos - Fobia       Fobia     0  \n",
       "841107                            NaN       Metro     0  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check nulls again\n",
    "song_df[song_df.isna().any(axis=1)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Remaining nulls on title can just be dropped \n",
    "song_df = song_df.dropna(subset=['title'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Nulls on release can be left as is, they do not stop us from recommending a song"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# -- Triplets_df --\n",
    "# data cleaning, handling duplicates and null values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>user_id</th>\n",
       "      <th>song_id</th>\n",
       "      <th>listen_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOAKIMP12A8C130995</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBBMDR12A8C13253B</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBXHDL12A81C204C0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBYHAJ12A6701BF1D</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SODACBL12A8C13C273</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999995</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJEYPO12AAA8C6B0E</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999996</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJJYDE12AF729FC16</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999997</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJKQSF12A6D4F5EE9</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999998</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJUXGA12AC961885C</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1999999</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJYOLS12A8C13C06F</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2000000 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          user_id             song_id  \\\n",
       "0        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOAKIMP12A8C130995   \n",
       "1        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBBMDR12A8C13253B   \n",
       "2        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBXHDL12A81C204C0   \n",
       "3        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBYHAJ12A6701BF1D   \n",
       "4        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SODACBL12A8C13C273   \n",
       "...                                           ...                 ...   \n",
       "1999995  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJEYPO12AAA8C6B0E   \n",
       "1999996  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJJYDE12AF729FC16   \n",
       "1999997  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJKQSF12A6D4F5EE9   \n",
       "1999998  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJUXGA12AC961885C   \n",
       "1999999  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJYOLS12A8C13C06F   \n",
       "\n",
       "         listen_count  \n",
       "0                   1  \n",
       "1                   2  \n",
       "2                   1  \n",
       "3                   1  \n",
       "4                   1  \n",
       "...               ...  \n",
       "1999995             2  \n",
       "1999996             4  \n",
       "1999997             3  \n",
       "1999998             1  \n",
       "1999999             1  \n",
       "\n",
       "[2000000 rows x 3 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "triplets_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# User-Pair duplicates need to be aggregated together to properly form the sparse matrix "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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></th>\n",
       "      <th>listen_count</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user_id</th>\n",
       "      <th>song_id</th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">00003a4459f33b92906be11abe0e93efc423c0ff</th>\n",
       "      <th>SOJJRVI12A6D4FBE49</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOKJWZB12A6D4F9487</th>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOMZHIH12A8AE45D00</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SONFEUF12AAF3B47E3</th>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOVMGXI12AF72A80B0</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">fffea3d509760c984e7d40789804c0e5e289cc86</th>\n",
       "      <th>SOUBKFT12A6701F07A</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOVPBLT12A6D4F5113</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOVRQIK12A6701F118</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOWEBRA12A6701F115</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOZMNAX12A58A77F88</th>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2000000 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                             listen_count\n",
       "user_id                                  song_id                         \n",
       "00003a4459f33b92906be11abe0e93efc423c0ff SOJJRVI12A6D4FBE49             1\n",
       "                                         SOKJWZB12A6D4F9487             4\n",
       "                                         SOMZHIH12A8AE45D00             3\n",
       "                                         SONFEUF12AAF3B47E3             3\n",
       "                                         SOVMGXI12AF72A80B0             1\n",
       "...                                                                   ...\n",
       "fffea3d509760c984e7d40789804c0e5e289cc86 SOUBKFT12A6701F07A             1\n",
       "                                         SOVPBLT12A6D4F5113             1\n",
       "                                         SOVRQIK12A6701F118             1\n",
       "                                         SOWEBRA12A6701F115             1\n",
       "                                         SOZMNAX12A58A77F88             1\n",
       "\n",
       "[2000000 rows x 1 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Run groupby on user_id and song_id to see how much the number of rows changes upon aggregation \n",
    "triplets_df.groupby(['user_id','song_id']).sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# From this we can logically deduce with certainty that there are no user-song pair duplicates\n",
    "# If there were then when aggregated, total rows would be less than 2,000,000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "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>user_id</th>\n",
       "      <th>song_id</th>\n",
       "      <th>listen_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [user_id, song_id, listen_count]\n",
       "Index: []"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# We can run a duplicates function to verify this with python as well\n",
    "triplets_df[triplets_df.duplicated(subset=['user_id','song_id'])]\n",
    "# As predicted no results returned"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "user_id         0\n",
       "song_id         0\n",
       "listen_count    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check for null values, there are none\n",
    "triplets_df.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count   2000000.00\n",
       "mean          3.05\n",
       "std           6.58\n",
       "min           1.00\n",
       "25%           1.00\n",
       "50%           1.00\n",
       "75%           3.00\n",
       "max        2213.00\n",
       "Name: listen_count, dtype: float64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Analyse listen_counts in triplets_df \n",
    "triplets_df['listen_count'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# We can see from this we have some quite extreme outliers, the highest listen_count is 2213 while the average is 3 and median 1\n",
    "# We will remove these from our dataset as it will mislead our model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# First we will calculate z-scores \n",
    "triplets_df['z-score'] = zscore(triplets_df['listen_count'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Remove any outliers with an absolute z score of more than 3\n",
    "triplets_df = triplets_df[triplets_df['z-score'].abs() <= 3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count   1973852.00\n",
       "mean          2.54\n",
       "std           3.04\n",
       "min           1.00\n",
       "25%           1.00\n",
       "50%           1.00\n",
       "75%           3.00\n",
       "max          22.00\n",
       "Name: listen_count, dtype: float64"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# View listen_count stats again\n",
    "triplets_df['listen_count'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The max listen_count for a particular song is now 22, with the mean dropping slightly to 2.54"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# -- EDA Section -- "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# -- EDA Contents --\n",
    "# 1. Songs with most listens (top 10) + visualisation\n",
    "# 2. Highest average number of listens (top 10)\n",
    "# 3. Songs with highest different number of listeners (unique listeners - top 10)\n",
    "# 4. Artists with most listens + visualisation (top 10)\n",
    "# 5. Artists with highest average number of plays\n",
    "# 6. Most polarising songs(largest std in listen count)\n",
    "# 7. Analysis by year\n",
    "# 8. Cosine Similarity\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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>user_id</th>\n",
       "      <th>song_id</th>\n",
       "      <th>listen_count</th>\n",
       "      <th>z-score</th>\n",
       "      <th>title</th>\n",
       "      <th>release</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOAKIMP12A8C130995</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>The Cove</td>\n",
       "      <td>Thicker Than Water</td>\n",
       "      <td>Jack Johnson</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBBMDR12A8C13253B</td>\n",
       "      <td>2</td>\n",
       "      <td>-0.16</td>\n",
       "      <td>Entre Dos Aguas</td>\n",
       "      <td>Flamenco Para Niños</td>\n",
       "      <td>Paco De Lucia</td>\n",
       "      <td>1976</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBXHDL12A81C204C0</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Stronger</td>\n",
       "      <td>Graduation</td>\n",
       "      <td>Kanye West</td>\n",
       "      <td>2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBYHAJ12A6701BF1D</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Constellations</td>\n",
       "      <td>In Between Dreams</td>\n",
       "      <td>Jack Johnson</td>\n",
       "      <td>2005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SODACBL12A8C13C273</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Learn To Fly</td>\n",
       "      <td>There Is Nothing Left To Lose</td>\n",
       "      <td>Foo Fighters</td>\n",
       "      <td>1999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973847</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJEYPO12AAA8C6B0E</td>\n",
       "      <td>2</td>\n",
       "      <td>-0.16</td>\n",
       "      <td>Ignorance (Album Version)</td>\n",
       "      <td>Ignorance</td>\n",
       "      <td>Paramore</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973848</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJJYDE12AF729FC16</td>\n",
       "      <td>4</td>\n",
       "      <td>0.15</td>\n",
       "      <td>Two Is Better Than One</td>\n",
       "      <td>Love Drunk</td>\n",
       "      <td>Boys Like Girls featuring Taylor Swift</td>\n",
       "      <td>2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973849</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJKQSF12A6D4F5EE9</td>\n",
       "      <td>3</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>What I've Done (Album Version)</td>\n",
       "      <td>What I've Done</td>\n",
       "      <td>Linkin Park</td>\n",
       "      <td>2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973850</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJUXGA12AC961885C</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Up</td>\n",
       "      <td>My Worlds</td>\n",
       "      <td>Justin Bieber</td>\n",
       "      <td>2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973851</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJYOLS12A8C13C06F</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Soil_ Soil (Album Version)</td>\n",
       "      <td>The Con</td>\n",
       "      <td>Tegan And Sara</td>\n",
       "      <td>2007</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1973852 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          user_id             song_id  \\\n",
       "0        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOAKIMP12A8C130995   \n",
       "1        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBBMDR12A8C13253B   \n",
       "2        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBXHDL12A81C204C0   \n",
       "3        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBYHAJ12A6701BF1D   \n",
       "4        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SODACBL12A8C13C273   \n",
       "...                                           ...                 ...   \n",
       "1973847  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJEYPO12AAA8C6B0E   \n",
       "1973848  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJJYDE12AF729FC16   \n",
       "1973849  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJKQSF12A6D4F5EE9   \n",
       "1973850  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJUXGA12AC961885C   \n",
       "1973851  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJYOLS12A8C13C06F   \n",
       "\n",
       "         listen_count  z-score                           title  \\\n",
       "0                   1    -0.31                        The Cove   \n",
       "1                   2    -0.16                 Entre Dos Aguas   \n",
       "2                   1    -0.31                        Stronger   \n",
       "3                   1    -0.31                  Constellations   \n",
       "4                   1    -0.31                    Learn To Fly   \n",
       "...               ...      ...                             ...   \n",
       "1973847             2    -0.16       Ignorance (Album Version)   \n",
       "1973848             4     0.15          Two Is Better Than One   \n",
       "1973849             3    -0.01  What I've Done (Album Version)   \n",
       "1973850             1    -0.31                              Up   \n",
       "1973851             1    -0.31      Soil_ Soil (Album Version)   \n",
       "\n",
       "                               release  \\\n",
       "0                   Thicker Than Water   \n",
       "1                  Flamenco Para Niños   \n",
       "2                           Graduation   \n",
       "3                    In Between Dreams   \n",
       "4        There Is Nothing Left To Lose   \n",
       "...                                ...   \n",
       "1973847                      Ignorance   \n",
       "1973848                     Love Drunk   \n",
       "1973849                 What I've Done   \n",
       "1973850                      My Worlds   \n",
       "1973851                        The Con   \n",
       "\n",
       "                                    artist_name  year  \n",
       "0                                  Jack Johnson     0  \n",
       "1                                 Paco De Lucia  1976  \n",
       "2                                    Kanye West  2007  \n",
       "3                                  Jack Johnson  2005  \n",
       "4                                  Foo Fighters  1999  \n",
       "...                                         ...   ...  \n",
       "1973847                                Paramore     0  \n",
       "1973848  Boys Like Girls featuring Taylor Swift  2009  \n",
       "1973849                             Linkin Park  2007  \n",
       "1973850                           Justin Bieber  2010  \n",
       "1973851                          Tegan And Sara  2007  \n",
       "\n",
       "[1973852 rows x 8 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Merge songs on triplets df \n",
    "merged_df = pd.merge(triplets_df, song_df, how='left', on='song_id')\n",
    "merged_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5019340"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "triplets_df['listen_count'].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5019340"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "merged_df['listen_count'].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "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>user_id</th>\n",
       "      <th>song_id</th>\n",
       "      <th>listen_count</th>\n",
       "      <th>z-score</th>\n",
       "      <th>title</th>\n",
       "      <th>release</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [user_id, song_id, listen_count, z-score, title, release, artist_name, year]\n",
       "Index: []"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check for no song-user pair duplicates in newly merged df\n",
    "merged_df[merged_df.duplicated(subset=['user_id','song_id'])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "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>song_id</th>\n",
       "      <th>listen_count</th>\n",
       "      <th>title</th>\n",
       "      <th>artist_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>317</th>\n",
       "      <td>SOAUWYT12A81C206F1</td>\n",
       "      <td>28929</td>\n",
       "      <td>Undo</td>\n",
       "      <td>Björk</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>614</th>\n",
       "      <td>SOBONKR12A58A7A7E0</td>\n",
       "      <td>27754</td>\n",
       "      <td>You're The One</td>\n",
       "      <td>Dwight Yoakam</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2220</th>\n",
       "      <td>SOFRQTD12A81C233C0</td>\n",
       "      <td>26943</td>\n",
       "      <td>Sehr kosmisch</td>\n",
       "      <td>Harmonia</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7416</th>\n",
       "      <td>SOSXLTC12AF72A7F54</td>\n",
       "      <td>23099</td>\n",
       "      <td>Revelry</td>\n",
       "      <td>Kings Of Leon</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>352</th>\n",
       "      <td>SOAXGDH12A8C13F8A1</td>\n",
       "      <td>22726</td>\n",
       "      <td>Dog Days Are Over (Radio Edit)</td>\n",
       "      <td>Florence + The Machine</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1664</th>\n",
       "      <td>SOEGIYH12A6D4FC0E3</td>\n",
       "      <td>19359</td>\n",
       "      <td>Horn Concerto No. 4 in E flat K495: II. Romanc...</td>\n",
       "      <td>Barry Tuckwell/Academy of St Martin-in-the-Fie...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5531</th>\n",
       "      <td>SONYKOW12AB01849C9</td>\n",
       "      <td>18485</td>\n",
       "      <td>Secrets</td>\n",
       "      <td>OneRepublic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1334</th>\n",
       "      <td>SODJWHY12A8C142CCE</td>\n",
       "      <td>13943</td>\n",
       "      <td>Hey_ Soul Sister</td>\n",
       "      <td>Train</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6246</th>\n",
       "      <td>SOPUCYA12A8C13A694</td>\n",
       "      <td>13876</td>\n",
       "      <td>Canada</td>\n",
       "      <td>Five Iron Frenzy</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4448</th>\n",
       "      <td>SOLFXKT12AB017E3E0</td>\n",
       "      <td>13154</td>\n",
       "      <td>Fireflies</td>\n",
       "      <td>Charttraxx Karaoke</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 song_id  listen_count  \\\n",
       "317   SOAUWYT12A81C206F1         28929   \n",
       "614   SOBONKR12A58A7A7E0         27754   \n",
       "2220  SOFRQTD12A81C233C0         26943   \n",
       "7416  SOSXLTC12AF72A7F54         23099   \n",
       "352   SOAXGDH12A8C13F8A1         22726   \n",
       "1664  SOEGIYH12A6D4FC0E3         19359   \n",
       "5531  SONYKOW12AB01849C9         18485   \n",
       "1334  SODJWHY12A8C142CCE         13943   \n",
       "6246  SOPUCYA12A8C13A694         13876   \n",
       "4448  SOLFXKT12AB017E3E0         13154   \n",
       "\n",
       "                                                  title  \\\n",
       "317                                                Undo   \n",
       "614                                      You're The One   \n",
       "2220                                      Sehr kosmisch   \n",
       "7416                                            Revelry   \n",
       "352                      Dog Days Are Over (Radio Edit)   \n",
       "1664  Horn Concerto No. 4 in E flat K495: II. Romanc...   \n",
       "5531                                            Secrets   \n",
       "1334                                   Hey_ Soul Sister   \n",
       "6246                                             Canada   \n",
       "4448                                          Fireflies   \n",
       "\n",
       "                                            artist_name  \n",
       "317                                               Björk  \n",
       "614                                       Dwight Yoakam  \n",
       "2220                                           Harmonia  \n",
       "7416                                      Kings Of Leon  \n",
       "352                              Florence + The Machine  \n",
       "1664  Barry Tuckwell/Academy of St Martin-in-the-Fie...  \n",
       "5531                                        OneRepublic  \n",
       "1334                                              Train  \n",
       "6246                                   Five Iron Frenzy  \n",
       "4448                                 Charttraxx Karaoke  "
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 1. Songs with most listens (top 10)\n",
    "songs_popular_10 = merged_df.groupby('song_id').agg({'listen_count': 'sum','title':'first','artist_name': 'first'}).reset_index().sort_values('listen_count', ascending=False).head(10)\n",
    "songs_popular_10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9]), <a list of 10 Text xticklabel objects>)"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualisation of top 10 listen counts\n",
    "plt.figure(figsize=(12, 6))\n",
    "top_10_df = triplets_df[triplets_df['song_id'].isin(songs_popular_10['song_id'].head(10))]\n",
    "top_10_df = pd.merge(top_10_df, song_df[['song_id','title']], how = 'left', on='song_id')\n",
    "sns.barplot(data = top_10_df[['title','listen_count']], x = 'title', y='listen_count')\n",
    "plt.xticks(rotation=90)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>song_id</th>\n",
       "      <th>listen_count</th>\n",
       "      <th>title</th>\n",
       "      <th>artist_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3374</th>\n",
       "      <td>SOINHRP12A8C1401FF</td>\n",
       "      <td>7.11</td>\n",
       "      <td>Novocaine</td>\n",
       "      <td>Kalafut</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5079</th>\n",
       "      <td>SOMVIOV12A6D4F719A</td>\n",
       "      <td>6.71</td>\n",
       "      <td>Tennessee (Pirate Radio Mix)</td>\n",
       "      <td>ARRESTED DEVELOPMENT</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2287</th>\n",
       "      <td>SOFWKCI12A8C13A22A</td>\n",
       "      <td>6.63</td>\n",
       "      <td>Bros Sto Rimagmeno Spiti (2003 Digital Remaster)</td>\n",
       "      <td>George Dalaras</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3987</th>\n",
       "      <td>SOJZSWQ12A8C13BD47</td>\n",
       "      <td>6.37</td>\n",
       "      <td>Canción Para Luchar</td>\n",
       "      <td>León Gieco</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>SOACBLB12AB01871C7</td>\n",
       "      <td>6.34</td>\n",
       "      <td>221</td>\n",
       "      <td>keller williams</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2659</th>\n",
       "      <td>SOGSDHY12AB017BF39</td>\n",
       "      <td>6.09</td>\n",
       "      <td>Working With Homesick</td>\n",
       "      <td>Homesick James</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>SOADQMO12A8C136FF9</td>\n",
       "      <td>6.07</td>\n",
       "      <td>Dumb Blood</td>\n",
       "      <td>Soft</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8908</th>\n",
       "      <td>SOWYRXW12A8C1351CA</td>\n",
       "      <td>6.03</td>\n",
       "      <td>Bedroom Suite</td>\n",
       "      <td>Joe Christmas</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1126</th>\n",
       "      <td>SOCWIJA12AB017FA92</td>\n",
       "      <td>5.90</td>\n",
       "      <td>Manuel</td>\n",
       "      <td>The Car Is On Fire</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3859</th>\n",
       "      <td>SOJSXJY12A8C13E32E</td>\n",
       "      <td>5.70</td>\n",
       "      <td>Clara meets Slope - Hard To Say</td>\n",
       "      <td>Clara Hill</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 song_id  listen_count  \\\n",
       "3374  SOINHRP12A8C1401FF          7.11   \n",
       "5079  SOMVIOV12A6D4F719A          6.71   \n",
       "2287  SOFWKCI12A8C13A22A          6.63   \n",
       "3987  SOJZSWQ12A8C13BD47          6.37   \n",
       "32    SOACBLB12AB01871C7          6.34   \n",
       "2659  SOGSDHY12AB017BF39          6.09   \n",
       "61    SOADQMO12A8C136FF9          6.07   \n",
       "8908  SOWYRXW12A8C1351CA          6.03   \n",
       "1126  SOCWIJA12AB017FA92          5.90   \n",
       "3859  SOJSXJY12A8C13E32E          5.70   \n",
       "\n",
       "                                                 title           artist_name  \n",
       "3374                                         Novocaine               Kalafut  \n",
       "5079                      Tennessee (Pirate Radio Mix)  ARRESTED DEVELOPMENT  \n",
       "2287  Bros Sto Rimagmeno Spiti (2003 Digital Remaster)        George Dalaras  \n",
       "3987                               Canción Para Luchar            León Gieco  \n",
       "32                                                 221       keller williams  \n",
       "2659                             Working With Homesick        Homesick James  \n",
       "61                                          Dumb Blood                  Soft  \n",
       "8908                                     Bedroom Suite         Joe Christmas  \n",
       "1126                                            Manuel    The Car Is On Fire  \n",
       "3859                   Clara meets Slope - Hard To Say            Clara Hill  "
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Highest average number of listens (top 10)\n",
    "merged_df.groupby('song_id').agg({'listen_count': 'mean','title':'first','artist_name': 'first'}).reset_index().sort_values('listen_count', ascending=False).head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "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>title</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>user_id_y</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3886</th>\n",
       "      <td>Sehr kosmisch</td>\n",
       "      <td>Harmonia</td>\n",
       "      <td>8165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57156</th>\n",
       "      <td>Dog Days Are Over (Radio Edit)</td>\n",
       "      <td>Florence + The Machine</td>\n",
       "      <td>6836</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50591</th>\n",
       "      <td>Undo</td>\n",
       "      <td>Björk</td>\n",
       "      <td>6565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64209</th>\n",
       "      <td>You're The One</td>\n",
       "      <td>Dwight Yoakam</td>\n",
       "      <td>6177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64210</th>\n",
       "      <td>You're The One</td>\n",
       "      <td>The Black Keys</td>\n",
       "      <td>6177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93137</th>\n",
       "      <td>Revelry</td>\n",
       "      <td>Kings Of Leon</td>\n",
       "      <td>5794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87157</th>\n",
       "      <td>Secrets</td>\n",
       "      <td>OneRepublic</td>\n",
       "      <td>5752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76573</th>\n",
       "      <td>Horn Concerto No. 4 in E flat K495: II. Romanc...</td>\n",
       "      <td>Barry Tuckwell/Academy of St Martin-in-the-Fie...</td>\n",
       "      <td>5110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82405</th>\n",
       "      <td>Fireflies</td>\n",
       "      <td>Charttraxx Karaoke</td>\n",
       "      <td>4752</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71855</th>\n",
       "      <td>Hey_ Soul Sister</td>\n",
       "      <td>Train</td>\n",
       "      <td>4718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>571351</th>\n",
       "      <td>Tive Sim</td>\n",
       "      <td>Cartola</td>\n",
       "      <td>4498</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    title  \\\n",
       "3886                                        Sehr kosmisch   \n",
       "57156                      Dog Days Are Over (Radio Edit)   \n",
       "50591                                                Undo   \n",
       "64209                                      You're The One   \n",
       "64210                                      You're The One   \n",
       "93137                                             Revelry   \n",
       "87157                                             Secrets   \n",
       "76573   Horn Concerto No. 4 in E flat K495: II. Romanc...   \n",
       "82405                                           Fireflies   \n",
       "71855                                    Hey_ Soul Sister   \n",
       "571351                                           Tive Sim   \n",
       "\n",
       "                                              artist_name  user_id_y  \n",
       "3886                                             Harmonia       8165  \n",
       "57156                              Florence + The Machine       6836  \n",
       "50591                                               Björk       6565  \n",
       "64209                                       Dwight Yoakam       6177  \n",
       "64210                                      The Black Keys       6177  \n",
       "93137                                       Kings Of Leon       5794  \n",
       "87157                                         OneRepublic       5752  \n",
       "76573   Barry Tuckwell/Academy of St Martin-in-the-Fie...       5110  \n",
       "82405                                  Charttraxx Karaoke       4752  \n",
       "71855                                               Train       4718  \n",
       "571351                                            Cartola       4498  "
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Top 10 songs with highest different number of listeners\n",
    "unique_user_counts = merged_df.groupby('title')['user_id'].nunique().reset_index()\n",
    "merged_with_counts = pd.merge(merged_df, unique_user_counts, on='title')\n",
    "top_10_songs_user = merged_with_counts[merged_with_counts['title'].isin(unique_user_counts.nlargest(10, 'user_id')['title'])]\n",
    "top_10_songs_user = top_10_songs_user.drop_duplicates(subset=['title', 'artist_name'])\n",
    "top_10_songs_user[['title', 'artist_name','user_id_y']].sort_values('user_id_y', ascending = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "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>artist_name</th>\n",
       "      <th>listen_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>649</th>\n",
       "      <td>Coldplay</td>\n",
       "      <td>65045</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1107</th>\n",
       "      <td>Florence + The Machine</td>\n",
       "      <td>51885</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1651</th>\n",
       "      <td>Kings Of Leon</td>\n",
       "      <td>48863</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2850</th>\n",
       "      <td>The Black Keys</td>\n",
       "      <td>46381</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1370</th>\n",
       "      <td>Jack Johnson</td>\n",
       "      <td>40183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1554</th>\n",
       "      <td>Justin Bieber</td>\n",
       "      <td>35236</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2205</th>\n",
       "      <td>OneRepublic</td>\n",
       "      <td>34696</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3174</th>\n",
       "      <td>Train</td>\n",
       "      <td>33593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2073</th>\n",
       "      <td>Muse</td>\n",
       "      <td>33580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>345</th>\n",
       "      <td>Björk</td>\n",
       "      <td>32412</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 artist_name  listen_count\n",
       "649                 Coldplay         65045\n",
       "1107  Florence + The Machine         51885\n",
       "1651           Kings Of Leon         48863\n",
       "2850          The Black Keys         46381\n",
       "1370            Jack Johnson         40183\n",
       "1554           Justin Bieber         35236\n",
       "2205             OneRepublic         34696\n",
       "3174                   Train         33593\n",
       "2073                    Muse         33580\n",
       "345                    Björk         32412"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Artists with most listens (top 10)\n",
    "artist_10 = merged_df.groupby('artist_name').agg({'listen_count': 'sum'}).reset_index().sort_values('listen_count', ascending=False).head(10)\n",
    "artist_10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x23b841f7e88>"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualisation of top 10 artists\n",
    "plt.figure(figsize=(12,6))\n",
    "plt.xticks(rotation=45)\n",
    "sns.barplot(data = artist_10, x = 'artist_name', y = 'listen_count')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "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>artist_name</th>\n",
       "      <th>listen_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1574</th>\n",
       "      <td>Kalafut</td>\n",
       "      <td>7.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1187</th>\n",
       "      <td>George Dalaras</td>\n",
       "      <td>6.63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1741</th>\n",
       "      <td>León Gieco</td>\n",
       "      <td>6.37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3368</th>\n",
       "      <td>keller williams</td>\n",
       "      <td>6.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1305</th>\n",
       "      <td>Homesick James</td>\n",
       "      <td>6.09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2670</th>\n",
       "      <td>Soft</td>\n",
       "      <td>6.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1457</th>\n",
       "      <td>Joe Christmas</td>\n",
       "      <td>6.03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2863</th>\n",
       "      <td>The Car Is On Fire</td>\n",
       "      <td>5.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>630</th>\n",
       "      <td>Clara Hill</td>\n",
       "      <td>5.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>493</th>\n",
       "      <td>CREMATORY</td>\n",
       "      <td>5.67</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             artist_name  listen_count\n",
       "1574             Kalafut          7.11\n",
       "1187      George Dalaras          6.63\n",
       "1741          León Gieco          6.37\n",
       "3368     keller williams          6.34\n",
       "1305      Homesick James          6.09\n",
       "2670                Soft          6.07\n",
       "1457       Joe Christmas          6.03\n",
       "2863  The Car Is On Fire          5.90\n",
       "630           Clara Hill          5.70\n",
       "493            CREMATORY          5.67"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Artists with highest average number of plays\n",
    "merged_df.groupby('artist_name').agg({'listen_count': 'mean'}).reset_index().sort_values('listen_count', ascending=False).head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "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>song_id</th>\n",
       "      <th>listen_count</th>\n",
       "      <th>title</th>\n",
       "      <th>artist_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1823</th>\n",
       "      <td>SOERPRE12A58A7E653</td>\n",
       "      <td>7.06</td>\n",
       "      <td>Scissors (Album Version)</td>\n",
       "      <td>Slipknot</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>SOACBLB12AB01871C7</td>\n",
       "      <td>6.03</td>\n",
       "      <td>221</td>\n",
       "      <td>keller williams</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3374</th>\n",
       "      <td>SOINHRP12A8C1401FF</td>\n",
       "      <td>5.94</td>\n",
       "      <td>Novocaine</td>\n",
       "      <td>Kalafut</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>SOADQMO12A8C136FF9</td>\n",
       "      <td>5.92</td>\n",
       "      <td>Dumb Blood</td>\n",
       "      <td>Soft</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>SOAFPSO12AF72A4521</td>\n",
       "      <td>5.90</td>\n",
       "      <td>Caroline</td>\n",
       "      <td>CREMATORY</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5240</th>\n",
       "      <td>SONGBWJ12A58A7AB4D</td>\n",
       "      <td>5.79</td>\n",
       "      <td>Rock 'n' Roll</td>\n",
       "      <td>Generation X</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5213</th>\n",
       "      <td>SONENUZ12A8C14230C</td>\n",
       "      <td>5.75</td>\n",
       "      <td>Tenderly</td>\n",
       "      <td>Euge Groove</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2287</th>\n",
       "      <td>SOFWKCI12A8C13A22A</td>\n",
       "      <td>5.74</td>\n",
       "      <td>Bros Sto Rimagmeno Spiti (2003 Digital Remaster)</td>\n",
       "      <td>George Dalaras</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9733</th>\n",
       "      <td>SOZEBAZ12AF72A80C8</td>\n",
       "      <td>5.62</td>\n",
       "      <td>Voices On A String (Album Version)</td>\n",
       "      <td>Thursday</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1126</th>\n",
       "      <td>SOCWIJA12AB017FA92</td>\n",
       "      <td>5.62</td>\n",
       "      <td>Manuel</td>\n",
       "      <td>The Car Is On Fire</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 song_id  listen_count  \\\n",
       "1823  SOERPRE12A58A7E653          7.06   \n",
       "32    SOACBLB12AB01871C7          6.03   \n",
       "3374  SOINHRP12A8C1401FF          5.94   \n",
       "61    SOADQMO12A8C136FF9          5.92   \n",
       "92    SOAFPSO12AF72A4521          5.90   \n",
       "5240  SONGBWJ12A58A7AB4D          5.79   \n",
       "5213  SONENUZ12A8C14230C          5.75   \n",
       "2287  SOFWKCI12A8C13A22A          5.74   \n",
       "9733  SOZEBAZ12AF72A80C8          5.62   \n",
       "1126  SOCWIJA12AB017FA92          5.62   \n",
       "\n",
       "                                                 title         artist_name  \n",
       "1823                          Scissors (Album Version)            Slipknot  \n",
       "32                                                 221     keller williams  \n",
       "3374                                         Novocaine             Kalafut  \n",
       "61                                          Dumb Blood                Soft  \n",
       "92                                            Caroline           CREMATORY  \n",
       "5240                                     Rock 'n' Roll        Generation X  \n",
       "5213                                          Tenderly         Euge Groove  \n",
       "2287  Bros Sto Rimagmeno Spiti (2003 Digital Remaster)      George Dalaras  \n",
       "9733                Voices On A String (Album Version)            Thursday  \n",
       "1126                                            Manuel  The Car Is On Fire  "
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Most polarising songs, largest std in listen count\n",
    "merged_df.groupby('song_id').agg({'listen_count': 'std','title':'first','artist_name': 'first'}).reset_index().sort_values('listen_count', ascending=False).head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "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>user_id</th>\n",
       "      <th>song_id</th>\n",
       "      <th>listen_count</th>\n",
       "      <th>z-score</th>\n",
       "      <th>title</th>\n",
       "      <th>release</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOAKIMP12A8C130995</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>The Cove</td>\n",
       "      <td>Thicker Than Water</td>\n",
       "      <td>Jack Johnson</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBBMDR12A8C13253B</td>\n",
       "      <td>2</td>\n",
       "      <td>-0.16</td>\n",
       "      <td>Entre Dos Aguas</td>\n",
       "      <td>Flamenco Para Niños</td>\n",
       "      <td>Paco De Lucia</td>\n",
       "      <td>1976</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBXHDL12A81C204C0</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Stronger</td>\n",
       "      <td>Graduation</td>\n",
       "      <td>Kanye West</td>\n",
       "      <td>2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBYHAJ12A6701BF1D</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Constellations</td>\n",
       "      <td>In Between Dreams</td>\n",
       "      <td>Jack Johnson</td>\n",
       "      <td>2005</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SODACBL12A8C13C273</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Learn To Fly</td>\n",
       "      <td>There Is Nothing Left To Lose</td>\n",
       "      <td>Foo Fighters</td>\n",
       "      <td>1999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973847</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJEYPO12AAA8C6B0E</td>\n",
       "      <td>2</td>\n",
       "      <td>-0.16</td>\n",
       "      <td>Ignorance (Album Version)</td>\n",
       "      <td>Ignorance</td>\n",
       "      <td>Paramore</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973848</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJJYDE12AF729FC16</td>\n",
       "      <td>4</td>\n",
       "      <td>0.15</td>\n",
       "      <td>Two Is Better Than One</td>\n",
       "      <td>Love Drunk</td>\n",
       "      <td>Boys Like Girls featuring Taylor Swift</td>\n",
       "      <td>2009</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973849</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJKQSF12A6D4F5EE9</td>\n",
       "      <td>3</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>What I've Done (Album Version)</td>\n",
       "      <td>What I've Done</td>\n",
       "      <td>Linkin Park</td>\n",
       "      <td>2007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973850</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJUXGA12AC961885C</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Up</td>\n",
       "      <td>My Worlds</td>\n",
       "      <td>Justin Bieber</td>\n",
       "      <td>2010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1973851</th>\n",
       "      <td>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</td>\n",
       "      <td>SOJYOLS12A8C13C06F</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Soil_ Soil (Album Version)</td>\n",
       "      <td>The Con</td>\n",
       "      <td>Tegan And Sara</td>\n",
       "      <td>2007</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1973852 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          user_id             song_id  \\\n",
       "0        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOAKIMP12A8C130995   \n",
       "1        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBBMDR12A8C13253B   \n",
       "2        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBXHDL12A81C204C0   \n",
       "3        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBYHAJ12A6701BF1D   \n",
       "4        b80344d063b5ccb3212f76538f3d9e43d87dca9e  SODACBL12A8C13C273   \n",
       "...                                           ...                 ...   \n",
       "1973847  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJEYPO12AAA8C6B0E   \n",
       "1973848  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJJYDE12AF729FC16   \n",
       "1973849  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJKQSF12A6D4F5EE9   \n",
       "1973850  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJUXGA12AC961885C   \n",
       "1973851  d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  SOJYOLS12A8C13C06F   \n",
       "\n",
       "         listen_count  z-score                           title  \\\n",
       "0                   1    -0.31                        The Cove   \n",
       "1                   2    -0.16                 Entre Dos Aguas   \n",
       "2                   1    -0.31                        Stronger   \n",
       "3                   1    -0.31                  Constellations   \n",
       "4                   1    -0.31                    Learn To Fly   \n",
       "...               ...      ...                             ...   \n",
       "1973847             2    -0.16       Ignorance (Album Version)   \n",
       "1973848             4     0.15          Two Is Better Than One   \n",
       "1973849             3    -0.01  What I've Done (Album Version)   \n",
       "1973850             1    -0.31                              Up   \n",
       "1973851             1    -0.31      Soil_ Soil (Album Version)   \n",
       "\n",
       "                               release  \\\n",
       "0                   Thicker Than Water   \n",
       "1                  Flamenco Para Niños   \n",
       "2                           Graduation   \n",
       "3                    In Between Dreams   \n",
       "4        There Is Nothing Left To Lose   \n",
       "...                                ...   \n",
       "1973847                      Ignorance   \n",
       "1973848                     Love Drunk   \n",
       "1973849                 What I've Done   \n",
       "1973850                      My Worlds   \n",
       "1973851                        The Con   \n",
       "\n",
       "                                    artist_name  year  \n",
       "0                                  Jack Johnson     0  \n",
       "1                                 Paco De Lucia  1976  \n",
       "2                                    Kanye West  2007  \n",
       "3                                  Jack Johnson  2005  \n",
       "4                                  Foo Fighters  1999  \n",
       "...                                         ...   ...  \n",
       "1973847                                Paramore     0  \n",
       "1973848  Boys Like Girls featuring Taylor Swift  2009  \n",
       "1973849                             Linkin Park  2007  \n",
       "1973850                           Justin Bieber  2010  \n",
       "1973851                          Tegan And Sara  2007  \n",
       "\n",
       "[1973852 rows x 8 columns]"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "merged_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Analysis by Decade\n",
    "def decade(row):\n",
    "    return (row['year']//10)*10\n",
    "\n",
    "# Create column for decade with apply\n",
    "merged_df['decade'] = merged_df.apply(decade, axis=1)     "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "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>user_id</th>\n",
       "      <th>song_id</th>\n",
       "      <th>listen_count</th>\n",
       "      <th>z-score</th>\n",
       "      <th>title</th>\n",
       "      <th>release</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>year</th>\n",
       "      <th>decade</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOAKIMP12A8C130995</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>The Cove</td>\n",
       "      <td>Thicker Than Water</td>\n",
       "      <td>Jack Johnson</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBBMDR12A8C13253B</td>\n",
       "      <td>2</td>\n",
       "      <td>-0.16</td>\n",
       "      <td>Entre Dos Aguas</td>\n",
       "      <td>Flamenco Para Niños</td>\n",
       "      <td>Paco De Lucia</td>\n",
       "      <td>1976</td>\n",
       "      <td>1970</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBXHDL12A81C204C0</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Stronger</td>\n",
       "      <td>Graduation</td>\n",
       "      <td>Kanye West</td>\n",
       "      <td>2007</td>\n",
       "      <td>2000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SOBYHAJ12A6701BF1D</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Constellations</td>\n",
       "      <td>In Between Dreams</td>\n",
       "      <td>Jack Johnson</td>\n",
       "      <td>2005</td>\n",
       "      <td>2000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>b80344d063b5ccb3212f76538f3d9e43d87dca9e</td>\n",
       "      <td>SODACBL12A8C13C273</td>\n",
       "      <td>1</td>\n",
       "      <td>-0.31</td>\n",
       "      <td>Learn To Fly</td>\n",
       "      <td>There Is Nothing Left To Lose</td>\n",
       "      <td>Foo Fighters</td>\n",
       "      <td>1999</td>\n",
       "      <td>1990</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    user_id             song_id  listen_count  \\\n",
       "0  b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOAKIMP12A8C130995             1   \n",
       "1  b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBBMDR12A8C13253B             2   \n",
       "2  b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBXHDL12A81C204C0             1   \n",
       "3  b80344d063b5ccb3212f76538f3d9e43d87dca9e  SOBYHAJ12A6701BF1D             1   \n",
       "4  b80344d063b5ccb3212f76538f3d9e43d87dca9e  SODACBL12A8C13C273             1   \n",
       "\n",
       "   z-score            title                        release    artist_name  \\\n",
       "0    -0.31         The Cove             Thicker Than Water   Jack Johnson   \n",
       "1    -0.16  Entre Dos Aguas            Flamenco Para Niños  Paco De Lucia   \n",
       "2    -0.31         Stronger                     Graduation     Kanye West   \n",
       "3    -0.31   Constellations              In Between Dreams   Jack Johnson   \n",
       "4    -0.31     Learn To Fly  There Is Nothing Left To Lose   Foo Fighters   \n",
       "\n",
       "   year  decade  \n",
       "0     0       0  \n",
       "1  1976    1970  \n",
       "2  2007    2000  \n",
       "3  2005    2000  \n",
       "4  1999    1990  "
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "merged_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "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>decade</th>\n",
       "      <th>listen_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1950</td>\n",
       "      <td>6427</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1960</td>\n",
       "      <td>67338</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1970</td>\n",
       "      <td>162892</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1980</td>\n",
       "      <td>259999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1990</td>\n",
       "      <td>652206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2000</td>\n",
       "      <td>2710059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2010</td>\n",
       "      <td>165160</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   decade  listen_count\n",
       "1    1950          6427\n",
       "2    1960         67338\n",
       "3    1970        162892\n",
       "4    1980        259999\n",
       "5    1990        652206\n",
       "6    2000       2710059\n",
       "7    2010        165160"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Total listen_counts by each decade\n",
    "listens_decades = merged_df[['listen_count','decade']].groupby('decade').sum()\n",
    "listens_decades = listens_decades.reset_index()\n",
    "listens_decades = listens_decades[listens_decades['decade']!=0]\n",
    "listens_decades"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x23b8473c248>"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualise total listens by decade\n",
    "listens_decades.plot(kind = 'line', x = 'decade', y = 'listen_count', figsize=(12,6) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "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>decade</th>\n",
       "      <th>user_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>66396</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1950</td>\n",
       "      <td>2008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1960</td>\n",
       "      <td>15267</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1970</td>\n",
       "      <td>30352</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1980</td>\n",
       "      <td>38256</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1990</td>\n",
       "      <td>58423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2000</td>\n",
       "      <td>73833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2010</td>\n",
       "      <td>25540</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   decade  user_id\n",
       "0       0    66396\n",
       "1    1950     2008\n",
       "2    1960    15267\n",
       "3    1970    30352\n",
       "4    1980    38256\n",
       "5    1990    58423\n",
       "6    2000    73833\n",
       "7    2010    25540"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Total unique users by decade \n",
    "unique_decade = merged_df[['user_id','decade']].groupby('decade').nunique().reset_index()\n",
    "unique_decade"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x23b847bd948>"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualisation of number of unique users by decade \n",
    "plt.figure(figsize=(12,6))\n",
    "sns.barplot(data = unique_decade[unique_decade['decade']!=0], x = 'decade', y = 'user_id' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "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>decade</th>\n",
       "      <th>song_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1950</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1960</td>\n",
       "      <td>147</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1970</td>\n",
       "      <td>326</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1980</td>\n",
       "      <td>561</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1990</td>\n",
       "      <td>1356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2000</td>\n",
       "      <td>5534</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2010</td>\n",
       "      <td>286</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   decade  song_id\n",
       "0       0     1773\n",
       "1    1950       17\n",
       "2    1960      147\n",
       "3    1970      326\n",
       "4    1980      561\n",
       "5    1990     1356\n",
       "6    2000     5534\n",
       "7    2010      286"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of different songs each decade\n",
    "no_songs_decade = merged_df[['song_id','decade']].groupby('decade').nunique().reset_index()\n",
    "no_songs_decade"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x23b8480dd48>"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualisation of number of different songs by each decade\n",
    "plt.figure(figsize=(12,6))\n",
    "sns.barplot(data = no_songs_decade[no_songs_decade['decade']!=0], x = 'decade', y = 'song_id' )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "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>year</th>\n",
       "      <th>listen_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>2.72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1954</td>\n",
       "      <td>2.82</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1955</td>\n",
       "      <td>1.54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1956</td>\n",
       "      <td>2.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1957</td>\n",
       "      <td>1.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1958</td>\n",
       "      <td>3.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1959</td>\n",
       "      <td>2.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1960</td>\n",
       "      <td>3.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1961</td>\n",
       "      <td>3.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1962</td>\n",
       "      <td>3.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1963</td>\n",
       "      <td>2.99</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1964</td>\n",
       "      <td>2.37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1965</td>\n",
       "      <td>2.49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1966</td>\n",
       "      <td>2.79</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1967</td>\n",
       "      <td>2.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1968</td>\n",
       "      <td>2.91</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1969</td>\n",
       "      <td>2.49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1970</td>\n",
       "      <td>2.52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1971</td>\n",
       "      <td>2.53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1972</td>\n",
       "      <td>2.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1973</td>\n",
       "      <td>2.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>1974</td>\n",
       "      <td>2.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>1975</td>\n",
       "      <td>2.90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>1976</td>\n",
       "      <td>2.69</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>1977</td>\n",
       "      <td>2.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1978</td>\n",
       "      <td>2.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1979</td>\n",
       "      <td>2.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1980</td>\n",
       "      <td>2.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>1981</td>\n",
       "      <td>2.31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1982</td>\n",
       "      <td>2.55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>1983</td>\n",
       "      <td>2.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>1984</td>\n",
       "      <td>2.51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>1985</td>\n",
       "      <td>2.51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>1986</td>\n",
       "      <td>2.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>1987</td>\n",
       "      <td>2.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>1988</td>\n",
       "      <td>2.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>1989</td>\n",
       "      <td>2.66</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>1990</td>\n",
       "      <td>3.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>1991</td>\n",
       "      <td>2.53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>1992</td>\n",
       "      <td>2.57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>1993</td>\n",
       "      <td>2.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>1994</td>\n",
       "      <td>2.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>1995</td>\n",
       "      <td>2.64</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>1996</td>\n",
       "      <td>2.56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>1997</td>\n",
       "      <td>2.63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>1998</td>\n",
       "      <td>2.59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>1999</td>\n",
       "      <td>2.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>2000</td>\n",
       "      <td>2.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>2001</td>\n",
       "      <td>2.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>2002</td>\n",
       "      <td>2.46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>2003</td>\n",
       "      <td>2.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>2004</td>\n",
       "      <td>2.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>2005</td>\n",
       "      <td>2.36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>2006</td>\n",
       "      <td>2.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>2007</td>\n",
       "      <td>2.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>2008</td>\n",
       "      <td>2.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>2009</td>\n",
       "      <td>2.59</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>2010</td>\n",
       "      <td>2.49</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    year  listen_count\n",
       "0      0          2.72\n",
       "1   1954          2.82\n",
       "2   1955          1.54\n",
       "3   1956          2.20\n",
       "4   1957          1.99\n",
       "5   1958          3.19\n",
       "6   1959          2.25\n",
       "7   1960          3.07\n",
       "8   1961          3.07\n",
       "9   1962          3.00\n",
       "10  1963          2.99\n",
       "11  1964          2.37\n",
       "12  1965          2.49\n",
       "13  1966          2.79\n",
       "14  1967          2.48\n",
       "15  1968          2.91\n",
       "16  1969          2.49\n",
       "17  1970          2.52\n",
       "18  1971          2.53\n",
       "19  1972          2.41\n",
       "20  1973          2.58\n",
       "21  1974          2.69\n",
       "22  1975          2.90\n",
       "23  1976          2.69\n",
       "24  1977          2.45\n",
       "25  1978          2.42\n",
       "26  1979          2.85\n",
       "27  1980          2.68\n",
       "28  1981          2.31\n",
       "29  1982          2.55\n",
       "30  1983          2.45\n",
       "31  1984          2.51\n",
       "32  1985          2.51\n",
       "33  1986          2.68\n",
       "34  1987          2.33\n",
       "35  1988          2.70\n",
       "36  1989          2.66\n",
       "37  1990          3.36\n",
       "38  1991          2.53\n",
       "39  1992          2.57\n",
       "40  1993          2.70\n",
       "41  1994          2.50\n",
       "42  1995          2.64\n",
       "43  1996          2.56\n",
       "44  1997          2.63\n",
       "45  1998          2.59\n",
       "46  1999          2.68\n",
       "47  2000          2.42\n",
       "48  2001          2.68\n",
       "49  2002          2.46\n",
       "50  2003          2.34\n",
       "51  2004          2.45\n",
       "52  2005          2.36\n",
       "53  2006          2.50\n",
       "54  2007          2.34\n",
       "55  2008          2.45\n",
       "56  2009          2.59\n",
       "57  2010          2.49"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# average listen count by year\n",
    "avg_lc_year = merged_df[['listen_count','year']].groupby('year').mean().reset_index()\n",
    "avg_lc_year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x23b84898d88>"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualisation of Average listen count \n",
    "plt.figure(figsize=(12,6))\n",
    "sns.lineplot(data = avg_lc_year[avg_lc_year['year']!=0], x = 'year', y='listen_count')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "# -- Analysis -- \n",
    "# Despite the 2000s decade having the highest listen count and containing a lot of the most popular artists, average listen counts are much higher for 1958, 1960, 1975 and 1990 \n",
    "# This is partly due to 1958, 1975 having fewer song counts and users overall so if there a few very keen listeners, for a small amount of songs it pushes the average up.\n",
    "# This is also the case for 1990 which has comparitivley few different songs, concentrating the listen counts over a smaller number and increasing the average\n",
    "# Interstingly though 1990 has a fairly similar amount of unique users to the 2000s, suggesting some of its individual songs are more popular from the 90s"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Analyse user and item relationships with Cosine Similarity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 1. item-item similarity "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create CSR Sparse Matrix\n",
    "from sklearn.metrics.pairwise import cosine_similarity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Ollie\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \n",
      "C:\\Users\\Ollie\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:3: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n"
     ]
    }
   ],
   "source": [
    "# Convert to categorical variables\n",
    "triplets_df['user_id'] = triplets_df['user_id'].astype('category')\n",
    "triplets_df['song_id'] = triplets_df['song_id'].astype('category')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get indicies for CSR matrix\n",
    "row_indices = triplets_df['user_id'].cat.codes\n",
    "col_indices = triplets_df['song_id'].cat.codes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define sparse matrix \n",
    "sparse_matrix = csr_matrix((triplets_df['listen_count'], (row_indices, col_indices)),shape=(triplets_df['user_id'].nunique(), triplets_df['song_id'].nunique()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get item-item cosine similarity \n",
    "item_item_similarity = cosine_similarity(sparse_matrix.T)\n",
    "# Get original song titles\n",
    "song_id = triplets_df['song_id'].cat.categories "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "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>SOAAAGQ12A8C1420C8</th>\n",
       "      <th>SOAACPJ12A81C21360</th>\n",
       "      <th>SOAACSG12AB018DC80</th>\n",
       "      <th>SOAAEJI12AB0188AB5</th>\n",
       "      <th>SOAAFAC12A67ADF7EB</th>\n",
       "      <th>SOAAFYH12A8C13717A</th>\n",
       "      <th>SOAAJMQ12A6D4F7D17</th>\n",
       "      <th>SOAAKPM12A58A77210</th>\n",
       "      <th>SOAALWN12A6D4F7FDA</th>\n",
       "      <th>SOAAMOW12AB018149B</th>\n",
       "      <th>...</th>\n",
       "      <th>SOZZKPR12A6D4F8147</th>\n",
       "      <th>SOZZLTY12A67AE0AD0</th>\n",
       "      <th>SOZZLZN12A8AE48D6D</th>\n",
       "      <th>SOZZRHE12A6702165F</th>\n",
       "      <th>SOZZTCU12AB0182C58</th>\n",
       "      <th>SOZZTNF12A8C139916</th>\n",
       "      <th>SOZZVWB12AB0189C30</th>\n",
       "      <th>SOZZWZV12A67AE140F</th>\n",
       "      <th>SOZZYAO12A6701FF36</th>\n",
       "      <th>SOZZZPV12A8C1444B5</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>SOAAAGQ12A8C1420C8</th>\n",
       "      <td>1.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOAACPJ12A81C21360</th>\n",
       "      <td>0.01</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOAACSG12AB018DC80</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOAAEJI12AB0188AB5</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOAAFAC12A67ADF7EB</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</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>SOZZTNF12A8C139916</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOZZVWB12AB0189C30</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOZZWZV12A67AE140F</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOZZYAO12A6701FF36</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOZZZPV12A8C1444B5</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10000 rows × 10000 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                    SOAAAGQ12A8C1420C8  SOAACPJ12A81C21360  \\\n",
       "SOAAAGQ12A8C1420C8                1.00                0.01   \n",
       "SOAACPJ12A81C21360                0.01                1.00   \n",
       "SOAACSG12AB018DC80                0.00                0.00   \n",
       "SOAAEJI12AB0188AB5                0.00                0.00   \n",
       "SOAAFAC12A67ADF7EB                0.00                0.00   \n",
       "...                                ...                 ...   \n",
       "SOZZTNF12A8C139916                0.00                0.00   \n",
       "SOZZVWB12AB0189C30                0.00                0.00   \n",
       "SOZZWZV12A67AE140F                0.00                0.00   \n",
       "SOZZYAO12A6701FF36                0.00                0.00   \n",
       "SOZZZPV12A8C1444B5                0.00                0.00   \n",
       "\n",
       "                    SOAACSG12AB018DC80  SOAAEJI12AB0188AB5  \\\n",
       "SOAAAGQ12A8C1420C8                0.00                0.00   \n",
       "SOAACPJ12A81C21360                0.00                0.00   \n",
       "SOAACSG12AB018DC80                1.00                0.00   \n",
       "SOAAEJI12AB0188AB5                0.00                1.00   \n",
       "SOAAFAC12A67ADF7EB                0.00                0.01   \n",
       "...                                ...                 ...   \n",
       "SOZZTNF12A8C139916                0.00                0.00   \n",
       "SOZZVWB12AB0189C30                0.00                0.00   \n",
       "SOZZWZV12A67AE140F                0.00                0.00   \n",
       "SOZZYAO12A6701FF36                0.00                0.00   \n",
       "SOZZZPV12A8C1444B5                0.00                0.01   \n",
       "\n",
       "                    SOAAFAC12A67ADF7EB  SOAAFYH12A8C13717A  \\\n",
       "SOAAAGQ12A8C1420C8                0.00                0.00   \n",
       "SOAACPJ12A81C21360                0.00                0.00   \n",
       "SOAACSG12AB018DC80                0.00                0.00   \n",
       "SOAAEJI12AB0188AB5                0.01                0.01   \n",
       "SOAAFAC12A67ADF7EB                1.00                0.00   \n",
       "...                                ...                 ...   \n",
       "SOZZTNF12A8C139916                0.00                0.00   \n",
       "SOZZVWB12AB0189C30                0.00                0.00   \n",
       "SOZZWZV12A67AE140F                0.00                0.01   \n",
       "SOZZYAO12A6701FF36                0.00                0.00   \n",
       "SOZZZPV12A8C1444B5                0.00                0.00   \n",
       "\n",
       "                    SOAAJMQ12A6D4F7D17  SOAAKPM12A58A77210  \\\n",
       "SOAAAGQ12A8C1420C8                0.00                0.00   \n",
       "SOAACPJ12A81C21360                0.00                0.00   \n",
       "SOAACSG12AB018DC80                0.00                0.00   \n",
       "SOAAEJI12AB0188AB5                0.00                0.00   \n",
       "SOAAFAC12A67ADF7EB                0.00                0.00   \n",
       "...                                ...                 ...   \n",
       "SOZZTNF12A8C139916                0.00                0.00   \n",
       "SOZZVWB12AB0189C30                0.00                0.00   \n",
       "SOZZWZV12A67AE140F                0.00                0.00   \n",
       "SOZZYAO12A6701FF36                0.00                0.00   \n",
       "SOZZZPV12A8C1444B5                0.00                0.00   \n",
       "\n",
       "                    SOAALWN12A6D4F7FDA  SOAAMOW12AB018149B  ...  \\\n",
       "SOAAAGQ12A8C1420C8                0.00                0.00  ...   \n",
       "SOAACPJ12A81C21360                0.00                0.02  ...   \n",
       "SOAACSG12AB018DC80                0.00                0.00  ...   \n",
       "SOAAEJI12AB0188AB5                0.00                0.00  ...   \n",
       "SOAAFAC12A67ADF7EB                0.00                0.00  ...   \n",
       "...                                ...                 ...  ...   \n",
       "SOZZTNF12A8C139916                0.00                0.00  ...   \n",
       "SOZZVWB12AB0189C30                0.00                0.00  ...   \n",
       "SOZZWZV12A67AE140F                0.00                0.00  ...   \n",
       "SOZZYAO12A6701FF36                0.00                0.00  ...   \n",
       "SOZZZPV12A8C1444B5                0.00                0.00  ...   \n",
       "\n",
       "                    SOZZKPR12A6D4F8147  SOZZLTY12A67AE0AD0  \\\n",
       "SOAAAGQ12A8C1420C8                0.00                0.00   \n",
       "SOAACPJ12A81C21360                0.00                0.00   \n",
       "SOAACSG12AB018DC80                0.00                0.00   \n",
       "SOAAEJI12AB0188AB5                0.00                0.00   \n",
       "SOAAFAC12A67ADF7EB                0.00                0.00   \n",
       "...                                ...                 ...   \n",
       "SOZZTNF12A8C139916                0.00                0.00   \n",
       "SOZZVWB12AB0189C30                0.00                0.01   \n",
       "SOZZWZV12A67AE140F                0.00                0.00   \n",
       "SOZZYAO12A6701FF36                0.00                0.00   \n",
       "SOZZZPV12A8C1444B5                0.00                0.00   \n",
       "\n",
       "                    SOZZLZN12A8AE48D6D  SOZZRHE12A6702165F  \\\n",
       "SOAAAGQ12A8C1420C8                0.00                0.00   \n",
       "SOAACPJ12A81C21360                0.00                0.00   \n",
       "SOAACSG12AB018DC80                0.00                0.00   \n",
       "SOAAEJI12AB0188AB5                0.00                0.00   \n",
       "SOAAFAC12A67ADF7EB                0.00                0.00   \n",
       "...                                ...                 ...   \n",
       "SOZZTNF12A8C139916                0.00                0.00   \n",
       "SOZZVWB12AB0189C30                0.00                0.00   \n",
       "SOZZWZV12A67AE140F                0.00                0.00   \n",
       "SOZZYAO12A6701FF36                0.00                0.00   \n",
       "SOZZZPV12A8C1444B5                0.00                0.00   \n",
       "\n",
       "                    SOZZTCU12AB0182C58  SOZZTNF12A8C139916  \\\n",
       "SOAAAGQ12A8C1420C8                0.00                0.00   \n",
       "SOAACPJ12A81C21360                0.00                0.00   \n",
       "SOAACSG12AB018DC80                0.00                0.00   \n",
       "SOAAEJI12AB0188AB5                0.00                0.00   \n",
       "SOAAFAC12A67ADF7EB                0.00                0.00   \n",
       "...                                ...                 ...   \n",
       "SOZZTNF12A8C139916                0.00                1.00   \n",
       "SOZZVWB12AB0189C30                0.00                0.00   \n",
       "SOZZWZV12A67AE140F                0.00                0.00   \n",
       "SOZZYAO12A6701FF36                0.00                0.00   \n",
       "SOZZZPV12A8C1444B5                0.00                0.00   \n",
       "\n",
       "                    SOZZVWB12AB0189C30  SOZZWZV12A67AE140F  \\\n",
       "SOAAAGQ12A8C1420C8                0.00                0.00   \n",
       "SOAACPJ12A81C21360                0.00                0.00   \n",
       "SOAACSG12AB018DC80                0.00                0.00   \n",
       "SOAAEJI12AB0188AB5                0.00                0.00   \n",
       "SOAAFAC12A67ADF7EB                0.00                0.00   \n",
       "...                                ...                 ...   \n",
       "SOZZTNF12A8C139916                0.00                0.00   \n",
       "SOZZVWB12AB0189C30                1.00                0.00   \n",
       "SOZZWZV12A67AE140F                0.00                1.00   \n",
       "SOZZYAO12A6701FF36                0.00                0.00   \n",
       "SOZZZPV12A8C1444B5                0.00                0.00   \n",
       "\n",
       "                    SOZZYAO12A6701FF36  SOZZZPV12A8C1444B5  \n",
       "SOAAAGQ12A8C1420C8                0.00                0.00  \n",
       "SOAACPJ12A81C21360                0.00                0.00  \n",
       "SOAACSG12AB018DC80                0.00                0.00  \n",
       "SOAAEJI12AB0188AB5                0.00                0.01  \n",
       "SOAAFAC12A67ADF7EB                0.00                0.00  \n",
       "...                                ...                 ...  \n",
       "SOZZTNF12A8C139916                0.00                0.00  \n",
       "SOZZVWB12AB0189C30                0.00                0.00  \n",
       "SOZZWZV12A67AE140F                0.00                0.00  \n",
       "SOZZYAO12A6701FF36                1.00                0.00  \n",
       "SOZZZPV12A8C1444B5                0.00                1.00  \n",
       "\n",
       "[10000 rows x 10000 columns]"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create a DataFrame for easier analysis and visualization\n",
    "item_similarity_df = pd.DataFrame(item_item_similarity, index=song_id, columns= song_id)\n",
    "item_similarity_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "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>SOAUWYT12A81C206F1</th>\n",
       "      <th>SOBONKR12A58A7A7E0</th>\n",
       "      <th>SOFRQTD12A81C233C0</th>\n",
       "      <th>SOSXLTC12AF72A7F54</th>\n",
       "      <th>SOAXGDH12A8C13F8A1</th>\n",
       "      <th>SOEGIYH12A6D4FC0E3</th>\n",
       "      <th>SONYKOW12AB01849C9</th>\n",
       "      <th>SODJWHY12A8C142CCE</th>\n",
       "      <th>SOPUCYA12A8C13A694</th>\n",
       "      <th>SOLFXKT12AB017E3E0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>SOAAAGQ12A8C1420C8</th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOAACPJ12A81C21360</th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOAACSG12AB018DC80</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOAAEJI12AB0188AB5</th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOAAFAC12A67ADF7EB</th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOZZTNF12A8C139916</th>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOZZVWB12AB0189C30</th>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOZZWZV12A67AE140F</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOZZYAO12A6701FF36</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SOZZZPV12A8C1444B5</th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10000 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                    SOAUWYT12A81C206F1  SOBONKR12A58A7A7E0  \\\n",
       "SOAAAGQ12A8C1420C8                0.01                0.01   \n",
       "SOAACPJ12A81C21360                0.01                0.00   \n",
       "SOAACSG12AB018DC80                0.00                0.01   \n",
       "SOAAEJI12AB0188AB5                0.01                0.01   \n",
       "SOAAFAC12A67ADF7EB                0.01                0.01   \n",
       "...                                ...                 ...   \n",
       "SOZZTNF12A8C139916                0.02                0.02   \n",
       "SOZZVWB12AB0189C30                0.02                0.02   \n",
       "SOZZWZV12A67AE140F                0.00                0.00   \n",
       "SOZZYAO12A6701FF36                0.00                0.00   \n",
       "SOZZZPV12A8C1444B5                0.01                0.01   \n",
       "\n",
       "                    SOFRQTD12A81C233C0  SOSXLTC12AF72A7F54  \\\n",
       "SOAAAGQ12A8C1420C8                0.00                0.02   \n",
       "SOAACPJ12A81C21360                0.00                0.01   \n",
       "SOAACSG12AB018DC80                0.02                0.01   \n",
       "SOAAEJI12AB0188AB5                0.00                0.01   \n",
       "SOAAFAC12A67ADF7EB                0.02                0.01   \n",
       "...                                ...                 ...   \n",
       "SOZZTNF12A8C139916                0.01                0.02   \n",
       "SOZZVWB12AB0189C30                0.00                0.01   \n",
       "SOZZWZV12A67AE140F                0.00                0.00   \n",
       "SOZZYAO12A6701FF36                0.00                0.00   \n",
       "SOZZZPV12A8C1444B5                0.00                0.00   \n",
       "\n",
       "                    SOAXGDH12A8C13F8A1  SOEGIYH12A6D4FC0E3  \\\n",
       "SOAAAGQ12A8C1420C8                0.01                0.03   \n",
       "SOAACPJ12A81C21360                0.01                0.01   \n",
       "SOAACSG12AB018DC80                0.00                0.01   \n",
       "SOAAEJI12AB0188AB5                0.01                0.00   \n",
       "SOAAFAC12A67ADF7EB                0.02                0.01   \n",
       "...                                ...                 ...   \n",
       "SOZZTNF12A8C139916                0.02                0.02   \n",
       "SOZZVWB12AB0189C30                0.00                0.01   \n",
       "SOZZWZV12A67AE140F                0.03                0.00   \n",
       "SOZZYAO12A6701FF36                0.00                0.00   \n",
       "SOZZZPV12A8C1444B5                0.02                0.01   \n",
       "\n",
       "                    SONYKOW12AB01849C9  SODJWHY12A8C142CCE  \\\n",
       "SOAAAGQ12A8C1420C8                0.01                0.00   \n",
       "SOAACPJ12A81C21360                0.00                0.01   \n",
       "SOAACSG12AB018DC80                0.00                0.02   \n",
       "SOAAEJI12AB0188AB5                0.00                0.00   \n",
       "SOAAFAC12A67ADF7EB                0.01                0.01   \n",
       "...                                ...                 ...   \n",
       "SOZZTNF12A8C139916                0.02                0.01   \n",
       "SOZZVWB12AB0189C30                0.01                0.00   \n",
       "SOZZWZV12A67AE140F                0.00                0.00   \n",
       "SOZZYAO12A6701FF36                0.00                0.00   \n",
       "SOZZZPV12A8C1444B5                0.02                0.00   \n",
       "\n",
       "                    SOPUCYA12A8C13A694  SOLFXKT12AB017E3E0  \n",
       "SOAAAGQ12A8C1420C8                0.00                0.00  \n",
       "SOAACPJ12A81C21360                0.02                0.00  \n",
       "SOAACSG12AB018DC80                0.00                0.00  \n",
       "SOAAEJI12AB0188AB5                0.00                0.00  \n",
       "SOAAFAC12A67ADF7EB                0.01                0.01  \n",
       "...                                ...                 ...  \n",
       "SOZZTNF12A8C139916                0.02                0.01  \n",
       "SOZZVWB12AB0189C30                0.01                0.00  \n",
       "SOZZWZV12A67AE140F                0.00                0.00  \n",
       "SOZZYAO12A6701FF36                0.00                0.00  \n",
       "SOZZZPV12A8C1444B5                0.00                0.01  \n",
       "\n",
       "[10000 rows x 10 columns]"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Filter for cosine similarities of the top 10 most popular songs, with the rest of the data set\n",
    "pop_song_list = songs_popular_10['song_id'].tolist()\n",
    "item_similarity_df_10 = item_similarity_df[pop_song_list]\n",
    "item_similarity_df_10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Ollie\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \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>song_id</th>\n",
       "      <th>title</th>\n",
       "      <th>release</th>\n",
       "      <th>artist_name</th>\n",
       "      <th>year</th>\n",
       "      <th>title_artist</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>SOQMMHC12AB0180CB8</td>\n",
       "      <td>Silent Night</td>\n",
       "      <td>Monster Ballads X-Mas</td>\n",
       "      <td>Faster Pussy cat</td>\n",
       "      <td>2003</td>\n",
       "      <td>Silent Night - Faster Pussy cat</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>SOVFVAK12A8C1350D9</td>\n",
       "      <td>Tanssi vaan</td>\n",
       "      <td>Karkuteillä</td>\n",
       "      <td>Karkkiautomaatti</td>\n",
       "      <td>1995</td>\n",
       "      <td>Tanssi vaan - Karkkiautomaatti</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>SOGTUKN12AB017F4F1</td>\n",
       "      <td>No One Could Ever</td>\n",
       "      <td>Butter</td>\n",
       "      <td>Hudson Mohawke</td>\n",
       "      <td>2006</td>\n",
       "      <td>No One Could Ever - Hudson Mohawke</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SOBNYVR12A8C13558C</td>\n",
       "      <td>Si Vos Querés</td>\n",
       "      <td>De Culo</td>\n",
       "      <td>Yerba Brava</td>\n",
       "      <td>2003</td>\n",
       "      <td>Si Vos Querés - Yerba Brava</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>SOHSBXH12A8C13B0DF</td>\n",
       "      <td>Tangle Of Aspens</td>\n",
       "      <td>Rene Ablaze Presents Winter Sessions</td>\n",
       "      <td>Der Mystic</td>\n",
       "      <td>0</td>\n",
       "      <td>Tangle Of Aspens - Der Mystic</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              song_id              title  \\\n",
       "0  SOQMMHC12AB0180CB8       Silent Night   \n",
       "1  SOVFVAK12A8C1350D9        Tanssi vaan   \n",
       "2  SOGTUKN12AB017F4F1  No One Could Ever   \n",
       "3  SOBNYVR12A8C13558C      Si Vos Querés   \n",
       "4  SOHSBXH12A8C13B0DF   Tangle Of Aspens   \n",
       "\n",
       "                                release       artist_name  year  \\\n",
       "0                 Monster Ballads X-Mas  Faster Pussy cat  2003   \n",
       "1                           Karkuteillä  Karkkiautomaatti  1995   \n",
       "2                                Butter    Hudson Mohawke  2006   \n",
       "3                               De Culo       Yerba Brava  2003   \n",
       "4  Rene Ablaze Presents Winter Sessions        Der Mystic     0   \n",
       "\n",
       "                         title_artist  \n",
       "0     Silent Night - Faster Pussy cat  \n",
       "1      Tanssi vaan - Karkkiautomaatti  \n",
       "2  No One Could Ever - Hudson Mohawke  \n",
       "3         Si Vos Querés - Yerba Brava  \n",
       "4       Tangle Of Aspens - Der Mystic  "
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get this df but with song titles and artist_name\n",
    "song_df['title_artist'] = song_df['title'] + ' - ' + song_df['artist_name'] \n",
    "song_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "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>Undo - Björk</th>\n",
       "      <th>You're The One - Dwight Yoakam</th>\n",
       "      <th>Sehr kosmisch - Harmonia</th>\n",
       "      <th>Revelry - Kings Of Leon</th>\n",
       "      <th>Dog Days Are Over (Radio Edit) - Florence + The Machine</th>\n",
       "      <th>Horn Concerto No. 4 in E flat K495: II. Romance (Andante cantabile) - Barry Tuckwell/Academy of St Martin-in-the-Fields/Sir Neville Marriner</th>\n",
       "      <th>Secrets - OneRepublic</th>\n",
       "      <th>Hey_ Soul Sister - Train</th>\n",
       "      <th>Canada - Five Iron Frenzy</th>\n",
       "      <th>Fireflies - Charttraxx Karaoke</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Orgelblut - Bohren &amp; Der Club Of Gore</th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Cearc Agus Coileach / The Hen And Cock - Mícheál Ó Súilleabháin</th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I Gotta Feeling (Edit) - The Black Eyed Peas</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Godlovesugly - Atmosphere</th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Rome Wasn't Built In A Day - Morcheeba</th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Le Million - Ste Strausz</th>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Recovered - Marco V</th>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Henrietta - The Fratellis</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sunset (Bird Of Prey) - Fatboy Slim</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fall Back - Blacks ft. P Money</th>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10000 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    Undo - Björk  \\\n",
       "Orgelblut - Bohren & Der Club Of Gore                       0.01   \n",
       "Cearc Agus Coileach / The Hen And Cock - Mícheá...          0.01   \n",
       "I Gotta Feeling (Edit) - The Black Eyed Peas                0.00   \n",
       "Godlovesugly - Atmosphere                                   0.01   \n",
       "Rome Wasn't Built In A Day - Morcheeba                      0.01   \n",
       "...                                                          ...   \n",
       "Le Million - Ste Strausz                                    0.02   \n",
       "Recovered - Marco V                                         0.02   \n",
       "Henrietta - The Fratellis                                   0.00   \n",
       "Sunset (Bird Of Prey) - Fatboy Slim                         0.00   \n",
       "Fall Back - Blacks ft. P Money                              0.01   \n",
       "\n",
       "                                                    You're The One - Dwight Yoakam  \\\n",
       "Orgelblut - Bohren & Der Club Of Gore                                         0.01   \n",
       "Cearc Agus Coileach / The Hen And Cock - Mícheá...                            0.00   \n",
       "I Gotta Feeling (Edit) - The Black Eyed Peas                                  0.01   \n",
       "Godlovesugly - Atmosphere                                                     0.01   \n",
       "Rome Wasn't Built In A Day - Morcheeba                                        0.01   \n",
       "...                                                                            ...   \n",
       "Le Million - Ste Strausz                                                      0.02   \n",
       "Recovered - Marco V                                                           0.02   \n",
       "Henrietta - The Fratellis                                                     0.00   \n",
       "Sunset (Bird Of Prey) - Fatboy Slim                                           0.00   \n",
       "Fall Back - Blacks ft. P Money                                                0.01   \n",
       "\n",
       "                                                    Sehr kosmisch - Harmonia  \\\n",
       "Orgelblut - Bohren & Der Club Of Gore                                   0.00   \n",
       "Cearc Agus Coileach / The Hen And Cock - Mícheá...                      0.00   \n",
       "I Gotta Feeling (Edit) - The Black Eyed Peas                            0.02   \n",
       "Godlovesugly - Atmosphere                                               0.00   \n",
       "Rome Wasn't Built In A Day - Morcheeba                                  0.02   \n",
       "...                                                                      ...   \n",
       "Le Million - Ste Strausz                                                0.01   \n",
       "Recovered - Marco V                                                     0.00   \n",
       "Henrietta - The Fratellis                                               0.00   \n",
       "Sunset (Bird Of Prey) - Fatboy Slim                                     0.00   \n",
       "Fall Back - Blacks ft. P Money                                          0.00   \n",
       "\n",
       "                                                    Revelry - Kings Of Leon  \\\n",
       "Orgelblut - Bohren & Der Club Of Gore                                  0.02   \n",
       "Cearc Agus Coileach / The Hen And Cock - Mícheá...                     0.01   \n",
       "I Gotta Feeling (Edit) - The Black Eyed Peas                           0.01   \n",
       "Godlovesugly - Atmosphere                                              0.01   \n",
       "Rome Wasn't Built In A Day - Morcheeba                                 0.01   \n",
       "...                                                                     ...   \n",
       "Le Million - Ste Strausz                                               0.02   \n",
       "Recovered - Marco V                                                    0.01   \n",
       "Henrietta - The Fratellis                                              0.00   \n",
       "Sunset (Bird Of Prey) - Fatboy Slim                                    0.00   \n",
       "Fall Back - Blacks ft. P Money                                         0.00   \n",
       "\n",
       "                                                    Dog Days Are Over (Radio Edit) - Florence + The Machine  \\\n",
       "Orgelblut - Bohren & Der Club Of Gore                                                            0.01         \n",
       "Cearc Agus Coileach / The Hen And Cock - Mícheá...                                               0.01         \n",
       "I Gotta Feeling (Edit) - The Black Eyed Peas                                                     0.00         \n",
       "Godlovesugly - Atmosphere                                                                        0.01         \n",
       "Rome Wasn't Built In A Day - Morcheeba                                                           0.02         \n",
       "...                                                                                               ...         \n",
       "Le Million - Ste Strausz                                                                         0.02         \n",
       "Recovered - Marco V                                                                              0.00         \n",
       "Henrietta - The Fratellis                                                                        0.03         \n",
       "Sunset (Bird Of Prey) - Fatboy Slim                                                              0.00         \n",
       "Fall Back - Blacks ft. P Money                                                                   0.02         \n",
       "\n",
       "                                                    Horn Concerto No. 4 in E flat K495: II. Romance (Andante cantabile) - Barry Tuckwell/Academy of St Martin-in-the-Fields/Sir Neville Marriner  \\\n",
       "Orgelblut - Bohren & Der Club Of Gore                                                            0.03                                                                                              \n",
       "Cearc Agus Coileach / The Hen And Cock - Mícheá...                                               0.01                                                                                              \n",
       "I Gotta Feeling (Edit) - The Black Eyed Peas                                                     0.01                                                                                              \n",
       "Godlovesugly - Atmosphere                                                                        0.00                                                                                              \n",
       "Rome Wasn't Built In A Day - Morcheeba                                                           0.01                                                                                              \n",
       "...                                                                                               ...                                                                                              \n",
       "Le Million - Ste Strausz                                                                         0.02                                                                                              \n",
       "Recovered - Marco V                                                                              0.01                                                                                              \n",
       "Henrietta - The Fratellis                                                                        0.00                                                                                              \n",
       "Sunset (Bird Of Prey) - Fatboy Slim                                                              0.00                                                                                              \n",
       "Fall Back - Blacks ft. P Money                                                                   0.01                                                                                              \n",
       "\n",
       "                                                    Secrets - OneRepublic  \\\n",
       "Orgelblut - Bohren & Der Club Of Gore                                0.01   \n",
       "Cearc Agus Coileach / The Hen And Cock - Mícheá...                   0.00   \n",
       "I Gotta Feeling (Edit) - The Black Eyed Peas                         0.00   \n",
       "Godlovesugly - Atmosphere                                            0.00   \n",
       "Rome Wasn't Built In A Day - Morcheeba                               0.01   \n",
       "...                                                                   ...   \n",
       "Le Million - Ste Strausz                                             0.02   \n",
       "Recovered - Marco V                                                  0.01   \n",
       "Henrietta - The Fratellis                                            0.00   \n",
       "Sunset (Bird Of Prey) - Fatboy Slim                                  0.00   \n",
       "Fall Back - Blacks ft. P Money                                       0.02   \n",
       "\n",
       "                                                    Hey_ Soul Sister - Train  \\\n",
       "Orgelblut - Bohren & Der Club Of Gore                                   0.00   \n",
       "Cearc Agus Coileach / The Hen And Cock - Mícheá...                      0.01   \n",
       "I Gotta Feeling (Edit) - The Black Eyed Peas                            0.02   \n",
       "Godlovesugly - Atmosphere                                               0.00   \n",
       "Rome Wasn't Built In A Day - Morcheeba                                  0.01   \n",
       "...                                                                      ...   \n",
       "Le Million - Ste Strausz                                                0.01   \n",
       "Recovered - Marco V                                                     0.00   \n",
       "Henrietta - The Fratellis                                               0.00   \n",
       "Sunset (Bird Of Prey) - Fatboy Slim                                     0.00   \n",
       "Fall Back - Blacks ft. P Money                                          0.00   \n",
       "\n",
       "                                                    Canada - Five Iron Frenzy  \\\n",
       "Orgelblut - Bohren & Der Club Of Gore                                    0.00   \n",
       "Cearc Agus Coileach / The Hen And Cock - Mícheá...                       0.02   \n",
       "I Gotta Feeling (Edit) - The Black Eyed Peas                             0.00   \n",
       "Godlovesugly - Atmosphere                                                0.00   \n",
       "Rome Wasn't Built In A Day - Morcheeba                                   0.01   \n",
       "...                                                                       ...   \n",
       "Le Million - Ste Strausz                                                 0.02   \n",
       "Recovered - Marco V                                                      0.01   \n",
       "Henrietta - The Fratellis                                                0.00   \n",
       "Sunset (Bird Of Prey) - Fatboy Slim                                      0.00   \n",
       "Fall Back - Blacks ft. P Money                                           0.00   \n",
       "\n",
       "                                                    Fireflies - Charttraxx Karaoke  \n",
       "Orgelblut - Bohren & Der Club Of Gore                                         0.00  \n",
       "Cearc Agus Coileach / The Hen And Cock - Mícheá...                            0.00  \n",
       "I Gotta Feeling (Edit) - The Black Eyed Peas                                  0.00  \n",
       "Godlovesugly - Atmosphere                                                     0.00  \n",
       "Rome Wasn't Built In A Day - Morcheeba                                        0.01  \n",
       "...                                                                            ...  \n",
       "Le Million - Ste Strausz                                                      0.01  \n",
       "Recovered - Marco V                                                           0.00  \n",
       "Henrietta - The Fratellis                                                     0.00  \n",
       "Sunset (Bird Of Prey) - Fatboy Slim                                           0.00  \n",
       "Fall Back - Blacks ft. P Money                                                0.01  \n",
       "\n",
       "[10000 rows x 10 columns]"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Create dictionary zipping song_id to now concatenated title and artist\n",
    "id_to_title_artist = dict(zip(song_df['song_id'], song_df['title_artist']))\n",
    "item_similarity_df_10 = item_similarity_df_10.rename(index=id_to_title_artist, columns=id_to_title_artist)\n",
    "item_similarity_df_10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Secrets - OneRepublic                                     1.00\n",
       "Sehr kosmisch - Harmonia                                  0.36\n",
       "Fireflies - Charttraxx Karaoke                            0.31\n",
       "Dog Days Are Over (Radio Edit) - Florence + The Machine   0.31\n",
       "All The Right Moves - OneRepublic                         0.29\n",
       "Marry Me - Train                                          0.28\n",
       "Hey_ Soul Sister - Train                                  0.28\n",
       "OMG - Usher featuring will.i.am                           0.28\n",
       "Use Somebody - Kings Of Leon                              0.25\n",
       "The Only Exception (Album Version) - Paramore             0.24\n",
       "Name: Secrets - OneRepublic, dtype: float64"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Filter for highest cosine similarity, try example with 'Secrets - One Republic'\n",
    "item_similarity_df_10['Secrets - OneRepublic'].sort_values(ascending = False).head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Upon seeing the highest cosine similarities, typically the higest 20 or 10 in the set are over 0.1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "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>Undo - Björk</th>\n",
       "      <th>You're The One - Dwight Yoakam</th>\n",
       "      <th>Sehr kosmisch - Harmonia</th>\n",
       "      <th>Revelry - Kings Of Leon</th>\n",
       "      <th>Dog Days Are Over (Radio Edit) - Florence + The Machine</th>\n",
       "      <th>Horn Concerto No. 4 in E flat K495: II. Romance (Andante cantabile) - Barry Tuckwell/Academy of St Martin-in-the-Fields/Sir Neville Marriner</th>\n",
       "      <th>Secrets - OneRepublic</th>\n",
       "      <th>Hey_ Soul Sister - Train</th>\n",
       "      <th>Canada - Five Iron Frenzy</th>\n",
       "      <th>Fireflies - Charttraxx Karaoke</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Undo - Björk</th>\n",
       "      <td>NaN</td>\n",
       "      <td>0.40</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.42</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.28</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dog Days Are Over (Radio Edit) - Florence + The Machine</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.23</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lucky (Album Version) - Jason Mraz &amp; Colbie Caillat</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.23</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>You're The One - Dwight Yoakam</th>\n",
       "      <td>0.40</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.37</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.24</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Bleed It Out [Live At Milton Keynes] - Linkin Park</th>\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>0.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Somebody To Love - Justin Bieber</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.21</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Hey_ Soul Sister - Train</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.23</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.28</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Horn Concerto No. 4 in E flat K495: II. Romance (Andante cantabile) - Barry Tuckwell/Academy of St Martin-in-the-Fields/Sir Neville Marriner</th>\n",
       "      <td>0.28</td>\n",
       "      <td>0.24</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.29</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sehr kosmisch - Harmonia</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.36</td>\n",
       "      <td>0.30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Vanilla Twilight - Owl City</th>\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>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Kiss With A Fist - Florence + The Machine</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.33</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I'm Not Calling You A Liar - Florence + The Machine</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Scientist - Coldplay</th>\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>0.21</td>\n",
       "      <td>0.20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>The Only Exception (Album Version) - Paramore</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.24</td>\n",
       "      <td>0.20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fireflies - Charttraxx Karaoke</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.32</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.23</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.32</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>All The Right Moves - OneRepublic</th>\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>0.29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I Kissed A Girl - Katy Perry</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.20</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Secrets - OneRepublic</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.36</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.31</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.28</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Between Two Lungs - Florence + The Machine</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.22</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Représente - Alliance Ethnik</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.24</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Rabbit Heart (Raise It Up) - Florence + The Machine</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>You've Got The Love - Florence + The Machine</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.34</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Pursuit Of Happiness (nightmare) - Kid Cudi / MGMT / Ratatat</th>\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>0.20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Billionaire [feat. Bruno Mars]  (Explicit Album Version) - Travie McCoy</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.25</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Revelry - Kings Of Leon</th>\n",
       "      <td>0.42</td>\n",
       "      <td>0.37</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.29</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Marry Me - Train</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.25</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.24</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.28</td>\n",
       "      <td>0.28</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Love Story - Taylor Swift</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.24</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.23</td>\n",
       "      <td>0.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>My Boy Builds Coffins - Florence + The Machine</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.27</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>OMG - Usher featuring will.i.am</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.30</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.20</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.28</td>\n",
       "      <td>0.27</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Drop The World - Lil Wayne / Eminem</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.24</td>\n",
       "      <td>0.22</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ain't Misbehavin - Sam Cooke</th>\n",
       "      <td>0.23</td>\n",
       "      <td>0.25</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.27</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Use Somebody - Kings Of Leon</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.21</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.22</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.25</td>\n",
       "      <td>0.25</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Cosmic Love - Florence + The Machine</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.33</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    Undo - Björk  \\\n",
       "Undo - Björk                                                 NaN   \n",
       "Dog Days Are Over (Radio Edit) - Florence + The...           NaN   \n",
       "Lucky (Album Version) - Jason Mraz & Colbie Cai...           NaN   \n",
       "You're The One - Dwight Yoakam                              0.40   \n",
       "Bleed It Out [Live At Milton Keynes] - Linkin Park           NaN   \n",
       "Somebody To Love - Justin Bieber                             NaN   \n",
       "Hey_ Soul Sister - Train                                     NaN   \n",
       "Horn Concerto No. 4 in E flat K495: II. Romance...          0.28   \n",
       "Sehr kosmisch - Harmonia                                     NaN   \n",
       "Vanilla Twilight - Owl City                                  NaN   \n",
       "Kiss With A Fist - Florence + The Machine                    NaN   \n",
       "I'm Not Calling You A Liar - Florence + The Mac...           NaN   \n",
       "The Scientist - Coldplay                                     NaN   \n",
       "The Only Exception (Album Version) - Paramore                NaN   \n",
       "Fireflies - Charttraxx Karaoke                               NaN   \n",
       "All The Right Moves - OneRepublic                            NaN   \n",
       "I Kissed A Girl - Katy Perry                                 NaN   \n",
       "Secrets - OneRepublic                                        NaN   \n",
       "Between Two Lungs - Florence + The Machine                   NaN   \n",
       "Représente - Alliance Ethnik                                 NaN   \n",
       "Rabbit Heart (Raise It Up) - Florence + The Mac...           NaN   \n",
       "You've Got The Love - Florence + The Machine                 NaN   \n",
       "Pursuit Of Happiness (nightmare) - Kid Cudi / M...           NaN   \n",
       "Billionaire [feat. Bruno Mars]  (Explicit Album...           NaN   \n",
       "Revelry - Kings Of Leon                                     0.42   \n",
       "Marry Me - Train                                             NaN   \n",
       "Love Story - Taylor Swift                                    NaN   \n",
       "My Boy Builds Coffins - Florence + The Machine               NaN   \n",
       "OMG - Usher featuring will.i.am                              NaN   \n",
       "Drop The World - Lil Wayne / Eminem                          NaN   \n",
       "Ain't Misbehavin - Sam Cooke                                0.23   \n",
       "Use Somebody - Kings Of Leon                                 NaN   \n",
       "Cosmic Love - Florence + The Machine                         NaN   \n",
       "\n",
       "                                                    You're The One - Dwight Yoakam  \\\n",
       "Undo - Björk                                                                  0.40   \n",
       "Dog Days Are Over (Radio Edit) - Florence + The...                             NaN   \n",
       "Lucky (Album Version) - Jason Mraz & Colbie Cai...                             NaN   \n",
       "You're The One - Dwight Yoakam                                                 NaN   \n",
       "Bleed It Out [Live At Milton Keynes] - Linkin Park                             NaN   \n",
       "Somebody To Love - Justin Bieber                                               NaN   \n",
       "Hey_ Soul Sister - Train                                                       NaN   \n",
       "Horn Concerto No. 4 in E flat K495: II. Romance...                            0.24   \n",
       "Sehr kosmisch - Harmonia                                                       NaN   \n",
       "Vanilla Twilight - Owl City                                                    NaN   \n",
       "Kiss With A Fist - Florence + The Machine                                      NaN   \n",
       "I'm Not Calling You A Liar - Florence + The Mac...                             NaN   \n",
       "The Scientist - Coldplay                                                       NaN   \n",
       "The Only Exception (Album Version) - Paramore                                  NaN   \n",
       "Fireflies - Charttraxx Karaoke                                                 NaN   \n",
       "All The Right Moves - OneRepublic                                              NaN   \n",
       "I Kissed A Girl - Katy Perry                                                   NaN   \n",
       "Secrets - OneRepublic                                                          NaN   \n",
       "Between Two Lungs - Florence + The Machine                                     NaN   \n",
       "Représente - Alliance Ethnik                                                   NaN   \n",
       "Rabbit Heart (Raise It Up) - Florence + The Mac...                             NaN   \n",
       "You've Got The Love - Florence + The Machine                                   NaN   \n",
       "Pursuit Of Happiness (nightmare) - Kid Cudi / M...                             NaN   \n",
       "Billionaire [feat. Bruno Mars]  (Explicit Album...                             NaN   \n",
       "Revelry - Kings Of Leon                                                       0.37   \n",
       "Marry Me - Train                                                               NaN   \n",
       "Love Story - Taylor Swift                                                      NaN   \n",
       "My Boy Builds Coffins - Florence + The Machine                                 NaN   \n",
       "OMG - Usher featuring will.i.am                                                NaN   \n",
       "Drop The World - Lil Wayne / Eminem                                            NaN   \n",
       "Ain't Misbehavin - Sam Cooke                                                  0.25   \n",
       "Use Somebody - Kings Of Leon                                                   NaN   \n",
       "Cosmic Love - Florence + The Machine                                           NaN   \n",
       "\n",
       "                                                    Sehr kosmisch - Harmonia  \\\n",
       "Undo - Björk                                                             NaN   \n",
       "Dog Days Are Over (Radio Edit) - Florence + The...                      0.30   \n",
       "Lucky (Album Version) - Jason Mraz & Colbie Cai...                      0.21   \n",
       "You're The One - Dwight Yoakam                                           NaN   \n",
       "Bleed It Out [Live At Milton Keynes] - Linkin Park                       NaN   \n",
       "Somebody To Love - Justin Bieber                                        0.21   \n",
       "Hey_ Soul Sister - Train                                                0.30   \n",
       "Horn Concerto No. 4 in E flat K495: II. Romance...                       NaN   \n",
       "Sehr kosmisch - Harmonia                                                 NaN   \n",
       "Vanilla Twilight - Owl City                                              NaN   \n",
       "Kiss With A Fist - Florence + The Machine                                NaN   \n",
       "I'm Not Calling You A Liar - Florence + The Mac...                       NaN   \n",
       "The Scientist - Coldplay                                                 NaN   \n",
       "The Only Exception (Album Version) - Paramore                           0.21   \n",
       "Fireflies - Charttraxx Karaoke                                          0.32   \n",
       "All The Right Moves - OneRepublic                                        NaN   \n",
       "I Kissed A Girl - Katy Perry                                            0.20   \n",
       "Secrets - OneRepublic                                                   0.36   \n",
       "Between Two Lungs - Florence + The Machine                               NaN   \n",
       "Représente - Alliance Ethnik                                             NaN   \n",
       "Rabbit Heart (Raise It Up) - Florence + The Mac...                       NaN   \n",
       "You've Got The Love - Florence + The Machine                             NaN   \n",
       "Pursuit Of Happiness (nightmare) - Kid Cudi / M...                       NaN   \n",
       "Billionaire [feat. Bruno Mars]  (Explicit Album...                      0.21   \n",
       "Revelry - Kings Of Leon                                                  NaN   \n",
       "Marry Me - Train                                                        0.25   \n",
       "Love Story - Taylor Swift                                               0.24   \n",
       "My Boy Builds Coffins - Florence + The Machine                           NaN   \n",
       "OMG - Usher featuring will.i.am                                         0.30   \n",
       "Drop The World - Lil Wayne / Eminem                                      NaN   \n",
       "Ain't Misbehavin - Sam Cooke                                             NaN   \n",
       "Use Somebody - Kings Of Leon                                            0.21   \n",
       "Cosmic Love - Florence + The Machine                                     NaN   \n",
       "\n",
       "                                                    Revelry - Kings Of Leon  \\\n",
       "Undo - Björk                                                           0.42   \n",
       "Dog Days Are Over (Radio Edit) - Florence + The...                      NaN   \n",
       "Lucky (Album Version) - Jason Mraz & Colbie Cai...                      NaN   \n",
       "You're The One - Dwight Yoakam                                         0.37   \n",
       "Bleed It Out [Live At Milton Keynes] - Linkin Park                      NaN   \n",
       "Somebody To Love - Justin Bieber                                        NaN   \n",
       "Hey_ Soul Sister - Train                                                NaN   \n",
       "Horn Concerto No. 4 in E flat K495: II. Romance...                     0.29   \n",
       "Sehr kosmisch - Harmonia                                                NaN   \n",
       "Vanilla Twilight - Owl City                                             NaN   \n",
       "Kiss With A Fist - Florence + The Machine                               NaN   \n",
       "I'm Not Calling You A Liar - Florence + The Mac...                      NaN   \n",
       "The Scientist - Coldplay                                                NaN   \n",
       "The Only Exception (Album Version) - Paramore                           NaN   \n",
       "Fireflies - Charttraxx Karaoke                                          NaN   \n",
       "All The Right Moves - OneRepublic                                       NaN   \n",
       "I Kissed A Girl - Katy Perry                                            NaN   \n",
       "Secrets - OneRepublic                                                   NaN   \n",
       "Between Two Lungs - Florence + The Machine                              NaN   \n",
       "Représente - Alliance Ethnik                                            NaN   \n",
       "Rabbit Heart (Raise It Up) - Florence + The Mac...                      NaN   \n",
       "You've Got The Love - Florence + The Machine                            NaN   \n",
       "Pursuit Of Happiness (nightmare) - Kid Cudi / M...                      NaN   \n",
       "Billionaire [feat. Bruno Mars]  (Explicit Album...                      NaN   \n",
       "Revelry - Kings Of Leon                                                 NaN   \n",
       "Marry Me - Train                                                        NaN   \n",
       "Love Story - Taylor Swift                                               NaN   \n",
       "My Boy Builds Coffins - Florence + The Machine                          NaN   \n",
       "OMG - Usher featuring will.i.am                                         NaN   \n",
       "Drop The World - Lil Wayne / Eminem                                     NaN   \n",
       "Ain't Misbehavin - Sam Cooke                                           0.27   \n",
       "Use Somebody - Kings Of Leon                                            NaN   \n",
       "Cosmic Love - Florence + The Machine                                    NaN   \n",
       "\n",
       "                                                    Dog Days Are Over (Radio Edit) - Florence + The Machine  \\\n",
       "Undo - Björk                                                                                      NaN         \n",
       "Dog Days Are Over (Radio Edit) - Florence + The...                                                NaN         \n",
       "Lucky (Album Version) - Jason Mraz & Colbie Cai...                                                NaN         \n",
       "You're The One - Dwight Yoakam                                                                    NaN         \n",
       "Bleed It Out [Live At Milton Keynes] - Linkin Park                                                NaN         \n",
       "Somebody To Love - Justin Bieber                                                                  NaN         \n",
       "Hey_ Soul Sister - Train                                                                         0.23         \n",
       "Horn Concerto No. 4 in E flat K495: II. Romance...                                                NaN         \n",
       "Sehr kosmisch - Harmonia                                                                         0.30         \n",
       "Vanilla Twilight - Owl City                                                                       NaN         \n",
       "Kiss With A Fist - Florence + The Machine                                                        0.33         \n",
       "I'm Not Calling You A Liar - Florence + The Mac...                                               0.30         \n",
       "The Scientist - Coldplay                                                                          NaN         \n",
       "The Only Exception (Album Version) - Paramore                                                    0.21         \n",
       "Fireflies - Charttraxx Karaoke                                                                   0.23         \n",
       "All The Right Moves - OneRepublic                                                                 NaN         \n",
       "I Kissed A Girl - Katy Perry                                                                      NaN         \n",
       "Secrets - OneRepublic                                                                            0.31         \n",
       "Between Two Lungs - Florence + The Machine                                                       0.22         \n",
       "Représente - Alliance Ethnik                                                                      NaN         \n",
       "Rabbit Heart (Raise It Up) - Florence + The Mac...                                               0.30         \n",
       "You've Got The Love - Florence + The Machine                                                     0.34         \n",
       "Pursuit Of Happiness (nightmare) - Kid Cudi / M...                                                NaN         \n",
       "Billionaire [feat. Bruno Mars]  (Explicit Album...                                                NaN         \n",
       "Revelry - Kings Of Leon                                                                           NaN         \n",
       "Marry Me - Train                                                                                 0.24         \n",
       "Love Story - Taylor Swift                                                                         NaN         \n",
       "My Boy Builds Coffins - Florence + The Machine                                                   0.27         \n",
       "OMG - Usher featuring will.i.am                                                                  0.20         \n",
       "Drop The World - Lil Wayne / Eminem                                                               NaN         \n",
       "Ain't Misbehavin - Sam Cooke                                                                      NaN         \n",
       "Use Somebody - Kings Of Leon                                                                     0.22         \n",
       "Cosmic Love - Florence + The Machine                                                             0.33         \n",
       "\n",
       "                                                    Horn Concerto No. 4 in E flat K495: II. Romance (Andante cantabile) - Barry Tuckwell/Academy of St Martin-in-the-Fields/Sir Neville Marriner  \\\n",
       "Undo - Björk                                                                                     0.28                                                                                              \n",
       "Dog Days Are Over (Radio Edit) - Florence + The...                                                NaN                                                                                              \n",
       "Lucky (Album Version) - Jason Mraz & Colbie Cai...                                                NaN                                                                                              \n",
       "You're The One - Dwight Yoakam                                                                   0.24                                                                                              \n",
       "Bleed It Out [Live At Milton Keynes] - Linkin Park                                                NaN                                                                                              \n",
       "Somebody To Love - Justin Bieber                                                                  NaN                                                                                              \n",
       "Hey_ Soul Sister - Train                                                                          NaN                                                                                              \n",
       "Horn Concerto No. 4 in E flat K495: II. Romance...                                                NaN                                                                                              \n",
       "Sehr kosmisch - Harmonia                                                                          NaN                                                                                              \n",
       "Vanilla Twilight - Owl City                                                                       NaN                                                                                              \n",
       "Kiss With A Fist - Florence + The Machine                                                         NaN                                                                                              \n",
       "I'm Not Calling You A Liar - Florence + The Mac...                                                NaN                                                                                              \n",
       "The Scientist - Coldplay                                                                          NaN                                                                                              \n",
       "The Only Exception (Album Version) - Paramore                                                     NaN                                                                                              \n",
       "Fireflies - Charttraxx Karaoke                                                                    NaN                                                                                              \n",
       "All The Right Moves - OneRepublic                                                                 NaN                                                                                              \n",
       "I Kissed A Girl - Katy Perry                                                                      NaN                                                                                              \n",
       "Secrets - OneRepublic                                                                             NaN                                                                                              \n",
       "Between Two Lungs - Florence + The Machine                                                        NaN                                                                                              \n",
       "Représente - Alliance Ethnik                                                                     0.24                                                                                              \n",
       "Rabbit Heart (Raise It Up) - Florence + The Mac...                                                NaN                                                                                              \n",
       "You've Got The Love - Florence + The Machine                                                      NaN                                                                                              \n",
       "Pursuit Of Happiness (nightmare) - Kid Cudi / M...                                                NaN                                                                                              \n",
       "Billionaire [feat. Bruno Mars]  (Explicit Album...                                                NaN                                                                                              \n",
       "Revelry - Kings Of Leon                                                                          0.29                                                                                              \n",
       "Marry Me - Train                                                                                  NaN                                                                                              \n",
       "Love Story - Taylor Swift                                                                         NaN                                                                                              \n",
       "My Boy Builds Coffins - Florence + The Machine                                                    NaN                                                                                              \n",
       "OMG - Usher featuring will.i.am                                                                   NaN                                                                                              \n",
       "Drop The World - Lil Wayne / Eminem                                                              0.24                                                                                              \n",
       "Ain't Misbehavin - Sam Cooke                                                                      NaN                                                                                              \n",
       "Use Somebody - Kings Of Leon                                                                      NaN                                                                                              \n",
       "Cosmic Love - Florence + The Machine                                                              NaN                                                                                              \n",
       "\n",
       "                                                    Secrets - OneRepublic  \\\n",
       "Undo - Björk                                                          NaN   \n",
       "Dog Days Are Over (Radio Edit) - Florence + The...                   0.31   \n",
       "Lucky (Album Version) - Jason Mraz & Colbie Cai...                   0.22   \n",
       "You're The One - Dwight Yoakam                                        NaN   \n",
       "Bleed It Out [Live At Milton Keynes] - Linkin Park                   0.21   \n",
       "Somebody To Love - Justin Bieber                                      NaN   \n",
       "Hey_ Soul Sister - Train                                             0.28   \n",
       "Horn Concerto No. 4 in E flat K495: II. Romance...                    NaN   \n",
       "Sehr kosmisch - Harmonia                                             0.36   \n",
       "Vanilla Twilight - Owl City                                           NaN   \n",
       "Kiss With A Fist - Florence + The Machine                             NaN   \n",
       "I'm Not Calling You A Liar - Florence + The Mac...                    NaN   \n",
       "The Scientist - Coldplay                                             0.21   \n",
       "The Only Exception (Album Version) - Paramore                        0.24   \n",
       "Fireflies - Charttraxx Karaoke                                       0.31   \n",
       "All The Right Moves - OneRepublic                                    0.29   \n",
       "I Kissed A Girl - Katy Perry                                          NaN   \n",
       "Secrets - OneRepublic                                                 NaN   \n",
       "Between Two Lungs - Florence + The Machine                            NaN   \n",
       "Représente - Alliance Ethnik                                          NaN   \n",
       "Rabbit Heart (Raise It Up) - Florence + The Mac...                    NaN   \n",
       "You've Got The Love - Florence + The Machine                          NaN   \n",
       "Pursuit Of Happiness (nightmare) - Kid Cudi / M...                   0.20   \n",
       "Billionaire [feat. Bruno Mars]  (Explicit Album...                    NaN   \n",
       "Revelry - Kings Of Leon                                               NaN   \n",
       "Marry Me - Train                                                     0.28   \n",
       "Love Story - Taylor Swift                                            0.23   \n",
       "My Boy Builds Coffins - Florence + The Machine                        NaN   \n",
       "OMG - Usher featuring will.i.am                                      0.28   \n",
       "Drop The World - Lil Wayne / Eminem                                  0.22   \n",
       "Ain't Misbehavin - Sam Cooke                                          NaN   \n",
       "Use Somebody - Kings Of Leon                                         0.25   \n",
       "Cosmic Love - Florence + The Machine                                  NaN   \n",
       "\n",
       "                                                    Hey_ Soul Sister - Train  \\\n",
       "Undo - Björk                                                             NaN   \n",
       "Dog Days Are Over (Radio Edit) - Florence + The...                      0.23   \n",
       "Lucky (Album Version) - Jason Mraz & Colbie Cai...                      0.23   \n",
       "You're The One - Dwight Yoakam                                           NaN   \n",
       "Bleed It Out [Live At Milton Keynes] - Linkin Park                       NaN   \n",
       "Somebody To Love - Justin Bieber                                         NaN   \n",
       "Hey_ Soul Sister - Train                                                 NaN   \n",
       "Horn Concerto No. 4 in E flat K495: II. Romance...                       NaN   \n",
       "Sehr kosmisch - Harmonia                                                0.30   \n",
       "Vanilla Twilight - Owl City                                              NaN   \n",
       "Kiss With A Fist - Florence + The Machine                                NaN   \n",
       "I'm Not Calling You A Liar - Florence + The Mac...                       NaN   \n",
       "The Scientist - Coldplay                                                0.20   \n",
       "The Only Exception (Album Version) - Paramore                           0.20   \n",
       "Fireflies - Charttraxx Karaoke                                          0.32   \n",
       "All The Right Moves - OneRepublic                                        NaN   \n",
       "I Kissed A Girl - Katy Perry                                             NaN   \n",
       "Secrets - OneRepublic                                                   0.28   \n",
       "Between Two Lungs - Florence + The Machine                               NaN   \n",
       "Représente - Alliance Ethnik                                             NaN   \n",
       "Rabbit Heart (Raise It Up) - Florence + The Mac...                       NaN   \n",
       "You've Got The Love - Florence + The Machine                             NaN   \n",
       "Pursuit Of Happiness (nightmare) - Kid Cudi / M...                       NaN   \n",
       "Billionaire [feat. Bruno Mars]  (Explicit Album...                      0.25   \n",
       "Revelry - Kings Of Leon                                                  NaN   \n",
       "Marry Me - Train                                                        0.28   \n",
       "Love Story - Taylor Swift                                               0.21   \n",
       "My Boy Builds Coffins - Florence + The Machine                           NaN   \n",
       "OMG - Usher featuring will.i.am                                         0.27   \n",
       "Drop The World - Lil Wayne / Eminem                                      NaN   \n",
       "Ain't Misbehavin - Sam Cooke                                             NaN   \n",
       "Use Somebody - Kings Of Leon                                            0.25   \n",
       "Cosmic Love - Florence + The Machine                                     NaN   \n",
       "\n",
       "                                                    Canada - Five Iron Frenzy  \\\n",
       "Undo - Björk                                                              NaN   \n",
       "Dog Days Are Over (Radio Edit) - Florence + The...                        NaN   \n",
       "Lucky (Album Version) - Jason Mraz & Colbie Cai...                        NaN   \n",
       "You're The One - Dwight Yoakam                                            NaN   \n",
       "Bleed It Out [Live At Milton Keynes] - Linkin Park                        NaN   \n",
       "Somebody To Love - Justin Bieber                                          NaN   \n",
       "Hey_ Soul Sister - Train                                                  NaN   \n",
       "Horn Concerto No. 4 in E flat K495: II. Romance...                        NaN   \n",
       "Sehr kosmisch - Harmonia                                                  NaN   \n",
       "Vanilla Twilight - Owl City                                               NaN   \n",
       "Kiss With A Fist - Florence + The Machine                                 NaN   \n",
       "I'm Not Calling You A Liar - Florence + The Mac...                        NaN   \n",
       "The Scientist - Coldplay                                                  NaN   \n",
       "The Only Exception (Album Version) - Paramore                             NaN   \n",
       "Fireflies - Charttraxx Karaoke                                            NaN   \n",
       "All The Right Moves - OneRepublic                                         NaN   \n",
       "I Kissed A Girl - Katy Perry                                              NaN   \n",
       "Secrets - OneRepublic                                                     NaN   \n",
       "Between Two Lungs - Florence + The Machine                                NaN   \n",
       "Représente - Alliance Ethnik                                              NaN   \n",
       "Rabbit Heart (Raise It Up) - Florence + The Mac...                        NaN   \n",
       "You've Got The Love - Florence + The Machine                              NaN   \n",
       "Pursuit Of Happiness (nightmare) - Kid Cudi / M...                        NaN   \n",
       "Billionaire [feat. Bruno Mars]  (Explicit Album...                        NaN   \n",
       "Revelry - Kings Of Leon                                                   NaN   \n",
       "Marry Me - Train                                                          NaN   \n",
       "Love Story - Taylor Swift                                                 NaN   \n",
       "My Boy Builds Coffins - Florence + The Machine                            NaN   \n",
       "OMG - Usher featuring will.i.am                                           NaN   \n",
       "Drop The World - Lil Wayne / Eminem                                       NaN   \n",
       "Ain't Misbehavin - Sam Cooke                                              NaN   \n",
       "Use Somebody - Kings Of Leon                                              NaN   \n",
       "Cosmic Love - Florence + The Machine                                      NaN   \n",
       "\n",
       "                                                    Fireflies - Charttraxx Karaoke  \n",
       "Undo - Björk                                                                   NaN  \n",
       "Dog Days Are Over (Radio Edit) - Florence + The...                            0.23  \n",
       "Lucky (Album Version) - Jason Mraz & Colbie Cai...                            0.21  \n",
       "You're The One - Dwight Yoakam                                                 NaN  \n",
       "Bleed It Out [Live At Milton Keynes] - Linkin Park                            0.21  \n",
       "Somebody To Love - Justin Bieber                                               NaN  \n",
       "Hey_ Soul Sister - Train                                                      0.32  \n",
       "Horn Concerto No. 4 in E flat K495: II. Romance...                             NaN  \n",
       "Sehr kosmisch - Harmonia                                                      0.32  \n",
       "Vanilla Twilight - Owl City                                                   0.27  \n",
       "Kiss With A Fist - Florence + The Machine                                      NaN  \n",
       "I'm Not Calling You A Liar - Florence + The Mac...                             NaN  \n",
       "The Scientist - Coldplay                                                       NaN  \n",
       "The Only Exception (Album Version) - Paramore                                 0.21  \n",
       "Fireflies - Charttraxx Karaoke                                                 NaN  \n",
       "All The Right Moves - OneRepublic                                              NaN  \n",
       "I Kissed A Girl - Katy Perry                                                   NaN  \n",
       "Secrets - OneRepublic                                                         0.31  \n",
       "Between Two Lungs - Florence + The Machine                                     NaN  \n",
       "Représente - Alliance Ethnik                                                   NaN  \n",
       "Rabbit Heart (Raise It Up) - Florence + The Mac...                             NaN  \n",
       "You've Got The Love - Florence + The Machine                                   NaN  \n",
       "Pursuit Of Happiness (nightmare) - Kid Cudi / M...                             NaN  \n",
       "Billionaire [feat. Bruno Mars]  (Explicit Album...                            0.21  \n",
       "Revelry - Kings Of Leon                                                        NaN  \n",
       "Marry Me - Train                                                              0.26  \n",
       "Love Story - Taylor Swift                                                     0.23  \n",
       "My Boy Builds Coffins - Florence + The Machine                                 NaN  \n",
       "OMG - Usher featuring will.i.am                                               0.27  \n",
       "Drop The World - Lil Wayne / Eminem                                           0.20  \n",
       "Ain't Misbehavin - Sam Cooke                                                   NaN  \n",
       "Use Somebody - Kings Of Leon                                                  0.25  \n",
       "Cosmic Love - Florence + The Machine                                           NaN  "
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Filter for only values greater that 0.1 and less than 1, (1.0 will certainly be a comparison with itself)\n",
    "similarity_enough = item_similarity_df_10[(item_similarity_df_10 < 0.999) & (item_similarity_df_10> 0.2)]\n",
    "# Drop rows where every value is NaN (only 1 or less than 0.1)\n",
    "similarity_enough = similarity_enough.dropna(how = 'all')\n",
    "similarity_enough"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x23b88381c08>"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualise songs with similarity over 0.2 with one of the top 10 using a heatmap\n",
    "df_custom = similarity_enough.copy()\n",
    "# Rename the hugely long title with a shorter version\n",
    "df_custom.rename(columns={'Horn Concerto No. 4 in E flat K495: II. Romance (Andante cantabile) - Barry Tuckwell/Academy of St Martin-in-the-Fields/Sir Neville Marriner': 'Horn Concerto no4 in e flat:Barry Tuckwell etc'}, inplace=True)\n",
    "df_custom.rename(index={'Horn Concerto No. 4 in E flat K495: II. Romance (Andante cantabile) - Barry Tuckwell/Academy of St Martin-in-the-Fields/Sir Neville Marriner': 'Horn Concerto no4 in e flat:Barry Tuckwell etc'}, inplace=True)\n",
    "plt.figure(figsize=(15, 10))\n",
    "sns.heatmap(data = df_custom, cmap = 'viridis')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 2. User-User Cosine similarity analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "76328"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# See number of different users\n",
    "triplets_df['user_id'].nunique()\n",
    "# This is too large to form a dataframe of every interaction, there would be 76,353^2 floats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "# However many of these users are not worth including \n",
    "# A lot of them have only a few listens, its not worth drawing conclusions or similarities between these users."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "user_id\n",
       "00003a4459f33b92906be11abe0e93efc423c0ff    17\n",
       "00005c6177188f12fb5e2e82cdbd93e8a3f35e64     6\n",
       "00030033e3a2f904a48ec1dd53019c9969b6ef1f    24\n",
       "0007235c769e610e3d339a17818a5708e41008d9    30\n",
       "0007c0e74728ca9ef0fe4eb7f75732e8026a278b    10\n",
       "Name: listen_count, dtype: int64"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# See how many listens each user has\n",
    "user_grouped = triplets_df.groupby('user_id')['listen_count'].sum()\n",
    "user_grouped.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count   76328.00\n",
       "mean       65.76\n",
       "std        85.60\n",
       "min         1.00\n",
       "25%        17.00\n",
       "50%        37.00\n",
       "75%        80.00\n",
       "max      2334.00\n",
       "Name: listen_count, dtype: float64"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Analyse describe stastics to see averages and quartile\n",
    "user_grouped.describe()\n",
    "\n",
    "# We can see that the average is 65 listens, while 50% have less than 37. \n",
    "# We can't really draw similarities very accuaractly from people with a very low listen count\n",
    "# If someone only has 1 listen of a song and another happens to have listened to only that song once, they will have perfect cosine similarity\n",
    "# In reality they might not have similar preferences at all. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Filter for users that are active enough such that they have at least the median number of listens\n",
    "active_users = user_grouped[user_grouped >= 37].index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get sample of 5,000 from all users that meet this criteria\n",
    "n_samples = 5000 \n",
    "sampled_users = np.random.choice(active_users, n_samples, replace=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Ollie\\anaconda3\\lib\\site-packages\\ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "# Convert the user IDs in the triplets_df to categorical variables\n",
    "triplets_df['user_id'] = triplets_df['user_id'].astype('category')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Map the sampled user IDs to their categorical codes (indices in the sparse matrix)\n",
    "sampled_user_indices = triplets_df['user_id'].cat.codes[triplets_df['user_id'].isin(sampled_users)].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Filter original sparse_matrix to get only sampled users of over 41 listens\n",
    "sampled_user_item_sparse_matrix = sparse_matrix[sampled_user_indices, :]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get cosine similarity using now filtered sparse_matrix\n",
    "sampled_user_user_similarity = cosine_similarity(sampled_user_item_sparse_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "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>969cc6fb74e076a68e36a04409cb9d3765757508</th>\n",
       "      <th>4bd88bfb25263a75bbdd467e74018f4ae570e5df</th>\n",
       "      <th>b64cdd1a0bd907e5e00b39e345194768e330d652</th>\n",
       "      <th>a820d2d4f16bbd53be9e41e0417dfb234bfdfba8</th>\n",
       "      <th>bd8475385f0aa78830fa6dfce9e7242164b035c8</th>\n",
       "      <th>2c42e6551311710ca5a839d62058820a42ead493</th>\n",
       "      <th>3f9ed694a79835c921ef6d94acd28f876c1d901e</th>\n",
       "      <th>b4e32cdb654ab914a0ac73d2bbc5cae142da405a</th>\n",
       "      <th>ea07020bb223c733ccc55aa925ebcc25c4d97377</th>\n",
       "      <th>95942345306393998eb3a051ae5fa3c4d5afbaa4</th>\n",
       "      <th>...</th>\n",
       "      <th>3fe95fa31600137672e0ede4a8277c3ac62d96b6</th>\n",
       "      <th>fcdfda9294154d265211750d8c61cd494bbecab8</th>\n",
       "      <th>6a76f7237d1e4988bf23ed184c10b7a9d2f96974</th>\n",
       "      <th>888411ede2dc6306b05ba9f75f4edf26b3be57eb</th>\n",
       "      <th>a4f10e6da623ce3bf470fedff15b9b2d0032f72a</th>\n",
       "      <th>9fbce142d6707dae18898b16ffb8c41f6f80a7f6</th>\n",
       "      <th>613148566a2fd83e881ca467dd4caee075cf6141</th>\n",
       "      <th>6af523b1dff77e4269f5f575698c539966665060</th>\n",
       "      <th>f8dd34d31f233ea3837420220c55cf0acdad6382</th>\n",
       "      <th>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>969cc6fb74e076a68e36a04409cb9d3765757508</th>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4bd88bfb25263a75bbdd467e74018f4ae570e5df</th>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b64cdd1a0bd907e5e00b39e345194768e330d652</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a820d2d4f16bbd53be9e41e0417dfb234bfdfba8</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bd8475385f0aa78830fa6dfce9e7242164b035c8</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.02</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.14</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.13</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>9fbce142d6707dae18898b16ffb8c41f6f80a7f6</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.14</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>613148566a2fd83e881ca467dd4caee075cf6141</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6af523b1dff77e4269f5f575698c539966665060</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f8dd34d31f233ea3837420220c55cf0acdad6382</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>1.00</td>\n",
       "      <td>0.16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.13</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.16</td>\n",
       "      <td>1.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5000 rows × 5000 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          969cc6fb74e076a68e36a04409cb9d3765757508  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      1.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.00   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          4bd88bfb25263a75bbdd467e74018f4ae570e5df  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      1.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.01   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.05   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.05   \n",
       "\n",
       "                                          b64cdd1a0bd907e5e00b39e345194768e330d652  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.01   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      1.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.01   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          a820d2d4f16bbd53be9e41e0417dfb234bfdfba8  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      1.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.00   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          bd8475385f0aa78830fa6dfce9e7242164b035c8  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.05   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.01   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      1.00   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.14   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.02   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.08   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.04   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.13   \n",
       "\n",
       "                                          2c42e6551311710ca5a839d62058820a42ead493  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.00   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          3f9ed694a79835c921ef6d94acd28f876c1d901e  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.02   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.06   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.10   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.03   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.07   \n",
       "\n",
       "                                          b4e32cdb654ab914a0ac73d2bbc5cae142da405a  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.01   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.05   \n",
       "\n",
       "                                          ea07020bb223c733ccc55aa925ebcc25c4d97377  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.02   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.07   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.03   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.08   \n",
       "\n",
       "                                          95942345306393998eb3a051ae5fa3c4d5afbaa4  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.02   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          ...  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508  ...   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df  ...   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652  ...   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8  ...   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8  ...   \n",
       "...                                       ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6  ...   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141  ...   \n",
       "6af523b1dff77e4269f5f575698c539966665060  ...   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382  ...   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  ...   \n",
       "\n",
       "                                          3fe95fa31600137672e0ede4a8277c3ac62d96b6  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.00   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          fcdfda9294154d265211750d8c61cd494bbecab8  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.00   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          6a76f7237d1e4988bf23ed184c10b7a9d2f96974  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.07   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.01   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.07   \n",
       "\n",
       "                                          888411ede2dc6306b05ba9f75f4edf26b3be57eb  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.02   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.03   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.04   \n",
       "\n",
       "                                          a4f10e6da623ce3bf470fedff15b9b2d0032f72a  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.02   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.00   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.01   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          9fbce142d6707dae18898b16ffb8c41f6f80a7f6  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.14   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      1.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.06   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.07   \n",
       "\n",
       "                                          613148566a2fd83e881ca467dd4caee075cf6141  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.02   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      1.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.04   \n",
       "\n",
       "                                          6af523b1dff77e4269f5f575698c539966665060  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.08   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      1.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.01   \n",
       "\n",
       "                                          f8dd34d31f233ea3837420220c55cf0acdad6382  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.04   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.06   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      1.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.16   \n",
       "\n",
       "                                          d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  \n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00  \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.05  \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00  \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00  \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.13  \n",
       "...                                                                            ...  \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.07  \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.04  \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.01  \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.16  \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      1.00  \n",
       "\n",
       "[5000 rows x 5000 columns]"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert to a DataFrame for easier handling\n",
    "sampled_user_ids = triplets_df['user_id'].cat.categories[sampled_user_indices]\n",
    "sampled_user_user_similarity_df = pd.DataFrame(sampled_user_user_similarity, index=sampled_user_ids, columns=sampled_user_ids)\n",
    "sampled_user_user_similarity_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "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>969cc6fb74e076a68e36a04409cb9d3765757508</th>\n",
       "      <th>4bd88bfb25263a75bbdd467e74018f4ae570e5df</th>\n",
       "      <th>b64cdd1a0bd907e5e00b39e345194768e330d652</th>\n",
       "      <th>a820d2d4f16bbd53be9e41e0417dfb234bfdfba8</th>\n",
       "      <th>bd8475385f0aa78830fa6dfce9e7242164b035c8</th>\n",
       "      <th>2c42e6551311710ca5a839d62058820a42ead493</th>\n",
       "      <th>3f9ed694a79835c921ef6d94acd28f876c1d901e</th>\n",
       "      <th>b4e32cdb654ab914a0ac73d2bbc5cae142da405a</th>\n",
       "      <th>ea07020bb223c733ccc55aa925ebcc25c4d97377</th>\n",
       "      <th>95942345306393998eb3a051ae5fa3c4d5afbaa4</th>\n",
       "      <th>...</th>\n",
       "      <th>3fe95fa31600137672e0ede4a8277c3ac62d96b6</th>\n",
       "      <th>fcdfda9294154d265211750d8c61cd494bbecab8</th>\n",
       "      <th>6a76f7237d1e4988bf23ed184c10b7a9d2f96974</th>\n",
       "      <th>888411ede2dc6306b05ba9f75f4edf26b3be57eb</th>\n",
       "      <th>a4f10e6da623ce3bf470fedff15b9b2d0032f72a</th>\n",
       "      <th>9fbce142d6707dae18898b16ffb8c41f6f80a7f6</th>\n",
       "      <th>613148566a2fd83e881ca467dd4caee075cf6141</th>\n",
       "      <th>6af523b1dff77e4269f5f575698c539966665060</th>\n",
       "      <th>f8dd34d31f233ea3837420220c55cf0acdad6382</th>\n",
       "      <th>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>...</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "      <td>4999.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.02</td>\n",
       "      <td>...</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>0.38</td>\n",
       "      <td>0.29</td>\n",
       "      <td>0.39</td>\n",
       "      <td>0.65</td>\n",
       "      <td>0.31</td>\n",
       "      <td>0.38</td>\n",
       "      <td>0.32</td>\n",
       "      <td>0.44</td>\n",
       "      <td>0.44</td>\n",
       "      <td>0.35</td>\n",
       "      <td>...</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0.26</td>\n",
       "      <td>0.46</td>\n",
       "      <td>0.42</td>\n",
       "      <td>0.18</td>\n",
       "      <td>0.58</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0.53</td>\n",
       "      <td>0.44</td>\n",
       "      <td>0.42</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>8 rows × 5000 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       969cc6fb74e076a68e36a04409cb9d3765757508  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.00   \n",
       "std                                        0.02   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.00   \n",
       "max                                        0.38   \n",
       "\n",
       "       4bd88bfb25263a75bbdd467e74018f4ae570e5df  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.01   \n",
       "std                                        0.02   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.00   \n",
       "max                                        0.29   \n",
       "\n",
       "       b64cdd1a0bd907e5e00b39e345194768e330d652  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.01   \n",
       "std                                        0.02   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.00   \n",
       "max                                        0.39   \n",
       "\n",
       "       a820d2d4f16bbd53be9e41e0417dfb234bfdfba8  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.00   \n",
       "std                                        0.01   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.00   \n",
       "max                                        0.65   \n",
       "\n",
       "       bd8475385f0aa78830fa6dfce9e7242164b035c8  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.02   \n",
       "std                                        0.04   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.01   \n",
       "75%                                        0.03   \n",
       "max                                        0.31   \n",
       "\n",
       "       2c42e6551311710ca5a839d62058820a42ead493  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.00   \n",
       "std                                        0.01   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.00   \n",
       "max                                        0.38   \n",
       "\n",
       "       3f9ed694a79835c921ef6d94acd28f876c1d901e  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.01   \n",
       "std                                        0.03   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.01   \n",
       "max                                        0.32   \n",
       "\n",
       "       b4e32cdb654ab914a0ac73d2bbc5cae142da405a  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.01   \n",
       "std                                        0.03   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.00   \n",
       "max                                        0.44   \n",
       "\n",
       "       ea07020bb223c733ccc55aa925ebcc25c4d97377  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.01   \n",
       "std                                        0.03   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.01   \n",
       "max                                        0.44   \n",
       "\n",
       "       95942345306393998eb3a051ae5fa3c4d5afbaa4  ...  \\\n",
       "count                                   4999.00  ...   \n",
       "mean                                       0.00  ...   \n",
       "std                                        0.02  ...   \n",
       "min                                        0.00  ...   \n",
       "25%                                        0.00  ...   \n",
       "50%                                        0.00  ...   \n",
       "75%                                        0.00  ...   \n",
       "max                                        0.35  ...   \n",
       "\n",
       "       3fe95fa31600137672e0ede4a8277c3ac62d96b6  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.00   \n",
       "std                                        0.01   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.00   \n",
       "max                                        0.40   \n",
       "\n",
       "       fcdfda9294154d265211750d8c61cd494bbecab8  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.01   \n",
       "std                                        0.02   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.00   \n",
       "max                                        0.26   \n",
       "\n",
       "       6a76f7237d1e4988bf23ed184c10b7a9d2f96974  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.02   \n",
       "std                                        0.04   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.01   \n",
       "max                                        0.46   \n",
       "\n",
       "       888411ede2dc6306b05ba9f75f4edf26b3be57eb  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.01   \n",
       "std                                        0.02   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.01   \n",
       "max                                        0.42   \n",
       "\n",
       "       a4f10e6da623ce3bf470fedff15b9b2d0032f72a  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.00   \n",
       "std                                        0.01   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.00   \n",
       "max                                        0.18   \n",
       "\n",
       "       9fbce142d6707dae18898b16ffb8c41f6f80a7f6  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.01   \n",
       "std                                        0.03   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.00   \n",
       "max                                        0.58   \n",
       "\n",
       "       613148566a2fd83e881ca467dd4caee075cf6141  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.01   \n",
       "std                                        0.01   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.01   \n",
       "max                                        0.40   \n",
       "\n",
       "       6af523b1dff77e4269f5f575698c539966665060  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.01   \n",
       "std                                        0.03   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.00   \n",
       "max                                        0.53   \n",
       "\n",
       "       f8dd34d31f233ea3837420220c55cf0acdad6382  \\\n",
       "count                                   4999.00   \n",
       "mean                                       0.02   \n",
       "std                                        0.05   \n",
       "min                                        0.00   \n",
       "25%                                        0.00   \n",
       "50%                                        0.00   \n",
       "75%                                        0.02   \n",
       "max                                        0.44   \n",
       "\n",
       "       d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92  \n",
       "count                                   4999.00  \n",
       "mean                                       0.03  \n",
       "std                                        0.06  \n",
       "min                                        0.00  \n",
       "25%                                        0.00  \n",
       "50%                                        0.00  \n",
       "75%                                        0.04  \n",
       "max                                        0.42  \n",
       "\n",
       "[8 rows x 5000 columns]"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Filter out self-comparisons\n",
    "sampled_user_user_similarity_df = sampled_user_user_similarity_df[sampled_user_user_similarity_df  <= 0.9999]\n",
    "sampled_user_user_similarity_df.describe()\n",
    "# We can see from .describe() that for some user's similarity is over 0.7, in some cases over 0.9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Take a random sample of 10 users and analyse similarity with all other users"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "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>651470cd4ef0fe6e2f430d166580405146f1d44f</th>\n",
       "      <th>7d4ae3106619dc7cb642e29dd3c9dd265a9150ca</th>\n",
       "      <th>2724239a7fc34793d2949824ecfb19904e866ae9</th>\n",
       "      <th>cd0cbb0e243f9f6ef66e57c5f82e758a7dd775c1</th>\n",
       "      <th>3a1148968bac3d3ec90fb3be0e4d376c0155a986</th>\n",
       "      <th>4446d24372beca4092c5444350f3bcaddcadd583</th>\n",
       "      <th>59d92d7581641055164bd571ff3ee46fbee24001</th>\n",
       "      <th>0ff371a0c21bd86cc9f56dce11b3a46619263af4</th>\n",
       "      <th>59bad0aea569a12d785e401d185f8814d128333e</th>\n",
       "      <th>78b67248d2522d41923d28f831ce02c669838926</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>969cc6fb74e076a68e36a04409cb9d3765757508</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4bd88bfb25263a75bbdd467e74018f4ae570e5df</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b64cdd1a0bd907e5e00b39e345194768e330d652</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.04</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>a820d2d4f16bbd53be9e41e0417dfb234bfdfba8</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>bd8475385f0aa78830fa6dfce9e7242164b035c8</th>\n",
       "      <td>0.03</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9fbce142d6707dae18898b16ffb8c41f6f80a7f6</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>613148566a2fd83e881ca467dd4caee075cf6141</th>\n",
       "      <td>0.03</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6af523b1dff77e4269f5f575698c539966665060</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.03</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>f8dd34d31f233ea3837420220c55cf0acdad6382</th>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92</th>\n",
       "      <td>0.16</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5000 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          651470cd4ef0fe6e2f430d166580405146f1d44f  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.03   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.03   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.16   \n",
       "\n",
       "                                          7d4ae3106619dc7cb642e29dd3c9dd265a9150ca  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.00   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          2724239a7fc34793d2949824ecfb19904e866ae9  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.01   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.02   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.03   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.02   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.02   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.11   \n",
       "\n",
       "                                          cd0cbb0e243f9f6ef66e57c5f82e758a7dd775c1  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.01   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.01   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          3a1148968bac3d3ec90fb3be0e4d376c0155a986  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.00   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          4446d24372beca4092c5444350f3bcaddcadd583  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.00   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.03   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          59d92d7581641055164bd571ff3ee46fbee24001  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.00   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          0ff371a0c21bd86cc9f56dce11b3a46619263af4  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.06   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.04   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.06   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.06   \n",
       "\n",
       "                                          59bad0aea569a12d785e401d185f8814d128333e  \\\n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00   \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00   \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00   \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00   \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.00   \n",
       "...                                                                            ...   \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00   \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00   \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00   \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00   \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00   \n",
       "\n",
       "                                          78b67248d2522d41923d28f831ce02c669838926  \n",
       "969cc6fb74e076a68e36a04409cb9d3765757508                                      0.00  \n",
       "4bd88bfb25263a75bbdd467e74018f4ae570e5df                                      0.00  \n",
       "b64cdd1a0bd907e5e00b39e345194768e330d652                                      0.00  \n",
       "a820d2d4f16bbd53be9e41e0417dfb234bfdfba8                                      0.00  \n",
       "bd8475385f0aa78830fa6dfce9e7242164b035c8                                      0.01  \n",
       "...                                                                            ...  \n",
       "9fbce142d6707dae18898b16ffb8c41f6f80a7f6                                      0.00  \n",
       "613148566a2fd83e881ca467dd4caee075cf6141                                      0.00  \n",
       "6af523b1dff77e4269f5f575698c539966665060                                      0.00  \n",
       "f8dd34d31f233ea3837420220c55cf0acdad6382                                      0.00  \n",
       "d8bfd4ec88f0f3773a9e022e3c1a0f1d3b7b6a92                                      0.00  \n",
       "\n",
       "[5000 rows x 10 columns]"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Get all users index\n",
    "all_users = sampled_user_user_similarity_df.index\n",
    "\n",
    "# Sample 10 random users for columns\n",
    "sampled_10 = np.random.choice(all_users, size=10, replace=False)\n",
    "\n",
    "# Create a filtered matrix with all users as rows and 10 sampled users as columns\n",
    "sample_10_matrix = sampled_user_user_similarity_df.loc[:, sampled_10]\n",
    "sample_10_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "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>651470cd4ef0fe6e2f430d166580405146f1d44f</th>\n",
       "      <th>7d4ae3106619dc7cb642e29dd3c9dd265a9150ca</th>\n",
       "      <th>2724239a7fc34793d2949824ecfb19904e866ae9</th>\n",
       "      <th>3a1148968bac3d3ec90fb3be0e4d376c0155a986</th>\n",
       "      <th>4446d24372beca4092c5444350f3bcaddcadd583</th>\n",
       "      <th>59bad0aea569a12d785e401d185f8814d128333e</th>\n",
       "      <th>78b67248d2522d41923d28f831ce02c669838926</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2c42e6551311710ca5a839d62058820a42ead493</th>\n",
       "      <td>NaN</td>\n",
       "      <td>0.38</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3d766c6bd78be4598a6c8946640097d929b0b09f</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.42</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52843949671b246e56ec8842cadb599e32d998ba</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.39</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2ecca069111af5716fa0be6cfdfb5e2f500cdab2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.39</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0cc972c46fa7bfe81500fc63a9186157c10712a4</th>\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>0.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>dbb98c22429724137982c79e49da372d859dcfe3</th>\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>0.31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3bd82072d48527e803d8c092a5efb9d582195e48</th>\n",
       "      <td>0.34</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7cd6f87c6cf9d06464814964adb7ae2839bdce49</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.31</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6bcace58344f65ea8611afb01287e85a8e18827b</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.33</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8dca390a049834fecd2511f489fbd7884d8cba0e</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.35</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b93e873853c4b0acb3cf11514e46eff902a76c14</th>\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>0.41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2df8db584e831559491197a21c82184ed93832a3</th>\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>0.43</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>fa0ac4daf9628b0fdf874c71f50e9619fdfcc5d7</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.35</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>e80582287da1f1a66282ba80b23df4d73979ac17</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.44</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>554e5cf51f2fc724ae47e1cac1ebab0f58de28db</th>\n",
       "      <td>0.33</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6cf576c6838b15c80a38287e880a48dd731aba99</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.35</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4bb3bf56ffeb05bf5bd49ca062e1d024f7452f3a</th>\n",
       "      <td>0.34</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>b5e27ce5e72536c3637d98f984f624ca8896a6f5</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.38</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                          651470cd4ef0fe6e2f430d166580405146f1d44f  \\\n",
       "2c42e6551311710ca5a839d62058820a42ead493                                       NaN   \n",
       "3d766c6bd78be4598a6c8946640097d929b0b09f                                       NaN   \n",
       "52843949671b246e56ec8842cadb599e32d998ba                                       NaN   \n",
       "2ecca069111af5716fa0be6cfdfb5e2f500cdab2                                       NaN   \n",
       "0cc972c46fa7bfe81500fc63a9186157c10712a4                                       NaN   \n",
       "dbb98c22429724137982c79e49da372d859dcfe3                                       NaN   \n",
       "3bd82072d48527e803d8c092a5efb9d582195e48                                      0.34   \n",
       "7cd6f87c6cf9d06464814964adb7ae2839bdce49                                       NaN   \n",
       "6bcace58344f65ea8611afb01287e85a8e18827b                                       NaN   \n",
       "8dca390a049834fecd2511f489fbd7884d8cba0e                                       NaN   \n",
       "b93e873853c4b0acb3cf11514e46eff902a76c14                                       NaN   \n",
       "2df8db584e831559491197a21c82184ed93832a3                                       NaN   \n",
       "fa0ac4daf9628b0fdf874c71f50e9619fdfcc5d7                                       NaN   \n",
       "e80582287da1f1a66282ba80b23df4d73979ac17                                       NaN   \n",
       "554e5cf51f2fc724ae47e1cac1ebab0f58de28db                                      0.33   \n",
       "6cf576c6838b15c80a38287e880a48dd731aba99                                       NaN   \n",
       "4bb3bf56ffeb05bf5bd49ca062e1d024f7452f3a                                      0.34   \n",
       "b5e27ce5e72536c3637d98f984f624ca8896a6f5                                       NaN   \n",
       "\n",
       "                                          7d4ae3106619dc7cb642e29dd3c9dd265a9150ca  \\\n",
       "2c42e6551311710ca5a839d62058820a42ead493                                      0.38   \n",
       "3d766c6bd78be4598a6c8946640097d929b0b09f                                       NaN   \n",
       "52843949671b246e56ec8842cadb599e32d998ba                                       NaN   \n",
       "2ecca069111af5716fa0be6cfdfb5e2f500cdab2                                       NaN   \n",
       "0cc972c46fa7bfe81500fc63a9186157c10712a4                                       NaN   \n",
       "dbb98c22429724137982c79e49da372d859dcfe3                                       NaN   \n",
       "3bd82072d48527e803d8c092a5efb9d582195e48                                       NaN   \n",
       "7cd6f87c6cf9d06464814964adb7ae2839bdce49                                       NaN   \n",
       "6bcace58344f65ea8611afb01287e85a8e18827b                                       NaN   \n",
       "8dca390a049834fecd2511f489fbd7884d8cba0e                                       NaN   \n",
       "b93e873853c4b0acb3cf11514e46eff902a76c14                                       NaN   \n",
       "2df8db584e831559491197a21c82184ed93832a3                                       NaN   \n",
       "fa0ac4daf9628b0fdf874c71f50e9619fdfcc5d7                                       NaN   \n",
       "e80582287da1f1a66282ba80b23df4d73979ac17                                       NaN   \n",
       "554e5cf51f2fc724ae47e1cac1ebab0f58de28db                                       NaN   \n",
       "6cf576c6838b15c80a38287e880a48dd731aba99                                       NaN   \n",
       "4bb3bf56ffeb05bf5bd49ca062e1d024f7452f3a                                       NaN   \n",
       "b5e27ce5e72536c3637d98f984f624ca8896a6f5                                       NaN   \n",
       "\n",
       "                                          2724239a7fc34793d2949824ecfb19904e866ae9  \\\n",
       "2c42e6551311710ca5a839d62058820a42ead493                                       NaN   \n",
       "3d766c6bd78be4598a6c8946640097d929b0b09f                                       NaN   \n",
       "52843949671b246e56ec8842cadb599e32d998ba                                       NaN   \n",
       "2ecca069111af5716fa0be6cfdfb5e2f500cdab2                                       NaN   \n",
       "0cc972c46fa7bfe81500fc63a9186157c10712a4                                       NaN   \n",
       "dbb98c22429724137982c79e49da372d859dcfe3                                       NaN   \n",
       "3bd82072d48527e803d8c092a5efb9d582195e48                                       NaN   \n",
       "7cd6f87c6cf9d06464814964adb7ae2839bdce49                                       NaN   \n",
       "6bcace58344f65ea8611afb01287e85a8e18827b                                      0.33   \n",
       "8dca390a049834fecd2511f489fbd7884d8cba0e                                       NaN   \n",
       "b93e873853c4b0acb3cf11514e46eff902a76c14                                       NaN   \n",
       "2df8db584e831559491197a21c82184ed93832a3                                       NaN   \n",
       "fa0ac4daf9628b0fdf874c71f50e9619fdfcc5d7                                      0.35   \n",
       "e80582287da1f1a66282ba80b23df4d73979ac17                                       NaN   \n",
       "554e5cf51f2fc724ae47e1cac1ebab0f58de28db                                       NaN   \n",
       "6cf576c6838b15c80a38287e880a48dd731aba99                                       NaN   \n",
       "4bb3bf56ffeb05bf5bd49ca062e1d024f7452f3a                                       NaN   \n",
       "b5e27ce5e72536c3637d98f984f624ca8896a6f5                                       NaN   \n",
       "\n",
       "                                          3a1148968bac3d3ec90fb3be0e4d376c0155a986  \\\n",
       "2c42e6551311710ca5a839d62058820a42ead493                                       NaN   \n",
       "3d766c6bd78be4598a6c8946640097d929b0b09f                                      0.42   \n",
       "52843949671b246e56ec8842cadb599e32d998ba                                      0.39   \n",
       "2ecca069111af5716fa0be6cfdfb5e2f500cdab2                                      0.39   \n",
       "0cc972c46fa7bfe81500fc63a9186157c10712a4                                       NaN   \n",
       "dbb98c22429724137982c79e49da372d859dcfe3                                       NaN   \n",
       "3bd82072d48527e803d8c092a5efb9d582195e48                                       NaN   \n",
       "7cd6f87c6cf9d06464814964adb7ae2839bdce49                                       NaN   \n",
       "6bcace58344f65ea8611afb01287e85a8e18827b                                       NaN   \n",
       "8dca390a049834fecd2511f489fbd7884d8cba0e                                       NaN   \n",
       "b93e873853c4b0acb3cf11514e46eff902a76c14                                       NaN   \n",
       "2df8db584e831559491197a21c82184ed93832a3                                       NaN   \n",
       "fa0ac4daf9628b0fdf874c71f50e9619fdfcc5d7                                       NaN   \n",
       "e80582287da1f1a66282ba80b23df4d73979ac17                                       NaN   \n",
       "554e5cf51f2fc724ae47e1cac1ebab0f58de28db                                       NaN   \n",
       "6cf576c6838b15c80a38287e880a48dd731aba99                                      0.35   \n",
       "4bb3bf56ffeb05bf5bd49ca062e1d024f7452f3a                                       NaN   \n",
       "b5e27ce5e72536c3637d98f984f624ca8896a6f5                                       NaN   \n",
       "\n",
       "                                          4446d24372beca4092c5444350f3bcaddcadd583  \\\n",
       "2c42e6551311710ca5a839d62058820a42ead493                                       NaN   \n",
       "3d766c6bd78be4598a6c8946640097d929b0b09f                                       NaN   \n",
       "52843949671b246e56ec8842cadb599e32d998ba                                       NaN   \n",
       "2ecca069111af5716fa0be6cfdfb5e2f500cdab2                                       NaN   \n",
       "0cc972c46fa7bfe81500fc63a9186157c10712a4                                       NaN   \n",
       "dbb98c22429724137982c79e49da372d859dcfe3                                       NaN   \n",
       "3bd82072d48527e803d8c092a5efb9d582195e48                                       NaN   \n",
       "7cd6f87c6cf9d06464814964adb7ae2839bdce49                                      0.31   \n",
       "6bcace58344f65ea8611afb01287e85a8e18827b                                       NaN   \n",
       "8dca390a049834fecd2511f489fbd7884d8cba0e                                       NaN   \n",
       "b93e873853c4b0acb3cf11514e46eff902a76c14                                       NaN   \n",
       "2df8db584e831559491197a21c82184ed93832a3                                       NaN   \n",
       "fa0ac4daf9628b0fdf874c71f50e9619fdfcc5d7                                       NaN   \n",
       "e80582287da1f1a66282ba80b23df4d73979ac17                                      0.44   \n",
       "554e5cf51f2fc724ae47e1cac1ebab0f58de28db                                       NaN   \n",
       "6cf576c6838b15c80a38287e880a48dd731aba99                                       NaN   \n",
       "4bb3bf56ffeb05bf5bd49ca062e1d024f7452f3a                                       NaN   \n",
       "b5e27ce5e72536c3637d98f984f624ca8896a6f5                                      0.38   \n",
       "\n",
       "                                          59bad0aea569a12d785e401d185f8814d128333e  \\\n",
       "2c42e6551311710ca5a839d62058820a42ead493                                       NaN   \n",
       "3d766c6bd78be4598a6c8946640097d929b0b09f                                       NaN   \n",
       "52843949671b246e56ec8842cadb599e32d998ba                                       NaN   \n",
       "2ecca069111af5716fa0be6cfdfb5e2f500cdab2                                       NaN   \n",
       "0cc972c46fa7bfe81500fc63a9186157c10712a4                                       NaN   \n",
       "dbb98c22429724137982c79e49da372d859dcfe3                                       NaN   \n",
       "3bd82072d48527e803d8c092a5efb9d582195e48                                       NaN   \n",
       "7cd6f87c6cf9d06464814964adb7ae2839bdce49                                       NaN   \n",
       "6bcace58344f65ea8611afb01287e85a8e18827b                                       NaN   \n",
       "8dca390a049834fecd2511f489fbd7884d8cba0e                                      0.35   \n",
       "b93e873853c4b0acb3cf11514e46eff902a76c14                                       NaN   \n",
       "2df8db584e831559491197a21c82184ed93832a3                                       NaN   \n",
       "fa0ac4daf9628b0fdf874c71f50e9619fdfcc5d7                                       NaN   \n",
       "e80582287da1f1a66282ba80b23df4d73979ac17                                       NaN   \n",
       "554e5cf51f2fc724ae47e1cac1ebab0f58de28db                                       NaN   \n",
       "6cf576c6838b15c80a38287e880a48dd731aba99                                       NaN   \n",
       "4bb3bf56ffeb05bf5bd49ca062e1d024f7452f3a                                       NaN   \n",
       "b5e27ce5e72536c3637d98f984f624ca8896a6f5                                       NaN   \n",
       "\n",
       "                                          78b67248d2522d41923d28f831ce02c669838926  \n",
       "2c42e6551311710ca5a839d62058820a42ead493                                       NaN  \n",
       "3d766c6bd78be4598a6c8946640097d929b0b09f                                       NaN  \n",
       "52843949671b246e56ec8842cadb599e32d998ba                                       NaN  \n",
       "2ecca069111af5716fa0be6cfdfb5e2f500cdab2                                       NaN  \n",
       "0cc972c46fa7bfe81500fc63a9186157c10712a4                                      0.30  \n",
       "dbb98c22429724137982c79e49da372d859dcfe3                                      0.31  \n",
       "3bd82072d48527e803d8c092a5efb9d582195e48                                       NaN  \n",
       "7cd6f87c6cf9d06464814964adb7ae2839bdce49                                       NaN  \n",
       "6bcace58344f65ea8611afb01287e85a8e18827b                                       NaN  \n",
       "8dca390a049834fecd2511f489fbd7884d8cba0e                                       NaN  \n",
       "b93e873853c4b0acb3cf11514e46eff902a76c14                                      0.41  \n",
       "2df8db584e831559491197a21c82184ed93832a3                                      0.43  \n",
       "fa0ac4daf9628b0fdf874c71f50e9619fdfcc5d7                                       NaN  \n",
       "e80582287da1f1a66282ba80b23df4d73979ac17                                       NaN  \n",
       "554e5cf51f2fc724ae47e1cac1ebab0f58de28db                                       NaN  \n",
       "6cf576c6838b15c80a38287e880a48dd731aba99                                       NaN  \n",
       "4bb3bf56ffeb05bf5bd49ca062e1d024f7452f3a                                       NaN  \n",
       "b5e27ce5e72536c3637d98f984f624ca8896a6f5                                       NaN  "
      ]
     },
     "execution_count": 107,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Filter for similarities over 0.3, drop any users pairs that are all null values\n",
    "sample_10_matrix = sample_10_matrix[sample_10_matrix > 0.3]\n",
    "sample_10_matrix = sample_10_matrix.dropna(how = 'all')\n",
    "sample_10_matrix = sample_10_matrix.dropna(how = 'all', axis = 1)\n",
    "sample_10_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x23b89b69488>"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Visualise the user similarities that are over 0.3 for each of the 10 sampled users\n",
    "plt.figure(figsize=(12, 8))\n",
    "sns.heatmap(data = sample_10_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [],
   "source": [
    "# NOTE: 3 of the sampled 10 are not on the heatmap\n",
    "# This is because it was filtered for 0.3 similarity at least, so they have no interactions with any other users large enough \n",
    "# Therefore these rows and columns where all nulls and so removed for the visualisation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [],
   "source": [
    "# -- EDA Complete -- "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save sparse matrix to apply in next notebook and construct SVD recommender system \n",
    "import scipy.sparse\n",
    "\n",
    "# Save the existing CSR matrix to a .npz file\n",
    "scipy.sparse.save_npz('sparse_matrix.npz', sparse_matrix)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save dfs for use in next notebook \n",
    "triplets_df.to_csv('triplets_df.csv', index = True)\n",
    "song_df.to_csv('songs_df.csv', index = True)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}