metric_at_customern_EDA.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# !cd metrics-customerN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# %%capture\n",
    "# !git clone https://github.com/Reclist/reclist/\n",
    "# !cd reclist ; pip install -e .\n",
    "\n",
    "# !git clone https://github.com/RecList/evalRS-KDD-2023\n",
    "# !cd evalRS-KDD-2023 ; git checkout readme_kdd_2023"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "!which python\n",
    "!pip install -r evalRS-KDD-2023/requirements.txt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/mayanksingh/VSCodeProjects/metrics-customerN/.venv/lib/python3.9/site-packages/urllib3/__init__.py:34: NotOpenSSLWarning: urllib3 v2.0 only supports OpenSSL 1.1.1+, currently the 'ssl' module is compiled with 'LibreSSL 2.8.3'. See: https://github.com/urllib3/urllib3/issues/3020\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "from EvalRSRunner import ChallengeDataset\n",
    "\n",
    "import gc\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import pyarrow.parquet as pq\n",
    "\n",
    "# sns.set_theme(\"notebook\")\n",
    "# Change notebook data format to be like 100.10 instead of scientific notation\n",
    "pd.options.display.float_format = '{:.2f}'.format"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.ticker as ticker\n",
    "import seaborn as sns\n",
    "from matplotlib.ticker import PercentFormatter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LFM dataset already downloaded. Skipping download.\n",
      "Loading dataset.\n",
      "Generating Train/Test Split.\n",
      "Generating dataset hashes.\n"
     ]
    }
   ],
   "source": [
    "dataset = ChallengeDataset(force_download=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((6869679, 6), (29722, 3))"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train, test = dataset.get_sample_train_test()\n",
    "train.shape, test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.4326548591280612"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# test is only 0.4% of the data, no need to combine with train for EDA analysis\n",
    "100.00*test.shape[0]/train.shape[0] "
   ]
  },
  {
   "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>user_id</th>\n",
       "      <th>artist_id</th>\n",
       "      <th>album_id</th>\n",
       "      <th>track_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>user_track_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>50967444</td>\n",
       "      <td>3059</td>\n",
       "      <td>114114</td>\n",
       "      <td>336613</td>\n",
       "      <td>1409538054</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>50967444</td>\n",
       "      <td>3059</td>\n",
       "      <td>114114</td>\n",
       "      <td>336156</td>\n",
       "      <td>1409537916</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>50967444</td>\n",
       "      <td>3059</td>\n",
       "      <td>114114</td>\n",
       "      <td>339139</td>\n",
       "      <td>1409537712</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>50967444</td>\n",
       "      <td>677</td>\n",
       "      <td>3035</td>\n",
       "      <td>674449</td>\n",
       "      <td>1409532988</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>50967444</td>\n",
       "      <td>677</td>\n",
       "      <td>3035</td>\n",
       "      <td>639652</td>\n",
       "      <td>1409532015</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    user_id  artist_id  album_id  track_id   timestamp  user_track_count\n",
       "0  50967444       3059    114114    336613  1409538054                 1\n",
       "1  50967444       3059    114114    336156  1409537916                 1\n",
       "2  50967444       3059    114114    339139  1409537712                 1\n",
       "4  50967444        677      3035    674449  1409532988                 1\n",
       "5  50967444        677      3035    639652  1409532015                 1"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "# convert the timestamp in a more convenient format\n",
    "train[\"timestamp\"] = pd.to_datetime(train[\"timestamp\"], unit=\"s\")"
   ]
  },
  {
   "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>user_id</th>\n",
       "      <th>artist_id</th>\n",
       "      <th>album_id</th>\n",
       "      <th>track_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>user_track_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>50967444</td>\n",
       "      <td>3059</td>\n",
       "      <td>114114</td>\n",
       "      <td>336613</td>\n",
       "      <td>2014-09-01 02:20:54</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>50967444</td>\n",
       "      <td>3059</td>\n",
       "      <td>114114</td>\n",
       "      <td>336156</td>\n",
       "      <td>2014-09-01 02:18:36</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>50967444</td>\n",
       "      <td>3059</td>\n",
       "      <td>114114</td>\n",
       "      <td>339139</td>\n",
       "      <td>2014-09-01 02:15:12</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>50967444</td>\n",
       "      <td>677</td>\n",
       "      <td>3035</td>\n",
       "      <td>674449</td>\n",
       "      <td>2014-09-01 00:56:28</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>50967444</td>\n",
       "      <td>677</td>\n",
       "      <td>3035</td>\n",
       "      <td>639652</td>\n",
       "      <td>2014-09-01 00:40:15</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    user_id  artist_id  album_id  track_id           timestamp  \\\n",
       "0  50967444       3059    114114    336613 2014-09-01 02:20:54   \n",
       "1  50967444       3059    114114    336156 2014-09-01 02:18:36   \n",
       "2  50967444       3059    114114    339139 2014-09-01 02:15:12   \n",
       "4  50967444        677      3035    674449 2014-09-01 00:56:28   \n",
       "5  50967444        677      3035    639652 2014-09-01 00:40:15   \n",
       "\n",
       "   user_track_count  \n",
       "0                 1  \n",
       "1                 1  \n",
       "2                 1  \n",
       "4                 1  \n",
       "5                 1  "
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "# A session id per customer based on a 6 hr window, if their last session was more than 6 hrs ago, then it is a new session\n",
    "\n",
    "train['session_id'] = train.groupby(['user_id', pd.Grouper(key='timestamp', freq='6H')]).ngroup()"
   ]
  },
  {
   "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>user_id</th>\n",
       "      <th>artist_id</th>\n",
       "      <th>album_id</th>\n",
       "      <th>track_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>user_track_count</th>\n",
       "      <th>session_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>50967444</td>\n",
       "      <td>3059</td>\n",
       "      <td>114114</td>\n",
       "      <td>336613</td>\n",
       "      <td>2014-09-01 02:20:54</td>\n",
       "      <td>1</td>\n",
       "      <td>1442994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>50967444</td>\n",
       "      <td>3059</td>\n",
       "      <td>114114</td>\n",
       "      <td>336156</td>\n",
       "      <td>2014-09-01 02:18:36</td>\n",
       "      <td>1</td>\n",
       "      <td>1442994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>50967444</td>\n",
       "      <td>3059</td>\n",
       "      <td>114114</td>\n",
       "      <td>339139</td>\n",
       "      <td>2014-09-01 02:15:12</td>\n",
       "      <td>1</td>\n",
       "      <td>1442994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>50967444</td>\n",
       "      <td>677</td>\n",
       "      <td>3035</td>\n",
       "      <td>674449</td>\n",
       "      <td>2014-09-01 00:56:28</td>\n",
       "      <td>1</td>\n",
       "      <td>1442994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>50967444</td>\n",
       "      <td>677</td>\n",
       "      <td>3035</td>\n",
       "      <td>639652</td>\n",
       "      <td>2014-09-01 00:40:15</td>\n",
       "      <td>1</td>\n",
       "      <td>1442994</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    user_id  artist_id  album_id  track_id           timestamp  \\\n",
       "0  50967444       3059    114114    336613 2014-09-01 02:20:54   \n",
       "1  50967444       3059    114114    336156 2014-09-01 02:18:36   \n",
       "2  50967444       3059    114114    339139 2014-09-01 02:15:12   \n",
       "4  50967444        677      3035    674449 2014-09-01 00:56:28   \n",
       "5  50967444        677      3035    639652 2014-09-01 00:40:15   \n",
       "\n",
       "   user_track_count  session_id  \n",
       "0                 1     1442994  \n",
       "1                 1     1442994  \n",
       "2                 1     1442994  \n",
       "4                 1     1442994  \n",
       "5                 1     1442994  "
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count   29722.00\n",
       "mean      231.13\n",
       "std       108.11\n",
       "min         9.00\n",
       "25%       149.00\n",
       "50%       237.00\n",
       "75%       315.00\n",
       "max       496.00\n",
       "Name: user_id, dtype: float64"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# user count\n",
    "train.groupby('user_id')['user_id'].count().describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count   29722.00\n",
       "mean      231.13\n",
       "std       108.11\n",
       "min         9.00\n",
       "25%       149.00\n",
       "50%       237.00\n",
       "75%       315.00\n",
       "max       496.00\n",
       "Name: session_id, dtype: float64"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Session per user\n",
    "train.groupby('user_id')['session_id'].count().describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "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>artist_id</th>\n",
       "      <th>album_id</th>\n",
       "      <th>track_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>user_track_count</th>\n",
       "      <th>session_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>37926426</th>\n",
       "      <td>384</td>\n",
       "      <td>4166</td>\n",
       "      <td>184583</td>\n",
       "      <td>562698</td>\n",
       "      <td>2013-12-29 23:42:55</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926425</th>\n",
       "      <td>384</td>\n",
       "      <td>289</td>\n",
       "      <td>8126</td>\n",
       "      <td>198978</td>\n",
       "      <td>2013-12-29 23:50:28</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926424</th>\n",
       "      <td>384</td>\n",
       "      <td>7990</td>\n",
       "      <td>39548</td>\n",
       "      <td>177872</td>\n",
       "      <td>2013-12-30 00:12:10</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926423</th>\n",
       "      <td>384</td>\n",
       "      <td>245</td>\n",
       "      <td>292362</td>\n",
       "      <td>6415</td>\n",
       "      <td>2013-12-30 00:18:47</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926421</th>\n",
       "      <td>384</td>\n",
       "      <td>6290</td>\n",
       "      <td>94625</td>\n",
       "      <td>277550</td>\n",
       "      <td>2013-12-30 00:27:02</td>\n",
       "      <td>1</td>\n",
       "      <td>1</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926123</th>\n",
       "      <td>384</td>\n",
       "      <td>67</td>\n",
       "      <td>3178231</td>\n",
       "      <td>4474</td>\n",
       "      <td>2014-04-15 00:09:46</td>\n",
       "      <td>1</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926122</th>\n",
       "      <td>384</td>\n",
       "      <td>67</td>\n",
       "      <td>3178231</td>\n",
       "      <td>4473</td>\n",
       "      <td>2014-04-15 00:13:38</td>\n",
       "      <td>1</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926121</th>\n",
       "      <td>384</td>\n",
       "      <td>67</td>\n",
       "      <td>3178231</td>\n",
       "      <td>4472</td>\n",
       "      <td>2014-04-15 00:16:47</td>\n",
       "      <td>1</td>\n",
       "      <td>45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926120</th>\n",
       "      <td>384</td>\n",
       "      <td>1513716</td>\n",
       "      <td>7625712</td>\n",
       "      <td>14892045</td>\n",
       "      <td>2014-04-16 11:33:18</td>\n",
       "      <td>16</td>\n",
       "      <td>46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926118</th>\n",
       "      <td>384</td>\n",
       "      <td>1513716</td>\n",
       "      <td>7625712</td>\n",
       "      <td>15832613</td>\n",
       "      <td>2014-04-16 11:39:22</td>\n",
       "      <td>15</td>\n",
       "      <td>46</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>202 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          user_id  artist_id  album_id  track_id           timestamp  \\\n",
       "37926426      384       4166    184583    562698 2013-12-29 23:42:55   \n",
       "37926425      384        289      8126    198978 2013-12-29 23:50:28   \n",
       "37926424      384       7990     39548    177872 2013-12-30 00:12:10   \n",
       "37926423      384        245    292362      6415 2013-12-30 00:18:47   \n",
       "37926421      384       6290     94625    277550 2013-12-30 00:27:02   \n",
       "...           ...        ...       ...       ...                 ...   \n",
       "37926123      384         67   3178231      4474 2014-04-15 00:09:46   \n",
       "37926122      384         67   3178231      4473 2014-04-15 00:13:38   \n",
       "37926121      384         67   3178231      4472 2014-04-15 00:16:47   \n",
       "37926120      384    1513716   7625712  14892045 2014-04-16 11:33:18   \n",
       "37926118      384    1513716   7625712  15832613 2014-04-16 11:39:22   \n",
       "\n",
       "          user_track_count  session_id  \n",
       "37926426                 1           0  \n",
       "37926425                 1           0  \n",
       "37926424                 2           1  \n",
       "37926423                 1           1  \n",
       "37926421                 1           1  \n",
       "...                    ...         ...  \n",
       "37926123                 1          45  \n",
       "37926122                 1          45  \n",
       "37926121                 1          45  \n",
       "37926120                16          46  \n",
       "37926118                15          46  \n",
       "\n",
       "[202 rows x 7 columns]"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train[train['user_id']==384].sort_values(by='timestamp')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Calculate track_playtime as time difference between two consecutive timestamp after sorting by user_id and timestamp and within the same session_id. The track_playtime should be for the last track played in the session.\n",
    "# # So the last track won't have a track playtime\n",
    "# train['playtime_mins'] = -train.sort_values(['user_id', 'timestamp']).groupby('user_id')['timestamp'].diff().dt.total_seconds()/60.00\n",
    "\n",
    "\n",
    "# Cannot calculate time spent on each track because there is only one unique row per user_id, track_id. The timestamp is probably when they first heard the track"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "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>artist_id</th>\n",
       "      <th>album_id</th>\n",
       "      <th>track_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>user_track_count</th>\n",
       "      <th>session_id</th>\n",
       "      <th>track_playtime</th>\n",
       "      <th>playtime_mins</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>37926426</th>\n",
       "      <td>384</td>\n",
       "      <td>4166</td>\n",
       "      <td>184583</td>\n",
       "      <td>562698</td>\n",
       "      <td>2013-12-29 23:42:55</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926425</th>\n",
       "      <td>384</td>\n",
       "      <td>289</td>\n",
       "      <td>8126</td>\n",
       "      <td>198978</td>\n",
       "      <td>2013-12-29 23:50:28</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7.55</td>\n",
       "      <td>-7.55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926424</th>\n",
       "      <td>384</td>\n",
       "      <td>7990</td>\n",
       "      <td>39548</td>\n",
       "      <td>177872</td>\n",
       "      <td>2013-12-30 00:12:10</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>21.70</td>\n",
       "      <td>-21.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926423</th>\n",
       "      <td>384</td>\n",
       "      <td>245</td>\n",
       "      <td>292362</td>\n",
       "      <td>6415</td>\n",
       "      <td>2013-12-30 00:18:47</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>6.62</td>\n",
       "      <td>-6.62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926421</th>\n",
       "      <td>384</td>\n",
       "      <td>6290</td>\n",
       "      <td>94625</td>\n",
       "      <td>277550</td>\n",
       "      <td>2013-12-30 00:27:02</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>8.25</td>\n",
       "      <td>-8.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926123</th>\n",
       "      <td>384</td>\n",
       "      <td>67</td>\n",
       "      <td>3178231</td>\n",
       "      <td>4474</td>\n",
       "      <td>2014-04-15 00:09:46</td>\n",
       "      <td>1</td>\n",
       "      <td>45</td>\n",
       "      <td>5.98</td>\n",
       "      <td>-5.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926122</th>\n",
       "      <td>384</td>\n",
       "      <td>67</td>\n",
       "      <td>3178231</td>\n",
       "      <td>4473</td>\n",
       "      <td>2014-04-15 00:13:38</td>\n",
       "      <td>1</td>\n",
       "      <td>45</td>\n",
       "      <td>3.87</td>\n",
       "      <td>-3.87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926121</th>\n",
       "      <td>384</td>\n",
       "      <td>67</td>\n",
       "      <td>3178231</td>\n",
       "      <td>4472</td>\n",
       "      <td>2014-04-15 00:16:47</td>\n",
       "      <td>1</td>\n",
       "      <td>45</td>\n",
       "      <td>3.15</td>\n",
       "      <td>-3.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926120</th>\n",
       "      <td>384</td>\n",
       "      <td>1513716</td>\n",
       "      <td>7625712</td>\n",
       "      <td>14892045</td>\n",
       "      <td>2014-04-16 11:33:18</td>\n",
       "      <td>16</td>\n",
       "      <td>46</td>\n",
       "      <td>2116.52</td>\n",
       "      <td>-2116.52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37926118</th>\n",
       "      <td>384</td>\n",
       "      <td>1513716</td>\n",
       "      <td>7625712</td>\n",
       "      <td>15832613</td>\n",
       "      <td>2014-04-16 11:39:22</td>\n",
       "      <td>15</td>\n",
       "      <td>46</td>\n",
       "      <td>6.07</td>\n",
       "      <td>-6.07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>202 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          user_id  artist_id  album_id  track_id           timestamp  \\\n",
       "37926426      384       4166    184583    562698 2013-12-29 23:42:55   \n",
       "37926425      384        289      8126    198978 2013-12-29 23:50:28   \n",
       "37926424      384       7990     39548    177872 2013-12-30 00:12:10   \n",
       "37926423      384        245    292362      6415 2013-12-30 00:18:47   \n",
       "37926421      384       6290     94625    277550 2013-12-30 00:27:02   \n",
       "...           ...        ...       ...       ...                 ...   \n",
       "37926123      384         67   3178231      4474 2014-04-15 00:09:46   \n",
       "37926122      384         67   3178231      4473 2014-04-15 00:13:38   \n",
       "37926121      384         67   3178231      4472 2014-04-15 00:16:47   \n",
       "37926120      384    1513716   7625712  14892045 2014-04-16 11:33:18   \n",
       "37926118      384    1513716   7625712  15832613 2014-04-16 11:39:22   \n",
       "\n",
       "          user_track_count  session_id  track_playtime  playtime_mins  \n",
       "37926426                 1           0             NaN            NaN  \n",
       "37926425                 1           0            7.55          -7.55  \n",
       "37926424                 2           1           21.70         -21.70  \n",
       "37926423                 1           1            6.62          -6.62  \n",
       "37926421                 1           1            8.25          -8.25  \n",
       "...                    ...         ...             ...            ...  \n",
       "37926123                 1          45            5.98          -5.98  \n",
       "37926122                 1          45            3.87          -3.87  \n",
       "37926121                 1          45            3.15          -3.15  \n",
       "37926120                16          46         2116.52       -2116.52  \n",
       "37926118                15          46            6.07          -6.07  \n",
       "\n",
       "[202 rows x 9 columns]"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train[train['user_id'] == 384].sort_values(['user_id', 'timestamp'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "user_session_df = train.groupby(['user_id', 'session_id'])['track_id'].count().reset_index()\n",
    "user_session_df.columns = ['user_id', 'session_id', 'tracks_listened']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "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>session_id</th>\n",
       "      <th>tracks_listened</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>384</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>384</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>384</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>384</td>\n",
       "      <td>3</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>384</td>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id  session_id  tracks_listened\n",
       "0      384           0                2\n",
       "1      384           1                6\n",
       "2      384           2                3\n",
       "3      384           3                7\n",
       "4      384           4                7"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_session_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(47, 3)"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_session_df[user_session_df['user_id']==384].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Session and tracks_listened count\n",
    "# user_summary = user_session_df.groupby(['user_id']).agg({'session_id':'nunique', 'tracks_listened': 'median'})\n",
    "# user_summary.columns = ['session_count', 'customern_tracks']\n",
    "# user_summary['customern_tracks'] = user_summary['customern_tracks'].astype(int)\n",
    "# user_summary = user_summary.reset_index()\n",
    "\n",
    "\n",
    "\n",
    "user_summary = user_session_df.groupby('user_id').agg({'tracks_listened': [np.median, np.std, np.mean], 'session_id': 'nunique'})\n",
    "user_summary.columns = user_summary.columns.droplevel()\n",
    "user_summary = user_summary.reset_index()\n",
    "user_summary.columns = ['user_id', 'customern_tracks', 'impr_std', 'impr_mean', 'session_count']\n",
    "user_summary['customern_tracks'] = user_summary['customern_tracks'].apply(int)\n",
    "user_summary['impr_std_round'] = user_summary['impr_std'].fillna(0).apply(np.round).apply(int)\n",
    "\n",
    "user_summary = user_summary[user_summary['session_count'] >= 3] # 99.5% of the users have >= 3 sessions\n",
    "\n",
    "\n",
    "user_summary['cv'] = np.round(user_summary['impr_std']/user_summary['impr_mean'], 2)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "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>customern_tracks</th>\n",
       "      <th>impr_std</th>\n",
       "      <th>impr_mean</th>\n",
       "      <th>session_count</th>\n",
       "      <th>impr_std_round</th>\n",
       "      <th>cv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>384</td>\n",
       "      <td>3</td>\n",
       "      <td>3.08</td>\n",
       "      <td>4.30</td>\n",
       "      <td>47</td>\n",
       "      <td>3</td>\n",
       "      <td>0.72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10879</td>\n",
       "      <td>6</td>\n",
       "      <td>7.11</td>\n",
       "      <td>7.52</td>\n",
       "      <td>29</td>\n",
       "      <td>7</td>\n",
       "      <td>0.95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12619</td>\n",
       "      <td>2</td>\n",
       "      <td>5.54</td>\n",
       "      <td>4.84</td>\n",
       "      <td>38</td>\n",
       "      <td>6</td>\n",
       "      <td>1.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>14308</td>\n",
       "      <td>3</td>\n",
       "      <td>5.42</td>\n",
       "      <td>5.00</td>\n",
       "      <td>39</td>\n",
       "      <td>5</td>\n",
       "      <td>1.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>15067</td>\n",
       "      <td>8</td>\n",
       "      <td>7.92</td>\n",
       "      <td>8.81</td>\n",
       "      <td>31</td>\n",
       "      <td>8</td>\n",
       "      <td>0.90</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29716</th>\n",
       "      <td>50081937</td>\n",
       "      <td>6</td>\n",
       "      <td>9.00</td>\n",
       "      <td>9.25</td>\n",
       "      <td>4</td>\n",
       "      <td>9</td>\n",
       "      <td>0.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29718</th>\n",
       "      <td>50085736</td>\n",
       "      <td>1</td>\n",
       "      <td>1.20</td>\n",
       "      <td>2.00</td>\n",
       "      <td>15</td>\n",
       "      <td>1</td>\n",
       "      <td>0.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29719</th>\n",
       "      <td>50086315</td>\n",
       "      <td>3</td>\n",
       "      <td>19.08</td>\n",
       "      <td>13.00</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>1.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29720</th>\n",
       "      <td>50900118</td>\n",
       "      <td>1</td>\n",
       "      <td>1.49</td>\n",
       "      <td>1.81</td>\n",
       "      <td>89</td>\n",
       "      <td>1</td>\n",
       "      <td>0.82</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29721</th>\n",
       "      <td>50967444</td>\n",
       "      <td>6</td>\n",
       "      <td>10.64</td>\n",
       "      <td>9.96</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>1.07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>29669 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        user_id  customern_tracks  impr_std  impr_mean  session_count  \\\n",
       "0           384                 3      3.08       4.30             47   \n",
       "1         10879                 6      7.11       7.52             29   \n",
       "2         12619                 2      5.54       4.84             38   \n",
       "3         14308                 3      5.42       5.00             39   \n",
       "4         15067                 8      7.92       8.81             31   \n",
       "...         ...               ...       ...        ...            ...   \n",
       "29716  50081937                 6      9.00       9.25              4   \n",
       "29718  50085736                 1      1.20       2.00             15   \n",
       "29719  50086315                 3     19.08      13.00              3   \n",
       "29720  50900118                 1      1.49       1.81             89   \n",
       "29721  50967444                 6     10.64       9.96             28   \n",
       "\n",
       "       impr_std_round   cv  \n",
       "0                   3 0.72  \n",
       "1                   7 0.95  \n",
       "2                   6 1.14  \n",
       "3                   5 1.08  \n",
       "4                   8 0.90  \n",
       "...               ...  ...  \n",
       "29716               9 0.97  \n",
       "29718               1 0.60  \n",
       "29719              19 1.47  \n",
       "29720               1 0.82  \n",
       "29721              11 1.07  \n",
       "\n",
       "[29669 rows x 7 columns]"
      ]
     },
     "execution_count": 178,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count   29669.00\n",
       "mean       48.63\n",
       "std        28.95\n",
       "min         3.00\n",
       "25%        28.00\n",
       "50%        43.00\n",
       "75%        63.00\n",
       "max       309.00\n",
       "Name: session_count, dtype: float64"
      ]
     },
     "execution_count": 179,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_summary['session_count'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 180,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot: ylabel='Frequency'>"
      ]
     },
     "execution_count": 180,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Distribution of user sessions\n",
    "user_summary['session_count'].plot(kind='hist', bins=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 181,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count   29669.00\n",
       "mean        3.58\n",
       "std         2.55\n",
       "min         1.00\n",
       "25%         2.00\n",
       "50%         3.00\n",
       "75%         4.00\n",
       "max        44.00\n",
       "Name: customern_tracks, dtype: float64"
      ]
     },
     "execution_count": 181,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_summary['customern_tracks'].describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot: ylabel='Frequency'>"
      ]
     },
     "execution_count": 182,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Distribution of customern_tracks (median tracks listened by a customer)\n",
    "\n",
    "# Distribution of user sessions\n",
    "\n",
    "threshold = 10\n",
    "\n",
    "# Create a pareto chart with less than threshold value of customern_tracks\n",
    "user_summary[user_summary['customern_tracks'] <= threshold]['customern_tracks'].plot(kind='hist', bins=threshold)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "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>customern_tracks</th>\n",
       "      <th>impr_std</th>\n",
       "      <th>impr_mean</th>\n",
       "      <th>session_count</th>\n",
       "      <th>impr_std_round</th>\n",
       "      <th>cv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>384</td>\n",
       "      <td>3</td>\n",
       "      <td>3.08</td>\n",
       "      <td>4.30</td>\n",
       "      <td>47</td>\n",
       "      <td>3</td>\n",
       "      <td>0.72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10879</td>\n",
       "      <td>6</td>\n",
       "      <td>7.11</td>\n",
       "      <td>7.52</td>\n",
       "      <td>29</td>\n",
       "      <td>7</td>\n",
       "      <td>0.95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12619</td>\n",
       "      <td>2</td>\n",
       "      <td>5.54</td>\n",
       "      <td>4.84</td>\n",
       "      <td>38</td>\n",
       "      <td>6</td>\n",
       "      <td>1.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>14308</td>\n",
       "      <td>3</td>\n",
       "      <td>5.42</td>\n",
       "      <td>5.00</td>\n",
       "      <td>39</td>\n",
       "      <td>5</td>\n",
       "      <td>1.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>15067</td>\n",
       "      <td>8</td>\n",
       "      <td>7.92</td>\n",
       "      <td>8.81</td>\n",
       "      <td>31</td>\n",
       "      <td>8</td>\n",
       "      <td>0.90</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29716</th>\n",
       "      <td>50081937</td>\n",
       "      <td>6</td>\n",
       "      <td>9.00</td>\n",
       "      <td>9.25</td>\n",
       "      <td>4</td>\n",
       "      <td>9</td>\n",
       "      <td>0.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29718</th>\n",
       "      <td>50085736</td>\n",
       "      <td>1</td>\n",
       "      <td>1.20</td>\n",
       "      <td>2.00</td>\n",
       "      <td>15</td>\n",
       "      <td>1</td>\n",
       "      <td>0.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29719</th>\n",
       "      <td>50086315</td>\n",
       "      <td>3</td>\n",
       "      <td>19.08</td>\n",
       "      <td>13.00</td>\n",
       "      <td>3</td>\n",
       "      <td>19</td>\n",
       "      <td>1.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29720</th>\n",
       "      <td>50900118</td>\n",
       "      <td>1</td>\n",
       "      <td>1.49</td>\n",
       "      <td>1.81</td>\n",
       "      <td>89</td>\n",
       "      <td>1</td>\n",
       "      <td>0.82</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29721</th>\n",
       "      <td>50967444</td>\n",
       "      <td>6</td>\n",
       "      <td>10.64</td>\n",
       "      <td>9.96</td>\n",
       "      <td>28</td>\n",
       "      <td>11</td>\n",
       "      <td>1.07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>29669 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        user_id  customern_tracks  impr_std  impr_mean  session_count  \\\n",
       "0           384                 3      3.08       4.30             47   \n",
       "1         10879                 6      7.11       7.52             29   \n",
       "2         12619                 2      5.54       4.84             38   \n",
       "3         14308                 3      5.42       5.00             39   \n",
       "4         15067                 8      7.92       8.81             31   \n",
       "...         ...               ...       ...        ...            ...   \n",
       "29716  50081937                 6      9.00       9.25              4   \n",
       "29718  50085736                 1      1.20       2.00             15   \n",
       "29719  50086315                 3     19.08      13.00              3   \n",
       "29720  50900118                 1      1.49       1.81             89   \n",
       "29721  50967444                 6     10.64       9.96             28   \n",
       "\n",
       "       impr_std_round   cv  \n",
       "0                   3 0.72  \n",
       "1                   7 0.95  \n",
       "2                   6 1.14  \n",
       "3                   5 1.08  \n",
       "4                   8 0.90  \n",
       "...               ...  ...  \n",
       "29716               9 0.97  \n",
       "29718               1 0.60  \n",
       "29719              19 1.47  \n",
       "29720               1 0.82  \n",
       "29721              11 1.07  \n",
       "\n",
       "[29669 rows x 7 columns]"
      ]
     },
     "execution_count": 183,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Summarising by CustomerN\n",
    "customern_df = user_summary.groupby('customern_tracks')['user_id'].count().reset_index()\n",
    "\n",
    "# Cumulative percentage of users\n",
    "customern_df['cum_perc_users'] = 100.00*customern_df['user_id'].cumsum()/customern_df['user_id'].sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "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>customern_tracks</th>\n",
       "      <th>user_id</th>\n",
       "      <th>cum_perc_users</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>3321</td>\n",
       "      <td>11.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>8677</td>\n",
       "      <td>40.44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>6529</td>\n",
       "      <td>62.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>4094</td>\n",
       "      <td>76.24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>2535</td>\n",
       "      <td>84.79</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   customern_tracks  user_id  cum_perc_users\n",
       "0                 1     3321           11.19\n",
       "1                 2     8677           40.44\n",
       "2                 3     6529           62.45\n",
       "3                 4     4094           76.24\n",
       "4                 5     2535           84.79"
      ]
     },
     "execution_count": 185,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "customern_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 186,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 1800x900 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "threshold = 10\n",
    "plt.figure(figsize=(18, 9))\n",
    "plot_df = customern_df[customern_df['customern_tracks'] <= threshold]\n",
    "# bar plot\n",
    "fig, ax = plt.subplots()\n",
    "ax.bar(plot_df['customern_tracks'], plot_df['user_id'], color='lightblue')\n",
    "\n",
    "# cum perc line\n",
    "ax2 = ax.twinx()\n",
    "ax2.plot(plot_df['customern_tracks'], plot_df['cum_perc_users'], color='darkblue')\n",
    "plt.xticks(np.arange(0, 10, 1))\n",
    "ax.set_xlabel('CustomerN \\n (Median Impressions viewed)')\n",
    "ax.set_ylabel('Customer Count')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "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>customern_tracks</th>\n",
       "      <th>impr_std</th>\n",
       "      <th>impr_mean</th>\n",
       "      <th>session_count</th>\n",
       "      <th>impr_std_round</th>\n",
       "      <th>cv</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>384</td>\n",
       "      <td>3</td>\n",
       "      <td>3.08</td>\n",
       "      <td>4.30</td>\n",
       "      <td>47</td>\n",
       "      <td>3</td>\n",
       "      <td>0.72</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>10879</td>\n",
       "      <td>6</td>\n",
       "      <td>7.11</td>\n",
       "      <td>7.52</td>\n",
       "      <td>29</td>\n",
       "      <td>7</td>\n",
       "      <td>0.95</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>12619</td>\n",
       "      <td>2</td>\n",
       "      <td>5.54</td>\n",
       "      <td>4.84</td>\n",
       "      <td>38</td>\n",
       "      <td>6</td>\n",
       "      <td>1.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>14308</td>\n",
       "      <td>3</td>\n",
       "      <td>5.42</td>\n",
       "      <td>5.00</td>\n",
       "      <td>39</td>\n",
       "      <td>5</td>\n",
       "      <td>1.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>15067</td>\n",
       "      <td>8</td>\n",
       "      <td>7.92</td>\n",
       "      <td>8.81</td>\n",
       "      <td>31</td>\n",
       "      <td>8</td>\n",
       "      <td>0.90</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_id  customern_tracks  impr_std  impr_mean  session_count  \\\n",
       "0      384                 3      3.08       4.30             47   \n",
       "1    10879                 6      7.11       7.52             29   \n",
       "2    12619                 2      5.54       4.84             38   \n",
       "3    14308                 3      5.42       5.00             39   \n",
       "4    15067                 8      7.92       8.81             31   \n",
       "\n",
       "   impr_std_round   cv  \n",
       "0               3 0.72  \n",
       "1               7 0.95  \n",
       "2               6 1.14  \n",
       "3               5 1.08  \n",
       "4               8 0.90  "
      ]
     },
     "execution_count": 187,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_summary.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 193,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot: xlabel='customern_tracks', ylabel='impr_std_round'>"
      ]
     },
     "execution_count": 193,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "threshold = 10\n",
    "plot_df = user_summary[user_summary['customern_tracks'] <= threshold]\n",
    "plt.figure(figsize=(6,4))\n",
    "sns.boxplot(x='customern_tracks',\n",
    "            y='impr_std_round',\n",
    "            data=plot_df,\n",
    "            boxprops={'facecolor':'lightblue'},\n",
    "            showfliers=False,\n",
    "            whis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 194,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot: xlabel='customern_tracks', ylabel='cv'>"
      ]
     },
     "execution_count": 194,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAhgAAAFzCAYAAAB8X3AUAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjYuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/P9b71AAAACXBIWXMAAA9hAAAPYQGoP6dpAAAqAElEQVR4nO3de3RU5b3/8c8Ekj0DBASBhGCAgCj3EEVYXM5BBYzIyVLboxSschE4trAEQgWi3FxewkUolKIULIIXQE+5tBVxilFggRS5mFMsitw8UEgArRoSTiaB7N8f/phmzG0IT7IzmfdrrVlrZs+zZ3+HCZlPnmc/z3bZtm0LAADAoAinCwAAALUPAQMAABhHwAAAAMYRMAAAgHEEDAAAYBwBAwAAGEfAAAAAxhEwAACAcXWdLqC6FRUV6ezZs4qOjpbL5XK6HAAAQoZt27p48aLi4uIUEVF+H0XYBYyzZ88qPj7e6TIAAAhZp0+f1k033VRum7ALGNHR0ZJ++Mdp2LChw9UAABA6cnJyFB8f7/8uLU/YBYyrwyINGzYkYAAAUAnBnGLASZ4AAMA4AgYAADCOgAEAAIwjYAAAAOMIGAAAwDgCBgAAMI6AAQAAjCNgAAAA4wgYAADAuLBbyfPHbNtWfn5+hW18Pp/xY1uWVeFqaG63m4uyAQBCTtgHjPz8fCUnJztdRpm8Xq88Ho/TZQAAcE0YIgEAAMaFfQ9GceMWrlCkZZXYbtu2LhcUGD9e3aioUoc/Cn0+rZgyzvjxAACoLgSMYiItS5GWu9TnotwMUwAAECyGSAAAgHEEDAAAYBwBAwAAGEfAAAAAxhEwAACAcQQMAABgHAEDAAAYR8AAAADGETAAAIBxBAwAAGAcAQMAABhHwAAAAMYRMAAAgHEEDAAAYJyjAWPnzp1KSUlRXFycXC6XNm/eXG77jRs3atCgQWrWrJkaNmyo3r17y+v1Vk+xAAAgaI4GjLy8PCUmJmrZsmVBtd+5c6cGDRqk9957TwcOHNBdd92llJQUffrpp1VcKQAAuBZ1nTz44MGDNXjw4KDbL168OODxiy++qD/+8Y/685//rKSkJMPVAQCAynI0YFyvoqIiXbx4UU2aNHG6FEfYtq38/PwK2/h8PqPHtSxLLperwnZutzuodgCA2iekA8ZLL72k3NxcPfzww2W28fl8AV+wOTk51VFatcjPz1dycrLTZZTJ6/XK4/E4XQYAwAEhO4tk7dq1evbZZ/XOO++oefPmZbZLT09Xo0aN/Lf4+PhqrBIAgPAUkj0Y69ev15gxY/Tf//3fGjhwYLlt09LSlJqa6n+ck5NTK0PGuIUrFGlZJbbbtq3LBQVGj1U3KqrMoY9Cn08rpowzejwAQOgJuYCxbt06jR49WuvXr9eQIUMqbG9ZlqxSvnhrm0jLUqTlLvW5KDfDFACA6uVowMjNzdWxY8f8j0+ePKnMzEw1adJErVq1Ulpams6cOaPXX39d0g/DIiNGjNCSJUvUq1cvZWdnS5I8Ho8aNWrkyHsAAAAlOXoOxv79+5WUlOSfYpqamqqkpCTNmjVLkpSVlaVTp075269YsUKXL1/W+PHj1aJFC/9t4sSJjtQPAABK52gPxp133inbtst8fvXq1QGPt2/fXrUFAQAAI0J2FgkAAKi5CBgAAMA4AgYAADCOgAEAAIwjYAAAAOMIGAAAwDgCBgAAMI6AAQAAjCNgAAAA4wgYAADAOAIGAAAwjoABAACMI2AAAADjCBgAAMA4AgYAADCOgAEAAIwjYAAAAOMIGAAAwDgCBgAAMI6AAQAAjCNgAAAA4wgYAADAOAIGAAAwjoABAACMq+t0AQg/tm0rPz+/wjY+n8/4sS3LksvlKreN2+2usA0AoHwEDFS7/Px8JScnO11Gmbxerzwej9NlAEBIY4gEAAAYRw8GHDVu4QpFWlaJ7bZt63JBgfHj1Y2KKnX4o9Dn04op44wfDwDCFQEDjoq0LEVa7lKfi3IzTAEAoYohEgAAYBwBAwAAGEfAAAAAxhEwAACAcQQMAABgHAEDAAAYR8AAAADGETAAAIBxBAwAAGAcAQMAABhHwAAAAMYRMAAAgHEEDAAAYBwBAwAAGEfAAAAAxjkaMHbu3KmUlBTFxcXJ5XJp8+bN5bbPysrS8OHDdcsttygiIkKTJk2qljoBAMC1cTRg5OXlKTExUcuWLQuqvc/nU7NmzTRjxgwlJiZWcXUAAKCy6jp58MGDB2vw4MFBt2/Tpo2WLFkiSVq1alVVlQUAAK6TowGjOvh8Pvl8Pv/jnJwcB6sBACA81PqTPNPT09WoUSP/LT4+3umSAACo9Wp9wEhLS9P333/vv50+fdrpkgAAqPVq/RCJZVmyLMvpMgAACCu1vgcDAABUP0d7MHJzc3Xs2DH/45MnTyozM1NNmjRRq1atlJaWpjNnzuj111/3t8nMzPTve+HCBWVmZioqKkqdOnWq7vIBAEAZHA0Y+/fv11133eV/nJqaKkkaMWKEVq9eraysLJ06dSpgn6SkJP/9AwcOaO3atWrdurW++uqraqkZAABUzNGAceedd8q27TKfX716dYlt5bUHAAA1A+dgAAAA4wgYAADAuFo/TRUwybZt5efnV9im+OqxJliWJZfLVWE7t9sdVDsAqGoEDOAa5OfnKzk52ekyyuT1euXxeJwuAwAYIgEAAObRgwFU0riFKxRZyiqxtm3rckGB0WPVjYoqc+ij0OfTiinjjB4PAK4XAQOopEjLUqTlLvW5KDfDFADCG0MkAADAOAIGAAAwjoABAACMI2AAAADjCBgAAMA4AgYAADCOgAEAAIwjYAAAAOMIGAAAwDgCBgAAMI6AAQAAjCNgAAAA47jYWQizbdt/v9Dnc7CSfyleR/H6AADhhYARwnzFvsxr4uW6fT6f6tWr53QZAAAHMEQCAACMowcjhFmW5b8/buEKRRZ77JRCn8/fm2LVgHoAAM4gYIQwl8vlvx9pWYq03A5WU1Lx+gAA4SXsAwYnSiIc2Lat/Pz8Ctv4quD/gGVZFYZNt9tNIAVqmbAPGJwoiXCQn5+v5ORkp8sok9frlcfjcboMAAZxkicAADAu7HswOFES4aasn3PbtnW5oMD48epGRZU6/FH85xxA7RP2AYMTJRFuyvs5j3IzTAHADIZIAACAcQQMAABgHAEDAAAYR8AAAADGETAAAIBxBAwAAGAcAQMAABhHwAAAAMYRMAAAgHEEDAAAYBwBAwAAGEfAAAAAxoX9xc5Q/Wzb9t8v9PkcrORfitdRvD4AQOUQMFDtfMW+zGvi5bp9Pp/q1avndBkAENIYIgEAAMY52oOxc+dOLViwQAcOHFBWVpY2bdqkBx54oNx9tm/frtTUVP39739XfHy8ZsyYoZEjR1ZLvTDDsiz//XELVyiy2GOnFPp8/t4UqwbUAwChztGAkZeXp8TERI0ePVo/+clPKmx/8uRJDRkyRE888YTeeustZWRkaMyYMWrRooWSk5OroWKY4HK5/PcjLUuRltvBakoqXh8AoHIcDRiDBw/W4MGDg26/fPlyJSQkaOHChZKkjh07ateuXfr1r39NwABqKdu2lZ+fX2Ebn+EThi3LCipsut1uQilQipA6yXPPnj0aOHBgwLbk5GRNmjSpzH18Pl/AL56cnJyqKg9AFcjPz6/Rf0B4vV55PB6nywBqnJA6yTM7O1sxMTEB22JiYpSTk6P/+7//K3Wf9PR0NWrUyH+Lj4+vjlIBAAhrIdWDURlpaWlKTU31P87JySFkACGqrJOCbdvW5YICo8eqGxVV5tBH8ZOCAZQupAJGbGyszp07F7Dt3LlzatiwYZldlJZlMSsAqCXKOyk4ys0wBVCTVGqIZNeuXabrCErv3r2VkZERsG3btm3q3bu3I/UAAIDSVSpg3H333UpISNDTTz+tw4cPV/rgubm5yszMVGZmpqQfpqFmZmbq1KlTkn4Y3njsscf87Z944gmdOHFCU6dO1RdffKGXX35Z77zzjiZPnlzpGgAAgHmVChhnz57VlClTtGPHDnXp0kXdu3fXggUL9I9//OOaXmf//v1KSkpSUlKSJCk1NVVJSUmaNWuWJCkrK8sfNiQpISFBW7Zs0bZt25SYmKiFCxfq1VdfrdFnmAMAEI4qdQ5G06ZNNWHCBE2YMEEnT57U2rVrtWbNGqWlpenf//3f9eGHHwb1OnfeeWe5F5ZavXp1qft8+umnlSkbAABUk+ueppqQkKDp06dr7ty56tq1q3bs2GGiLgAAEMKuK2Ds3r1bv/zlL9WiRQsNHz5cXbp00ZYtW0zVBgAAQlSlhkjS0tK0fv16nT17VoMGDdKSJUt0//33c4lrAPj/nFriXApumXOWOEdVq1TA2Llzp6ZOnaqHHnpITZs2NV0TAIQ8ljhHuKvUEMmQIUNkWVaJcLFq1SrNmzfPSGEAACB0VaoHY8WKFVq7dm2J7Z07d9bPfvYzTZs27boLA4DaojqXOJfKXuacJc5RnSoVMLKzs9WiRYsS25s1a6asrKzrLgoAahOWOEc4qtQQSXx8vHbv3l1i++7duxUXF3fdRQEAgNBWqR6MsWPHatKkSSosLNTdd98tScrIyNDUqVM1ZcoUowUCAIDQU6mA8dRTT+mbb77RL3/5SxX8//FDt9utadOmKS0tzWiBAAAg9FQqYLhcLs2bN08zZ87U559/Lo/Ho/bt23NZdAAAIKmSAeOqBg0a6I477jBVCwAAqCWu+1okAAAAP3ZdPRhAuCl+9d/CKljiuTKK11He1YmBYDm1zHkwS5xLLHMeKggYwDUo/gu1Ji5Y5PP5uCYQrhvLnMMEhkgAAIBx9GAA16D4TKmyln+ubsWXf2YmF0yrzmXOy1riXGKZ81BEwACuQfFffuUt/+wUxqVhGsuco7IYIgEAAMbRgwEAqBWcmv0iBTcDJtxmvxAwAAC1ArNfahaGSAAAgHH0YAAAap3qnP0ilT0DJpxnvxAwAAC1DrNfnMcQCQAAMI6AAQAAjCNgAAAA4wgYAADAOAIGAAAwjoABAACMY5pqLVFYxtK3TlzxEAAAAkYtEa4LuQAAaiaGSAAAgHH0YIQwt9str9dbbpuquHJgMFcNlH6oDwAQnggYIczlcgV1Zb569epVQzUAAPwLAQMIA7Zt++/XlBNxi9dRvD4AtQMBAwgDxYfJauIJwT6fj542oJbhJE8AAGAcPRhAGLAsy39/3MIViiz22CmFPp+/N8WqAfUAMIuAAYSB4rN+Ii1LkVbNmuETzKwkAKGFIRIAAGAcPRgAADjMtm3l5+dX2MbJdY2utaeRgAEAgMPy8/OVnJzsdBll8nq9Qa27VBxDJAAAwLga0YOxbNkyLViwQNnZ2UpMTNTSpUvVs2fPUtsWFhYqPT1da9as0ZkzZ3Trrbdq3rx5uvfee6u5agAAzCtrppcTV8e+nnVzHA8Yb7/9tlJTU7V8+XL16tVLixcvVnJyso4cOaLmzZuXaD9jxgy9+eabWrlypTp06CCv16sHH3xQH3/8sZKSkhx4BwAAmFPeTK8o97UNUzjJ8YCxaNEijR07VqNGjZIkLV++XFu2bNGqVas0ffr0Eu3feOMNPfPMM7rvvvskSb/4xS/0wQcfaOHChXrzzTertXZcv7KWra6KpC6VndZryvLZKIllzoHQ5GjAKCgo0IEDB5SWlubfFhERoYEDB2rPnj2l7uPz+UpcpdPj8WjXrl1lti9+1m1OTo6BymFKTVy2GjULy5wDocnRkzy//vprXblyRTExMQHbY2JilJ2dXeo+ycnJWrRokY4ePaqioiJt27ZNGzduVFZWVqnt09PT1ahRI/8tPj7e+PsAAACBHB8iuVZLlizR2LFj1aFDB7lcLrVr106jRo3SqlWrSm2flpam1NRU/+OcnBxChsPcbre8Xm+5bapivrcU3JzvH/eQwVkscw6EJkcDRtOmTVWnTh2dO3cuYPu5c+cUGxtb6j7NmjXT5s2blZ+fr2+++UZxcXGaPn262rZtW2p7y7L4BVDDuFyuoOZT0+0MiWXOgVDl6BBJVFSUbr/9dmVkZPi3FRUVKSMjQ7179y53X7fbrZYtW+ry5cvasGGD7r///qouFwAABMnxIZLU1FSNGDFCPXr0UM+ePbV48WLl5eX5Z5U89thjatmypdLT0yVJe/fu1ZkzZ9S9e3edOXNGc+bMUVFRkaZOnerk2wAAAMU4HjCGDh2qCxcuaNasWcrOzlb37t31/vvv+0/8PHXqlCIi/tXRkp+frxkzZujEiRNq0KCB7rvvPr3xxhu64YYbHHoHAADgxxwPGJI0YcIETZgwodTntm/fHvC4f//+Onz4cDVUBQAAKotrkQAAAOMIGAAAwLgaMUQCALUNS5wj3BEwAKAKsMQ5wh1DJAAAwDh6MACgCrDEOcIdAQMAqgBLnCPcMUQCAACMI2AAAADjCBgAAMA4AgYAADCOgAEAAIwjYAAAAOOYpgpUUlnLP9u2rcsFBUaPVTcqqsxphTVlGWrUHixzDhMIGEAl1cTlnwETWOYcJjBEAgAAjKMHA7gGbrdbXq+33Da2bQf8BWiCZVlBrbzodtes1SIRmljmHCYQMIBr4HK55PF4KmxH9y1CGcucwwSGSAAAgHH0YBRTnbMCpLJnBtSUs7YBAKgsAkYxNfFsaQAAQhFDJAAAwLiw78FwalaAFNzMAGYFAABCUdgHDGYFAABgXtgHDABA7cAS5zULAQMAUCuwxHnNwkmeAADAOHowAAC1Akuc1ywEDABArcAS5zULQyQAAMA4ejCAMMOS+ACqAwEDCDM18ex6ALUPQyQAAMA4ejCAMMCS+ACqGwEDCAMsiQ+gujFEAgAAjCNgAAAA4wgYAADAOAIGAAAwjoABAACMI2AAAADjmKYKIGRU5zLnZS1xXl4dQGXZtu2/X1N+vorXUby+YBEwAIQMljlHbVV8kbua+HPu8/mueZ0chkgAAIBxNaIHY9myZVqwYIGys7OVmJiopUuXqmfPnmW2X7x4sV555RWdOnVKTZs21X/+538qPT2d5YaBWsipZc6DWeJcYplzmGFZlv/+uIUrFFnssVMKfT5/b4pViXocDxhvv/22UlNTtXz5cvXq1UuLFy9WcnKyjhw5oubNm5dov3btWk2fPl2rVq1Snz599OWXX2rkyJFyuVxatGiRA+8AQFVimXOEg+JhNtKyFGnVrOAaTNj+MceHSBYtWqSxY8dq1KhR6tSpk5YvX6569epp1apVpbb/+OOP1bdvXw0fPlxt2rTRPffco2HDhumTTz6p5soBAEBZHA0YBQUFOnDggAYOHOjfFhERoYEDB2rPnj2l7tOnTx8dOHDAHyhOnDih9957T/fdd1+p7X0+n3JycgJuAACgajk6RPL111/rypUriomJCdgeExOjL774otR9hg8frq+//lr9+vX7YWra5ct64okn9PTTT5faPj09Xc8++6zx2gEAQNkcHyK5Vtu3b9eLL76ol19+WQcPHtTGjRu1ZcsWPffcc6W2T0tL0/fff++/nT59uporBgAg/Djag9G0aVPVqVNH586dC9h+7tw5xcbGlrrPzJkz9eijj2rMmDGSpK5duyovL0/jxo3TM888o4iIwMxkWValzn4FAACV52gPRlRUlG6//XZlZGT4txUVFSkjI0O9e/cudZ9Lly6VCBF16tSRVLmVxgAAgHmOT1NNTU3ViBEj1KNHD/Xs2VOLFy9WXl6eRo0aJUl67LHH1LJlS6Wnp0uSUlJStGjRIiUlJalXr146duyYZs6cqZSUFH/QAAAAznI8YAwdOlQXLlzQrFmzlJ2dre7du+v999/3n/h56tSpgB6LGTNmyOVyacaMGTpz5oyaNWumlJQUvfDCC069BQAA8COOBwxJmjBhgiZMmFDqc9u3bw94XLduXc2ePVuzZ8+uhsoAAEBlhNwsEgAAUPMRMAAAgHE1YogEAGqzwjIuxGbbti4XFBg/Xt2oqFKvHVFWHUBVIGAAQBW7ekVKIJwwRAIAAIyjBwMAqoDb7ZbX6y23jW3b8lXBsIVlWRVeXtvtrlmXA0ftQ8AAgCrgcrnk8XgqbFevXr1qqAaofgyRAAAA4+jBAACUqTpnwJQ1+6W8OlBzETAAAGViBgwqiyESAABgHD0YAIAATs2ACWb2i8QMmFBBwAAABGAGDExgiAQAABhHwAAAAMYRMAAAgHEEDAAAYBwBAwAAGEfAAAAAxjFNFQBQ61TnEudS2cuch/MS5wQMAECtwxLnzmOIBAAAGEcPBgCgVnBqiXMpuGXOw22JcwIGAKBWYInzmoUhEgAAYBwBAwAAGEfAAAAAxhEwAACAcQQMAABgHAEDAAAYxzRVAABqkOpc5rysJc7LqyPo176uvQEAgFG1ZZlzhkgAAIBx9GAAAOAwp5Y5D2aJc6lyy5wTMAAAcFhtXOacIRIAAGAcAQMAABhHwAAAAMYRMAAAgHEEDAAAYBwBAwAAGEfAAAAAxhEwAACAcQQMAABgHAEDAAAYF3ZLhdu2LUnKyclxuBIAAELL1e/Oq9+l5Qm7gHHx4kVJUnx8vMOVAAAQmi5evKhGjRqV28ZlBxNDapGioiKdPXtW0dHRQV1B7lrl5OQoPj5ep0+fVsOGDY2/flUJ1bql0K09VOuWQrf2UK1bCt3aQ7VuKXRrr8q6bdvWxYsXFRcXp4iI8s+yCLsejIiICN10001VfpyGDRuG1A/kVaFatxS6tYdq3VLo1h6qdUuhW3uo1i2Fbu1VVXdFPRdXcZInAAAwjoABAACMI2AYZlmWZs+eLcuynC7lmoRq3VLo1h6qdUuhW3uo1i2Fbu2hWrcUurXXlLrD7iRPAABQ9ejBAAAAxhEwAACAcQQMAABgHAEDAAAYR8AwZOfOnUpJSVFcXJxcLpc2b97sdElBSU9P1x133KHo6Gg1b95cDzzwgI4cOeJ0WRV65ZVX1K1bN/9CMr1799bWrVudLuuazZ07Vy6XS5MmTXK6lArNmTNHLpcr4NahQwenywramTNn9POf/1w33nijPB6Punbtqv379ztdVrnatGlT4t/c5XJp/PjxTpdWoStXrmjmzJlKSEiQx+NRu3bt9NxzzwV1DQunXbx4UZMmTVLr1q3l8XjUp08f7du3z+mySqjoe8e2bc2aNUstWrSQx+PRwIEDdfTo0Wqrj4BhSF5enhITE7Vs2TKnS7kmO3bs0Pjx4/XXv/5V27ZtU2Fhoe655x7l5eU5XVq5brrpJs2dO1cHDhzQ/v37dffdd+v+++/X3//+d6dLC9q+ffv0u9/9Tt26dXO6lKB17txZWVlZ/tuuXbucLiko3377rfr27avIyEht3bpVhw8f1sKFC9W4cWOnSyvXvn37Av69t23bJkl66KGHHK6sYvPmzdMrr7yi3/72t/r88881b948zZ8/X0uXLnW6tAqNGTNG27Zt0xtvvKFDhw7pnnvu0cCBA3XmzBmnSwtQ0ffO/Pnz9Zvf/EbLly/X3r17Vb9+fSUnJys/P796CrRhnCR706ZNTpdRKefPn7cl2Tt27HC6lGvWuHFj+9VXX3W6jKBcvHjRbt++vb1t2za7f//+9sSJE50uqUKzZ8+2ExMTnS6jUqZNm2b369fP6TKu28SJE+127drZRUVFTpdSoSFDhtijR48O2PaTn/zEfuSRRxyqKDiXLl2y69SpY7/77rsB22+77Tb7mWeecaiqiv34e6eoqMiOjY21FyxY4N/23Xff2ZZl2evWrauWmujBQIDvv/9ektSkSROHKwnelStXtH79euXl5al3795OlxOU8ePHa8iQIRo4cKDTpVyTo0ePKi4uTm3bttUjjzyiU6dOOV1SUP70pz+pR48eeuihh9S8eXMlJSVp5cqVTpd1TQoKCvTmm29q9OjRVXKhRtP69OmjjIwMffnll5Kk//mf/9GuXbs0ePBghysr3+XLl3XlyhW53e6A7R6PJ2R67CTp5MmTys7ODvgd06hRI/Xq1Ut79uyplhrC7mJnKFtRUZEmTZqkvn37qkuXLk6XU6FDhw6pd+/eys/PV4MGDbRp0yZ16tTJ6bIqtH79eh08eLBGjumWp1evXlq9erVuvfVWZWVl6dlnn9W//du/6bPPPlN0dLTT5ZXrxIkTeuWVV5Samqqnn35a+/bt05NPPqmoqCiNGDHC6fKCsnnzZn333XcaOXKk06UEZfr06crJyVGHDh1Up04dXblyRS+88IIeeeQRp0srV3R0tHr37q3nnntOHTt2VExMjNatW6c9e/bo5ptvdrq8oGVnZ0uSYmJiArbHxMT4n6tqBAz4jR8/Xp999lnIpPRbb71VmZmZ+v777/WHP/xBI0aM0I4dO2p0yDh9+rQmTpyobdu2lfgLqaYr/pdnt27d1KtXL7Vu3VrvvPOOHn/8cQcrq1hRUZF69OihF198UZKUlJSkzz77TMuXLw+ZgPH73/9egwcPVlxcnNOlBOWdd97RW2+9pbVr16pz587KzMzUpEmTFBcXV+P/zd944w2NHj1aLVu2VJ06dXTbbbdp2LBhOnDggNOlhRSGSCBJmjBhgt5991199NFH1XI5exOioqJ088036/bbb1d6eroSExO1ZMkSp8sq14EDB3T+/Hnddtttqlu3rurWrasdO3boN7/5jerWrasrV644XWLQbrjhBt1yyy06duyY06VUqEWLFiWCZ8eOHUNmiOd///d/9cEHH2jMmDFOlxK0p556StOnT9fPfvYzde3aVY8++qgmT56s9PR0p0urULt27bRjxw7l5ubq9OnT+uSTT1RYWKi2bds6XVrQYmNjJUnnzp0L2H7u3Dn/c1WNgBHmbNvWhAkTtGnTJn344YdKSEhwuqRKKyoqks/nc7qMcg0YMECHDh1SZmam/9ajRw898sgjyszMVJ06dZwuMWi5ubk6fvy4WrRo4XQpFerbt2+J6ddffvmlWrdu7VBF1+a1115T8+bNNWTIEKdLCdqlS5cUERH4FVOnTh0VFRU5VNG1q1+/vlq0aKFvv/1WXq9X999/v9MlBS0hIUGxsbHKyMjwb8vJydHevXur7Vw1hkgMyc3NDfhL7uTJk8rMzFSTJk3UqlUrBysr3/jx47V27Vr98Y9/VHR0tH9srlGjRvJ4PA5XV7a0tDQNHjxYrVq10sWLF7V27Vpt375dXq/X6dLKFR0dXeL8lvr16+vGG2+s8ee9/OpXv1JKSopat26ts2fPavbs2apTp46GDRvmdGkVmjx5svr06aMXX3xRDz/8sD755BOtWLFCK1ascLq0ChUVFem1117TiBEjVLdu6PzKTklJ0QsvvKBWrVqpc+fO+vTTT7Vo0SKNHj3a6dIq5PV6Zdu2br31Vh07dkxPPfWUOnTooFGjRjldWoCKvncmTZqk559/Xu3bt1dCQoJmzpypuLg4PfDAA9VTYLXMVQkDH330kS2pxG3EiBFOl1au0mqWZL/22mtOl1au0aNH261bt7ajoqLsZs2a2QMGDLD/8pe/OF1WpYTKNNWhQ4faLVq0sKOiouyWLVvaQ4cOtY8dO+Z0WUH785//bHfp0sW2LMvu0KGDvWLFCqdLCorX67Ul2UeOHHG6lGuSk5NjT5w40W7VqpXtdrvttm3b2s8884zt8/mcLq1Cb7/9tt22bVs7KirKjo2NtcePH29/9913TpdVQkXfO0VFRfbMmTPtmJgY27Ise8CAAdX6c8Tl2gEAgHGcgwEAAIwjYAAAAOMIGAAAwDgCBgAAMI6AAQAAjCNgAAAA4wgYAADAOAIGABg0Z84cde/e3ekyAMcRMIAwtn37drlcLn333XdOl1KlXC6XNm/e7HQZQFghYACosWzb1uXLl6vlWAUFBdVyHCBcEDCAEFBUVKT58+fr5ptvlmVZatWqlV544YVSeyAyMzPlcrn01VdfSfrhUt8pKSlq3Lix6tevr86dO+u9997TV199pbvuukuS1LhxY7lcLo0cOVKS5PP59OSTT6p58+Zyu93q16+f9u3b5z/G1eN6vV4lJSXJ4/Ho7rvv1vnz57V161Z17NhRDRs21PDhw3Xp0qWA95Genq6EhAR5PB4lJibqD3/4Q4nX3bp1q26//XZZlqVdu3bpzjvv1JNPPqmpU6eqSZMmio2N1Zw5c4L6t2vTpo0k6cEHH5TL5fI/vjqU8eqrryohIUFut1uS9P7776tfv3664YYbdOONN+o//uM/dPz48YDX/Mc//qFhw4apSZMmql+/vnr06KG9e/eWevzjx4+rbdu2mjBhgmzbLvPzAGqb0Lk0HxDG0tLStHLlSv36179Wv379lJWVpS+++CKofcePH6+CggLt3LlT9evX1+HDh9WgQQPFx8drw4YN+ulPf6ojR46oYcOG/ivoTp06VRs2bNCaNWvUunVrzZ8/X8nJyTp27JiaNGnif+05c+bot7/9rerVq6eHH35YDz/8sCzL0tq1a5Wbm6sHH3xQS5cu1bRp0yRJ6enpevPNN7V8+XK1b99eO3fu1M9//nM1a9ZM/fv397/u9OnT9dJLL6lt27Zq3LixJGnNmjVKTU3V3r17tWfPHo0cOVJ9+/bVoEGDyn3/+/btU/PmzfXaa6/p3nvvVZ06dfzPHTt2TBs2bNDGjRv92/Py8pSamqpu3bopNzdXs2bN0oMPPqjMzExFREQoNzdX/fv3V8uWLfWnP/1JsbGxOnjwYKmXIf/b3/6m5ORkPf7443r++efL/TyAWqfaLqsGoFJycnJsy7LslStXlnju6tUUv/32W/+2Tz/91JZknzx50rZt2+7atas9Z86cUl+7tP1zc3PtyMhI+6233vJvKygosOPi4uz58+cH7PfBBx/426Snp9uS7OPHj/u3/dd//ZednJxs27Zt5+fn2/Xq1bM//vjjgBoef/xxe9iwYQGvu3nz5oA2/fv3t/v16xew7Y477rCnTZtW6vv6MUn2pk2bArbNnj3bjoyMtM+fP1/uvhcuXLAl2YcOHbJt27Z/97vf2dHR0fY333xTavvZs2fbiYmJ9u7du+3GjRvbL730UsDz5X0eQG3CEAlQw33++efy+XwaMGBApfZ/8skn9fzzz6tv376aPXu2/va3v5Xb/vjx4yosLFTfvn392yIjI9WzZ099/vnnAW27devmvx8TE6N69eqpbdu2AdvOnz8v6YfegkuXLmnQoEFq0KCB//b666+XGILo0aNHibqKH0uSWrRo4X/tymrdurWaNWsWsO3o0aMaNmyY2rZtq4YNG/qHVE6dOiXphyGopKSkgJ6cHzt16pQGDRqkWbNmacqUKQHPXevnAYQqAgZQw10dtihNRMQP/4Vt2/ZvKywsDGgzZswYnThxQo8++qgOHTqkHj16aOnSpUZqi4yM9N93uVwBj69uuzp0kJubK0nasmWLMjMz/bfDhw8HnIchSfXr1y/3WD9+7coq7TgpKSn65z//qZUrV2rv3r3+cyuungRa3udxVbNmzdSzZ0+tW7dOOTk5Ac9V5ecB1CQEDKCGa9++vTwejzIyMko8d/Wv76ysLP+2zMzMEu3i4+P1xBNPaOPGjZoyZYpWrlwpSYqKipIkXblyxd+2Xbt2ioqK0u7du/3bCgsLtW/fPnXq1KnS76NTp06yLEunTp3SzTffHHCLj4+v9OsGIzIyMuA9luWbb77RkSNHNGPGDA0YMEAdO3bUt99+G9CmW7duyszM1D//+c8yX8fj8ejdd9+V2+1WcnKyLl68GPB8WZ8HUJsQMIAazu12a9q0aZo6dap/OOGvf/2rfv/73/u/nOfMmaOjR49qy5YtWrhwYcD+kyZNktfr1cmTJ3Xw4EF99NFH6tixo6QfhghcLpfeffddXbhwQbm5uapfv75+8Ytf6KmnntL777+vw4cPa+zYsbp06ZIef/zxSr+P6Oho/epXv9LkyZO1Zs0aHT9+XAcPHtTSpUu1Zs2a6/o3qkibNm2UkZGh7OzsEoGhuMaNG+vGG2/UihUrdOzYMX344YdKTU0NaDNs2DDFxsbqgQce0O7du3XixAlt2LBBe/bsCWhXv359bdmyRXXr1tXgwYP9PTjlfR5AbULAAELAzJkzNWXKFM2aNUsdO3bU0KFDdf78eUVGRmrdunX64osv1K1bN82bN88/W+GqK1euaPz48erYsaPuvfde3XLLLXr55ZclSS1bttSzzz6r6dOnKyYmRhMmTJAkzZ07Vz/96U/16KOP6rbbbtOxY8fk9Xr9Mzoq67nnntPMmTOVnp7ur2fLli1KSEi4rtetyMKFC7Vt2zbFx8crKSmpzHYRERFav369Dhw4oC5dumjy5MlasGBBQJuoqCj95S9/UfPmzXXfffepa9eumjt3bsDslKsaNGigrVu3yrZtDRkyRHl5eeV+HkBt4rKLD94CAAAYQA8GAAAwjoABIKS99dZbAdNei986d+7sdHlA2GKIBEBIu3jxos6dO1fqc5GRkWrdunU1VwRAImAAAIAqwBAJAAAwjoABAACMI2AAAADjCBgAAMA4AgYAADCOgAEAAIwjYAAAAOMIGAAAwLj/B53HPE2Rzs2kAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "threshold = 10\n",
    "plot_df = user_summary[user_summary['customern_tracks'] <= threshold]\n",
    "plt.figure(figsize=(6,4))\n",
    "sns.boxplot(x='customern_tracks',\n",
    "            y='cv',\n",
    "            data=plot_df,\n",
    "            boxprops={'facecolor':'lightblue'},\n",
    "            showfliers=False,\n",
    "            whis=0)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "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": "3.9.6"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}