Lab_2_Data_Exploration_and_BQML_Model_Training.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "281c2f8a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Copyright 2023 Google LLC\n",
    "#\n",
    "# Licensed under the Apache License, Version 2.0 (the \"License\");\n",
    "# you may not use this file except in compliance with the License.\n",
    "# You may obtain a copy of the License at\n",
    "#\n",
    "#     https://www.apache.org/licenses/LICENSE-2.0\n",
    "#\n",
    "# Unless required by applicable law or agreed to in writing, software\n",
    "# distributed under the License is distributed on an \"AS IS\" BASIS,\n",
    "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
    "# See the License for the specific language governing permissions and\n",
    "# limitations under the License."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6693e950-502b-4f4e-aaac-34b44b015bff",
   "metadata": {
    "id": "6693e950-502b-4f4e-aaac-34b44b015bff"
   },
   "source": [
    "# **Lab 2:** Data Exploration and Model Training"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "101df100-8730-4286-9a02-2819e1e0855c",
   "metadata": {
    "id": "101df100-8730-4286-9a02-2819e1e0855c"
   },
   "source": [
    "### **Step 1:** Preparation and Authentication\n",
    "Set the right parameters and authenticate your user."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "988eb9c8-1d06-4c2f-8947-2a843177201b",
   "metadata": {
    "id": "988eb9c8-1d06-4c2f-8947-2a843177201b"
   },
   "outputs": [],
   "source": [
    "project_id =  \"\" \n",
    "team_name  = \"\" \n",
    "location   = \"us\" #This is currently necessary\n",
    "dataset_name = \"datathon_ds_{}\".format(team_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "209ec9e2-e26c-4f02-84ec-206bcb1a3fcb",
   "metadata": {
    "id": "209ec9e2-e26c-4f02-84ec-206bcb1a3fcb"
   },
   "outputs": [],
   "source": [
    "from google.cloud import bigquery\n",
    "from google.cloud.bigquery import Client, QueryJobConfig\n",
    "\n",
    "client = bigquery.Client(project=project_id)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1jjaJyFyG00m",
   "metadata": {
    "id": "1jjaJyFyG00m"
   },
   "source": [
    "### **Step 2:** Data Exploration\n",
    "Get a feeling of the data and explore different facets and features"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4dc1d19a-e397-401d-85b8-5e3f7a441e7f",
   "metadata": {
    "id": "4dc1d19a-e397-401d-85b8-5e3f7a441e7f"
   },
   "source": [
    "To remember: This is how the final training dataset looks like\n",
    "\n",
    "| User Demographic Data | User Behavioral Data | Label\n",
    "| --- | --- | --- |\n",
    "| country | cnt_user_engagement | churned\n",
    "| operating_system | cnt_level_start_quickplay |  |\n",
    "| language | cnt_level_end_quickplay |  |\n",
    "|  |  cnt_level_complete_quickplay |  |\n",
    "|  |  cnt_level_reset_quickplay |  |\n",
    "|  |  cnt_post_score |  |\n",
    "|  |  cnt_spend_virtual_currency |  |\n",
    "|  |  cnt_ad_reward |  |\n",
    "|  |  cnt_challenge_a_friend |  |\n",
    "|  |  cnt_completed_5_levels |  |\n",
    "|  |  cnt_use_extra_steps |  |\n",
    "|  |  user_first_engagement |  | **bold text**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "cf4e64a1-44d3-45ef-b310-5194d987a7c1",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 265
    },
    "executionInfo": {
     "elapsed": 7998,
     "status": "ok",
     "timestamp": 1656934987706,
     "user": {
      "displayName": "Amrit Raj",
      "userId": "11987938039820858613"
     },
     "user_tz": -120
    },
    "id": "cf4e64a1-44d3-45ef-b310-5194d987a7c1",
    "outputId": "b37dee16-8d19-4d00-c380-7f64fe609261"
   },
   "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_pseudo_id</th>\n",
       "      <th>country</th>\n",
       "      <th>operating_system</th>\n",
       "      <th>language</th>\n",
       "      <th>cnt_user_engagement</th>\n",
       "      <th>cnt_level_start_quickplay</th>\n",
       "      <th>cnt_level_end_quickplay</th>\n",
       "      <th>cnt_level_complete_quickplay</th>\n",
       "      <th>cnt_level_reset_quickplay</th>\n",
       "      <th>cnt_post_score</th>\n",
       "      <th>cnt_spend_virtual_currency</th>\n",
       "      <th>cnt_ad_reward</th>\n",
       "      <th>cnt_challenge_a_friend</th>\n",
       "      <th>cnt_completed_5_levels</th>\n",
       "      <th>cnt_use_extra_steps</th>\n",
       "      <th>user_first_engagement</th>\n",
       "      <th>churned</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4F349267084BB476E7D496AA3422388B</td>\n",
       "      <td>United States</td>\n",
       "      <td>ANDROID</td>\n",
       "      <td>en-us</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1529769862166004</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>F94E005EB49AFB8DA0A78030D4ABB44B</td>\n",
       "      <td>United States</td>\n",
       "      <td>ANDROID</td>\n",
       "      <td>en-us</td>\n",
       "      <td>8</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1530299690915001</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>56C7AB443CB21553FF8690752E58BC79</td>\n",
       "      <td>United States</td>\n",
       "      <td>ANDROID</td>\n",
       "      <td>en-us</td>\n",
       "      <td>22</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>10</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1533950048297006</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>D9A79CD45804EB82483DAB809D7B417C</td>\n",
       "      <td>Germany</td>\n",
       "      <td>ANDROID</td>\n",
       "      <td>de-de</td>\n",
       "      <td>7</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1531053918021001</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0CAAA1B6F161BA4D363EB7A2DEF64A08</td>\n",
       "      <td>United States</td>\n",
       "      <td>ANDROID</td>\n",
       "      <td>en-us</td>\n",
       "      <td>64</td>\n",
       "      <td>32</td>\n",
       "      <td>32</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>28</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1529334029639003</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     user_pseudo_id        country operating_system language  \\\n",
       "0  4F349267084BB476E7D496AA3422388B  United States          ANDROID    en-us   \n",
       "1  F94E005EB49AFB8DA0A78030D4ABB44B  United States          ANDROID    en-us   \n",
       "2  56C7AB443CB21553FF8690752E58BC79  United States          ANDROID    en-us   \n",
       "3  D9A79CD45804EB82483DAB809D7B417C        Germany          ANDROID    de-de   \n",
       "4  0CAAA1B6F161BA4D363EB7A2DEF64A08  United States          ANDROID    en-us   \n",
       "\n",
       "   cnt_user_engagement  cnt_level_start_quickplay  cnt_level_end_quickplay  \\\n",
       "0                    7                          2                        1   \n",
       "1                    8                          4                        2   \n",
       "2                   22                          0                        0   \n",
       "3                    7                          2                        1   \n",
       "4                   64                         32                       32   \n",
       "\n",
       "   cnt_level_complete_quickplay  cnt_level_reset_quickplay  cnt_post_score  \\\n",
       "0                             0                          0               0   \n",
       "1                             2                          1               2   \n",
       "2                             0                          0              10   \n",
       "3                             1                          0               1   \n",
       "4                             0                          0              28   \n",
       "\n",
       "   cnt_spend_virtual_currency  cnt_ad_reward  cnt_challenge_a_friend  \\\n",
       "0                           0              0                       0   \n",
       "1                           0              0                       0   \n",
       "2                           0              0                       0   \n",
       "3                           1              0                       0   \n",
       "4                           0              0                       0   \n",
       "\n",
       "   cnt_completed_5_levels  cnt_use_extra_steps  user_first_engagement  churned  \n",
       "0                       0                    0       1529769862166004        0  \n",
       "1                       0                    0       1530299690915001        0  \n",
       "2                       0                    0       1533950048297006        1  \n",
       "3                       0                    1       1531053918021001        0  \n",
       "4                       0                    0       1529334029639003        0  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "query = f\"\"\"SELECT * FROM `{dataset_name}.cc_train_dataset`\"\"\"\n",
    "job = client.query(query)\n",
    "df = job.to_dataframe()\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "N6bPe48HHdiF",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 476
    },
    "executionInfo": {
     "elapsed": 22,
     "status": "ok",
     "timestamp": 1656924726894,
     "user": {
      "displayName": "Amrit Raj",
      "userId": "11987938039820858613"
     },
     "user_tz": -120
    },
    "id": "N6bPe48HHdiF",
    "outputId": "dd1801cc-3ca2-4a0b-d419-ecf787f80ad2"
   },
   "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_pseudo_id</th>\n",
       "      <th>country</th>\n",
       "      <th>operating_system</th>\n",
       "      <th>language</th>\n",
       "      <th>cnt_user_engagement</th>\n",
       "      <th>cnt_level_start_quickplay</th>\n",
       "      <th>cnt_level_end_quickplay</th>\n",
       "      <th>cnt_level_complete_quickplay</th>\n",
       "      <th>cnt_level_reset_quickplay</th>\n",
       "      <th>cnt_post_score</th>\n",
       "      <th>cnt_spend_virtual_currency</th>\n",
       "      <th>cnt_ad_reward</th>\n",
       "      <th>cnt_challenge_a_friend</th>\n",
       "      <th>cnt_completed_5_levels</th>\n",
       "      <th>cnt_use_extra_steps</th>\n",
       "      <th>user_first_engagement</th>\n",
       "      <th>churned</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>6386</td>\n",
       "      <td>6386</td>\n",
       "      <td>6183</td>\n",
       "      <td>6386</td>\n",
       "      <td>6386.000000</td>\n",
       "      <td>6386.000000</td>\n",
       "      <td>6386.000000</td>\n",
       "      <td>6386.000000</td>\n",
       "      <td>6386.000000</td>\n",
       "      <td>6386.000000</td>\n",
       "      <td>6386.000000</td>\n",
       "      <td>6386.000000</td>\n",
       "      <td>6386.000000</td>\n",
       "      <td>6386.000000</td>\n",
       "      <td>6386.000000</td>\n",
       "      <td>6.386000e+03</td>\n",
       "      <td>6386.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>6386</td>\n",
       "      <td>118</td>\n",
       "      <td>2</td>\n",
       "      <td>138</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>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>4F349267084BB476E7D496AA3422388B</td>\n",
       "      <td>United States</td>\n",
       "      <td>IOS</td>\n",
       "      <td>en-us</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>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>1</td>\n",
       "      <td>3506</td>\n",
       "      <td>3152</td>\n",
       "      <td>3837</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>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>30.538209</td>\n",
       "      <td>7.765581</td>\n",
       "      <td>4.938459</td>\n",
       "      <td>2.093642</td>\n",
       "      <td>1.941121</td>\n",
       "      <td>4.872847</td>\n",
       "      <td>0.447072</td>\n",
       "      <td>0.059192</td>\n",
       "      <td>0.018948</td>\n",
       "      <td>0.133104</td>\n",
       "      <td>0.401660</td>\n",
       "      <td>-1.356218e+15</td>\n",
       "      <td>0.234262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>52.468331</td>\n",
       "      <td>26.602648</td>\n",
       "      <td>16.759185</td>\n",
       "      <td>10.233683</td>\n",
       "      <td>16.480642</td>\n",
       "      <td>14.401616</td>\n",
       "      <td>1.877727</td>\n",
       "      <td>0.675517</td>\n",
       "      <td>0.251139</td>\n",
       "      <td>0.363327</td>\n",
       "      <td>1.757354</td>\n",
       "      <td>2.995785e+12</td>\n",
       "      <td>0.423570</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.528787e+15</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.529774e+15</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.531470e+15</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>34.000000</td>\n",
       "      <td>7.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>5.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.535155e+15</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>928.000000</td>\n",
       "      <td>832.000000</td>\n",
       "      <td>655.000000</td>\n",
       "      <td>595.000000</td>\n",
       "      <td>751.000000</td>\n",
       "      <td>595.000000</td>\n",
       "      <td>52.000000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>9.000000</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>52.000000</td>\n",
       "      <td>1.538620e+15</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          user_pseudo_id        country operating_system  \\\n",
       "count                               6386           6386             6183   \n",
       "unique                              6386            118                2   \n",
       "top     4F349267084BB476E7D496AA3422388B  United States              IOS   \n",
       "freq                                   1           3506             3152   \n",
       "mean                                 NaN            NaN              NaN   \n",
       "std                                  NaN            NaN              NaN   \n",
       "min                                  NaN            NaN              NaN   \n",
       "25%                                  NaN            NaN              NaN   \n",
       "50%                                  NaN            NaN              NaN   \n",
       "75%                                  NaN            NaN              NaN   \n",
       "max                                  NaN            NaN              NaN   \n",
       "\n",
       "       language  cnt_user_engagement  cnt_level_start_quickplay  \\\n",
       "count      6386          6386.000000                6386.000000   \n",
       "unique      138                  NaN                        NaN   \n",
       "top       en-us                  NaN                        NaN   \n",
       "freq       3837                  NaN                        NaN   \n",
       "mean        NaN            30.538209                   7.765581   \n",
       "std         NaN            52.468331                  26.602648   \n",
       "min         NaN             1.000000                   0.000000   \n",
       "25%         NaN             6.000000                   0.000000   \n",
       "50%         NaN            14.000000                   2.000000   \n",
       "75%         NaN            34.000000                   7.000000   \n",
       "max         NaN           928.000000                 832.000000   \n",
       "\n",
       "        cnt_level_end_quickplay  cnt_level_complete_quickplay  \\\n",
       "count               6386.000000                   6386.000000   \n",
       "unique                      NaN                           NaN   \n",
       "top                         NaN                           NaN   \n",
       "freq                        NaN                           NaN   \n",
       "mean                   4.938459                      2.093642   \n",
       "std                   16.759185                     10.233683   \n",
       "min                    0.000000                      0.000000   \n",
       "25%                    0.000000                      0.000000   \n",
       "50%                    1.000000                      0.000000   \n",
       "75%                    4.000000                      1.000000   \n",
       "max                  655.000000                    595.000000   \n",
       "\n",
       "        cnt_level_reset_quickplay  cnt_post_score  cnt_spend_virtual_currency  \\\n",
       "count                 6386.000000     6386.000000                 6386.000000   \n",
       "unique                        NaN             NaN                         NaN   \n",
       "top                           NaN             NaN                         NaN   \n",
       "freq                          NaN             NaN                         NaN   \n",
       "mean                     1.941121        4.872847                    0.447072   \n",
       "std                     16.480642       14.401616                    1.877727   \n",
       "min                      0.000000        0.000000                    0.000000   \n",
       "25%                      0.000000        0.000000                    0.000000   \n",
       "50%                      0.000000        1.000000                    0.000000   \n",
       "75%                      0.000000        5.000000                    0.000000   \n",
       "max                    751.000000      595.000000                   52.000000   \n",
       "\n",
       "        cnt_ad_reward  cnt_challenge_a_friend  cnt_completed_5_levels  \\\n",
       "count     6386.000000             6386.000000             6386.000000   \n",
       "unique            NaN                     NaN                     NaN   \n",
       "top               NaN                     NaN                     NaN   \n",
       "freq              NaN                     NaN                     NaN   \n",
       "mean         0.059192                0.018948                0.133104   \n",
       "std          0.675517                0.251139                0.363327   \n",
       "min          0.000000                0.000000                0.000000   \n",
       "25%          0.000000                0.000000                0.000000   \n",
       "50%          0.000000                0.000000                0.000000   \n",
       "75%          0.000000                0.000000                0.000000   \n",
       "max         29.000000                9.000000                4.000000   \n",
       "\n",
       "        cnt_use_extra_steps  user_first_engagement      churned  \n",
       "count           6386.000000           6.386000e+03  6386.000000  \n",
       "unique                  NaN                    NaN          NaN  \n",
       "top                     NaN                    NaN          NaN  \n",
       "freq                    NaN                    NaN          NaN  \n",
       "mean               0.401660          -1.356218e+15     0.234262  \n",
       "std                1.757354           2.995785e+12     0.423570  \n",
       "min                0.000000           1.528787e+15     0.000000  \n",
       "25%                0.000000           1.529774e+15     0.000000  \n",
       "50%                0.000000           1.531470e+15     0.000000  \n",
       "75%                0.000000           1.535155e+15     0.000000  \n",
       "max               52.000000           1.538620e+15     1.000000  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe(include='all')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "dfd30813-76c3-4b69-8e71-3ec148ab9065",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 266
    },
    "executionInfo": {
     "elapsed": 481,
     "status": "ok",
     "timestamp": 1656924731145,
     "user": {
      "displayName": "Amrit Raj",
      "userId": "11987938039820858613"
     },
     "user_tz": -120
    },
    "id": "dfd30813-76c3-4b69-8e71-3ec148ab9065",
    "outputId": "7fadd15a-40c4-44d3-fc69-02f5d291b127"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:ylabel='operating_system'>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig, axes = plt.subplots(nrows=1, ncols=2)\n",
    "plt.subplots_adjust(right=2.0)\n",
    "\n",
    "# which country do users belong to?\n",
    "df.country.value_counts(sort=True)[:5].plot(kind='pie',ax=axes[0])\n",
    "# which operating system is widely used?\n",
    "df.operating_system.value_counts().plot(kind='pie', ax=axes[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "c0e57848-c796-41f0-a410-541b9a8bc962",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 266
    },
    "executionInfo": {
     "elapsed": 472,
     "status": "ok",
     "timestamp": 1656924734420,
     "user": {
      "displayName": "Amrit Raj",
      "userId": "11987938039820858613"
     },
     "user_tz": -120
    },
    "id": "c0e57848-c796-41f0-a410-541b9a8bc962",
    "outputId": "ffab019b-52d8-415e-93ff-5e0f042d6fe7"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:ylabel='churned'>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# How many users churned?\n",
    "df.churned.value_counts().plot(kind='pie')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "433e28c9-cd11-418b-8731-61ba2f61660c",
   "metadata": {
    "id": "433e28c9-cd11-418b-8731-61ba2f61660c"
   },
   "source": [
    "## **Step 3:** Train Propensity Churn Model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d74ce070-0b26-4aea-b164-231fdbd87322",
   "metadata": {
    "id": "d74ce070-0b26-4aea-b164-231fdbd87322"
   },
   "source": [
    "Which model should we use for our use case?\n",
    "* List of classification [models](https://cloud.google.com/bigquery-ml/docs/reference/standard-sql/bigqueryml-syntax-create-glm) in BQ to explore (Tip: Keep in mind the training time for more complex models ⏰)\n",
    "  *  Logistic Regression\n",
    "  *  Boosted Trees\n",
    "  *  Deep Neural Networks\n",
    "  *  Wide and Deep Networks\n",
    "  *  AutoML tables\n",
    "\n",
    "Feel free to switch the BQ UI to run your queries there.\n",
    "Save your progress by saving your queries through the editor.\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "Jpmc5_QW1MPx",
   "metadata": {
    "id": "Jpmc5_QW1MPx"
   },
   "source": [
    "**Now it is your turn to train a model usign BQML !!**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5qfOrzcH1SV6",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 49
    },
    "executionInfo": {
     "elapsed": 98014,
     "status": "ok",
     "timestamp": 1656925014844,
     "user": {
      "displayName": "Amrit Raj",
      "userId": "11987938039820858613"
     },
     "user_tz": -120
    },
    "id": "5qfOrzcH1SV6",
    "outputId": "8140bb36-118f-4e17-a7fd-54490836ed26"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-2ad31470-f814-4e4e-a173-47fe963e7354\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-2ad31470-f814-4e4e-a173-47fe963e7354')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-2ad31470-f814-4e4e-a173-47fe963e7354 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-2ad31470-f814-4e4e-a173-47fe963e7354');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: []\n",
       "Index: []"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%bigquery --project $project_id\n",
    "\n",
    "CREATE MODEL `<model-name>`\n",
    "OPTIONS(MODEL_TYPE='<type>',\n",
    "        INPUT_LABEL_COLS = ['<label>'])\n",
    "AS\n",
    "SELECT * EXCEPT (user_pseudo_id)\n",
    "FROM `<train-dataset>`"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "OMNBrx0fD4Iu",
   "metadata": {
    "id": "OMNBrx0fD4Iu"
   },
   "source": [
    "### **Feature Engineering**: How about we create new features from the user_first_engagement timestamp field?\n",
    "\n",
    "Let's use the [TRANSFORM](https://cloud.google.com/bigquery-ml/docs/bigqueryml-transform) method in BigQuery ML to extract month, day of year, day of week and hour from the user_first_enagagement field.\n",
    "\n",
    "Using the TRANSFORM clause, you can specify all preprocessing during model creation. The preprocessing is **automatically** applied during the prediction and evaluation phases of machine learning 😎"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "N6qhC-erIyGG",
   "metadata": {
    "id": "N6qhC-erIyGG"
   },
   "source": [
    "Hint: Use the [EXTRACT](https://cloud.google.com/bigquery/docs/reference/standard-sql/date_functions#extract) and [TIMESTAMP_MICROS](https://cloud.google.com/bigquery/docs/reference/standard-sql/timestamp_functions#timestamp_micros) functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "FdZrvf4YFLhq",
   "metadata": {
    "id": "FdZrvf4YFLhq"
   },
   "outputs": [],
   "source": [
    "%%bigquery --project $project_id\n",
    "CREATE MODEL `<model-name>`\n",
    "\n",
    "TRANSFORM(\n",
    "  EXTRACT(<..>) as month,\n",
    "  EXTRACT(<..>) as julianday,\n",
    "  EXTRACT(<..>) as dayofweek,\n",
    "  EXTRACT(<..>) as hour,\n",
    "  * EXCEPT(user_first_engagement, user_pseudo_id)\n",
    ")\n",
    "\n",
    "OPTIONS(MODEL_TYPE='<model-type>',\n",
    "        INPUT_LABEL_COLS = ['<label>'])\n",
    "AS\n",
    "SELECT *\n",
    "FROM `<train-dataset>`"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "edd5fa1b-0886-4d92-ab7c-32e614af87fc",
   "metadata": {
    "id": "edd5fa1b-0886-4d92-ab7c-32e614af87fc"
   },
   "source": [
    "## Step 5: Evaluate Model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ik4jjBigbAXL",
   "metadata": {
    "id": "ik4jjBigbAXL"
   },
   "source": [
    "**Let**'s evaluate our model performance on the previously created evaluation view. Evaluation datasets are a useful way to tune your models accuracy before testing in production."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "uFnloB8bCRpN",
   "metadata": {
    "id": "uFnloB8bCRpN"
   },
   "outputs": [],
   "source": [
    "%%bigquery --project $project_id\n",
    "SELECT\n",
    "  *\n",
    "FROM\n",
    "  ML.EVALUATE(MODEL `<model-name>`,\n",
    "    (\n",
    "    SELECT * EXCEPT(user_pseudo_id)\n",
    "    FROM\n",
    "      `<eval-dataset>`)\n",
    "    )"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "provenance": [
    {
     "file_id": "1vbWqTaE2kgNjYLhxOkD_U8pCM6xdkoxm",
     "timestamp": 1656942192443
    },
    {
     "file_id": "1rcZkkOxAcnNCydTXfFxdBRbw5Z5zncl2",
     "timestamp": 1656923027322
    }
   ]
  },
  "environment": {
   "kernel": "python3",
   "name": "common-cpu.m107",
   "type": "gcloud",
   "uri": "gcr.io/deeplearning-platform-release/base-cpu:m107"
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.8.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}