Model_creation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.callbacks import ReduceLROnPlateau, EarlyStopping\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense,InputLayer, Dropout\n",
    "from tensorflow.keras import metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>id</th>\n",
       "      <th>created_at</th>\n",
       "      <th>campaign_var_1</th>\n",
       "      <th>campaign_var_2</th>\n",
       "      <th>products_purchased</th>\n",
       "      <th>signup_date</th>\n",
       "      <th>user_activity_var_1</th>\n",
       "      <th>user_activity_var_2</th>\n",
       "      <th>user_activity_var_3</th>\n",
       "      <th>user_activity_var_4</th>\n",
       "      <th>user_activity_var_5</th>\n",
       "      <th>user_activity_var_6</th>\n",
       "      <th>user_activity_var_7</th>\n",
       "      <th>user_activity_var_8</th>\n",
       "      <th>user_activity_var_9</th>\n",
       "      <th>user_activity_var_10</th>\n",
       "      <th>user_activity_var_11</th>\n",
       "      <th>user_activity_var_12</th>\n",
       "      <th>buy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>2021-01-01</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2020-09-24</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>2021-01-01</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2020-09-19</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>2021-01-01</td>\n",
       "      <td>9</td>\n",
       "      <td>3</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2021-08-11</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>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>2021-01-01</td>\n",
       "      <td>6</td>\n",
       "      <td>7</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2017-10-04</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>2021-01-01</td>\n",
       "      <td>4</td>\n",
       "      <td>6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2020-06-08</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id  created_at  campaign_var_1  campaign_var_2  products_purchased  \\\n",
       "0   1  2021-01-01               1               2                 2.0   \n",
       "1   2  2021-01-01               2               1                 2.0   \n",
       "2   3  2021-01-01               9               3                 3.0   \n",
       "3   4  2021-01-01               6               7                 2.0   \n",
       "4   5  2021-01-01               4               6                 NaN   \n",
       "\n",
       "  signup_date  user_activity_var_1  user_activity_var_2  user_activity_var_3  \\\n",
       "0  2020-09-24                    0                    0                    0   \n",
       "1  2020-09-19                    1                    0                    1   \n",
       "2  2021-08-11                    1                    0                    0   \n",
       "3  2017-10-04                    0                    0                    0   \n",
       "4  2020-06-08                    0                    0                    0   \n",
       "\n",
       "   user_activity_var_4  user_activity_var_5  user_activity_var_6  \\\n",
       "0                    0                    0                    0   \n",
       "1                    0                    0                    0   \n",
       "2                    0                    0                    0   \n",
       "3                    0                    0                    0   \n",
       "4                    0                    0                    0   \n",
       "\n",
       "   user_activity_var_7  user_activity_var_8  user_activity_var_9  \\\n",
       "0                    1                    1                    0   \n",
       "1                    1                    0                    0   \n",
       "2                    0                    0                    0   \n",
       "3                    0                    0                    0   \n",
       "4                    1                    0                    0   \n",
       "\n",
       "   user_activity_var_10  user_activity_var_11  user_activity_var_12  buy  \n",
       "0                     0                     0                     0    0  \n",
       "1                     0                     0                     0    0  \n",
       "2                     0                     0                     0    0  \n",
       "3                     0                     0                     0    0  \n",
       "4                     0                     1                     0    0  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load training data\n",
    "\n",
    "df = pd.read_csv(\"train_wn75k28.csv\")\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of Training Data  (39161, 19)\n"
     ]
    }
   ],
   "source": [
    "print(\"Shape of Training Data \",df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>user_activity_var_1</th>\n",
       "      <th>user_activity_var_2</th>\n",
       "      <th>user_activity_var_4</th>\n",
       "      <th>user_activity_var_5</th>\n",
       "      <th>user_activity_var_8</th>\n",
       "      <th>user_activity_var_9</th>\n",
       "      <th>user_activity_var_11</th>\n",
       "      <th>buy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_activity_var_1  user_activity_var_2  user_activity_var_4  \\\n",
       "0                    0                    0                    0   \n",
       "1                    1                    0                    0   \n",
       "2                    1                    0                    0   \n",
       "3                    0                    0                    0   \n",
       "4                    0                    0                    0   \n",
       "\n",
       "   user_activity_var_5  user_activity_var_8  user_activity_var_9  \\\n",
       "0                    0                    1                    0   \n",
       "1                    0                    0                    0   \n",
       "2                    0                    0                    0   \n",
       "3                    0                    0                    0   \n",
       "4                    0                    0                    0   \n",
       "\n",
       "   user_activity_var_11  buy  \n",
       "0                     0    0  \n",
       "1                     0    0  \n",
       "2                     0    0  \n",
       "3                     0    0  \n",
       "4                     1    0  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#data = df[['campaign_var_1', 'campaign_var_2','user_activity_var_1','user_activity_var_2','user_activity_var_3','user_activity_var_4','user_activity_var_5','user_activity_var_6','user_activity_var_7','user_activity_var_8','user_activity_var_9','user_activity_var_10','user_activity_var_11','user_activity_var_12','buy']]\n",
    "\n",
    "\n",
    "data = df[['user_activity_var_1','user_activity_var_2','user_activity_var_4','user_activity_var_5','user_activity_var_8','user_activity_var_9','user_activity_var_11','buy']]\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of pos examples  (1998, 8)\n",
      "Shape of neg_examples  (37163, 8)\n"
     ]
    }
   ],
   "source": [
    "neg_examples = data[data['buy']==0]\n",
    "pos_examples = data[data['buy']==1]\n",
    "\n",
    "# Shuffle Neg Examples\n",
    "pos_examples = pos_examples.sample(frac=1).reset_index(drop=True)\n",
    "neg_examples = neg_examples.sample(frac=1).reset_index(drop=True)\n",
    "\n",
    "print(\"Shape of pos examples \", pos_examples.shape)\n",
    "print(\"Shape of neg_examples \", neg_examples.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## CREATION OF TRAIN AND VALIDATION SETS"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Using 80:20 split ratio.\n",
    "- Training set contains 80% of both pos and neg examples each.\n",
    "- Validation set contains 20% of both pos and neg examples each."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "pos_size = int(len(pos_examples) * 0.8)\n",
    "neg_size = int(len(neg_examples) * 0.8)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. Creating validation data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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_activity_var_1</th>\n",
       "      <th>user_activity_var_2</th>\n",
       "      <th>user_activity_var_4</th>\n",
       "      <th>user_activity_var_5</th>\n",
       "      <th>user_activity_var_8</th>\n",
       "      <th>user_activity_var_9</th>\n",
       "      <th>user_activity_var_11</th>\n",
       "      <th>buy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\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",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_activity_var_1  user_activity_var_2  user_activity_var_4  \\\n",
       "0                    0                    0                    0   \n",
       "1                    0                    0                    0   \n",
       "2                    0                    0                    0   \n",
       "3                    1                    0                    0   \n",
       "4                    1                    0                    0   \n",
       "\n",
       "   user_activity_var_5  user_activity_var_8  user_activity_var_9  \\\n",
       "0                    0                    0                    0   \n",
       "1                    0                    1                    0   \n",
       "2                    0                    0                    0   \n",
       "3                    0                    0                    0   \n",
       "4                    0                    0                    0   \n",
       "\n",
       "   user_activity_var_11  buy  \n",
       "0                     0    0  \n",
       "1                     0    0  \n",
       "2                     0    0  \n",
       "3                     0    0  \n",
       "4                     0    0  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Creating validation data frame\n",
    "\n",
    "val_pos = pos_examples.iloc[pos_size:len(pos_examples)]\n",
    "val_neg = neg_examples.iloc[neg_size:len(neg_examples)]\n",
    "\n",
    "val_data = pd.concat([val_pos, val_neg], ignore_index=True,axis=0)\n",
    "val_data = val_data.sample(frac=1).reset_index(drop=True)\n",
    "val_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7833, 8)\n"
     ]
    }
   ],
   "source": [
    "print(val_data.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. Creating Training set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1598, 8)\n",
      "(29730, 8)\n"
     ]
    }
   ],
   "source": [
    "train_pos = pos_examples.iloc[0:pos_size]\n",
    "train_neg = neg_examples.iloc[0:neg_size]\n",
    "\n",
    "print(train_pos.shape)\n",
    "print(train_neg.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Since the data set is imbalanced. Using hybrid sampling technique\n",
    "\n",
    "size_pos_train_examples = 3000\n",
    "size_neg_train_examples = 10000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3000, 8)\n"
     ]
    }
   ],
   "source": [
    "# Generating random numbers with replacement\n",
    "pos_rand_index = np.random.randint(low=0, high=train_pos.shape[0], size=size_pos_train_examples)\n",
    "\n",
    "train_pos_data = train_pos.iloc[pos_rand_index]\n",
    "print(train_pos_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10000, 8)\n"
     ]
    }
   ],
   "source": [
    "rng = np.random.default_rng()\n",
    "neg_rand_index = rng.choice(train_neg.shape[0], size=size_neg_train_examples, replace=False)\n",
    "\n",
    "train_neg_data = train_neg.iloc[neg_rand_index]\n",
    "print(train_neg_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(13000, 8)\n"
     ]
    }
   ],
   "source": [
    "train_data = pd.concat([train_pos_data, train_neg_data], ignore_index=True,axis=0)\n",
    "print(train_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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_activity_var_1</th>\n",
       "      <th>user_activity_var_2</th>\n",
       "      <th>user_activity_var_4</th>\n",
       "      <th>user_activity_var_5</th>\n",
       "      <th>user_activity_var_8</th>\n",
       "      <th>user_activity_var_9</th>\n",
       "      <th>user_activity_var_11</th>\n",
       "      <th>buy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\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",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_activity_var_1  user_activity_var_2  user_activity_var_4  \\\n",
       "0                    1                    0                    0   \n",
       "1                    0                    0                    0   \n",
       "2                    1                    0                    0   \n",
       "3                    0                    0                    0   \n",
       "4                    1                    0                    0   \n",
       "\n",
       "   user_activity_var_5  user_activity_var_8  user_activity_var_9  \\\n",
       "0                    0                    0                    0   \n",
       "1                    0                    0                    0   \n",
       "2                    0                    1                    0   \n",
       "3                    0                    0                    0   \n",
       "4                    0                    0                    0   \n",
       "\n",
       "   user_activity_var_11  buy  \n",
       "0                     0    0  \n",
       "1                     0    0  \n",
       "2                     0    0  \n",
       "3                     1    1  \n",
       "4                     0    0  "
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Shuffle dataset\n",
    "\n",
    "train_data = train_data.sample(frac=1).reset_index(drop=True)\n",
    "train_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(13000, 8)\n",
      "(7833, 8)\n"
     ]
    }
   ],
   "source": [
    "train_data = train_data.to_numpy()\n",
    "val_data = val_data.to_numpy()\n",
    "\n",
    "print(train_data.shape)\n",
    "print(val_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_x = train_data[:,0:train_data.shape[1]-1]\n",
    "train_y = train_data[:,-1]\n",
    "\n",
    "val_x = val_data[:, 0:val_data.shape[1]-1]\n",
    "val_y = val_data[:,-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(13000, 7)\n",
      "(13000,)\n"
     ]
    }
   ],
   "source": [
    "# Print training data shape\n",
    "\n",
    "print(train_x.shape)\n",
    "print(train_y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(7833, 7)\n",
      "(7833,)\n"
     ]
    }
   ],
   "source": [
    "# Print validation data shape\n",
    "\n",
    "print(val_x.shape)\n",
    "print(val_y.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense (Dense)                (None, 512)               4096      \n",
      "_________________________________________________________________\n",
      "dropout (Dropout)            (None, 512)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 128)               65664     \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 2)                 258       \n",
      "=================================================================\n",
      "Total params: 70,018\n",
      "Trainable params: 70,018\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-06-05 21:29:05.003568: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  SSE4.1 SSE4.2 AVX AVX2 AVX512F FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2022-06-05 21:29:05.004644: I tensorflow/core/common_runtime/process_util.cc:146] Creating new thread pool with default inter op setting: 2. Tune using inter_op_parallelism_threads for best performance.\n"
     ]
    }
   ],
   "source": [
    "model=Sequential()\n",
    "model.add(InputLayer(input_shape=(7,)))\n",
    "\n",
    "model.add(Dense(512,activation='relu'))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(Dense(128,activation='relu'))\n",
    "model.add(Dropout(0.1))\n",
    "model.add(Dense(2,activation='softmax'))\n",
    "\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Callbacks\n",
    "\n",
    "#Learning Rate Annealer\n",
    "lrr= ReduceLROnPlateau(monitor='val_accuracy', factor=.01, patience=3, min_lr=1e-5)\n",
    "\n",
    "# Early Stoppping\n",
    "EarlyStop = EarlyStopping( monitor='val_loss', patience=3, restore_best_weights=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2022-06-05 21:29:05.091279: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:185] None of the MLIR Optimization Passes are enabled (registered 2)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "407/407 [==============================] - 3s 5ms/step - loss: 0.3045 - accuracy: 0.8839 - val_loss: 0.1446 - val_accuracy: 0.9614\n",
      "Epoch 2/10\n",
      "407/407 [==============================] - 2s 4ms/step - loss: 0.2785 - accuracy: 0.8953 - val_loss: 0.1451 - val_accuracy: 0.9612\n",
      "Epoch 3/10\n",
      "407/407 [==============================] - 4s 10ms/step - loss: 0.2762 - accuracy: 0.8960 - val_loss: 0.1317 - val_accuracy: 0.9732\n",
      "Epoch 4/10\n",
      "407/407 [==============================] - 5s 12ms/step - loss: 0.2751 - accuracy: 0.8967 - val_loss: 0.1478 - val_accuracy: 0.9732\n",
      "Epoch 5/10\n",
      "407/407 [==============================] - 3s 7ms/step - loss: 0.2752 - accuracy: 0.8960 - val_loss: 0.1198 - val_accuracy: 0.9734\n",
      "Epoch 6/10\n",
      "407/407 [==============================] - 2s 6ms/step - loss: 0.2733 - accuracy: 0.8982 - val_loss: 0.1759 - val_accuracy: 0.9307\n",
      "Epoch 7/10\n",
      "407/407 [==============================] - 3s 6ms/step - loss: 0.2736 - accuracy: 0.8967 - val_loss: 0.1518 - val_accuracy: 0.9606\n",
      "Epoch 8/10\n",
      "407/407 [==============================] - 3s 7ms/step - loss: 0.2735 - accuracy: 0.8974 - val_loss: 0.1747 - val_accuracy: 0.9606\n"
     ]
    }
   ],
   "source": [
    "#model.compile(optimizer='adam',loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),metrics=['accuracy'])\n",
    "\n",
    "model.compile(optimizer='adam',loss=tf.keras.losses.SparseCategoricalCrossentropy(),metrics=['accuracy'])\n",
    "history = model.fit(x=train_x, y=train_y, epochs=10, validation_data=(val_x,val_y),callbacks=[lrr, EarlyStop]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plotting model accuracy for different epochs\n",
    "plt.figure(1, figsize = (10, 10))  \n",
    "plt.subplot(211)  \n",
    "plt.plot(history.history['accuracy'])  \n",
    "plt.plot(history.history['val_accuracy'])  \n",
    "plt.title('Model Accuracy')  \n",
    "plt.ylabel('Accuracy')  \n",
    "plt.xlabel('Epoch')  \n",
    "plt.legend(['train', 'validation'], loc='upper left')   \n",
    "\n",
    "# plotting model loss for different epochs\n",
    "plt.subplot(212)  \n",
    "plt.plot(history.history['loss'])  \n",
    "plt.plot(history.history['val_loss'])  \n",
    "plt.title('Model Loss')  \n",
    "plt.ylabel('Loss')  \n",
    "plt.xlabel('Epoch')  \n",
    "plt.legend(['train', 'validation'], loc='upper left')  \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate model on validation data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import recall_score\n",
    "from sklearn.metrics import f1_score\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred = model.predict(val_x)\n",
    "y_pred = np.argmax(pred, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy: 0.9734456785395124\n"
     ]
    }
   ],
   "source": [
    "# Accuracy\n",
    "\n",
    "accuracy = accuracy_score(y_true=val_y, y_pred=y_pred)\n",
    "print('Accuracy:', accuracy)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# confusion Matrix\n",
    "\n",
    "cf_matrix = confusion_matrix(y_true=val_y, y_pred=y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<AxesSubplot:>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "label_names = ['True Neg','False Pos','False Neg','True Pos']\n",
    "label_counts = cf_matrix.flatten()\n",
    "\n",
    "labels = [f\"{v1}\\n{v2}\" for v1, v2 in zip(label_names,label_counts)]\n",
    "labels = np.asarray(labels).reshape(2,2)\n",
    "\n",
    "sns.heatmap(cf_matrix, annot=labels,fmt='', cmap='Blues',yticklabels=['True Label'],xticklabels=['Predicted label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precision:  0.82\n"
     ]
    }
   ],
   "source": [
    "# Precison Score\n",
    "\n",
    "precision = precision_score(y_true=val_y, y_pred=y_pred)\n",
    "print('Precision: ',precision)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Recall:  0.615\n"
     ]
    }
   ],
   "source": [
    "# Recall Score\n",
    "\n",
    "recall = recall_score(y_true=val_y, y_pred=y_pred)\n",
    "print('Recall: ',recall)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Score:  0.7028571428571427\n"
     ]
    }
   ],
   "source": [
    "# F1 Score\n",
    "\n",
    "f1Score = f1_score(y_true=val_y, y_pred=y_pred)\n",
    "print('F1 Score: ',f1Score)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Get Predictions from test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "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>id</th>\n",
       "      <th>created_at</th>\n",
       "      <th>campaign_var_1</th>\n",
       "      <th>campaign_var_2</th>\n",
       "      <th>products_purchased</th>\n",
       "      <th>signup_date</th>\n",
       "      <th>user_activity_var_1</th>\n",
       "      <th>user_activity_var_2</th>\n",
       "      <th>user_activity_var_3</th>\n",
       "      <th>user_activity_var_4</th>\n",
       "      <th>user_activity_var_5</th>\n",
       "      <th>user_activity_var_6</th>\n",
       "      <th>user_activity_var_7</th>\n",
       "      <th>user_activity_var_8</th>\n",
       "      <th>user_activity_var_9</th>\n",
       "      <th>user_activity_var_10</th>\n",
       "      <th>user_activity_var_11</th>\n",
       "      <th>user_activity_var_12</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>39162</td>\n",
       "      <td>2022-01-01</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2021-08-17</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>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>39163</td>\n",
       "      <td>2022-01-01</td>\n",
       "      <td>4</td>\n",
       "      <td>7</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2020-05-21</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>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>39164</td>\n",
       "      <td>2022-01-01</td>\n",
       "      <td>8</td>\n",
       "      <td>7</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>39165</td>\n",
       "      <td>2022-01-01</td>\n",
       "      <td>9</td>\n",
       "      <td>8</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2020-06-22</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>39166</td>\n",
       "      <td>2022-01-01</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2021-03-10</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>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id  created_at  campaign_var_1  campaign_var_2  products_purchased  \\\n",
       "0  39162  2022-01-01               2               2                 NaN   \n",
       "1  39163  2022-01-01               4               7                 3.0   \n",
       "2  39164  2022-01-01               8               7                 NaN   \n",
       "3  39165  2022-01-01               9               8                 2.0   \n",
       "4  39166  2022-01-01               4               5                 2.0   \n",
       "\n",
       "  signup_date  user_activity_var_1  user_activity_var_2  user_activity_var_3  \\\n",
       "0  2021-08-17                    1                    1                    0   \n",
       "1  2020-05-21                    1                    0                    0   \n",
       "2         NaN                    0                    0                    0   \n",
       "3  2020-06-22                    0                    0                    0   \n",
       "4  2021-03-10                    1                    0                    0   \n",
       "\n",
       "   user_activity_var_4  user_activity_var_5  user_activity_var_6  \\\n",
       "0                    0                    0                    1   \n",
       "1                    0                    0                    0   \n",
       "2                    0                    1                    1   \n",
       "3                    0                    1                    1   \n",
       "4                    0                    0                    0   \n",
       "\n",
       "   user_activity_var_7  user_activity_var_8  user_activity_var_9  \\\n",
       "0                    0                    0                    0   \n",
       "1                    0                    1                    0   \n",
       "2                    0                    0                    0   \n",
       "3                    1                    0                    0   \n",
       "4                    0                    0                    0   \n",
       "\n",
       "   user_activity_var_10  user_activity_var_11  user_activity_var_12  \n",
       "0                     0                     1                     0  \n",
       "1                     0                     0                     0  \n",
       "2                     0                     0                     0  \n",
       "3                     0                     2                     0  \n",
       "4                     0                     0                     0  "
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df = pd.read_csv(\"test_Wf7sxXF.csv\")\n",
    "test_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of Test set  (13184, 18)\n"
     ]
    }
   ],
   "source": [
    "print(\"Shape of Test set \",test_df.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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_activity_var_1</th>\n",
       "      <th>user_activity_var_2</th>\n",
       "      <th>user_activity_var_4</th>\n",
       "      <th>user_activity_var_5</th>\n",
       "      <th>user_activity_var_8</th>\n",
       "      <th>user_activity_var_9</th>\n",
       "      <th>user_activity_var_11</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\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",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   user_activity_var_1  user_activity_var_2  user_activity_var_4  \\\n",
       "0                    1                    1                    0   \n",
       "1                    1                    0                    0   \n",
       "2                    0                    0                    0   \n",
       "3                    0                    0                    0   \n",
       "4                    1                    0                    0   \n",
       "\n",
       "   user_activity_var_5  user_activity_var_8  user_activity_var_9  \\\n",
       "0                    0                    0                    0   \n",
       "1                    0                    1                    0   \n",
       "2                    1                    0                    0   \n",
       "3                    1                    0                    0   \n",
       "4                    0                    0                    0   \n",
       "\n",
       "   user_activity_var_11  \n",
       "0                     1  \n",
       "1                     0  \n",
       "2                     0  \n",
       "3                     2  \n",
       "4                     0  "
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_x = test_df[['user_activity_var_1','user_activity_var_2','user_activity_var_4','user_activity_var_5','user_activity_var_8','user_activity_var_9','user_activity_var_11']]\n",
    "test_x.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_id = test_df['id'].to_numpy()\n",
    "test_x = test_x.to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(13184,)\n",
      "(13184, 7)\n"
     ]
    }
   ],
   "source": [
    "print(test_id.shape)\n",
    "print(test_x.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred = model.predict(test_x)\n",
    "y_pred = np.argmax(pred, axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "result_df= pd.DataFrame({\"id\":test_id,\"buy\":y_pred})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "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>id</th>\n",
       "      <th>buy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>39162</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>39163</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>39164</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>39165</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>39166</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id  buy\n",
       "0  39162    1\n",
       "1  39163    0\n",
       "2  39164    0\n",
       "3  39165    1\n",
       "4  39166    0"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "result_df.to_csv('Ashish_result.csv', index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}