Project2_MultiTask_Bandit_Omar.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "IMDtFpsBbZPF",
   "metadata": {
    "id": "IMDtFpsBbZPF"
   },
   "source": [
    "<h4><i><font color='red'>The project has many components, and we encourage you to complete as many as you can! That said, we highly encourage you to submit your work even if your notebook is only partially completed - the TA can help review your work and provide tips on any places that you got stuck or have further questions!</font></i></h4>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "wIHxLWre9THR",
   "metadata": {
    "id": "wIHxLWre9THR"
   },
   "source": [
    "<font size=\"6\">**Table of Contents:**</font>  \n",
    "<br>\n",
    "<font size=\"5\">\n",
    "<u>Single- and Multi-task Models:</u>  \n",
    "&emsp;&emsp;1\\. [Introduction](#scrollTo=ETXp_Ixa9xVH&uniqifier=1)  \n",
    "&emsp;&emsp;2\\. [Setup](#scrollTo=CcKGfQtEDte3&uniqifier=1)  \n",
    "&emsp;&emsp;3\\. [Simple Feature Preprocessing](#scrollTo=8k4D0ZUPEXJ5&uniqifier=1)  \n",
    "&emsp;&emsp;4\\. [Negative Sampling](#scrollTo=ZOyl7LPbGeFQ&uniqifier=1)  \n",
    "&emsp;&emsp;5\\. [Single-Task Ranker](#scrollTo=ADo5s7YWHOoR&uniqifier=1)  \n",
    "&emsp;&emsp;6\\. [Evaluation](#scrollTo=MtAo2yv4H4jg&uniqifier=1)  \n",
    "&emsp;&emsp;7\\. [Multi-Task Ranker](#scrollTo=NI-XuLBcJld0&uniqifier=1)  \n",
    "</font>\n",
    "<font size=\"4\">\n",
    "&emsp;&emsp;&emsp;&emsp;a\\. [TODO 2.1: Add a Second Label](#scrollTo=ibq1uMfTL2H9&uniqifier=1)  \n",
    "&emsp;&emsp;&emsp;&emsp;b\\. [Model Training](#scrollTo=aMIjykLIMHZW&uniqifier=1)  \n",
    "&emsp;&emsp;&emsp;&emsp;c\\. [Freezing Layers for Secondary Tasks](#scrollTo=09LGM73ePNZK&uniqifier=1)  \n",
    "</font>\n",
    "<font size=\"5\">\n",
    "&emsp;&emsp;8\\. [TODO 2.2: Model Improvement (Optional)](#scrollTo=NTgBnh2-PrNL&uniqifier=1)  \n",
    "<br>\n",
    "<u>Bandits for Recommendations:</u>  \n",
    "&emsp;&emsp;1\\. [Introduction](#scrollTo=MAFpygeZumu1&uniqifier=1)  \n",
    "&emsp;&emsp;2\\. [Setup](#scrollTo=mGMwRjXuDxm9&uniqifier=1)  \n",
    "&emsp;&emsp;3\\. [Implement ϵ-greedy Policy](#scrollTo=s_bLy59XFaz7&uniqifier=1)  \n",
    "&emsp;&emsp;4\\. [Policy Evaluation](#scrollTo=NHsNS6B1JTrO&uniqifier=1)  \n",
    "&emsp;&emsp;5\\. [TODO 2.3: Implement UCB1](#scrollTo=aM2H5Jm4OqFb&uniqifier=1)  \n",
    "</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "NWpqDtwiKSyN",
   "metadata": {
    "id": "NWpqDtwiKSyN"
   },
   "source": [
    "# **Single- and Multi-task Models:**\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ETXp_Ixa9xVH",
   "metadata": {
    "id": "ETXp_Ixa9xVH"
   },
   "source": [
    "## **1. Introduction**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "837f0383-ee06-41a4-b431-032c431dfcbb",
   "metadata": {
    "id": "837f0383-ee06-41a4-b431-032c431dfcbb"
   },
   "source": [
    "Congratulations on making it to week 2! 👏 In the first week of this course, we covered the basics of how to design personalized recommendation systems. We then provided some system design examples for large scale recommenders from corporations like Spotify and YouTube, as well as techniques for candidate generation, specifically the two-tower model being used at Twitter and Pinterest.\n",
    "\n",
    "This week, we'll cover details of ML approaches for recommendations: Multi-task recommenders and contextual bandits.  \n",
    "  \n",
    "<br>We first begin by developing a simple multi-task model, and then cover a contextual bandit model. For ease of use, we will re-use our H&M dataset from week 1 for the Single- and Multi-task models, and will switch to using Yahoo! news dataset for the contextual bandit model in the second section of this notebook.  \n",
    "  \n",
    "<br>**Single-task Recommender:** We will start by deveoping a single task model that makes predictions about whether the user will purchase the article or not.\n",
    "  \n",
    "\n",
    "<br>**Multi-task Recommender:** We then expand the scope to consider auxilliary prediction tasks in order to improve the recommendation performance.\n",
    "\n",
    "Specifically, we make the following two predictions using a multi-task setup:\n",
    "1. Predicting which specific item the user will buy, as done with the single-task model.  \n",
    "2. Predicting which category the user will make purchase in.  \n",
    "\n",
    "  \n",
    "Example architecture:\n",
    "<div>\n",
    "<img src=\"https://miro.medium.com/max/3688/1*rrIJOpJO8fkFECNHlwq-jQ.png\" width=\"500\"/>\n",
    "</div>  \n",
    "  \n",
    "<br>**Negative Sampling:**\n",
    "It is important to note that the dataset is only of positive cases -- cases wherein the user bought an article. To train our models, we will have to resort to negative sampling to develop a mixed dataset comprising of positive and negative examples. We will use random negative sampling in this notebook.  \n",
    "  \n",
    "<br>**Evaluation:**\n",
    "We will evaluate the ranker performance on various ranking metrics.\n",
    "  \n",
    "<br>**This notebook benefits from using a GPU!**  \n",
    "You can use it for free, in Google Colab, by going to the menu bar above, selecting `Runtime`, then `Change runtime type`, then selecting `GPU` as the `Hardware accelerator`."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "CcKGfQtEDte3",
   "metadata": {
    "id": "CcKGfQtEDte3"
   },
   "source": [
    "## **2. Setup**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "07a2a936",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>.container { width:100% !important; }</style>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from IPython.display import display, HTML\n",
    "display(HTML(\"<style>.container { width:100% !important; }</style>\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "add75169-0279-469c-98c2-36e6fc28ebdd",
   "metadata": {
    "id": "add75169-0279-469c-98c2-36e6fc28ebdd"
   },
   "outputs": [],
   "source": [
    "# Installs and imports\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "from matplotlib import pyplot as plt\n",
    "from tqdm.notebook import tqdm\n",
    "from typing import Dict, Text\n",
    "import tensorflow as tf\n",
    "import random\n",
    "from datetime import timedelta\n",
    "import gc\n",
    "from sklearn.metrics import confusion_matrix, ConfusionMatrixDisplay\n",
    "from IPython.display import clear_output\n",
    "\n",
    "# Download data\n",
    "!pip install -U --no-cache-dir gdown --pre\n",
    "!gdown --no-cookies --folder https://drive.google.com/drive/folders/1j9QpkSKwqFfpgohbVxdPDKuwmwJXZ6yz?usp=sharing\n",
    "\n",
    "clear_output()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "vvZex8kEBj-Q",
   "metadata": {
    "id": "vvZex8kEBj-Q"
   },
   "outputs": [],
   "source": [
    "# Read data into memory\n",
    "article_df = pd.read_csv(\"hmdata/articles.csv.zip\")\n",
    "customer_df = pd.read_csv(\"hmdata/customers.csv.zip\")\n",
    "transaction_df = pd.read_csv('hmdata/transactions_train.csv.zip')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "3081bc7b-ab90-41c0-9c21-35dfe59b4228",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "3081bc7b-ab90-41c0-9c21-35dfe59b4228",
    "outputId": "2508a4a5-b062-4c91-9900-a4e3b47c7d62"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Split into train and test as we did in week 1\n",
    "\n",
    "N_DAYS_TRAIN = 45\n",
    "N_DAYS_TEST = 7\n",
    "\n",
    "max_date = transaction_df['t_dat'].max()\n",
    "train = transaction_df[(transaction_df['t_dat']>=((pd.to_datetime(max_date) - timedelta(days=N_DAYS_TRAIN+N_DAYS_TEST)).date().strftime('%Y-%m-%d')))\n",
    "                       & (transaction_df['t_dat']<((pd.to_datetime(max_date) - timedelta(days=N_DAYS_TEST)).date().strftime('%Y-%m-%d')))]\n",
    "test = transaction_df[(transaction_df['t_dat']>=((pd.to_datetime(max_date) - timedelta(days=N_DAYS_TEST)).date().strftime('%Y-%m-%d')))]\n",
    "\n",
    "# Delete transaction_df from the namespace to free up some memory\n",
    "transaction_df = None\n",
    "del transaction_df\n",
    "gc.collect()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8k4D0ZUPEXJ5",
   "metadata": {
    "id": "8k4D0ZUPEXJ5"
   },
   "source": [
    "## **3. Simple Feature Preprocessing**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2b073d71-0ffc-407c-9338-1a87cee00b07",
   "metadata": {
    "id": "2b073d71-0ffc-407c-9338-1a87cee00b07"
   },
   "source": [
    "In this section we focus on extracting features for training the ranking models. We focus primarily on extracting relatively simple features, and leave sophisticated feature modeling for next week.\n",
    "\n",
    "For users specifically, we will focus on extracting the following features:\n",
    "1. Number of unique articles purchased\n",
    "2. Number of unique days active\n",
    "3. Number of purchased items in total\n",
    "4. Min, Max, and Mean purchase price\n",
    "5. Number of unique product group names purchased from\n",
    "6. Number of unique department names purchased from\n",
    "7. Number of unique section names purchased from\n",
    "\n",
    "For articles we will only use `product_group_name` for simplicity. In practice, we'd use many more features. We encourage you to experiment with more or different features for both customers and articles. It does not need preprocessing so we can simply append it to the dataframe before training.  \n",
    "<br>  \n",
    "To avoid leakage we will only calculate features based on the training set. However, note that even this method can introduce a small amount of leakage since any user's features will have signal from all purchases in their training set history. For example, when predicting whether a user will buy their first product, their features will have information up to their most recent purchase. The effects of this sort of leakage are somewhat mitigated though, since the test set will not be affected by it and we're mostly concerned with performance on the test set.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "-LEuRXauErI3",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 288
    },
    "id": "-LEuRXauErI3",
    "outputId": "545c49a8-07bb-4389-805b-27cf2d7264ef"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(330712, 10)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>customer_id</th>\n",
       "      <th>num_articles</th>\n",
       "      <th>days_active</th>\n",
       "      <th>total_purchases</th>\n",
       "      <th>min_price</th>\n",
       "      <th>max_price</th>\n",
       "      <th>mean_price</th>\n",
       "      <th>num_unique_product_groups</th>\n",
       "      <th>num_unique_departments</th>\n",
       "      <th>num_unique_sections</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>00000dbacae5abe5e23885899a1fa44253a17956c6d1c3...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0.050831</td>\n",
       "      <td>0.050831</td>\n",
       "      <td>0.050831</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>00006413d8573cd20ed7128e53b7b13819fe5cfc2d801f...</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0.020322</td>\n",
       "      <td>0.047441</td>\n",
       "      <td>0.032186</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0000757967448a6cb83efb3ea7a3fb9d418ac7adf2379d...</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0.033881</td>\n",
       "      <td>0.042356</td>\n",
       "      <td>0.038119</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>00009d946eec3ea54add5ba56d5210ea898def4b46c685...</td>\n",
       "      <td>18</td>\n",
       "      <td>3</td>\n",
       "      <td>26</td>\n",
       "      <td>0.013542</td>\n",
       "      <td>0.050831</td>\n",
       "      <td>0.031147</td>\n",
       "      <td>3</td>\n",
       "      <td>11</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0000b2f1829e23b24feec422ef13df3ccedaedc85368e6...</td>\n",
       "      <td>6</td>\n",
       "      <td>2</td>\n",
       "      <td>7</td>\n",
       "      <td>0.002949</td>\n",
       "      <td>0.033881</td>\n",
       "      <td>0.017482</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         customer_id  num_articles  \\\n",
       "0  00000dbacae5abe5e23885899a1fa44253a17956c6d1c3...             1   \n",
       "1  00006413d8573cd20ed7128e53b7b13819fe5cfc2d801f...             4   \n",
       "2  0000757967448a6cb83efb3ea7a3fb9d418ac7adf2379d...             2   \n",
       "3  00009d946eec3ea54add5ba56d5210ea898def4b46c685...            18   \n",
       "4  0000b2f1829e23b24feec422ef13df3ccedaedc85368e6...             6   \n",
       "\n",
       "   days_active  total_purchases  min_price  max_price  mean_price  \\\n",
       "0            1                1   0.050831   0.050831    0.050831   \n",
       "1            1                4   0.020322   0.047441    0.032186   \n",
       "2            1                2   0.033881   0.042356    0.038119   \n",
       "3            3               26   0.013542   0.050831    0.031147   \n",
       "4            2                7   0.002949   0.033881    0.017482   \n",
       "\n",
       "   num_unique_product_groups  num_unique_departments  num_unique_sections  \n",
       "0                          1                       1                    1  \n",
       "1                          3                       4                    3  \n",
       "2                          1                       1                    1  \n",
       "3                          3                      11                   11  \n",
       "4                          2                       4                    4  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "customer_features_df = train.merge(article_df, on='article_id', how='outer')\n",
    "customer_features_df = customer_features_df.groupby('customer_id').agg(\n",
    "    num_articles=('article_id', 'nunique'),\n",
    "    days_active=('t_dat', 'nunique'),\n",
    "    total_purchases=('article_id', 'size'),\n",
    "    min_price=('price', 'min'),\n",
    "    max_price=('price', 'max'),\n",
    "    mean_price=('price', 'mean'),\n",
    "    num_unique_product_groups=('product_group_name', 'nunique'),\n",
    "    num_unique_departments=('department_name', 'nunique'),\n",
    "    num_unique_sections=('section_name', 'nunique')\n",
    ").reset_index()\n",
    "\n",
    "print(customer_features_df.shape)\n",
    "customer_features_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ZOyl7LPbGeFQ",
   "metadata": {
    "id": "ZOyl7LPbGeFQ"
   },
   "source": [
    "## **4. Negative Sampling**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9ff60881-c2ab-4c1f-b69a-67c28becb6cf",
   "metadata": {
    "id": "9ff60881-c2ab-4c1f-b69a-67c28becb6cf"
   },
   "source": [
    "In this dataset only have purchase information, also known as positive examples. This is generally the case -- we often only observe positive interactions from users. However, to train a model, we would need access to both positive and negative samples.\n",
    "\n",
    "To generate negative samples we focus on random sampling of negative examples per user-article transaction, but other methods exist to sample negatives, namely:\n",
    "1. Biased sampling from similar categories\n",
    "2. In/Cross-batch negative sampling\n",
    "\n",
    "We'll continue with the random sampling strategy:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "FbTOTvMrNUYD",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 99,
     "referenced_widgets": [
      "f96675512a1f4f47be2df206b1c4ada1",
      "3fdaac08daf847fcad219d7f6522021c",
      "c7aadd2d77644094961c0f426c8105b6",
      "300449ac26444b3587071a0225b8fb5b",
      "2c0812e3b51847e8b2cfbe926f59473f",
      "ad0a7b5d47da48f8af4e4587e8ce6dde",
      "9a7f950631c744919489f5645752ea07",
      "3cc7cd57d0f24c069db2c96af4f384c7",
      "a664a6b775b8401da8158a5c99cfca11",
      "6dc688fdbe26455f9f1ab5419f8e062f",
      "92b8c6b085eb41848275b673cb842b33",
      "112f1c82867e4c819fd87e707f027a8c",
      "87edeed6fe334333ae83fce4682c23da",
      "968de4d2649b46d3ab2906ee4b5a51d9",
      "02b0bbe556984c599ad1e9076b3a6fe8",
      "2fed36e5d12f4bf58d391f3f9b41b25b",
      "a5e83e39df5f4feb894b4eb88622cedd",
      "8400ad754357400db79dddfaa07ba210",
      "5778f8578b494d259037a3514b254735",
      "3499b6797ece417abf99198fda3fffcf",
      "2a353946d4874cc68c21d94b2dbc9c5f",
      "1931a2f097194bb48c788ab2ecacceab"
     ]
    },
    "id": "FbTOTvMrNUYD",
    "outputId": "dc00c5c6-ea20-419b-fea8-ed80efe14677"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "71880b6697514f78925356642d622d40",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1769097 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1a774c369a4a43aa82f32f5be63d2fc3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/266364 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "((5307291, 3), (799092, 3))"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Generate N negative samples per row (user-article transaction)\n",
    "\n",
    "def generate_negative_samples(df, article_list, num_neg=1, negative_sample_multiplier=20):\n",
    "    \"\"\"\n",
    "    Generates `num_neg` negative samples from `article_list` per row (customer_id, article_id interaction) in `df`.\n",
    "\n",
    "    Args:\n",
    "        df (DataFrame): Positive transactions between customers and articles. Expects features `customer_id` and\n",
    "        `article_id`.\n",
    "        article_list (list): Article IDs for negative sampling.\n",
    "        num_neg (int): Number of negative samples to generate per row.\n",
    "        negative_sample_multiplier (int): Multiplier that allows us to randomly sample negative candidates all at\n",
    "        once.\n",
    "            Necessary to be larger than 1 in most cases since we will filter out some candidates.\n",
    "\n",
    "    Raises:\n",
    "        Warning in case we did not generate the intended number of negative samples per row.\n",
    "\n",
    "    Returns:\n",
    "        negative_df (DataFrame): Original, positive samples concatenated with generated, negative samples.\n",
    "    \"\"\"\n",
    "    # Create lookup of positive samples. We don't want to include negative samples that are actually positives\n",
    "    customer2positives = dict(df.groupby('customer_id')['article_id'].apply(lambda x: x.values))\n",
    "    customer2positives = {k:set(v) for k,v in customer2positives.items()}\n",
    "\n",
    "    # Sample random negatives all at once\n",
    "    np.random.seed(42)\n",
    "    random_sample_size = (len(df), num_neg*negative_sample_multiplier)\n",
    "    random_samples = np.random.choice(article_list, \n",
    "                                      size=random_sample_size,\n",
    "                                      replace=True)\n",
    "    \n",
    "    # Generate negative samples\n",
    "    customers, articles, labels = [], [], []\n",
    "    for n, (customer_id, article_id) in enumerate(tqdm(zip(df['customer_id'], df['article_id']), total=len(df))):\n",
    "        customers.append(customer_id)\n",
    "        articles.append(article_id)\n",
    "        labels.append(1)\n",
    "        count, temp_negatives = 0, set()\n",
    "        for negative_id in random_samples[n]:\n",
    "            if (negative_id not in customer2positives[customer_id]) \\\n",
    "                & (negative_id not in temp_negatives):\n",
    "                customers.append(customer_id)\n",
    "                articles.append(negative_id)\n",
    "                labels.append(0)\n",
    "                temp_negatives.add(negative_id)\n",
    "                count += 1\n",
    "            if count == num_neg:\n",
    "                break\n",
    "\n",
    "    # Create result dataframe of positives and negatives, and data quality check\n",
    "    negative_df = pd.DataFrame({'customer_id':customers, 'article_id':articles, 'label':labels})\n",
    "    if len(negative_df) != (len(df)*num_neg+len(df)):\n",
    "        print('WARNING: Not enough samples generated. Increase `negative_sample_multiplier`.')\n",
    "    return negative_df\n",
    "\n",
    "\n",
    "# Note: We only generate negative samples from the training set to avoid leakage\n",
    "articles_set = list(train.article_id.unique())\n",
    "train_negative = generate_negative_samples(train, articles_set, num_neg=2)\n",
    "test_negative = generate_negative_samples(test, articles_set, num_neg=2)\n",
    "\n",
    "train_negative.shape, test_negative.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "ca-aLkIwn1_T",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ca-aLkIwn1_T",
    "outputId": "176afc17-4f77-47bc-be5b-c4c793ad326f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((5307291, 12), (448893, 12))"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add customer features to both positive and negative samples\n",
    "# Note: We effective remove any `cold-start` customers from test because we generated customer features based on training set interactions.\n",
    "    # If we did not do this, we'd have to fill those null features and we'd then get the same recommendations for each new user.\n",
    "    # In practice we'd aim to describe customers based on their inherent features such as age, gender, etc. so as to not have to rely solely on their purchase behavior.\n",
    "    # Or we could, alternatively, default to the same list of most popular items for new users until we had interactions for them.\n",
    "\n",
    "train_negative = train_negative.merge(customer_features_df, on='customer_id', how='inner')\n",
    "test_negative = test_negative.merge(customer_features_df, on='customer_id', how='inner')\n",
    "\n",
    "train_negative.shape, test_negative.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "YXOOS-45qAKt",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "YXOOS-45qAKt",
    "outputId": "d07e6094-1567-4ff4-a569-5626a8d375a9"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((5307291, 15), (448893, 15))"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add article features to both positive and negative samples\n",
    "\n",
    "train_negative = train_negative.merge(article_df[['article_id','product_group_name', 'department_name','section_name']], on='article_id', how='left')\n",
    "test_negative = test_negative.merge(article_df[['article_id','product_group_name', 'department_name','section_name']], on='article_id', how='left')\n",
    "\n",
    "train_negative.shape, test_negative.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "YTAC_Ka9qRMR",
   "metadata": {
    "id": "YTAC_Ka9qRMR"
   },
   "outputs": [],
   "source": [
    "# Check for nulls\n",
    "\n",
    "assert train_negative.isnull().sum().sum() == 0\n",
    "assert test_negative.isnull().sum().sum() == 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "6L-HAEgoqcaC",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 339
    },
    "id": "6L-HAEgoqcaC",
    "outputId": "1b2f652c-c70b-4967-af5e-04e668facd23"
   },
   "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>customer_id</th>\n",
       "      <th>article_id</th>\n",
       "      <th>label</th>\n",
       "      <th>num_articles</th>\n",
       "      <th>days_active</th>\n",
       "      <th>total_purchases</th>\n",
       "      <th>min_price</th>\n",
       "      <th>max_price</th>\n",
       "      <th>mean_price</th>\n",
       "      <th>num_unique_product_groups</th>\n",
       "      <th>num_unique_departments</th>\n",
       "      <th>num_unique_sections</th>\n",
       "      <th>product_group_name</th>\n",
       "      <th>department_name</th>\n",
       "      <th>section_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>743123001</td>\n",
       "      <td>1</td>\n",
       "      <td>30</td>\n",
       "      <td>7</td>\n",
       "      <td>33</td>\n",
       "      <td>0.00422</td>\n",
       "      <td>0.06778</td>\n",
       "      <td>0.022176</td>\n",
       "      <td>6</td>\n",
       "      <td>18</td>\n",
       "      <td>9</td>\n",
       "      <td>Swimwear</td>\n",
       "      <td>Swimwear</td>\n",
       "      <td>Womens Swimwear, beachwear</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>870123001</td>\n",
       "      <td>0</td>\n",
       "      <td>30</td>\n",
       "      <td>7</td>\n",
       "      <td>33</td>\n",
       "      <td>0.00422</td>\n",
       "      <td>0.06778</td>\n",
       "      <td>0.022176</td>\n",
       "      <td>6</td>\n",
       "      <td>18</td>\n",
       "      <td>9</td>\n",
       "      <td>Garment Lower body</td>\n",
       "      <td>Tops Girls</td>\n",
       "      <td>Kids Sports</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>212629047</td>\n",
       "      <td>0</td>\n",
       "      <td>30</td>\n",
       "      <td>7</td>\n",
       "      <td>33</td>\n",
       "      <td>0.00422</td>\n",
       "      <td>0.06778</td>\n",
       "      <td>0.022176</td>\n",
       "      <td>6</td>\n",
       "      <td>18</td>\n",
       "      <td>9</td>\n",
       "      <td>Garment Full body</td>\n",
       "      <td>Basic 1</td>\n",
       "      <td>Divided Basics</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>743123001</td>\n",
       "      <td>1</td>\n",
       "      <td>30</td>\n",
       "      <td>7</td>\n",
       "      <td>33</td>\n",
       "      <td>0.00422</td>\n",
       "      <td>0.06778</td>\n",
       "      <td>0.022176</td>\n",
       "      <td>6</td>\n",
       "      <td>18</td>\n",
       "      <td>9</td>\n",
       "      <td>Swimwear</td>\n",
       "      <td>Swimwear</td>\n",
       "      <td>Womens Swimwear, beachwear</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>904590001</td>\n",
       "      <td>0</td>\n",
       "      <td>30</td>\n",
       "      <td>7</td>\n",
       "      <td>33</td>\n",
       "      <td>0.00422</td>\n",
       "      <td>0.06778</td>\n",
       "      <td>0.022176</td>\n",
       "      <td>6</td>\n",
       "      <td>18</td>\n",
       "      <td>9</td>\n",
       "      <td>Garment Full body</td>\n",
       "      <td>Kids Girl Dresses</td>\n",
       "      <td>Kids Girl</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         customer_id  article_id  label  \\\n",
       "0  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   743123001      1   \n",
       "1  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   870123001      0   \n",
       "2  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   212629047      0   \n",
       "3  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   743123001      1   \n",
       "4  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   904590001      0   \n",
       "\n",
       "   num_articles  days_active  total_purchases  min_price  max_price  \\\n",
       "0            30            7               33    0.00422    0.06778   \n",
       "1            30            7               33    0.00422    0.06778   \n",
       "2            30            7               33    0.00422    0.06778   \n",
       "3            30            7               33    0.00422    0.06778   \n",
       "4            30            7               33    0.00422    0.06778   \n",
       "\n",
       "   mean_price  num_unique_product_groups  num_unique_departments  \\\n",
       "0    0.022176                          6                      18   \n",
       "1    0.022176                          6                      18   \n",
       "2    0.022176                          6                      18   \n",
       "3    0.022176                          6                      18   \n",
       "4    0.022176                          6                      18   \n",
       "\n",
       "   num_unique_sections  product_group_name    department_name  \\\n",
       "0                    9            Swimwear           Swimwear   \n",
       "1                    9  Garment Lower body         Tops Girls   \n",
       "2                    9   Garment Full body            Basic 1   \n",
       "3                    9            Swimwear           Swimwear   \n",
       "4                    9   Garment Full body  Kids Girl Dresses   \n",
       "\n",
       "                 section_name  \n",
       "0  Womens Swimwear, beachwear  \n",
       "1                 Kids Sports  \n",
       "2              Divided Basics  \n",
       "3  Womens Swimwear, beachwear  \n",
       "4                   Kids Girl  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Inspect the resulting dataset\n",
    "\n",
    "train_negative.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ADo5s7YWHOoR",
   "metadata": {
    "id": "ADo5s7YWHOoR"
   },
   "source": [
    "## **5. Single-Task Ranker**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "857e7128-8eee-4b15-a6dd-14b8ba62301c",
   "metadata": {
    "id": "857e7128-8eee-4b15-a6dd-14b8ba62301c"
   },
   "source": [
    "We begin by training a single-task ranker for the task of predicting whether the user will purchase a given article or not:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "00vOUEIG0jsI",
   "metadata": {
    "id": "00vOUEIG0jsI"
   },
   "outputs": [],
   "source": [
    "def train_single_task_ranker(customer_numerical_features=[],\n",
    "                             article_categorical_feature='',\n",
    "                             n_epochs=5,\n",
    "                             batch_size=1024*4):\n",
    "    \"\"\"\n",
    "    Args:\n",
    "        customer_numerical_features (list): Numerical features used to describe the customer.\n",
    "        article_categorical_feature (str): Single, categorical feature used to describe the article.\n",
    "            In practice, we'd use many more features to describe the article.\n",
    "        n_epochs (int): Number of epochs.\n",
    "        batch_size (int): Batch size.\n",
    "\n",
    "    Returns:\n",
    "        model (tf.Keras.Model): Trained model on train_negative.\n",
    "    \"\"\"\n",
    "\n",
    "    # Assemble training data\n",
    "    X1 = np.asarray(train_negative[customer_numerical_features].values)\n",
    "    X2 = train_negative[article_categorical_feature].astype(str)\n",
    "    y1 = np.asarray(train_negative['label'].values)\n",
    "\n",
    "\n",
    "    # Numerical customer features\n",
    "    input1_layer = tf.keras.Input((len(customer_numerical_features)))\n",
    "    input1 = tf.keras.layers.Dense(50, activation='relu')(input1_layer)\n",
    "\n",
    "    # Categorical article feature\n",
    "    unique_group_names = train_negative[article_categorical_feature].unique().astype(str)\n",
    "    input2_layer = tf.keras.Input((1, ), dtype=tf.string)\n",
    "    input2 = tf.keras.layers.StringLookup(vocabulary=unique_group_names, mask_token=None)(input2_layer)\n",
    "    input2 = tf.keras.layers.Embedding(len(unique_group_names) + 1, 50)(input2)\n",
    "    input2 = tf.keras.layers.Dense(32, activation='relu')(input2)\n",
    "    input2 = tf.keras.layers.Flatten()(input2)\n",
    "\n",
    "    # Combine inputs\n",
    "    combined_layer = tf.keras.layers.concatenate([input1, input2])\n",
    "    combined_layer = tf.keras.layers.Dense(50)(combined_layer)\n",
    "    combined_layer = tf.keras.layers.Dense(1, activation='sigmoid')(combined_layer)\n",
    "\n",
    "\n",
    "    # Compile model\n",
    "    bce = tf.keras.losses.BinaryCrossentropy(from_logits=False)\n",
    "    metric = tf.keras.metrics.BinaryAccuracy()\n",
    "    model = tf.keras.Model([input1_layer, input2_layer], combined_layer)\n",
    "    model.compile(loss=bce,\n",
    "                  optimizer='adam',\n",
    "                  metrics=[metric])\n",
    "\n",
    "    # Train model\n",
    "    model.fit([X1, X2], [y1], epochs=n_epochs, verbose=1, batch_size=batch_size)\n",
    "\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "XTNpoyLq3Inl",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "XTNpoyLq3Inl",
    "outputId": "2bc61478-1e18-48f2-8b57-5d43fa4e2488"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-11 19:24:58.683566: W tensorflow/tsl/platform/profile_utils/cpu_utils.cc:128] Failed to get CPU frequency: 0 Hz\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1296/1296 [==============================] - 14s 11ms/step - loss: 0.6275 - binary_accuracy: 0.6678\n",
      "Epoch 2/5\n",
      "1296/1296 [==============================] - 14s 11ms/step - loss: 0.6264 - binary_accuracy: 0.6683\n",
      "Epoch 3/5\n",
      "1296/1296 [==============================] - 14s 11ms/step - loss: 0.6262 - binary_accuracy: 0.6683\n",
      "Epoch 4/5\n",
      "1296/1296 [==============================] - 14s 11ms/step - loss: 0.6261 - binary_accuracy: 0.6684\n",
      "Epoch 5/5\n",
      "1296/1296 [==============================] - 14s 11ms/step - loss: 0.6260 - binary_accuracy: 0.6684\n"
     ]
    }
   ],
   "source": [
    "# Use customer features and product_group_name to predict whether the customer bought the product or not\n",
    "\n",
    "customer_numerical_features = ['num_articles', 'days_active', 'total_purchases', 'min_price','max_price',\n",
    "                               'mean_price', 'num_unique_product_groups', 'num_unique_departments', 'num_unique_sections']\n",
    "article_categorical_feature = 'product_group_name'\n",
    "single_task_model = train_single_task_ranker(customer_numerical_features=customer_numerical_features,\n",
    "                                             article_categorical_feature=article_categorical_feature)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "MtAo2yv4H4jg",
   "metadata": {
    "id": "MtAo2yv4H4jg"
   },
   "source": [
    "## **6. Evaluation**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "Z5KWMkNs9laZ",
   "metadata": {
    "id": "Z5KWMkNs9laZ"
   },
   "source": [
    "At this point we must ask ourselves whether the model learned to rank our known positive examples above the randomly generated negative samples. During the training process we saw that the binary accuracy did not increase every much between the first and last epochs, and it was also very close to simply predicting the majority class (0) for all instances (67% is 2/3, which is the distribution of 2 generated negatives to every 1 positive). This can be attributed to any or all of several choices:\n",
    "1. Number of generated negatives\n",
    "2. Method for generating negatives\n",
    "3. Model architecture\n",
    "4. Poor model fitting\n",
    "5. Lack of information in our features to describe customers or articles\n",
    "\n",
    "Normally this would be a big red flag - we want to do better than a naive prediction! But here we must consider the following instead: Our true metric/task is ranking, not accuracy. By using an integer-based ranking metric, we coarsely remove the true rank based on the original floating-point predictions. So instead we'll focus on the Mean Reciprocal Rank ([MRR](https://en.wikipedia.org/wiki/Mean_reciprocal_rank)) of the model to give us an idea for how well it does at ranking.\n",
    "<br><br>\n",
    "_Ultimately_, however, we're interested in whether the user will interact with any of the top-k candidates we serve to them from the combination of both the Candidate Generation and Ranking steps. In that case, we'd want to use a metric like Mean Average Precision ([MAP](https://en.wikipedia.org/wiki/Evaluation_measures_(information_retrieval)#Mean_average_precision), [MAPK](https://github.com/benhamner/Metrics/blob/master/Python/ml_metrics/average_precision.py)). This metric, however, is more compute-intensive to calculate and we'd like to show you that there are multiple ways to evaluate a ranker, so we'll be using MRR in this case (and another metric later on!).\n",
    "<br><br>\n",
    "In summary, know that accuracy was used to train the model but we're instead interested in its ability to rank. We will use MRR to understand its current capability to rank given a random negative sampling strategy."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "5isqe-TPu-wN",
   "metadata": {
    "id": "5isqe-TPu-wN"
   },
   "outputs": [],
   "source": [
    "# Evaluate model\n",
    "\n",
    "def evaluate(df, model, \n",
    "             customer_numerical_features=[], \n",
    "             article_categorical_feature='',\n",
    "             single_task=True):\n",
    "    \"\"\"\n",
    "    Args:\n",
    "        df (DataFrame): Train or test dataframe to evaluate.\n",
    "        model (tf.keras.Model): Trained model for inference.\n",
    "        customer_numerical_features (list): Numerical features used to describe the customer during training.\n",
    "        article_categorical_feature (str): Single, categorical feature used to describe the article during training.\n",
    "        single_task (bool): True for single-task, False for multi-task.\n",
    "\n",
    "    Returns:\n",
    "        None\n",
    "    \"\"\"\n",
    "\n",
    "    # Make predictions\n",
    "    X1 = np.asarray(df[customer_numerical_features].values)\n",
    "    X2 = df[article_categorical_feature].astype(str)\n",
    "    if single_task:\n",
    "        preds = model.predict([X1, X2], batch_size=1024*10, verbose=1)\n",
    "    else:\n",
    "        preds, _ = model.predict([X1, X2], batch_size=1024*10, verbose=1)\n",
    "    preds_rounded = (preds>=0.5).astype(int).flatten()\n",
    "\n",
    "    # Calculate MRR\n",
    "    df['prediction_score'] = preds\n",
    "    df.sort_values(by=['customer_id', 'prediction_score'], ascending=False, inplace=True)\n",
    "    reciprocal_rank = df.groupby('customer_id').apply(lambda x: 1/(np.where(x['label']==1)[0][0]+1))\n",
    "    df.drop('prediction_score', axis=1, inplace=True)\n",
    "    print(f'MRR: {reciprocal_rank.mean():.4f}')\n",
    "    print('Distribution:')\n",
    "    display(reciprocal_rank.describe())\n",
    "    \n",
    "    # Show distribution of labels and compare to model accuracy\n",
    "    print('\\nBaseline:')\n",
    "    print(df['label'].value_counts()/len(df), end='\\n\\n')\n",
    "    print(f\"Model Accuracy: {sum(preds_rounded == df['label'].values)/len(preds_rounded):.2f}\", end='\\n\\n')\n",
    "\n",
    "    # Show confusion matrix\n",
    "    print('Confusion Matrix:')\n",
    "    truth = df['label'].values\n",
    "    cm = confusion_matrix(truth, preds_rounded)\n",
    "    cmd = ConfusionMatrixDisplay(cm, display_labels=['0','1'])\n",
    "    cmd.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "QYw0_fr8gUf5",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 664
    },
    "id": "QYw0_fr8gUf5",
    "outputId": "773930f0-5c5f-4a91-d90f-fc953003dcb6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "519/519 [==============================] - 6s 11ms/step\n",
      "MRR: 0.6059\n",
      "Distribution:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    330712.000000\n",
       "mean          0.605853\n",
       "std           0.334009\n",
       "min           0.019231\n",
       "25%           0.333333\n",
       "50%           0.500000\n",
       "75%           1.000000\n",
       "max           1.000000\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Baseline:\n",
      "0    0.666667\n",
      "1    0.333333\n",
      "Name: label, dtype: float64\n",
      "\n",
      "Model Accuracy: 0.66\n",
      "\n",
      "Confusion Matrix:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Evaluate the training set with the single-task ranker\n",
    "\n",
    "evaluate(train_negative, single_task_model, \n",
    "         customer_numerical_features=customer_numerical_features, \n",
    "         article_categorical_feature=article_categorical_feature,\n",
    "         single_task=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "eEvs_YUqzfIp",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 653
    },
    "id": "eEvs_YUqzfIp",
    "outputId": "1149eb0e-0d4b-4e68-da5e-c455e8312b13"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "44/44 [==============================] - 0s 2ms/step\n",
      "MRR: 0.5947\n",
      "Distribution:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    42395.000000\n",
       "mean         0.594709\n",
       "std          0.321536\n",
       "min          0.027027\n",
       "25%          0.333333\n",
       "50%          0.500000\n",
       "75%          1.000000\n",
       "max          1.000000\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Baseline:\n",
      "0    0.666667\n",
      "1    0.333333\n",
      "Name: label, dtype: float64\n",
      "\n",
      "Model Accuracy: 0.66\n",
      "\n",
      "Confusion Matrix:\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAhUAAAGwCAYAAAAe3Ze+AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAA9hAAAPYQGoP6dpAABGtUlEQVR4nO3de1wU9f4/8NcusMt1QURABBHFG6mQaESlZXHEMsu0b2pWeD8amEpe0xBT8/y08nI0rSyxjpZ2SkssioP3RE2UUhPygqLiAoqwgsLC7vz+QCY3L7DujIjzej4e8zjtzHtm38uDw759fz6fGZUgCAKIiIiIbKSu7wSIiIjo/sCigoiIiCTBooKIiIgkwaKCiIiIJMGigoiIiCTBooKIiIgkwaKCiIiIJGFf3wnYwmw2Iy8vD25ublCpVPWdDhERWUkQBFy+fBl+fn5Qq+X7d255eTmMRqPN19FoNHB0dJQgo/tTgy4q8vLyEBAQUN9pEBGRjc6cOQN/f39Zrl1eXo6gQFfoC0w2X8vX1xc5OTksLG6hQRcVbm5uAIDTB1pA58qRHLo/vdCmY32nQCSbKlRiF34Q/57LwWg0Ql9gwumMFtC53fl3heGyGYHhp2A0GllU3EKDLipqhjx0rmqbflGI7mX2Kof6ToFIPtceFHE3hrBd3VRwdbvz9zGDw+y1adBFBRERUV2ZBDNMNjztyiSYpUvmPsWigoiIFMEMAWbceVVhy7lKwTEDIiIikgQ7FUREpAhmmGHLAIZtZysDiwoiIlIEkyDAJNz5EIYt5yoFhz+IiIhIEuxUEBGRInCipvxYVBARkSKYIcDEokJWHP4gIiIiSbBTQUREisDhD/mxqCAiIkXg6g/5cfiDiIiIJMFOBRERKYL52mbL+XR7LCqIiEgRTDau/rDlXKVgUUFERIpgEmDjU0qly+V+xTkVREREJAl2KoiISBE4p0J+LCqIiEgRzFDBBJVN59PtcfiDiIiIJMFOBRERKYJZqN5sOZ9uj0UFEREpgsnG4Q9bzlUKDn8QERGRJNipICIiRWCnQn4sKoiISBHMggpmwYbVHzacqxQc/iAiIiJJsFNBRESKwOEP+bGoICIiRTBBDZMNDXqThLncr1hUEBGRIgg2zqkQOKeiVpxTQURERJJgp4KIiBSBcyrkx6KCiIgUwSSoYRJsmFPB23TXisMfREREJAl2KoiISBHMUMFsw7+lzWCrojYsKoiISBE4p0J+HP4gIiIiSbBTQUREimD7RE0Of9SGRQURESlC9ZwKGx4oxuGPWnH4g4iIiCTBTgURESmC2cZnf3D1R+1YVBARkSJwToX8WFQQEZEimKHmfSpkxjkVREREJAl2KoiISBFMggomGx5fbsu5SsGigoiIFMFk40RNE4c/asXhDyIiIpIEOxVERKQIZkENsw2rP8xc/VErFhVERKQIHP6QH4c/iIiISBLsVBARkSKYYdsKDrN0qdy3WFQQEZEi2H7zKzb3a8OfEBEREUmCnQoiIlIE25/9wX+H14Y/ISIiUgQzVDZv1pg3bx66du0KNzc3eHt7o2/fvsjOzraIeeKJJ6BSqSy20aNHW8Tk5uaid+/ecHZ2hre3NyZNmoSqqiqLmG3btqFz587QarUIDg5GUlLSDfksW7YMLVq0gKOjIyIiIrBv3z6L4+Xl5YiNjUXjxo3h6uqK/v37Iz8/36rPzKKCiIgUoaZTYctmje3btyM2NhZ79uxBamoqKisr0bNnT5SVlVnEjRw5EufPnxe3+fPn/5WzyYTevXvDaDRi9+7dWL16NZKSkpCQkCDG5OTkoHfv3ujRowcyMzMxfvx4jBgxAj/99JMYs27dOsTHx2PmzJk4cOAAQkNDER0djYKCAjFmwoQJ2LRpE77++mts374deXl56Nevn1WfWSUIDfduHgaDAe7u7rj0Z0vo3Fgf0f0p2i+svlMgkk2VUIlt+A4lJSXQ6XSyvEfNd8XC/Y/AyfXOR/2vllZhQpfdd5xrYWEhvL29sX37dnTv3h1AdaciLCwMixYtuuk5P/74I5599lnk5eXBx8cHALBixQpMmTIFhYWF0Gg0mDJlCjZv3ozDhw+L5w0cOBDFxcVISUkBAERERKBr165YunQpAMBsNiMgIABjx47F1KlTUVJSgiZNmmDt2rV48cUXAQBZWVlo37490tPT8fDDD9fpM/KbmIiIFKHm5le2bEB1kXL9VlFRUaf3LykpAQB4enpa7F+zZg28vLzQoUMHTJs2DVeuXBGPpaeno2PHjmJBAQDR0dEwGAw4cuSIGBMVFWVxzejoaKSnpwMAjEYjMjIyLGLUajWioqLEmIyMDFRWVlrEtGvXDs2bNxdj6oITNYmISBHMggpmW+5Tce3cgIAAi/0zZ85EYmLi7c81mzF+/Hg8+uij6NChg7j/5ZdfRmBgIPz8/PD7779jypQpyM7OxrfffgsA0Ov1FgUFAPG1Xq+/bYzBYMDVq1dx6dIlmEymm8ZkZWWJ19BoNPDw8LghpuZ96oJFBRERkRXOnDljMfyh1WprPSc2NhaHDx/Grl27LPaPGjVK/O+OHTuiadOmeOqpp3DixAm0atVKuqTvEg5/EBGRIphtHPqoufmVTqez2GorKuLi4pCcnIytW7fC39//trEREREAgOPHjwMAfH19b1iBUfPa19f3tjE6nQ5OTk7w8vKCnZ3dTWOuv4bRaERxcfEtY+qCRQURESlCzVNKbdmsIQgC4uLisGHDBmzZsgVBQUG1npOZmQkAaNq0KQAgMjIShw4dslilkZqaCp1Oh5CQEDEmLS3N4jqpqamIjIwEAGg0GoSHh1vEmM1mpKWliTHh4eFwcHCwiMnOzkZubq4YUxcc/iAiIpJBbGws1q5di++++w5ubm7i3AR3d3c4OTnhxIkTWLt2LZ555hk0btwYv//+OyZMmIDu3bujU6dOAICePXsiJCQEr776KubPnw+9Xo8ZM2YgNjZW7JCMHj0aS5cuxeTJkzFs2DBs2bIF69evx+bNm8Vc4uPjERMTgy5duuChhx7CokWLUFZWhqFDh4o5DR8+HPHx8fD09IROp8PYsWMRGRlZ55UfAIsKIiJSCBNUMFl5A6u/n2+N5cuXA6heNnq9VatWYciQIdBoNPjf//4nfsEHBASgf//+mDFjhhhrZ2eH5ORkjBkzBpGRkXBxcUFMTAzeeecdMSYoKAibN2/GhAkTsHjxYvj7+2PlypWIjo4WYwYMGIDCwkIkJCRAr9cjLCwMKSkpFpM3Fy5cCLVajf79+6OiogLR0dH48MMPrfrMvE8F0T2O96mg+9ndvE/FrL1RcLThPhXlpVWYGfE/WXNt6PhNTERERJLg8AcRESmCCdYPYfz9fLo9FhVERKQId7KC4+/n0+2xqCAiIkXgo8/lx58QERERSYKdCiIiUgQBKphtmFMh2HCuUrCoICIiReDwh/z4EyIiIiJJsFNBRESKINWjz+nWWFQQEZEi1Dxt1Jbz6fb4EyIiIiJJsFNBRESKwOEP+bGoICIiRTBDDbMNDXpbzlUK/oSIiIhIEuxUEBGRIpgEFUw2DGHYcq5SsKggIiJF4JwK+bGoICIiRRBsfEqpwDtq1oo/ISIiIpIEOxVERKQIJqhgsuGhYLacqxQsKoiISBHMgm3zIsyChMncpzj8QURERJJgp+I+tWl1Y2z+3Av5ZzQAgMC25Rg8QY+uT16W5f0EAfh8gS9S1jZGqcEOIV3K8Ma/zqBZS6MY89pDIcg/q7E4b9i0PAwYWyBLTqRMA+Ly8egzJQgIroCxXI0/9jvj07lNcfaEoxjjoDVj1Mw8PPFcMRy0AjK2ueHf05qh+IKDGNMm9AqGvXUerTtdgSCokJ3phE/n+OHkH07XvZuAF0cX4unBF+HtXwlDkR2SV3vhyyU+d/ETU12ZbZyoacu5SsGi4j7VpGklhr2Vh2ZBFRAEFVK/boTEoUFY9vOfaNG23OrrffGeL/LPajBxUe5Nj69f5o3vPmuCiYtOw7e5EavnN8VbL7fCJ9uyoHH8q2f42qTzeHrwRfG1s6vZ+g9HdBudIsuwKckLf2Y6w85ewJCp5/Hulycx8vG2qLhqBwAYnZiHh6IMmPPPQJQZ7BA79xwSPj2F+OdbAwAcnU2Yu+Yk9qTqsPSt1rCzA16dqMfctSfxSpcQmKqqW+hjZuch/PHL+GS2H3KOOsLNwwRdI1O9fXa6PTNUMNswL8KWc5Xinii7li1bhhYtWsDR0RERERHYt29ffafU4D3c04CHnrqMZi2N8G9VgaFT9XB0MSMrwxkAUFpih4VvBuClDh3wQpuOmPx/rXDiiGMtV705QQA2rmyCQeP0eKSXAS1DyjF5yWlczHfA7hR3i1gnVzM8vavEzdGZRQVJa/rglkhd74nTfzri5B9OeH98c/j4V6J1p6sAAGc3E6IHFeGjRD/89osbjh9yxgfxAXig6xW061wGAAgIroDO04TPF/ji7AlHnP7TEf/5wAee3lXw8TdeiynHs69dQOLQFtjzszvyz2hx/JAzDuxwq7fPTlTf6r2oWLduHeLj4zFz5kwcOHAAoaGhiI6ORkEBW+JSMZmAbRs9UHFFjfZdqv9ozhnVAsUX7DFnzQksTclGcIermPpSMAyX7Ky+vj5Xg6ICB3TuViruc9GZ0e7BKzia4WIRu36pN158oANe/0cbfP1hE5iqbPtsRLVx0VV3Di4XV/9ut+50BQ4aAQd3/vXlf+a4I/LPOqB9+BUAwNkTWpQU2SF6UBHsHczQOJrRa1ARTv+phf7akOLDPQ04n6tFRJQBq/ccxeq9f2D8e2fg5sFf6ntVzR01bdno9up9+OODDz7AyJEjMXToUADAihUrsHnzZnz22WeYOnVqPWfXsOUcdcT4Pq1hrFDDycWMhE9zENimAof3uiA70xnrfj8MjbZ6aGLUzDzs/skduzZ74JlXLtZyZUtFBdW/Rh5NKi32ezSpFI8BwPPDCxHc8SrcPKrwx34XrJrXFEUFDvhnYp6Nn5To5lQqAaNnncPhfc44nV09F8LTuwrGChXKDJYFdHGhPTy9q3+Hr5bZYVL/Vkj87BReHp8PAMjL0eKtQS1hNlV/sTRtboRPMyO6PVuCBW8EQG0H/HPWOcz4+DSmvNTqLn5KqivOqZBfvRYVRqMRGRkZmDZtmrhPrVYjKioK6enpN8RXVFSgoqJCfG0wGO5Kng2Vf6sKfJiajSuX7bAz2QPvjQvEgm+P4eQfTigvU+P/HuhgEW8sVyPvVPW/wg7tdcGMwS3FY1WVKgiCCjuT/xrOGDf/LJ7sd6nO+fT/Z6H43y1DyuHgIGDxlAAMnXZeLG6IpBT37jkEtivHm32DrTpP42hG/PtnceRXF8x7PRBqu+oJmbO/yMHYZ1rDWK6GSi1A4yhgwbjmOHdSCwBY+GYAlv10DP6tyi0mhhIpRb0WFRcuXIDJZIKPj+VMaR8fH2RlZd0QP2/ePMyaNetupdfgOWgENAuqHv9t3ekqsjOdsXFlE/g2N8LTpxLz/3v8hnNcr7WK23S6gg9Ts8X9333aBBf0Dhg+/a+uQqMm1W1eT+/q/y0udEBjn79av8WFDmj1wNVb5te28xWYqlTIP6NBQHDFLeOI7kTs3LOI+IcBb77QChfO/7XqqKjAHhqtABedyaJb4dGkCkUF1as/erxwCT4BRozvEwzhWsv7X7FO+OboEURGl2D7d41QVOCAqkqIBQUA5B6rLiS8m1WyqLgHmWHjsz84UbNWDaqXM23aNJSUlIjbmTNn6julBkUQgEqjGsEdr6CowAF29kCzIKPF5t64uqjQOgkW+908THB2NVvsq1m54dvcCE/vShzc5Sq+V9llNbIOOqN9eNkt8zl5xAlqtQAPL45Bk5QExM49i0d6lWDy/7VC/hmtxdFjvzuj0qjCg4/9tbzav1U5fPwrcfTaRGatkxlmc/X/Z2qYzSoIAqC+9lfzyK8usHcAmgb+VRD7t6z+778vnaZ7g3Bt9cedbgKLilrVa6fCy8sLdnZ2yM/Pt9ifn58PX1/fG+K1Wi20Wu0N++lGn73bFF2fNKBJs0pcLVVj64ZG+H23K+auPYHO3UvRPrwMs4YGYcSMPDRrVYGLenvsS9Ph0adL0Cb01t2Fm1GpgL4jCvHlYh80C6oQl5Q29qnEI71KAAB/7HdG1kEXhD5yGc6uZhzNcMGKmX54sv8luHlwCR5JJ+7dc+jxwiUkDg3C1VI1Gl2b61N22Q7GcjWuXLbDT196YlRiHi4X26Psshqxc89V/44eqJ5YfHCHG0bOOI+4d8/hu8+8oFYDL8UVwFQF/PaL67UYVxz73QnxH5zBipl+UKmq3ztju6tF94LuHXxKqfzqtajQaDQIDw9HWloa+vbtCwAwm81IS0tDXFxcfabW4BVfsMeCNwJRVGAPZzcTgtqXY+7aEwh/vHqFxpz/nETSv5ri/fgAlFy0R6MmVej4cOkddw1eii1A+RU1Fk8OQKnBDg90LcPcNSfFe1Q4aARs/84D/3nfF5VGFXwDjOg3qhD9RhXWcmUi6/QZUj3R+L1vT1jsf298AFLXewIAViT6wSwAb39yCg5aAfu3uWHptGZi7Jnjjpg5JAiD4/VYtOkYBLMKxw87YfrgluIQiSCokBAThNg55/DetydQfkWN/Vvd8PEsv7v0SYnuPSpBEOp1hty6desQExODjz76CA899BAWLVqE9evXIysr64a5Fn9nMBjg7u6OS3+2hM6tQY3kENVZtF9YfadAJJsqoRLb8B1KSkqg0+lkeY+a74oXUofCweXOh6Yqy4zY8I9Vsuba0NX7ktIBAwagsLAQCQkJ0Ov1CAsLQ0pKSq0FBRERkTU4/CG/ei8qACAuLo7DHURERA3cPVFUEBERyY3P/pAfiwoiIlIEDn/Ij7MbiYiISBLsVBARkSKwUyE/FhVERKQILCrkx+EPIiIikgQ7FUREpAjsVMiPRQURESmCANuWhdbr7acbCBYVRESkCOxUyI9zKoiIiEgS7FQQEZEisFMhPxYVRESkCCwq5MfhDyIiIpIEOxVERKQI7FTIj0UFEREpgiCoINhQGNhyrlJw+IOIiIgkwU4FEREpghkqm25+Zcu5SsGigoiIFIFzKuTH4Q8iIiIZzJs3D127doWbmxu8vb3Rt29fZGdnW8SUl5cjNjYWjRs3hqurK/r374/8/HyLmNzcXPTu3RvOzs7w9vbGpEmTUFVVZRGzbds2dO7cGVqtFsHBwUhKSrohn2XLlqFFixZwdHREREQE9u3bZ3UutWFRQUREilAzUdOWzRrbt29HbGws9uzZg9TUVFRWVqJnz54oKysTYyZMmIBNmzbh66+/xvbt25GXl4d+/fqJx00mE3r37g2j0Yjdu3dj9erVSEpKQkJCghiTk5OD3r17o0ePHsjMzMT48eMxYsQI/PTTT2LMunXrEB8fj5kzZ+LAgQMIDQ1FdHQ0CgoK6pxLXagEQWiwz0gxGAxwd3fHpT9bQufG+ojuT9F+YfWdApFsqoRKbMN3KCkpgU6nk+U9ar4runw7HvYu2ju+TlVZBfb3W4QzZ85Y5KrVaqHV1n7dwsJCeHt7Y/v27ejevTtKSkrQpEkTrF27Fi+++CIAICsrC+3bt0d6ejoefvhh/Pjjj3j22WeRl5cHHx8fAMCKFSswZcoUFBYWQqPRYMqUKdi8eTMOHz4svtfAgQNRXFyMlJQUAEBERAS6du2KpUuXAgDMZjMCAgIwduxYTJ06tU651AW/iYmISBGk6lQEBATA3d1d3ObNm1en9y8pKQEAeHp6AgAyMjJQWVmJqKgoMaZdu3Zo3rw50tPTAQDp6eno2LGjWFAAQHR0NAwGA44cOSLGXH+NmpiaaxiNRmRkZFjEqNVqREVFiTF1yaUuOFGTiIjICjfrVNTGbDZj/PjxePTRR9GhQwcAgF6vh0ajgYeHh0Wsj48P9Hq9GHN9QVFzvObY7WIMBgOuXr2KS5cuwWQy3TQmKyurzrnUBYsKIiJSBMHG1R81nQqdTmf1UE1sbCwOHz6MXbt23fH7NwQc/iAiIkUQAAiCDdsdvm9cXBySk5OxdetW+Pv7i/t9fX1hNBpRXFxsEZ+fnw9fX18x5u8rMGpe1xaj0+ng5OQELy8v2NnZ3TTm+mvUlktdsKggIiKSgSAIiIuLw4YNG7BlyxYEBQVZHA8PD4eDgwPS0tLEfdnZ2cjNzUVkZCQAIDIyEocOHbJYpZGamgqdToeQkBAx5vpr1MTUXEOj0SA8PNwixmw2Iy0tTYypSy51weEPIiJSBDNUUN3FO2rGxsZi7dq1+O677+Dm5ibOTXB3d4eTkxPc3d0xfPhwxMfHw9PTEzqdDmPHjkVkZKS42qJnz54ICQnBq6++ivnz50Ov12PGjBmIjY0V53KMHj0aS5cuxeTJkzFs2DBs2bIF69evx+bNm8Vc4uPjERMTgy5duuChhx7CokWLUFZWhqFDh4o51ZZLXbCoICIiRbjbDxRbvnw5AOCJJ56w2L9q1SoMGTIEALBw4UKo1Wr0798fFRUViI6OxocffijG2tnZITk5GWPGjEFkZCRcXFwQExODd955R4wJCgrC5s2bMWHCBCxevBj+/v5YuXIloqOjxZgBAwagsLAQCQkJ0Ov1CAsLQ0pKisXkzdpyqQvep4LoHsf7VND97G7ep6LT1xNh53zn96kwXanA7//3nqy5NnTsVBARkSKYBRVUfPaHrFhUEBGRItSs4rDlfLo9jhkQERGRJNipICIiRbjbEzWViEUFEREpAosK+bGoICIiReBETflxTgURERFJgp0KIiJSBK7+kB+LCiIiUoTqosKWORUSJnOf4vAHERERSYKdCiIiUgSu/pAfiwoiIlIE4dpmy/l0exz+ICIiIkmwU0FERIrA4Q/5saggIiJl4PiH7FhUEBGRMtjYqQA7FbXinAoiIiKSBDsVRESkCLyjpvxYVBARkSJwoqb8OPxBREREkmCngoiIlEFQ2TbZkp2KWrGoICIiReCcCvlx+IOIiIgkwU4FEREpA29+JTsWFUREpAhc/SG/OhUV33//fZ0v+Nxzz91xMkRERNRw1amo6Nu3b50uplKpYDKZbMmHiIhIPhzCkFWdigqz2Sx3HkRERLLi8If8bFr9UV5eLlUeRERE8hIk2Oi2rC4qTCYTZs+ejWbNmsHV1RUnT54EALz99tv49NNPJU+QiIiIGgari4q5c+ciKSkJ8+fPh0ajEfd36NABK1eulDQ5IiIi6agk2Oh2rC4qPv/8c3z88ccYPHgw7OzsxP2hoaHIysqSNDkiIiLJcPhDdlYXFefOnUNwcPAN+81mMyorKyVJioiIiBoeq4uKkJAQ7Ny584b9//3vf/Hggw9KkhQREZHk2KmQndV31ExISEBMTAzOnTsHs9mMb7/9FtnZ2fj888+RnJwsR45ERES241NKZWd1p+L555/Hpk2b8L///Q8uLi5ISEjA0aNHsWnTJvzjH/+QI0ciIiJqAO7o2R/dunVDamqq1LkQERHJho8+l98dP1Bs//79OHr0KIDqeRbh4eGSJUVERCQ5PqVUdlYXFWfPnsWgQYPwyy+/wMPDAwBQXFyMRx55BF999RX8/f2lzpGIiIgaAKvnVIwYMQKVlZU4evQoioqKUFRUhKNHj8JsNmPEiBFy5EhERGS7momatmx0W1Z3KrZv347du3ejbdu24r62bdvi3//+N7p16yZpckRERFJRCdWbLefT7VldVAQEBNz0Jlcmkwl+fn6SJEVERCQ5zqmQndXDHwsWLMDYsWOxf/9+cd/+/fsxbtw4vPfee5ImR0RERA1HnToVjRo1gkr111hSWVkZIiIiYG9ffXpVVRXs7e0xbNgw9O3bV5ZEiYiIbMKbX8muTkXFokWLZE6DiIhIZhz+kF2dioqYmBi58yAiIqIG7o5vfgUA5eXlMBqNFvt0Op1NCREREcmCnQrZWT1Rs6ysDHFxcfD29oaLiwsaNWpksREREd2T+JRS2VldVEyePBlbtmzB8uXLodVqsXLlSsyaNQt+fn74/PPP5ciRiIiIGgCrhz82bdqEzz//HE888QSGDh2Kbt26ITg4GIGBgVizZg0GDx4sR55ERES24eoP2VndqSgqKkLLli0BVM+fKCoqAgA89thj2LFjh7TZERERSaTmjpq2bHR7VhcVLVu2RE5ODgCgXbt2WL9+PYDqDkbNA8aIiIiUbseOHejTpw/8/PygUqmwceNGi+NDhgyBSqWy2Hr16mURU1RUhMGDB0On08HDwwPDhw9HaWmpRczvv/+Obt26wdHREQEBAZg/f/4NuXz99ddo164dHB0d0bFjR/zwww8WxwVBQEJCApo2bQonJydERUXh2LFjVn9mq4uKoUOH4rfffgMATJ06FcuWLYOjoyMmTJiASZMmWZ0AERHRXXGXJ2qWlZUhNDQUy5Ytu2VMr169cP78eXH78ssvLY4PHjwYR44cQWpqKpKTk7Fjxw6MGjVKPG4wGNCzZ08EBgYiIyMDCxYsQGJiIj7++GMxZvfu3Rg0aBCGDx+OgwcPom/fvujbty8OHz4sxsyfPx9LlizBihUrsHfvXri4uCA6Ohrl5eVWfWaVIAg2NXROnz6NjIwMBAcHo1OnTrZcymoGgwHu7u649GdL6Nysro+IGoRov7D6ToFINlVCJbbhO5SUlMh2S4Ka74rm/28O1E6Od3wd89Vy5E6ZcUe5qlQqbNiwweKu00OGDEFxcfENHYwaR48eRUhICH799Vd06dIFAJCSkoJnnnkGZ8+ehZ+fH5YvX47p06dDr9dDo9EAqP4H/8aNG5GVlQUAGDBgAMrKypCcnCxe++GHH0ZYWBhWrFgBQRDg5+eHN998ExMnTgQAlJSUwMfHB0lJSRg4cGCdP6fN38SBgYHo16/fXS8oiIiIrKGCjXMqrl3HYDBYbBUVFXec07Zt2+Dt7Y22bdtizJgxuHjxongsPT0dHh4eYkEBAFFRUVCr1di7d68Y0717d7GgAIDo6GhkZ2fj0qVLYkxUVJTF+0ZHRyM9PR0AkJOTA71ebxHj7u6OiIgIMaau6rT6Y8mSJXW+4BtvvGFVAkRERA1JQECAxeuZM2ciMTHR6uv06tUL/fr1Q1BQEE6cOIG33noLTz/9NNLT02FnZwe9Xg9vb2+Lc+zt7eHp6Qm9Xg8A0Ov1CAoKsojx8fERjzVq1Ah6vV7cd33M9de4/rybxdRVnYqKhQsX1uliKpWqXoqK0G+GQe145y0tontZMPbUdwpE9weJlpSeOXPGYvhDq9Xe0eWuH1bo2LEjOnXqhFatWmHbtm146qmn7jzPelSnoqJmtQcREVGDJdFtunU6nSzzP1q2bAkvLy8cP34cTz31FHx9fVFQUGARU1VVhaKiIvj6+gIAfH19kZ+fbxFT87q2mOuP1+xr2rSpRUxYWJhVn4GzG4mIiO4BZ8+excWLF8Uv9sjISBQXFyMjI0OM2bJlC8xmMyIiIsSYHTt2oLKyUoxJTU1F27ZtxUdnREZGIi0tzeK9UlNTERkZCQAICgqCr6+vRYzBYMDevXvFmLpiUUFERMpwl5eUlpaWIjMzE5mZmQCqu/6ZmZnIzc1FaWkpJk2ahD179uDUqVNIS0vD888/j+DgYERHRwMA2rdvj169emHkyJHYt28ffvnlF8TFxWHgwIHw8/MDALz88svQaDQYPnw4jhw5gnXr1mHx4sWIj48X8xg3bhxSUlLw/vvvIysrC4mJidi/fz/i4uIAVE9dGD9+PObMmYPvv/8ehw4dwmuvvQY/Pz+L1Sp1YdNTSomIiBoKW++Kae25+/fvR48ePcTXNV/0MTExWL58OX7//XesXr0axcXF8PPzQ8+ePTF79myLORpr1qxBXFwcnnrqKajVavTv399i8YS7uzt+/vlnxMbGIjw8HF5eXkhISLC4l8UjjzyCtWvXYsaMGXjrrbfQunVrbNy4ER06dBBjJk+ejLKyMowaNQrFxcV47LHHkJKSAkcr5yvafJ+K+iSuPZ43hxM16b4VHM+JmnT/upv3qWgxd65N3xXm8nKcmj5d1lwbOnYqiIhIGSSaqEm3dkdzKnbu3IlXXnkFkZGROHfuHADgiy++wK5duyRNjoiISDJ3eU6FElldVHzzzTeIjo6Gk5MTDh48KN5JrKSkBO+++67kCRIREVHDYHVRMWfOHKxYsQKffPIJHBwcxP2PPvooDhw4IGlyREREUuGjz+Vn9ZyK7OxsdO/e/Yb97u7uKC4uliInIiIi6Ul0R026Nas7Fb6+vjh+/PgN+3ft2oWWLVtKkhQREZHkOKdCdlYXFSNHjsS4ceOwd+9eqFQq5OXlYc2aNZg4cSLGjBkjR45ERETUAFg9/DF16lSYzWY89dRTuHLlCrp37w6tVouJEydi7NixcuRIRERks7t98yslsrqoUKlUmD59OiZNmoTjx4+jtLQUISEhcHV1lSM/IiIiafA+FbK745tfaTQahISESJkLERERNWBWFxU9evSASnXrGbBbtmyxKSEiIiJZ2LoslJ2KWlldVPz92eqVlZXIzMzE4cOHERMTI1VeRERE0uLwh+ysLioWLlx40/2JiYkoLS21OSEiIiJqmO7o2R8388orr+Czzz6T6nJERETS4n0qZCfZU0rT09Otfu46ERHR3cIlpfKzuqjo16+fxWtBEHD+/Hns378fb7/9tmSJERERUcNidVHh7u5u8VqtVqNt27Z455130LNnT8kSIyIioobFqqLCZDJh6NCh6NixIxo1aiRXTkRERNLj6g/ZWTVR087ODj179uTTSImIqMHho8/lZ/Xqjw4dOuDkyZNy5EJEREQNmNVFxZw5czBx4kQkJyfj/PnzMBgMFhsREdE9i8tJZVXnORXvvPMO3nzzTTzzzDMAgOeee87idt2CIEClUsFkMkmfJRERka04p0J2dS4qZs2ahdGjR2Pr1q1y5kNEREQNVJ2LCkGoLtEef/xx2ZIhIiKSC29+JT+rlpTe7umkRERE9zQOf8jOqqKiTZs2tRYWRUVFNiVEREREDZNVRcWsWbNuuKMmERFRQ8DhD/lZVVQMHDgQ3t7ecuVCREQkHw5/yK7O96ngfAoiIiK6HatXfxARETVI7FTIrs5FhdlsljMPIiIiWXFOhfysfvQ5ERFRg8ROheysfvYHERER0c2wU0FERMrAToXsWFQQEZEicE6F/Dj8QURERJJgp4KIiJSBwx+yY1FBRESKwOEP+XH4g4iIiCTBTgURESkDhz9kx6KCiIiUgUWF7Dj8QURERJJgp4KIiBRBdW2z5Xy6PRYVRESkDBz+kB2LCiIiUgQuKZUf51QQERGRJNipICIiZeDwh+xYVBARkXKwMJAVhz+IiIhIEuxUEBGRInCipvxYVBARkTJwToXsOPxBREREkmBRQUREilAz/GHLZo0dO3agT58+8PPzg0qlwsaNGy2OC4KAhIQENG3aFE5OToiKisKxY8csYoqKijB48GDodDp4eHhg+PDhKC0ttYj5/fff0a1bNzg6OiIgIADz58+/IZevv/4a7dq1g6OjIzp27IgffvjB6lzqgkUFEREpgyDBZoWysjKEhoZi2bJlNz0+f/58LFmyBCtWrMDevXvh4uKC6OholJeXizGDBw/GkSNHkJqaiuTkZOzYsQOjRo0SjxsMBvTs2ROBgYHIyMjAggULkJiYiI8//liM2b17NwYNGoThw4fj4MGD6Nu3L/r27YvDhw9blUtdqARBaLCjRAaDAe7u7mg+bw7Ujo71nQ6RLILj99R3CkSyqRIqsQ3foaSkBDqdTpb3qPmu6Dj8Xdhp7vy7wmQsx6FP38KZM2csctVqtdBqtbc9V6VSYcOGDejbty+A6s6An58f3nzzTUycOBEAUFJSAh8fHyQlJWHgwIE4evQoQkJC8Ouvv6JLly4AgJSUFDzzzDM4e/Ys/Pz8sHz5ckyfPh16vR4ajQYAMHXqVGzcuBFZWVkAgAEDBqCsrAzJycliPg8//DDCwsKwYsWKOuVSV+xUEBGRIkg1/BEQEAB3d3dxmzdvntW55OTkQK/XIyoqStzn7u6OiIgIpKenAwDS09Ph4eEhFhQAEBUVBbVajb1794ox3bt3FwsKAIiOjkZ2djYuXbokxlz/PjUxNe9Tl1zqiqs/iIhIGSRa/XGzToW19Ho9AMDHx8div4+Pj3hMr9fD29vb4ri9vT08PT0tYoKCgm64Rs2xRo0aQa/X1/o+teVSVywqiIhIGSQqKnQ6nWxDNQ0dhz+IiIjuMl9fXwBAfn6+xf78/HzxmK+vLwoKCiyOV1VVoaioyCLmZte4/j1uFXP98dpyqSsWFUREpAh3e0np7QQFBcHX1xdpaWniPoPBgL179yIyMhIAEBkZieLiYmRkZIgxW7ZsgdlsRkREhBizY8cOVFZWijGpqalo27YtGjVqJMZc/z41MTXvU5dc6opFBRERKcNdXlJaWlqKzMxMZGZmAqieEJmZmYnc3FyoVCqMHz8ec+bMwffff49Dhw7htddeg5+fn7hCpH379ujVqxdGjhyJffv24ZdffkFcXBwGDhwIPz8/AMDLL78MjUaD4cOH48iRI1i3bh0WL16M+Ph4MY9x48YhJSUF77//PrKyspCYmIj9+/cjLi4OAOqUS11xTgUREZEM9u/fjx49eoiva77oY2JikJSUhMmTJ6OsrAyjRo1CcXExHnvsMaSkpMDxulskrFmzBnFxcXjqqaegVqvRv39/LFmyRDzu7u6On3/+GbGxsQgPD4eXlxcSEhIs7mXxyCOPYO3atZgxYwbeeusttG7dGhs3bkSHDh3EmLrkUhe8TwXRPY73qaD72d28T0XYq3Ntvk9F5hfTZc21oWOngoiIlIEPFJMd51QQERGRJNipICIiRbB1BYeUqz/uVywqiIhIGTj8ITsOfxAREZEk2KkgIiJF4PCH/FhUEBGRMnD4Q3YsKoiISBHYqZAf51QQERGRJNipICIiZeDwh+xYVBARkWJwCENeHP4gIiIiSbBTQUREyiAI1Zst59NtsaggIiJF4OoP+XH4g4iIiCTBTgURESkDV3/IjkUFEREpgspcvdlyPt0ehz+IiIhIEuxU3Ge6NsnDyHa/oYPnBfg4XcHonT2Rei7oprGzu+zAy8FHMftAJJL+7CTub+FWjKmhexDeJB8OahOyixtj4aEu2FPQTIw5MfCjG643bvdTSM4NBgCEe53HlNC9aKkrhpNdFc5dccOXx9tj1XXvM7r9QUT756ClrhgVJjscuOCL//dbBHIue0j00yAl6hBRiv97vRCtO15BY98qJA5rgfQU9+siBLw2KR+9Xr4IV50Jf+x3wZKp/sjL0YoRwR2vYPj082gTegVmkwq7fnDHR4l+KL9iJ8aEPXYZMZP1aNGuHOVX1Pjf142w6l9NYTap7uKnJatw+EN27FTcZ5ztq5BV3BiJ+x+7bVzPZjkIa1wA/RXnG46t7JYCe7WAV7Y8i74/9cfRYk980j0FXo5XLOIm730CERtfFbefz7YQj12tcsAXxzpgUNpz6PnjACw78iDiO/2Kga3+EGMivPPwn+MP4MXUvnht27OwV5ux+onNcLKrtO2HQIrm6GzGySOOWPqW/02PvxRbiOeHFeLfU/0x7tnWKL+ixrtrT8JBW93b9vSpxL++Oom8HC3GPdsa0we3RGDbckxcdEa8RsuQq5j9RQ72b3VDbM82eHd0IB7uacDw6efvymekO1Oz+sOWjW6vXouKHTt2oE+fPvDz84NKpcLGjRvrM537wvbzzfHBoYfw8y26EwDg41SGhPBfEJ/+JKoEy1+BRpqrCNKVYMXRMGSXNMapUncs+C0CzvZVaONeZBFrMGpwodxZ3IzmvxpffxR7YVNuMI4ZPHGuzA3fnW6Dnef90aWJXowZur03vslpi2MGT2QVN8bkvU+gmUspOngWSvTTICXav1WH1fObYrdFd6KGgL4jCvHlYh+k/+SOnKNOmP9GczT2qcQjvUoAABFRBlRVqbD0rWY4e8IRf/7mjCVT/NHt2RL4tagAADz+XDFyjjpizUJf5J3S4tAeV6yc0xR9Yi7AycV0Fz8tWaXmPhW2bHRb9VpUlJWVITQ0FMuWLavPNBRFBQHvP7wFK7NCcczgecPxS0ZHnDB4oF+LP+FkVwk7lRmDWh3FhXInHC5qYhGbGL4Lv76wGt/+41u8GJSF2/UGQzwuoLNXPvYVNL1ljJuDEQBQYnS8sw9HVAvf5kY09qnCgZ1u4r4rl+2QddAZ7cOrO3EOWjOqKlUQhL+GMYzl1X8qH3iorDpGI6CywvLPp7FcDa2TgNadrsr9MYjuWfU6p+Lpp5/G008/Xef4iooKVFRUiK8NBoMcad3X/tk+E1WCGkl/drhFhAqvbe2NFd1+wu8vfgazoMLFCicM3fYMDJV/jTkvPNQF6fnNcLXKHo/5nsU7XXbBxb4Sq491tLjaruf+A0/tVdirBCw5Eo71J9vf4l0FzHhwN/YX+uLPkhuLHSIpeHpXAQCKCy3/9BUX2sPTu3rY7bddbvjnzDy8OKYAG1d6wdHZjGFvnb92fnXM/u1u6DuyEE/0vYQd33ugkXcVBk/Ir47x4fDdvYo3v5Jfg5qoOW/ePMyaNau+02iwOjQqxJA2h/DcT/0B3GoymYDE8F24WO6EgWnPo9xkh5daZuHj7il44ecXUFjuAgBYeiRcPOOPYi8421diZPvfbigqBqY9B2f7SjzYuACTQvfi9GV3bLo2mfN6s8J3oY1HEQb873mpPi7RHTn9pyPeG98co2bmYdi08zCZVPjuMy8UFdiL3YsD292wcrYf3vjXWUxekotKoxprFnmj48NlELjs8N7FiZqya1BFxbRp0xAfHy++NhgMCAgIqMeMGpYuTc6jseNV7HxujbjPXi3grbA9GNr2EB7fNBiP+JzDk3656PztEJRWaQAAMzOa4DHfs+gX9Cc+OvrgTa+dedEbYzscgEZtgtH81wz5s2U6AMCfJY3h5XgVb3TYf0NRMbPzLjzZ7DQGpj0H/VVXqT82kaiooPpPnkeTKhQVOIj7PZpU4cQRJ/H11g2NsHVDI3h4VaL8ihqCAPQbVYjzpzVizLcfN8G3H3vB06cKpSV28PE3Yvhbepw//VdHj0hpGlRRodVqodXy/7B3auOpNtidbzkjftXjm7HxVBv8N6ctAMDRrro9bP5bJ8MsqKC+Te8vpNFFFFdoLQqKv1OpBGjsrp/EJmBm51/Q0z8Hg7c8JxYgRHLR52pwMd8eDz52GSevFRHOria0e/AKkj9vfEN88YXqwqPnwIuorFDjwA63v0WoUJRfHdPjhWIUnHPA8UNOoHsThz/k16CKCqqds30lAl1LxNf+LpfR3uMCio1anL/ihuK/TYKsEtQoLHcS7w1x8IIPSiq1WBCxFf8+Eo5ykx0GtjwKf5fL2JoXCAB40u8UvByvIvOiDypMdnjU9yzGhBzEyqy/7kHxSvBh5F1xxUlDIwBAV+/zGNHuN3x+3VyOWeG78FzgcfxzZzRKqxzEJauXKzWoMPFXk+6Mo7MJfkFG8bVvgBEtH7iKy8V2KDynwcaVTTBoXAHO5Wihz9UgZrIeF/MdLFaLPDf0Av7Y74yrZXbo3P0yRrydh8/ebYoyw19F84tjCrB/qxsEswqPPlOCl2ILMHd0IMxm3qfinsWnlMqOf7nvMx09C7H2yU3i6xmd0wEA3+S0weS9PWo9/5LRCcO2PYP4Tvvwnx6bYK8241hJI4zeFY2s4up/yVWZ1Xil9RFMfzAdKgg4XeqOdw9G4qsTf03CVKuASZ32wd/1MkxmNU6X6jD/twh8eTxEjHmldfU9K758apNFDpP3PoFvrnVOiKzVJvQqFnxzQnw9elYeAODndY3w/oTmWL+sCRydzRg3/yxcdSYc+dUF0we3tFjN0TbsCl59Uw9HFzPOHtdiyWR/pH1jOYG4a4/LGPRGPhw0Ak7+4YTEoS2wfyu7baRsKkGov9KrtLQUx48fBwA8+OCD+OCDD9CjRw94enqiefPmtZ5vMBjg7u6O5vPmQO3IZYh0fwqO31PfKRDJpkqoxDZ8h5KSEuh08hRlNd8VkU+/A3uHO/+uqKosR/qPCbLm2tDVa6di//796NHjr38910zCjImJQVJSUj1lRURE9yWu/pBdvRYVTzzxBOqxUUJEREQS4pwKIiJSBK7+kB+LCiIiUgazUL3Zcj7dFosKIiJSBs6pkB0ffU5ERESSYKeCiIgUQQUb51RIlsn9i0UFEREpA++oKTsOfxAREZEk2KkgIiJF4JJS+bGoICIiZeDqD9lx+IOIiIgkwU4FEREpgkoQoLJhsqUt5yoFiwoiIlIG87XNlvPptjj8QURERJJgp4KIiBSBwx/yY1FBRETKwNUfsmNRQUREysA7asqOcyqIiIhIEuxUEBGRIvCOmvJjUUFERMrA4Q/ZcfiDiIiIJMGigoiIFEFltn2zRmJiIlQqlcXWrl078Xh5eTliY2PRuHFjuLq6on///sjPz7e4Rm5uLnr37g1nZ2d4e3tj0qRJqKqqsojZtm0bOnfuDK1Wi+DgYCQlJd2Qy7Jly9CiRQs4OjoiIiIC+/bts+7D1BGLCiIiUoaa4Q9bNis98MADOH/+vLjt2rVLPDZhwgRs2rQJX3/9NbZv3468vDz069dPPG4ymdC7d28YjUbs3r0bq1evRlJSEhISEsSYnJwc9O7dGz169EBmZibGjx+PESNG4KeffhJj1q1bh/j4eMycORMHDhxAaGgooqOjUVBQcIc/yFtjUUFERCQTe3t7+Pr6ipuXlxcAoKSkBJ9++ik++OADPPnkkwgPD8eqVauwe/du7NmzBwDw888/448//sB//vMfhIWF4emnn8bs2bOxbNkyGI1GAMCKFSsQFBSE999/H+3bt0dcXBxefPFFLFy4UMzhgw8+wMiRIzF06FCEhIRgxYoVcHZ2xmeffSb552VRQUREyiBIsAEwGAwWW0VFxS3f8tixY/Dz80PLli0xePBg5ObmAgAyMjJQWVmJqKgoMbZdu3Zo3rw50tPTAQDp6eno2LEjfHx8xJjo6GgYDAYcOXJEjLn+GjUxNdcwGo3IyMiwiFGr1YiKihJjpMSigoiIFKHmNt22bAAQEBAAd3d3cZs3b95N3y8iIgJJSUlISUnB8uXLkZOTg27duuHy5cvQ6/XQaDTw8PCwOMfHxwd6vR4AoNfrLQqKmuM1x24XYzAYcPXqVVy4cAEmk+mmMTXXkBKXlBIREVnhzJkz0Ol04mutVnvTuKefflr8706dOiEiIgKBgYFYv349nJycZM+zPrBTQUREyiDRRE2dTmex3aqo+DsPDw+0adMGx48fh6+vL4xGI4qLiy1i8vPz4evrCwDw9fW9YTVIzevaYnQ6HZycnODl5QU7O7ubxtRcQ0osKoiISBkEAGYbNhvvfVVaWooTJ06gadOmCA8Ph4ODA9LS0sTj2dnZyM3NRWRkJAAgMjIShw4dslilkZqaCp1Oh5CQEDHm+mvUxNRcQ6PRIDw83CLGbDYjLS1NjJEShz+IiEgR7vajzydOnIg+ffogMDAQeXl5mDlzJuzs7DBo0CC4u7tj+PDhiI+Ph6enJ3Q6HcaOHYvIyEg8/PDDAICePXsiJCQEr776KubPnw+9Xo8ZM2YgNjZW7I6MHj0aS5cuxeTJkzFs2DBs2bIF69evx+bNm8U84uPjERMTgy5duuChhx7CokWLUFZWhqFDh97xz+JWWFQQERHJ4OzZsxg0aBAuXryIJk2a4LHHHsOePXvQpEkTAMDChQuhVqvRv39/VFRUIDo6Gh9++KF4vp2dHZKTkzFmzBhERkbCxcUFMTExeOedd8SYoKAgbN68GRMmTMDixYvh7++PlStXIjo6WowZMGAACgsLkZCQAL1ej7CwMKSkpNwweVMKKkFouDczNxgMcHd3R/N5c6B2dKzvdIhkERy/p75TIJJNlVCJbfgOJSUlFpMfpVTzXfFk2FTY29Vt/sPNVJkqsCXzX7Lm2tCxU0FERMrAB4rJjhM1iYiISBLsVBARkTKYAahsPJ9ui0UFEREpwt1e/aFEHP4gIiIiSbBTQUREysCJmrJjUUFERMrAokJ2HP4gIiIiSbBTQUREysBOhexYVBARkTJwSansWFQQEZEicEmp/DingoiIiCTBTgURESkD51TIjkUFEREpg1kAVDYUBmYWFbXh8AcRERFJgp0KIiJSBg5/yI5FBRERKYSNRQVYVNSGwx9EREQkCXYqiIhIGTj8ITsWFUREpAxmATYNYXD1R604/EFERESSYKeCiIiUQTBXb7acT7fFooKIiJSBcypkx6KCiIiUgXMqZMc5FURERCQJdiqIiEgZOPwhOxYVRESkDAJsLCoky+S+xeEPIiIikgQ7FUREpAwc/pAdiwoiIlIGsxmADfeaMPM+FbXh8AcRERFJgp0KIiJSBg5/yI5FBRERKQOLCtlx+IOIiIgkwU4FEREpA2/TLTsWFUREpAiCYIZgw5NGbTlXKVhUEBGRMgiCbd0GzqmoFedUEBERkSTYqSAiImUQbJxTwU5FrVhUEBGRMpjNgMqGeRGcU1ErDn8QERGRJNipICIiZeDwh+xYVBARkSIIZjMEG4Y/uKS0dhz+ICIiIkmwU0FERMrA4Q/ZsaggIiJlMAuAikWFnDj8QURERJJgp4KIiJRBEADYcp8Kdipqw6KCiIgUQTALEGwY/hBYVNSKRQURESmDYIZtnQouKa0N51QQERGRJNipICIiReDwh/xYVBARkTJw+EN2DbqoqKkazeXl9ZwJkXyqhMr6ToFINlWo/v2+G12AKlTadO+rmlzp1lRCA+7nnD17FgEBAfWdBhER2ejMmTPw9/eX5drl5eUICgqCXq+3+Vq+vr7IycmBo6OjBJndfxp0UWE2m5GXlwc3NzeoVKr6TkcRDAYDAgICcObMGeh0uvpOh0hS/P2++wRBwOXLl+Hn5we1Wr61A+Xl5TAajTZfR6PRsKC4jQY9/KFWq2WrbOn2dDod/+jSfYu/33eXu7u77O/h6OjIYuAu4JJSIiIikgSLCiIiIpIEiwqyilarxcyZM6HVaus7FSLJ8febyDYNeqImERER3TvYqSAiIiJJsKggIiIiSbCoICIiIkmwqCAiIiJJsKigOlu2bBlatGgBR0dHREREYN++ffWdEpEkduzYgT59+sDPzw8qlQobN26s75SIGiQWFVQn69atQ3x8PGbOnIkDBw4gNDQU0dHRKCgoqO/UiGxWVlaG0NBQLFu2rL5TIWrQuKSU6iQiIgJdu3bF0qVLAVQ/dyUgIABjx47F1KlT6zk7IumoVCps2LABffv2re9UiBocdiqoVkajERkZGYiKihL3qdVqREVFIT09vR4zIyKiewmLCqrVhQsXYDKZ4OPjY7Hfx8dHkkcJExHR/YFFBREREUmCRQXVysvLC3Z2dsjPz7fYn5+fD19f33rKioiI7jUsKqhWGo0G4eHhSEtLE/eZzWakpaUhMjKyHjMjIqJ7iX19J0ANQ3x8PGJiYtClSxc89NBDWLRoEcrKyjB06ND6To3IZqWlpTh+/Lj4OicnB5mZmfD09ETz5s3rMTOihoVLSqnOli5digULFkCv1yMsLAxLlixBREREfadFZLNt27ahR48eN+yPiYlBUlLS3U+IqIFiUUFERESS4JwKIiIikgSLCiIiIpIEiwoiIiKSBIsKIiIikgSLCiIiIpIEiwoiIiKSBIsKIiIikgSLCiIiIpIEiwoiGw0ZMgR9+/YVXz/xxBMYP378Xc9j27ZtUKlUKC4uvmWMSqXCxo0b63zNxMREhIWF2ZTXqVOnoFKpkJmZadN1iOjex6KC7ktDhgyBSqWCSqWCRqNBcHAw3nnnHVRVVcn+3t9++y1mz55dp9i6FAJERA0FHyhG961evXph1apVqKiowA8//IDY2Fg4ODhg2rRpN8QajUZoNBpJ3tfT01OS6xARNTTsVNB9S6vVwtfXF4GBgRgzZgyioqLw/fffA/hryGLu3Lnw8/ND27ZtAQBnzpzBSy+9BA8PD3h6euL555/HqVOnxGuaTCbEx8fDw8MDjRs3xuTJk/H3x+f8ffijoqICU6ZMQUBAALRaLYKDg/Hpp5/i1KlT4kOsGjVqBJVKhSFDhgCofrT8vHnzEBQUBCcnJ4SGhuK///2vxfv88MMPaNOmDZycnNCjRw+LPOtqypQpaNOmDZydndGyZUu8/fbbqKysvCHuo48+QkBAAJydnfHSSy+hpKTE4vjKlSvRvn17ODo6ol27dvjwww+tzoWIGj4WFaQYTk5OMBqN4uu0tDRkZ2cjNTUVycnJqKysRHR0NNzc3LBz50788ssvcHV1Ra9evcTz3n//fSQlJeGzzz7Drl27UFRUhA0bNtz2fV977TV8+eWXWLJkCY4ePYqPPvoIrq6uCAgIwDfffAMAyM7Oxvnz57F48WIAwLx58/D5559jxYoVOHLkCCZMmIBXXnkF27dvB1Bd/PTr1w99+vRBZmYmRowYgalTp1r9M3Fzc0NSUhL++OMPLF68GJ988gkWLlxoEXP8+HGsX78emzZtQkpKCg4ePIjXX39dPL5mzRokJCRg7ty5OHr0KN599128/fbbWL16tdX5EFEDJxDdh2JiYoTnn39eEARBMJvNQmpqqqDVaoWJEyeKx318fISKigrxnC+++EJo27atYDabxX0VFRWCk5OT8NNPPwmCIAhNmzYV5s+fLx6vrKwU/P39xfcSBEF4/PHHhXHjxgmCIAjZ2dkCACE1NfWmeW7dulUAIFy6dEncV15eLjg7Owu7d++2iB0+fLgwaNAgQRAEYdq0aUJISIjF8SlTptxwrb8DIGzYsOGWxxcsWCCEh4eLr2fOnCnY2dkJZ8+eFff9+OOPglqtFs6fPy8IgiC0atVKWLt2rcV1Zs+eLURGRgqCIAg5OTkCAOHgwYO3fF8iuj9wTgXdt5KTk+Hq6orKykqYzWa8/PLLSExMFI937NjRYh7Fb7/9huPHj8PNzc3iOuXl5Thx4gRKSkpw/vx5REREiMfs7e3RpUuXG4ZAamRmZsLOzg6PP/54nfM+fvw4rly5gn/84x8W+41GIx588EEAwNGjRy3yAIDIyMg6v0eNdevWYcmSJThx4gRKS0tRVVUFnU5nEdO8eXM0a9bM4n3MZjOys7Ph5uaGEydOYPjw4Rg5cqQYU1VVBXd3d6vzIaKGjUUF3bd69OiB5cuXQ6PRwM/PD/b2lr/uLi4uFq9LS0sRHh6ONWvW3HCtJk2a3FEOTk5OVp9TWloKANi8ebPFlzlQPU9EKunp6Rg8eDBmzZqF6OhouLu746uvvsL7779vda6ffPLJDUWOnZ2dZLkSUcPAooLuWy4uLggODq5zfOfOnbFu3Tp4e3vf8K/1Gk2bNsXevXvRvXt3ANX/Is/IyEDnzp1vGt+xY0eYzWZs374dUVFRNxyv6ZSYTCZxX0hICLRaLXJzc2/Z4Wjfvr046bTGnj17av+Q19m9ezcCAwMxffp0cd/p06dviMvNzUVeXh78/PzE91Gr1Wjbti18fHzg5+eHkydPYvDgwVa9PxHdfzhRk+iawYMHw8vLC88//zx27tyJnJwcbNu2DW+88QbOnj0LABg3bhz+9a9/YePGjcjKysLrr79+23tMtGjRAjExMRg2bBg2btwoXnP9+vUAgMDAQKhUKiQnJ6OwsBClpaVwc3PDxIkTMWHCBKxevRonTpzAgQMH8O9//1uc/Dh69GgcO3YMkyZNQnZ2NtauXYukpCSrPm/r1q2Rm5uLr776CidOnMCSJUtuOunU0dERMTEx+O2337Bz50688cYbeOmll+Dr6wsAmDVrFubNm4clS5bgzz//xKFDh7Bq1Sp88MEHVuVDRA0fiwqia5ydnbFjxw40b94c/fr1Q/v27TF8+HCUl5eLnYs333wTr776KmJiYhAZGQk3Nze88MILt73u8uXL8eKLL+L1119Hu3btMHLkSJSVlQEAmjVrhlmzZmHq1Knw8fFBXFwcAGD27Nl4++23MW/ePLRv3x69evXC5s2bERQUBKB6nsM333yDjRs3IjQ0FCtWrMC7775r1ed97rnnMGHCBMTFxSEsLAy7d+/G22+/fUNccHAw+vXrh2eeeQY9e/ZEp06dLJaMjhgxAitXrsSqVavQsWNHPP7440hKShJzJSLlUAm3mmFGREREZAV2KoiIiEgSLCqIiIhIEiwqiIiISBIsKoiIiEgSLCqIiIhIEiwqiIiISBIsKoiIiEgSLCqIiIhIEiwqiIiISBIsKoiIiEgSLCqIiIhIEv8fbCP7/LdDF+cAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Evaluate the testing set with the single-task ranker\n",
    "\n",
    "evaluate(test_negative, single_task_model, \n",
    "         customer_numerical_features=customer_numerical_features, \n",
    "         article_categorical_feature=article_categorical_feature,\n",
    "         single_task=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "qg5JrRUGI2RE",
   "metadata": {
    "id": "qg5JrRUGI2RE"
   },
   "source": [
    "We see that our test MRR is higher than our train MRR, which is unusual, but can likely be attributed to the 5 choices listed above. Given that MRR is bound between 0-1, and random assignment would give us MRR ~0.61 (given by `(1/np.random.choice([1,2,3], size=100_000)).mean()`), we can see that our model has learned to rank!"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "NI-XuLBcJld0",
   "metadata": {
    "id": "NI-XuLBcJld0"
   },
   "source": [
    "## **7. Multi-Task Ranker**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "65ada5f3-8157-4088-95b8-8e17989b2d52",
   "metadata": {
    "id": "65ada5f3-8157-4088-95b8-8e17989b2d52"
   },
   "source": [
    "We can now add another task to create a multi-task model. To do so, we first have to decide what task 2 should be. By looking at the data we have, we hypothesize that being able to predict whether or not a customer would purchase an article from certain product category would be a good learnt model to have. Indeed, being able to predict which category would the user purchase in would help us narrow down the list of recommendations we want to surface to users.\n",
    "\n",
    "Based on this intuition, we want to develop a multi-task model with 2 tasks:\n",
    "1. Task 1 (`label`): Predict whether or not a user would purchase an article, our original task.\n",
    "2. Task 2 (`label2`): Predict whether or not a user would purchase any article from a given category (`product_group_name`).\n",
    "\n",
    "The training data we have created thus far is solely based on prediction task 1: predicting whether or not a user would purchase a given article. Correspondingly, in train_negative, we have collected positive and negative examples for this task and have added a `label` column as the final label to train the model.\n",
    "\n",
    "To add task 2, for each row, we will need to identify whether the user has made any purchase in this category or not. If the user has made a purchase in this category then we will assign label2 = 1, else label2 = 0. Please note that for the same training example it may happen that label1 = 0 but label2 = 1; this would happen in cases wherein the customer did not purchase this specific article but did purchase any other article from this category."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ibq1uMfTL2H9",
   "metadata": {
    "id": "ibq1uMfTL2H9"
   },
   "source": [
    "### **a. TODO 2.1: Add a Second Label**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&line=3&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "Y7GhTAwoL2rl",
   "metadata": {
    "id": "Y7GhTAwoL2rl"
   },
   "source": [
    "Your task is to write the function that adds this `label2` column to the train_negative dataframe, to enable us to train a multi-task model. We intended that \"product category\" be `product_group_name`, but feel free to use another feature if you wish. We implemented our own solution and left the result so you could check it against yours."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "39b6f339-ccae-4cb8-949d-dc7947333bf1",
   "metadata": {
    "id": "39b6f339-ccae-4cb8-949d-dc7947333bf1"
   },
   "outputs": [],
   "source": [
    "def add_label2_to_train_negative(df, catgory_col='product_group_name', selected_category='Accessories'):\n",
    "    \"\"\"\n",
    "    Args:\n",
    "        train_negative (DataFrame): The `train_negative` DataFrame we've been using above.\n",
    "\n",
    "    Returns:\n",
    "        train_negative (DataFrame): The `train_negative` DataFrame we've been using above with a new integer column, label2.\n",
    "            Which is the label telling us whether or not a user would purchase any article from a given product_group_name.\n",
    "    \"\"\"\n",
    "\n",
    "    # Your code goes here\n",
    "    pos_customers = df.loc[(df['label']==1)&(df[catgory_col]==selected_category), 'customer_id'].unique()\n",
    "    df['label2'] = df.customer_id.isin(pos_customers).astype('int')\n",
    "\n",
    "    if 'label2' not in train_negative.columns:\n",
    "        raise AssertionError('`label2` must be a column in train_negative.')\n",
    "    if train_negative['label2'].dtype != int:\n",
    "        raise AssertionError('`label2` must be an integer.')\n",
    "    return train_negative"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "lMCoVTJ1ZbjD",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "lMCoVTJ1ZbjD",
    "outputId": "49356281-2b3e-45b1-8b3f-9015c0e3d0e8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Distribution of purchases (label):\n",
      "0    3538194\n",
      "1    1769097\n",
      "Name: label, dtype: int64\n",
      "\n",
      "Distribution of product_group_name purchases (label2):\n",
      "0    3967473\n",
      "1    1339818\n",
      "Name: label2, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "train_negative = add_label2_to_train_negative(train_negative)\n",
    "\n",
    "print('Distribution of purchases (label):')\n",
    "print(train_negative['label'].value_counts())\n",
    "\n",
    "print('\\nDistribution of product_group_name purchases (label2):')\n",
    "print(train_negative['label2'].value_counts())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aMIjykLIMHZW",
   "metadata": {
    "id": "aMIjykLIMHZW"
   },
   "source": [
    "### **b. Model Training**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&line=3&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f2642a24-090e-4180-9be0-e9bb27a89695",
   "metadata": {
    "id": "f2642a24-090e-4180-9be0-e9bb27a89695"
   },
   "source": [
    "Now that we have both labels we can construct a simple multi-task model with a shared bottom, which we call the common layers. Building on top of the common layers, we will have two separate modules: task1 layers and task2 layers. Layers of task 1, \"task1_layer\", are dedicated to learning weights that help us perform better on task 1 (`label`, purchase the item). Correspondingly, layers of task 2, \"task2_layer\", are dedicated to learning weights that help us perform better on task 2 (`label2`, purchase a given category).  \n",
    "<br>\n",
    "A couple notes:  \n",
    "* Your results may vary due to variations in model training between runs or your implementation of the second label.\n",
    "* This dataset is quite complex and large, so changing the architecture and/or hyperparameters may significantly change performance. We've designed these models to train quickly and act as baselines.  \n",
    "<br><br>\n",
    "\n",
    "Below we implement such a multi-task model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "tTsQp9GtCGtz",
   "metadata": {
    "id": "tTsQp9GtCGtz"
   },
   "outputs": [],
   "source": [
    "def train_multi_task_ranker(customer_numerical_features=[],\n",
    "                            article_categorical_feature='',\n",
    "                            n_epochs=5,\n",
    "                            batch_size=1024*4):\n",
    "    \"\"\"\n",
    "    Args:\n",
    "        customer_numerical_features (list): Numerical features used to describe the customer.\n",
    "        article_categorical_feature (str): Single, categorical feature used to describe the article.\n",
    "            In practice, we'd use many more features to describe the article.\n",
    "        n_epochs (int): Number of epochs.\n",
    "        batch_size (int): Batch size.\n",
    "\n",
    "    Returns:\n",
    "        model (tf.Keras.Model): Trained model on train_negative.\n",
    "    \"\"\"\n",
    "    \n",
    "    # Assemble training data\n",
    "    X1 = np.asarray(train_negative[customer_numerical_features].values)\n",
    "    X2 = train_negative[article_categorical_feature].astype(str)\n",
    "    y1 = np.asarray(train_negative['label'].values)\n",
    "    y2 = np.asarray(train_negative['label2'].values)\n",
    "\n",
    "\n",
    "    # Numerical customer features\n",
    "    input1_layer = tf.keras.Input((len(customer_numerical_features)))\n",
    "    input1 = tf.keras.layers.Dense(50, activation='relu')(input1_layer)\n",
    "\n",
    "    # Categorical article feature\n",
    "    unique_group_names = train_negative[article_categorical_feature].unique().astype(str)\n",
    "    input2_layer = tf.keras.Input((1, ), dtype=tf.string)\n",
    "    input2 = tf.keras.layers.StringLookup(vocabulary=unique_group_names, mask_token=None)(input2_layer)\n",
    "    input2 = tf.keras.layers.Embedding(len(unique_group_names) + 1, 50)(input2)\n",
    "    input2 = tf.keras.layers.Dense(32, activation='relu')(input2)\n",
    "    input2 = tf.keras.layers.Flatten()(input2)\n",
    "\n",
    "    # Combine inputs (task 1, label)\n",
    "    combined_layer = tf.keras.layers.concatenate([input1, input2])\n",
    "    combined_layer1 = tf.keras.layers.Dense(50, name=\"task1_layer\")(combined_layer)\n",
    "    combined_layer1 = tf.keras.layers.Dense(1, activation='sigmoid')(combined_layer1)\n",
    "\n",
    "    # Combine inputs (task 2, label2)\n",
    "    combined_layer2 = tf.keras.layers.Dense(50, name=\"task2_layer\")(combined_layer)\n",
    "    combined_layer2 = tf.keras.layers.Dense(1, activation='sigmoid')(combined_layer2)\n",
    "\n",
    "    # Compile model\n",
    "    bce = tf.keras.losses.BinaryCrossentropy(from_logits=False)\n",
    "    metric = tf.keras.metrics.BinaryAccuracy()\n",
    "    model = tf.keras.Model([input1_layer, input2_layer], \n",
    "                           [combined_layer1, combined_layer2])\n",
    "    model.compile(loss=bce,\n",
    "                  optimizer='adam',\n",
    "                  metrics=[metric])\n",
    "\n",
    "    # Train model\n",
    "    model.fit([X1, X2], [y1, y2], epochs=n_epochs, verbose=1, batch_size=batch_size)\n",
    "\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "H7F3PQN-CsST",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "H7F3PQN-CsST",
    "outputId": "015b31f7-a9e3-47f8-92fa-84682a1f627e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "1296/1296 [==============================] - 15s 11ms/step - loss: 1.0836 - dense_6_loss: 0.6379 - dense_7_loss: 0.4458 - dense_6_binary_accuracy: 0.6659 - dense_7_binary_accuracy: 0.7908\n",
      "Epoch 2/5\n",
      "1296/1296 [==============================] - 14s 11ms/step - loss: 1.0535 - dense_6_loss: 0.6266 - dense_7_loss: 0.4269 - dense_6_binary_accuracy: 0.6681 - dense_7_binary_accuracy: 0.7980\n",
      "Epoch 3/5\n",
      "1296/1296 [==============================] - 14s 11ms/step - loss: 1.0462 - dense_6_loss: 0.6266 - dense_7_loss: 0.4196 - dense_6_binary_accuracy: 0.6682 - dense_7_binary_accuracy: 0.8011\n",
      "Epoch 4/5\n",
      "1296/1296 [==============================] - 14s 11ms/step - loss: 1.0415 - dense_6_loss: 0.6265 - dense_7_loss: 0.4150 - dense_6_binary_accuracy: 0.6682 - dense_7_binary_accuracy: 0.8037\n",
      "Epoch 5/5\n",
      "1296/1296 [==============================] - 14s 11ms/step - loss: 1.0392 - dense_6_loss: 0.6263 - dense_7_loss: 0.4129 - dense_6_binary_accuracy: 0.6683 - dense_7_binary_accuracy: 0.8049\n"
     ]
    }
   ],
   "source": [
    "# Use the same features as we did with the single-task ranker to predict:\n",
    "    # 1) Whether the customer bought the product\n",
    "    # 2) Whether the customer bought a product in that category\n",
    "\n",
    "multi_task_model = train_multi_task_ranker(customer_numerical_features=customer_numerical_features,\n",
    "                                           article_categorical_feature=article_categorical_feature)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "aiDfXd1WDOcu",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 664
    },
    "id": "aiDfXd1WDOcu",
    "outputId": "de6b7ac2-e76a-48f6-861e-e083c4e2f440"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "519/519 [==============================] - 6s 11ms/step\n",
      "MRR: 0.6056\n",
      "Distribution:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    330712.000000\n",
       "mean          0.605611\n",
       "std           0.334024\n",
       "min           0.019231\n",
       "25%           0.333333\n",
       "50%           0.500000\n",
       "75%           1.000000\n",
       "max           1.000000\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Baseline:\n",
      "0    0.666667\n",
      "1    0.333333\n",
      "Name: label, dtype: float64\n",
      "\n",
      "Model Accuracy: 0.67\n",
      "\n",
      "Confusion Matrix:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Evaluate the training set with the multi-task ranker\n",
    "\n",
    "evaluate(train_negative, multi_task_model, \n",
    "         customer_numerical_features=customer_numerical_features, \n",
    "         article_categorical_feature=article_categorical_feature,\n",
    "         single_task=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "K7ILMPlEDQb8",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 653
    },
    "id": "K7ILMPlEDQb8",
    "outputId": "2501c6d2-dcae-48b9-8ff7-e3d696bfff23"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "44/44 [==============================] - 0s 4ms/step\n",
      "MRR: 0.5946\n",
      "Distribution:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    42395.000000\n",
       "mean         0.594556\n",
       "std          0.321411\n",
       "min          0.027027\n",
       "25%          0.333333\n",
       "50%          0.500000\n",
       "75%          1.000000\n",
       "max          1.000000\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Baseline:\n",
      "0    0.666667\n",
      "1    0.333333\n",
      "Name: label, dtype: float64\n",
      "\n",
      "Model Accuracy: 0.67\n",
      "\n",
      "Confusion Matrix:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Evaluate the testing set with the multi-task ranker\n",
    "\n",
    "evaluate(test_negative, multi_task_model, \n",
    "         customer_numerical_features=customer_numerical_features, \n",
    "         article_categorical_feature=article_categorical_feature,\n",
    "         single_task=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4aM-qdfG40Il",
   "metadata": {
    "id": "4aM-qdfG40Il"
   },
   "source": [
    "Given the same architecture, apart from an extra tower for the second label, MRR performance has unfortunately decreased (although accuracy increased, but still not better than majority voting). In general, though, giving the model access to more information in terms of additional labels is helpful for well-tuned neural networks. We have not gone through the tuning process for this model (especially given that we've trained a small, arbitrary number of epochs) for the purposes of highlighting architecture and evaluation choices. You'll have a chance to add more features to the model below so you can improve/change the training process!"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "09LGM73ePNZK",
   "metadata": {
    "id": "09LGM73ePNZK"
   },
   "source": [
    "### **c. Freezing Layers for Secondary Tasks**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&line=3&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d0478c09-91fa-404d-84f5-34ae7fe9f359",
   "metadata": {
    "id": "d0478c09-91fa-404d-84f5-34ae7fe9f359"
   },
   "source": [
    "\n",
    "Often when training a multi-task model, one needs to freeze parts of the neural network and only train one specific task module, while keeping the weights of the other task module fixed. In Keras we can do this by freezing layers. The below example demonstrates how one could go about freezing layers in a neural network, and train only the trainable parts of a nueral network."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "3b40d34e-fa1c-4e09-b29a-7df832a3746b",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "3b40d34e-fa1c-4e09-b29a-7df832a3746b",
    "outputId": "d9a8171f-5bfe-44e7-c72b-68f745cee855"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<keras.engine.functional.Functional object at 0x2e1c718e0> model_1 True\n",
      "<keras.engine.input_layer.InputLayer object at 0x339fe75b0> input_4 True\n",
      "<keras.layers.preprocessing.string_lookup.StringLookup object at 0x33f061e50> string_lookup_1 True\n",
      "<keras.layers.core.embedding.Embedding object at 0x34fea4040> embedding_1 True\n",
      "<keras.engine.input_layer.InputLayer object at 0x33f077160> input_3 True\n",
      "<keras.layers.core.dense.Dense object at 0x34fea49a0> dense_5 True\n",
      "<keras.layers.core.dense.Dense object at 0x33f0774f0> dense_4 True\n",
      "<keras.layers.reshaping.flatten.Flatten object at 0x33f069340> flatten_1 True\n",
      "<keras.layers.merging.concatenate.Concatenate object at 0x29fc00b50> concatenate_1 True\n",
      "<keras.layers.core.dense.Dense object at 0x2e1c323a0> task1_layer True\n",
      "<keras.layers.core.dense.Dense object at 0x2e1c32520> task2_layer True\n",
      "<keras.layers.core.dense.Dense object at 0x2e1c325b0> dense_6 True\n",
      "<keras.layers.core.dense.Dense object at 0x2e1c3a670> dense_7 True\n"
     ]
    }
   ],
   "source": [
    "# Show the training status of each layer of the models we have\n",
    "\n",
    "for k,v in multi_task_model._get_trainable_state().items():\n",
    "    print(k, k.name, v)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "4842f73e-3367-4b47-890d-93936b18b212",
   "metadata": {
    "id": "4842f73e-3367-4b47-890d-93936b18b212"
   },
   "outputs": [],
   "source": [
    "# Freeze the task2_layer so that no training example affects the module dedicated to task 2\n",
    "\n",
    "for layer in multi_task_model.layers:\n",
    "    if layer.name == \"task2_layer\":\n",
    "        layer.trainable = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "2b01277c-eb22-47b7-8bb7-7d1a106543c8",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "2b01277c-eb22-47b7-8bb7-7d1a106543c8",
    "outputId": "358cb6e2-0162-40ec-f683-ace7f8315fcb"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<keras.engine.functional.Functional object at 0x2e1c718e0> model_1 True\n",
      "<keras.engine.input_layer.InputLayer object at 0x339fe75b0> input_4 True\n",
      "<keras.layers.preprocessing.string_lookup.StringLookup object at 0x33f061e50> string_lookup_1 True\n",
      "<keras.layers.core.embedding.Embedding object at 0x34fea4040> embedding_1 True\n",
      "<keras.engine.input_layer.InputLayer object at 0x33f077160> input_3 True\n",
      "<keras.layers.core.dense.Dense object at 0x34fea49a0> dense_5 True\n",
      "<keras.layers.core.dense.Dense object at 0x33f0774f0> dense_4 True\n",
      "<keras.layers.reshaping.flatten.Flatten object at 0x33f069340> flatten_1 True\n",
      "<keras.layers.merging.concatenate.Concatenate object at 0x29fc00b50> concatenate_1 True\n",
      "<keras.layers.core.dense.Dense object at 0x2e1c323a0> task1_layer True\n",
      "<keras.layers.core.dense.Dense object at 0x2e1c32520> task2_layer False\n",
      "<keras.layers.core.dense.Dense object at 0x2e1c325b0> dense_6 True\n",
      "<keras.layers.core.dense.Dense object at 0x2e1c3a670> dense_7 True\n"
     ]
    }
   ],
   "source": [
    "# Inspect trainable states and confirm the task2_layer is frozen\n",
    "\n",
    "for k,v in multi_task_model._get_trainable_state().items():\n",
    "    print(k, k.name, v)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "NTgBnh2-PrNL",
   "metadata": {
    "id": "NTgBnh2-PrNL"
   },
   "source": [
    "## **8. TODO 2.2: Model Improvement (Optional)**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "48cccb53-490d-48fe-89f9-99895bdfc50d",
   "metadata": {
    "id": "48cccb53-490d-48fe-89f9-99895bdfc50d"
   },
   "source": [
    "The above gives a high level view on how one could train single- and multi-task models.\n",
    "\n",
    "Let's re-iterate on what we've accomplished for this part of the project: \n",
    "- Trained and evaluated a single-task ranker.\n",
    "- Added a second label task.\n",
    "- Trained and evaluated a multi-task ranker.\n",
    "- Demonstrated how to freeze layers for training specific tasks in a multi-task scenario.  \n",
    "  \n",
    "<br>Your optional task is to make changes to either the single- or multi-task models to see if you can improve the performance of the models:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "13b54c33",
   "metadata": {},
   "source": [
    "## Train model based on Article Description tfidf features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 183,
   "id": "k1JDYp0r5bcK",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "id": "k1JDYp0r5bcK",
    "outputId": "9b2a5ea0-f133-403e-af4b-ee52b04558f3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: seaborn in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (0.11.2)\n",
      "Requirement already satisfied: tqdm in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (4.64.1)\n",
      "Requirement already satisfied: texthero==1.0.5 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (1.0.5)\n",
      "Requirement already satisfied: scikit-learn in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (1.0.2)\n",
      "Requirement already satisfied: nltk in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from texthero==1.0.5) (3.8.1)\n",
      "Requirement already satisfied: plotly-express in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from texthero==1.0.5) (0.4.1)\n",
      "Requirement already satisfied: pandas>=0.23 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from seaborn) (1.3.4)\n",
      "Requirement already satisfied: matplotlib>=2.2 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from seaborn) (3.4.3)\n",
      "Requirement already satisfied: numpy>=1.15 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from seaborn) (1.24.1)\n",
      "Requirement already satisfied: scipy>=1.0 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from seaborn) (1.8.0)\n",
      "Requirement already satisfied: threadpoolctl>=2.0.0 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from scikit-learn) (3.1.0)\n",
      "Requirement already satisfied: joblib>=0.11 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from scikit-learn) (1.2.0)\n",
      "Requirement already satisfied: pyparsing>=2.2.1 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from matplotlib>=2.2->seaborn) (3.0.9)\n",
      "Requirement already satisfied: pillow>=6.2.0 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from matplotlib>=2.2->seaborn) (9.4.0)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from matplotlib>=2.2->seaborn) (1.4.4)\n",
      "Requirement already satisfied: cycler>=0.10 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from matplotlib>=2.2->seaborn) (0.11.0)\n",
      "Requirement already satisfied: python-dateutil>=2.7 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from matplotlib>=2.2->seaborn) (2.8.2)\n",
      "Requirement already satisfied: pytz>=2017.3 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from pandas>=0.23->seaborn) (2022.7.1)\n",
      "Requirement already satisfied: click in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from nltk->texthero==1.0.5) (8.1.3)\n",
      "Requirement already satisfied: regex>=2021.8.3 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from nltk->texthero==1.0.5) (2022.10.31)\n",
      "Requirement already satisfied: patsy>=0.5 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from plotly-express->texthero==1.0.5) (0.5.3)\n",
      "Requirement already satisfied: statsmodels>=0.9.0 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from plotly-express->texthero==1.0.5) (0.13.2)\n",
      "Requirement already satisfied: plotly>=4.1.0 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from plotly-express->texthero==1.0.5) (5.13.1)\n",
      "Requirement already satisfied: six in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from patsy>=0.5->plotly-express->texthero==1.0.5) (1.16.0)\n",
      "Requirement already satisfied: tenacity>=6.2.0 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from plotly>=4.1.0->plotly-express->texthero==1.0.5) (8.2.2)\n",
      "Requirement already satisfied: packaging>=21.3 in /Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages (from statsmodels>=0.9.0->plotly-express->texthero==1.0.5) (21.3)\n"
     ]
    }
   ],
   "source": [
    "!pip install seaborn tqdm texthero==1.0.5 scikit-learn \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "YOzspNysG_tr",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "YOzspNysG_tr",
    "outputId": "45ce0ba4-0ba2-4a6d-c644-cca2c7315d1c"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<spacy.lang.en.English at 0x2f48e2070>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import texthero as hero\n",
    "import spacy\n",
    "spacy.load('en_core_web_sm')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "86d162d3",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tqdm import tqdm\n",
    "# from tqdm.auto import tqdm  # for notebooks\n",
    "tqdm.pandas()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c86425ed",
   "metadata": {},
   "source": [
    "## Generate Article TFIDF Vectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "lbP6WyRSHBFr",
   "metadata": {
    "id": "lbP6WyRSHBFr"
   },
   "outputs": [],
   "source": [
    "article_df['tfidf'] = (\n",
    "   article_df['detail_desc']\n",
    "   .pipe(hero.clean)\n",
    "   .pipe(hero.do_tfidf)\n",
    "   .pipe(hero.do_pca, n_components=20)\n",
    ").apply(np.array)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "id": "25e62f09",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1769097/1769097 [00:01<00:00, 898583.95it/s]\n",
      "100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 266364/266364 [00:00<00:00, 915522.83it/s]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "((5307291, 4), (799092, 4))"
      ]
     },
     "execution_count": 206,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Generate N negative samples per row (user-article transaction)\n",
    "\n",
    "def generate_negative_samples(df, article_list, num_neg=1, negative_sample_multiplier=20):\n",
    "    \"\"\"\n",
    "    Generates `num_neg` negative samples from `article_list` per row (customer_id, article_id interaction) in `df`.\n",
    "\n",
    "    Args:\n",
    "        df (DataFrame): Positive transactions between customers and articles. Expects features `customer_id` and\n",
    "        `article_id`.\n",
    "        article_list (list): Article IDs for negative sampling.\n",
    "        num_neg (int): Number of negative samples to generate per row.\n",
    "        negative_sample_multiplier (int): Multiplier that allows us to randomly sample negative candidates all at\n",
    "        once.\n",
    "            Necessary to be larger than 1 in most cases since we will filter out some candidates.\n",
    "\n",
    "    Raises:\n",
    "        Warning in case we did not generate the intended number of negative samples per row.\n",
    "\n",
    "    Returns:\n",
    "        negative_df (DataFrame): Original, positive samples concatenated with generated, negative samples.\n",
    "    \"\"\"\n",
    "    # Create lookup of positive samples. We don't want to include negative samples that are actually positives\n",
    "    customer2positives = dict(df.groupby('customer_id')['article_id'].apply(lambda x: x.values))\n",
    "    customer2positives = {k:set(v) for k,v in customer2positives.items()}\n",
    "\n",
    "    # Sample random negatives all at once\n",
    "    np.random.seed(42)\n",
    "    random_sample_size = (len(df), num_neg*negative_sample_multiplier)\n",
    "    random_samples = np.random.choice(article_list, \n",
    "                                      size=random_sample_size,\n",
    "                                      replace=True)\n",
    "    \n",
    "    # Generate negative samples\n",
    "    customers, articles, date, labels = [], [], [], []\n",
    "    for n, (customer_id, article_id, t_dat) in enumerate(tqdm(zip(df['customer_id'], df['article_id'], df['t_dat']), total=len(df))):\n",
    "        customers.append(customer_id)\n",
    "        articles.append(article_id)\n",
    "        date.append(t_dat)\n",
    "        labels.append(1)\n",
    "        count, temp_negatives = 0, set()\n",
    "        for negative_id in random_samples[n]:\n",
    "            if (negative_id not in customer2positives[customer_id]) \\\n",
    "                & (negative_id not in temp_negatives):\n",
    "                customers.append(customer_id)\n",
    "                articles.append(negative_id)\n",
    "                date.append(t_dat)\n",
    "                labels.append(0)\n",
    "                temp_negatives.add(negative_id)\n",
    "                count += 1\n",
    "            if count == num_neg:\n",
    "                break\n",
    "\n",
    "    # Create result dataframe of positives and negatives, and data quality check\n",
    "    negative_df = pd.DataFrame({'customer_id':customers, 'article_id':articles, 't_dat': date, 'label':labels})\n",
    "    if len(negative_df) != (len(df)*num_neg+len(df)):\n",
    "        print('WARNING: Not enough samples generated. Increase `negative_sample_multiplier`.')\n",
    "    return negative_df\n",
    "\n",
    "\n",
    "# Note: We only generate negative samples from the training set to avoid leakage\n",
    "articles_set = list(train.article_id.unique())\n",
    "train_negative = generate_negative_samples(train, articles_set, num_neg=2)\n",
    "test_negative = generate_negative_samples(test, articles_set, num_neg=2)\n",
    "\n",
    "train_negative.shape, test_negative.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 205,
   "id": "480b70a6",
   "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>customer_id</th>\n",
       "      <th>article_id</th>\n",
       "      <th>t_dat</th>\n",
       "      <th>label</th>\n",
       "      <th>prev_items_tfidf_x</th>\n",
       "      <th>tfidf_x</th>\n",
       "      <th>prev_items_tfidf_y</th>\n",
       "      <th>tfidf_y</th>\n",
       "      <th>product_group_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ffffcf35913a0bee60e8741cb2b4e78b8a98ee5ff2e6a1...</td>\n",
       "      <td>762846027</td>\n",
       "      <td>2020-09-09</td>\n",
       "      <td>1</td>\n",
       "      <td>[0.6918655795839387, -0.35663200882050616, -0....</td>\n",
       "      <td>[0.19067515101280055, 0.055564732377888, -0.31...</td>\n",
       "      <td>[0.691865577858334, -0.35666563392571254, -0.2...</td>\n",
       "      <td>[0.19067515101280055, 0.055564732377888, -0.31...</td>\n",
       "      <td>Garment Upper body</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ffffcf35913a0bee60e8741cb2b4e78b8a98ee5ff2e6a1...</td>\n",
       "      <td>282832022</td>\n",
       "      <td>2020-09-09</td>\n",
       "      <td>0</td>\n",
       "      <td>[0.6918655795839387, -0.35663200882050616, -0....</td>\n",
       "      <td>[-0.14741212571534487, -0.08153015208128013, -...</td>\n",
       "      <td>[0.691865577858334, -0.35666563392571254, -0.2...</td>\n",
       "      <td>[-0.14741212571534487, -0.08153015208128013, -...</td>\n",
       "      <td>Garment Upper body</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ffffcf35913a0bee60e8741cb2b4e78b8a98ee5ff2e6a1...</td>\n",
       "      <td>840033006</td>\n",
       "      <td>2020-09-09</td>\n",
       "      <td>0</td>\n",
       "      <td>[0.6918655795839387, -0.35663200882050616, -0....</td>\n",
       "      <td>[0.20458352993581452, 0.08828260844679149, 0.2...</td>\n",
       "      <td>[0.691865577858334, -0.35666563392571254, -0.2...</td>\n",
       "      <td>[0.20458352993581452, 0.08828260844679149, 0.2...</td>\n",
       "      <td>Garment Upper body</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ffffcf35913a0bee60e8741cb2b4e78b8a98ee5ff2e6a1...</td>\n",
       "      <td>809238003</td>\n",
       "      <td>2020-09-08</td>\n",
       "      <td>0</td>\n",
       "      <td>[-0.44182159437733715, -0.0022738383333123723,...</td>\n",
       "      <td>[0.325056539190052, -0.07741629366772365, -0.3...</td>\n",
       "      <td>[-0.44182159438002244, -0.0022739558543872335,...</td>\n",
       "      <td>[0.325056539190052, -0.07741629366772365, -0.3...</td>\n",
       "      <td>Garment Upper body</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ffffcf35913a0bee60e8741cb2b4e78b8a98ee5ff2e6a1...</td>\n",
       "      <td>746282001</td>\n",
       "      <td>2020-09-08</td>\n",
       "      <td>0</td>\n",
       "      <td>[-0.44182159437733715, -0.0022738383333123723,...</td>\n",
       "      <td>[-0.003907417025550813, 0.12183911621887447, -...</td>\n",
       "      <td>[-0.44182159438002244, -0.0022739558543872335,...</td>\n",
       "      <td>[-0.003907417025550813, 0.12183911621887447, -...</td>\n",
       "      <td>Garment Upper body</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5307286</th>\n",
       "      <td>00006413d8573cd20ed7128e53b7b13819fe5cfc2d801f...</td>\n",
       "      <td>870640001</td>\n",
       "      <td>2020-08-12</td>\n",
       "      <td>0</td>\n",
       "      <td>[-0.44182159437733715, -0.0022738383333123723,...</td>\n",
       "      <td>[-0.2681135879098105, -0.40141412067644494, -0...</td>\n",
       "      <td>[-0.44182159438002244, -0.0022739558543872335,...</td>\n",
       "      <td>[-0.2681135879098105, -0.40141412067644494, -0...</td>\n",
       "      <td>Garment Lower body</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5307287</th>\n",
       "      <td>00006413d8573cd20ed7128e53b7b13819fe5cfc2d801f...</td>\n",
       "      <td>825783008</td>\n",
       "      <td>2020-08-12</td>\n",
       "      <td>0</td>\n",
       "      <td>[-0.44182159437733715, -0.0022738383333123723,...</td>\n",
       "      <td>[0.24303098110870489, -0.12804459561348136, 0....</td>\n",
       "      <td>[-0.44182159438002244, -0.0022739558543872335,...</td>\n",
       "      <td>[0.24303098110870489, -0.12804459561348136, 0....</td>\n",
       "      <td>Accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5307288</th>\n",
       "      <td>00000dbacae5abe5e23885899a1fa44253a17956c6d1c3...</td>\n",
       "      <td>568601043</td>\n",
       "      <td>2020-09-05</td>\n",
       "      <td>1</td>\n",
       "      <td>[-0.44182159437733715, -0.0022738383333123723,...</td>\n",
       "      <td>[-0.16790096697298518, 0.009406178804481543, -...</td>\n",
       "      <td>[-0.44182159438002244, -0.0022739558543872335,...</td>\n",
       "      <td>[-0.16790096697298518, 0.009406178804481543, -...</td>\n",
       "      <td>Garment Upper body</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5307289</th>\n",
       "      <td>00000dbacae5abe5e23885899a1fa44253a17956c6d1c3...</td>\n",
       "      <td>919418002</td>\n",
       "      <td>2020-09-05</td>\n",
       "      <td>0</td>\n",
       "      <td>[-0.44182159437733715, -0.0022738383333123723,...</td>\n",
       "      <td>[0.22787903246699484, 0.0828959295300294, 0.17...</td>\n",
       "      <td>[-0.44182159438002244, -0.0022739558543872335,...</td>\n",
       "      <td>[0.22787903246699484, 0.0828959295300294, 0.17...</td>\n",
       "      <td>Garment Full body</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5307290</th>\n",
       "      <td>00000dbacae5abe5e23885899a1fa44253a17956c6d1c3...</td>\n",
       "      <td>796560003</td>\n",
       "      <td>2020-09-05</td>\n",
       "      <td>0</td>\n",
       "      <td>[-0.44182159437733715, -0.0022738383333123723,...</td>\n",
       "      <td>[0.2225992237383261, 0.06412773704708585, 0.18...</td>\n",
       "      <td>[-0.44182159438002244, -0.0022739558543872335,...</td>\n",
       "      <td>[0.2225992237383261, 0.06412773704708585, 0.18...</td>\n",
       "      <td>Garment Upper body</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5307291 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                               customer_id  article_id  \\\n",
       "0        ffffcf35913a0bee60e8741cb2b4e78b8a98ee5ff2e6a1...   762846027   \n",
       "1        ffffcf35913a0bee60e8741cb2b4e78b8a98ee5ff2e6a1...   282832022   \n",
       "2        ffffcf35913a0bee60e8741cb2b4e78b8a98ee5ff2e6a1...   840033006   \n",
       "3        ffffcf35913a0bee60e8741cb2b4e78b8a98ee5ff2e6a1...   809238003   \n",
       "4        ffffcf35913a0bee60e8741cb2b4e78b8a98ee5ff2e6a1...   746282001   \n",
       "...                                                    ...         ...   \n",
       "5307286  00006413d8573cd20ed7128e53b7b13819fe5cfc2d801f...   870640001   \n",
       "5307287  00006413d8573cd20ed7128e53b7b13819fe5cfc2d801f...   825783008   \n",
       "5307288  00000dbacae5abe5e23885899a1fa44253a17956c6d1c3...   568601043   \n",
       "5307289  00000dbacae5abe5e23885899a1fa44253a17956c6d1c3...   919418002   \n",
       "5307290  00000dbacae5abe5e23885899a1fa44253a17956c6d1c3...   796560003   \n",
       "\n",
       "              t_dat  label                                 prev_items_tfidf_x  \\\n",
       "0        2020-09-09      1  [0.6918655795839387, -0.35663200882050616, -0....   \n",
       "1        2020-09-09      0  [0.6918655795839387, -0.35663200882050616, -0....   \n",
       "2        2020-09-09      0  [0.6918655795839387, -0.35663200882050616, -0....   \n",
       "3        2020-09-08      0  [-0.44182159437733715, -0.0022738383333123723,...   \n",
       "4        2020-09-08      0  [-0.44182159437733715, -0.0022738383333123723,...   \n",
       "...             ...    ...                                                ...   \n",
       "5307286  2020-08-12      0  [-0.44182159437733715, -0.0022738383333123723,...   \n",
       "5307287  2020-08-12      0  [-0.44182159437733715, -0.0022738383333123723,...   \n",
       "5307288  2020-09-05      1  [-0.44182159437733715, -0.0022738383333123723,...   \n",
       "5307289  2020-09-05      0  [-0.44182159437733715, -0.0022738383333123723,...   \n",
       "5307290  2020-09-05      0  [-0.44182159437733715, -0.0022738383333123723,...   \n",
       "\n",
       "                                                   tfidf_x  \\\n",
       "0        [0.19067515101280055, 0.055564732377888, -0.31...   \n",
       "1        [-0.14741212571534487, -0.08153015208128013, -...   \n",
       "2        [0.20458352993581452, 0.08828260844679149, 0.2...   \n",
       "3        [0.325056539190052, -0.07741629366772365, -0.3...   \n",
       "4        [-0.003907417025550813, 0.12183911621887447, -...   \n",
       "...                                                    ...   \n",
       "5307286  [-0.2681135879098105, -0.40141412067644494, -0...   \n",
       "5307287  [0.24303098110870489, -0.12804459561348136, 0....   \n",
       "5307288  [-0.16790096697298518, 0.009406178804481543, -...   \n",
       "5307289  [0.22787903246699484, 0.0828959295300294, 0.17...   \n",
       "5307290  [0.2225992237383261, 0.06412773704708585, 0.18...   \n",
       "\n",
       "                                        prev_items_tfidf_y  \\\n",
       "0        [0.691865577858334, -0.35666563392571254, -0.2...   \n",
       "1        [0.691865577858334, -0.35666563392571254, -0.2...   \n",
       "2        [0.691865577858334, -0.35666563392571254, -0.2...   \n",
       "3        [-0.44182159438002244, -0.0022739558543872335,...   \n",
       "4        [-0.44182159438002244, -0.0022739558543872335,...   \n",
       "...                                                    ...   \n",
       "5307286  [-0.44182159438002244, -0.0022739558543872335,...   \n",
       "5307287  [-0.44182159438002244, -0.0022739558543872335,...   \n",
       "5307288  [-0.44182159438002244, -0.0022739558543872335,...   \n",
       "5307289  [-0.44182159438002244, -0.0022739558543872335,...   \n",
       "5307290  [-0.44182159438002244, -0.0022739558543872335,...   \n",
       "\n",
       "                                                   tfidf_y  product_group_name  \n",
       "0        [0.19067515101280055, 0.055564732377888, -0.31...  Garment Upper body  \n",
       "1        [-0.14741212571534487, -0.08153015208128013, -...  Garment Upper body  \n",
       "2        [0.20458352993581452, 0.08828260844679149, 0.2...  Garment Upper body  \n",
       "3        [0.325056539190052, -0.07741629366772365, -0.3...  Garment Upper body  \n",
       "4        [-0.003907417025550813, 0.12183911621887447, -...  Garment Upper body  \n",
       "...                                                    ...                 ...  \n",
       "5307286  [-0.2681135879098105, -0.40141412067644494, -0...  Garment Lower body  \n",
       "5307287  [0.24303098110870489, -0.12804459561348136, 0....         Accessories  \n",
       "5307288  [-0.16790096697298518, 0.009406178804481543, -...  Garment Upper body  \n",
       "5307289  [0.22787903246699484, 0.0828959295300294, 0.17...   Garment Full body  \n",
       "5307290  [0.2225992237383261, 0.06412773704708585, 0.18...  Garment Upper body  \n",
       "\n",
       "[5307291 rows x 9 columns]"
      ]
     },
     "execution_count": 205,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_negative"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "14728833",
   "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>customer_id</th>\n",
       "      <th>article_id</th>\n",
       "      <th>t_dat</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>000058a12d5b43e67d225668fa1f8d618c13dc232df0ca...</td>\n",
       "      <td>794321007</td>\n",
       "      <td>2020-09-15</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>000058a12d5b43e67d225668fa1f8d618c13dc232df0ca...</td>\n",
       "      <td>870123001</td>\n",
       "      <td>2020-09-15</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>000058a12d5b43e67d225668fa1f8d618c13dc232df0ca...</td>\n",
       "      <td>212629047</td>\n",
       "      <td>2020-09-15</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>00040239317e877c77ac6e79df42eb2633ad38fcac09fc...</td>\n",
       "      <td>875272011</td>\n",
       "      <td>2020-09-15</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>00040239317e877c77ac6e79df42eb2633ad38fcac09fc...</td>\n",
       "      <td>904590001</td>\n",
       "      <td>2020-09-15</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>799087</th>\n",
       "      <td>fff4d3a8b1f3b60af93e78c30a7cb4cf75edaf2590d3e5...</td>\n",
       "      <td>871088001</td>\n",
       "      <td>2020-09-22</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>799088</th>\n",
       "      <td>fff4d3a8b1f3b60af93e78c30a7cb4cf75edaf2590d3e5...</td>\n",
       "      <td>587026038</td>\n",
       "      <td>2020-09-22</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>799089</th>\n",
       "      <td>fffef3b6b73545df065b521e19f64bf6fe93bfd450ab20...</td>\n",
       "      <td>898573003</td>\n",
       "      <td>2020-09-22</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>799090</th>\n",
       "      <td>fffef3b6b73545df065b521e19f64bf6fe93bfd450ab20...</td>\n",
       "      <td>802070001</td>\n",
       "      <td>2020-09-22</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>799091</th>\n",
       "      <td>fffef3b6b73545df065b521e19f64bf6fe93bfd450ab20...</td>\n",
       "      <td>864703014</td>\n",
       "      <td>2020-09-22</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>799092 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              customer_id  article_id  \\\n",
       "0       000058a12d5b43e67d225668fa1f8d618c13dc232df0ca...   794321007   \n",
       "1       000058a12d5b43e67d225668fa1f8d618c13dc232df0ca...   870123001   \n",
       "2       000058a12d5b43e67d225668fa1f8d618c13dc232df0ca...   212629047   \n",
       "3       00040239317e877c77ac6e79df42eb2633ad38fcac09fc...   875272011   \n",
       "4       00040239317e877c77ac6e79df42eb2633ad38fcac09fc...   904590001   \n",
       "...                                                   ...         ...   \n",
       "799087  fff4d3a8b1f3b60af93e78c30a7cb4cf75edaf2590d3e5...   871088001   \n",
       "799088  fff4d3a8b1f3b60af93e78c30a7cb4cf75edaf2590d3e5...   587026038   \n",
       "799089  fffef3b6b73545df065b521e19f64bf6fe93bfd450ab20...   898573003   \n",
       "799090  fffef3b6b73545df065b521e19f64bf6fe93bfd450ab20...   802070001   \n",
       "799091  fffef3b6b73545df065b521e19f64bf6fe93bfd450ab20...   864703014   \n",
       "\n",
       "             t_dat  label  \n",
       "0       2020-09-15      1  \n",
       "1       2020-09-15      0  \n",
       "2       2020-09-15      0  \n",
       "3       2020-09-15      1  \n",
       "4       2020-09-15      0  \n",
       "...            ...    ...  \n",
       "799087  2020-09-22      0  \n",
       "799088  2020-09-22      0  \n",
       "799089  2020-09-22      1  \n",
       "799090  2020-09-22      0  \n",
       "799091  2020-09-22      0  \n",
       "\n",
       "[799092 rows x 4 columns]"
      ]
     },
     "execution_count": 83,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_negative"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 184,
   "id": "dda681c7",
   "metadata": {},
   "outputs": [],
   "source": [
    "customer_new_features_df = train.merge(article_df[['detail_desc','article_id','tfidf','product_group_name']], on='article_id', how='outer')\n",
    "\n",
    "#print(customer_features_df.shape)\n",
    "#customer_features_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "id": "1150bbba",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 330712/330712 [01:36<00:00, 3441.26it/s]\n"
     ]
    }
   ],
   "source": [
    "z = customer_new_features_df\n",
    "z = z.sort_values('t_dat', ascending=True)\n",
    "z['user_prev_items'] = z.groupby('customer_id')['detail_desc'].progress_apply(lambda x:\n",
    "                                                                               x.shift(4, fill_value='') + ' ' +\n",
    "                                                                               x.shift(3, fill_value='') + ' ' +\n",
    "                                                                               x.shift(2, fill_value='') + ' ' +\n",
    "                                                                               x.shift(1, fill_value=''))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "id": "1073f593",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 1841054/1841054 [00:00<00:00, 1924080.26it/s]\n"
     ]
    }
   ],
   "source": [
    "z['user_prev_items'] = z['user_prev_items'].fillna('')\n",
    "z['user_prev_items'] = z.user_prev_items.progress_apply(lambda concat_chains: concat_chains.strip())\n",
    "z.loc[z['user_prev_items'] == '','user_prev_items'] = 'nothing'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "id": "d40284ce",
   "metadata": {},
   "outputs": [],
   "source": [
    "## making sure we have a single entry per customer per date (multiple entries can't be merged with the train_negatives, will mess things up)\n",
    "z = z.drop_duplicates(subset=['customer_id','t_dat'], keep='first')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "id": "fa0f3259",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages/texthero/preprocessing.py:46: FutureWarning: The default value of regex will change from True to False in a future version.\n",
      "  ----------\n",
      "/Users/omar.mohamed.2/miniforge3/envs/tfcpu/lib/python3.9/site-packages/texthero/preprocessing.py:81: FutureWarning: The default value of regex will change from True to False in a future version.\n",
      "  By default, only removes \"blocks\" of digits. For instance, `1234 falcon9` becomes ` falcon9`.\n"
     ]
    }
   ],
   "source": [
    "## create another tfidf feature for user previous articles tfidf \n",
    "z['prev_items_tfidf'] = (\n",
    "   z['user_prev_items']\n",
    "   .pipe(hero.clean)\n",
    "   .pipe(hero.do_tfidf)\n",
    "   .pipe(hero.do_pca, n_components=20)\n",
    ").apply(np.array)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "id": "b4a8dfb2",
   "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>customer_id</th>\n",
       "      <th>article_id</th>\n",
       "      <th>t_dat</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>000058a12d5b43e67d225668fa1f8d618c13dc232df0ca...</td>\n",
       "      <td>794321007</td>\n",
       "      <td>2020-09-15</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>000058a12d5b43e67d225668fa1f8d618c13dc232df0ca...</td>\n",
       "      <td>870123001</td>\n",
       "      <td>2020-09-15</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>000058a12d5b43e67d225668fa1f8d618c13dc232df0ca...</td>\n",
       "      <td>212629047</td>\n",
       "      <td>2020-09-15</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>00040239317e877c77ac6e79df42eb2633ad38fcac09fc...</td>\n",
       "      <td>875272011</td>\n",
       "      <td>2020-09-15</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>00040239317e877c77ac6e79df42eb2633ad38fcac09fc...</td>\n",
       "      <td>904590001</td>\n",
       "      <td>2020-09-15</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         customer_id  article_id       t_dat  \\\n",
       "0  000058a12d5b43e67d225668fa1f8d618c13dc232df0ca...   794321007  2020-09-15   \n",
       "1  000058a12d5b43e67d225668fa1f8d618c13dc232df0ca...   870123001  2020-09-15   \n",
       "2  000058a12d5b43e67d225668fa1f8d618c13dc232df0ca...   212629047  2020-09-15   \n",
       "3  00040239317e877c77ac6e79df42eb2633ad38fcac09fc...   875272011  2020-09-15   \n",
       "4  00040239317e877c77ac6e79df42eb2633ad38fcac09fc...   904590001  2020-09-15   \n",
       "\n",
       "   label  \n",
       "0      1  \n",
       "1      0  \n",
       "2      0  \n",
       "3      1  \n",
       "4      0  "
      ]
     },
     "execution_count": 207,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_negative.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 191,
   "id": "706f38ee",
   "metadata": {},
   "outputs": [],
   "source": [
    "## keeping last status of the customer in the training data --> proxy for status of customer before testing data\n",
    "## to make sure we have single row per customer to be merged with test_negative\n",
    "x = z.drop_duplicates(subset=['customer_id'], keep='last')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "id": "bec455a0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((5307291, 5), (448893, 5))"
      ]
     },
     "execution_count": 208,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_negative = train_negative.merge(z[['customer_id','prev_items_tfidf','t_dat']], on=['customer_id','t_dat'], how='inner')\n",
    "\n",
    "test_negative = test_negative.merge(x[['customer_id','prev_items_tfidf']], on=['customer_id'], how='inner')\n",
    "\n",
    "train_negative.shape, test_negative.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 209,
   "id": "e86bfb71",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((5307291, 7), (448893, 7))"
      ]
     },
     "execution_count": 209,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Add article features to both positive and negative samples\n",
    "\n",
    "train_negative = train_negative.merge(article_df[['article_id','tfidf','product_group_name']], on='article_id', how='left')\n",
    "test_negative = test_negative.merge(article_df[['article_id','tfidf','product_group_name']], on='article_id', how='left')\n",
    "\n",
    "train_negative.shape, test_negative.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 210,
   "id": "75ea19da",
   "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>customer_id</th>\n",
       "      <th>article_id</th>\n",
       "      <th>t_dat</th>\n",
       "      <th>label</th>\n",
       "      <th>prev_items_tfidf</th>\n",
       "      <th>tfidf</th>\n",
       "      <th>product_group_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>743123001</td>\n",
       "      <td>2020-08-01</td>\n",
       "      <td>1</td>\n",
       "      <td>[-0.44182159437998064, -0.00227395595104499, -...</td>\n",
       "      <td>[-0.08924477997284976, 0.34677323796692383, 0....</td>\n",
       "      <td>Swimwear</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>870123001</td>\n",
       "      <td>2020-08-01</td>\n",
       "      <td>0</td>\n",
       "      <td>[-0.44182159437998064, -0.00227395595104499, -...</td>\n",
       "      <td>[-0.14017426197529928, -0.2568790705576088, -0...</td>\n",
       "      <td>Garment Lower body</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>212629047</td>\n",
       "      <td>2020-08-01</td>\n",
       "      <td>0</td>\n",
       "      <td>[-0.44182159437998064, -0.00227395595104499, -...</td>\n",
       "      <td>[-0.06570801950846239, 0.13733480094385445, 0....</td>\n",
       "      <td>Garment Full body</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>743123001</td>\n",
       "      <td>2020-08-01</td>\n",
       "      <td>1</td>\n",
       "      <td>[-0.44182159437998064, -0.00227395595104499, -...</td>\n",
       "      <td>[-0.08924477997284976, 0.34677323796692383, 0....</td>\n",
       "      <td>Swimwear</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>904590001</td>\n",
       "      <td>2020-08-01</td>\n",
       "      <td>0</td>\n",
       "      <td>[-0.44182159437998064, -0.00227395595104499, -...</td>\n",
       "      <td>[0.10412189339607347, 0.12203929040686069, -0....</td>\n",
       "      <td>Garment Full body</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         customer_id  article_id       t_dat  \\\n",
       "0  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   743123001  2020-08-01   \n",
       "1  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   870123001  2020-08-01   \n",
       "2  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   212629047  2020-08-01   \n",
       "3  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   743123001  2020-08-01   \n",
       "4  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   904590001  2020-08-01   \n",
       "\n",
       "   label                                   prev_items_tfidf  \\\n",
       "0      1  [-0.44182159437998064, -0.00227395595104499, -...   \n",
       "1      0  [-0.44182159437998064, -0.00227395595104499, -...   \n",
       "2      0  [-0.44182159437998064, -0.00227395595104499, -...   \n",
       "3      1  [-0.44182159437998064, -0.00227395595104499, -...   \n",
       "4      0  [-0.44182159437998064, -0.00227395595104499, -...   \n",
       "\n",
       "                                               tfidf  product_group_name  \n",
       "0  [-0.08924477997284976, 0.34677323796692383, 0....            Swimwear  \n",
       "1  [-0.14017426197529928, -0.2568790705576088, -0...  Garment Lower body  \n",
       "2  [-0.06570801950846239, 0.13733480094385445, 0....   Garment Full body  \n",
       "3  [-0.08924477997284976, 0.34677323796692383, 0....            Swimwear  \n",
       "4  [0.10412189339607347, 0.12203929040686069, -0....   Garment Full body  "
      ]
     },
     "execution_count": 210,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_negative.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 195,
   "id": "925fd327",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['t_dat', 'customer_id', 'article_id', 'price', 'sales_channel_id'], dtype='object')"
      ]
     },
     "execution_count": 195,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 211,
   "id": "e8710c25",
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_single_task_ranker(customer_numerical_features=[],\n",
    "                             article_numerical_features=[],\n",
    "                             n_epochs=5,\n",
    "                             batch_size=1024*4):\n",
    "    \"\"\"\n",
    "    Args:\n",
    "        customer_numerical_features (list): Numerical features used to describe the customer.\n",
    "        article_categorical_feature (str): Single, categorical feature used to describe the article.\n",
    "            In practice, we'd use many more features to describe the article.\n",
    "        n_epochs (int): Number of epochs.\n",
    "        batch_size (int): Batch size.\n",
    "\n",
    "    Returns:\n",
    "        model (tf.Keras.Model): Trained model on train_negative.\n",
    "    \"\"\"\n",
    "\n",
    "    # Assemble training data\n",
    "    X1= np.array([np.array(x[0]) for x in train_negative[customer_numerical_features].values])\n",
    "    X2= np.array([np.array(x[0]) for x in train_negative[article_numerical_features].values])\n",
    "    #X1 = train_negative[customer_numerical_features].values.astype('float32')\n",
    "    #X2 = train_negative[article_numerical_features].values.astype('float32')\n",
    "    y1 = np.asarray(train_negative['label'].values)\n",
    "\n",
    "\n",
    "    # Numerical customer features\n",
    "    input1_layer = tf.keras.Input(20)\n",
    "    input1 = tf.keras.layers.Dense(20, activation='relu')(input1_layer)\n",
    "    \n",
    "    \n",
    "    # Numerical article features\n",
    "    input2_layer = tf.keras.Input(20)\n",
    "    input2 = tf.keras.layers.Dense(20, activation='relu')(input2_layer)\n",
    "\n",
    "    # Combine inputs\n",
    "    combined_layer = tf.keras.layers.concatenate([input1, input2])\n",
    "    combined_layer = tf.keras.layers.Dense(50, activation='relu')(combined_layer)\n",
    "    combined_layer = tf.keras.layers.Dense(50, activation='relu')(combined_layer)\n",
    "    combined_layer = tf.keras.layers.Dense(1, activation='sigmoid')(combined_layer)\n",
    "\n",
    "\n",
    "    # Compile model\n",
    "    bce = tf.keras.losses.BinaryCrossentropy(from_logits=False)\n",
    "    metric = tf.keras.metrics.BinaryAccuracy()\n",
    "    model = tf.keras.Model([input1_layer, input2_layer], combined_layer)\n",
    "    model.compile(loss=bce,\n",
    "                  optimizer='adam',\n",
    "                  metrics=[metric])\n",
    "\n",
    "    # Train model\n",
    "    model.fit([X1, X2], [y1], epochs=n_epochs, verbose=1, batch_size=batch_size)\n",
    "\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 212,
   "id": "f0d5657d",
   "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>customer_id</th>\n",
       "      <th>article_id</th>\n",
       "      <th>t_dat</th>\n",
       "      <th>label</th>\n",
       "      <th>prev_items_tfidf</th>\n",
       "      <th>tfidf</th>\n",
       "      <th>product_group_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>743123001</td>\n",
       "      <td>2020-08-01</td>\n",
       "      <td>1</td>\n",
       "      <td>[-0.44182159437998064, -0.00227395595104499, -...</td>\n",
       "      <td>[-0.08924477997284976, 0.34677323796692383, 0....</td>\n",
       "      <td>Swimwear</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>870123001</td>\n",
       "      <td>2020-08-01</td>\n",
       "      <td>0</td>\n",
       "      <td>[-0.44182159437998064, -0.00227395595104499, -...</td>\n",
       "      <td>[-0.14017426197529928, -0.2568790705576088, -0...</td>\n",
       "      <td>Garment Lower body</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>212629047</td>\n",
       "      <td>2020-08-01</td>\n",
       "      <td>0</td>\n",
       "      <td>[-0.44182159437998064, -0.00227395595104499, -...</td>\n",
       "      <td>[-0.06570801950846239, 0.13733480094385445, 0....</td>\n",
       "      <td>Garment Full body</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>743123001</td>\n",
       "      <td>2020-08-01</td>\n",
       "      <td>1</td>\n",
       "      <td>[-0.44182159437998064, -0.00227395595104499, -...</td>\n",
       "      <td>[-0.08924477997284976, 0.34677323796692383, 0....</td>\n",
       "      <td>Swimwear</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...</td>\n",
       "      <td>904590001</td>\n",
       "      <td>2020-08-01</td>\n",
       "      <td>0</td>\n",
       "      <td>[-0.44182159437998064, -0.00227395595104499, -...</td>\n",
       "      <td>[0.10412189339607347, 0.12203929040686069, -0....</td>\n",
       "      <td>Garment Full body</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         customer_id  article_id       t_dat  \\\n",
       "0  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   743123001  2020-08-01   \n",
       "1  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   870123001  2020-08-01   \n",
       "2  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   212629047  2020-08-01   \n",
       "3  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   743123001  2020-08-01   \n",
       "4  00075ef36696a7b4ed8c83e22a4bf7ea7c90ee110991ec...   904590001  2020-08-01   \n",
       "\n",
       "   label                                   prev_items_tfidf  \\\n",
       "0      1  [-0.44182159437998064, -0.00227395595104499, -...   \n",
       "1      0  [-0.44182159437998064, -0.00227395595104499, -...   \n",
       "2      0  [-0.44182159437998064, -0.00227395595104499, -...   \n",
       "3      1  [-0.44182159437998064, -0.00227395595104499, -...   \n",
       "4      0  [-0.44182159437998064, -0.00227395595104499, -...   \n",
       "\n",
       "                                               tfidf  product_group_name  \n",
       "0  [-0.08924477997284976, 0.34677323796692383, 0....            Swimwear  \n",
       "1  [-0.14017426197529928, -0.2568790705576088, -0...  Garment Lower body  \n",
       "2  [-0.06570801950846239, 0.13733480094385445, 0....   Garment Full body  \n",
       "3  [-0.08924477997284976, 0.34677323796692383, 0....            Swimwear  \n",
       "4  [0.10412189339607347, 0.12203929040686069, -0....   Garment Full body  "
      ]
     },
     "execution_count": 212,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_negative.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "id": "474a4752",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(549028, 10)"
      ]
     },
     "execution_count": 213,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "z.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "id": "7cdbdb7b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "1296/1296 [==============================] - 4s 3ms/step - loss: 0.5820 - binary_accuracy: 0.6951\n",
      "Epoch 2/20\n",
      "1296/1296 [==============================] - 3s 2ms/step - loss: 0.5538 - binary_accuracy: 0.7202\n",
      "Epoch 3/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5436 - binary_accuracy: 0.7286\n",
      "Epoch 4/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5365 - binary_accuracy: 0.7341\n",
      "Epoch 5/20\n",
      "1296/1296 [==============================] - 4s 3ms/step - loss: 0.5315 - binary_accuracy: 0.7377\n",
      "Epoch 6/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5276 - binary_accuracy: 0.7402\n",
      "Epoch 7/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5244 - binary_accuracy: 0.7427\n",
      "Epoch 8/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5219 - binary_accuracy: 0.7447\n",
      "Epoch 9/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5196 - binary_accuracy: 0.7465\n",
      "Epoch 10/20\n",
      "1296/1296 [==============================] - 3s 2ms/step - loss: 0.5177 - binary_accuracy: 0.7481\n",
      "Epoch 11/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5157 - binary_accuracy: 0.7497\n",
      "Epoch 12/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5141 - binary_accuracy: 0.7508\n",
      "Epoch 13/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5127 - binary_accuracy: 0.7519\n",
      "Epoch 14/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5115 - binary_accuracy: 0.7529\n",
      "Epoch 15/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5104 - binary_accuracy: 0.7538\n",
      "Epoch 16/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5091 - binary_accuracy: 0.7548\n",
      "Epoch 17/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5081 - binary_accuracy: 0.7554\n",
      "Epoch 18/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5073 - binary_accuracy: 0.7559\n",
      "Epoch 19/20\n",
      "1296/1296 [==============================] - 3s 2ms/step - loss: 0.5064 - binary_accuracy: 0.7565\n",
      "Epoch 20/20\n",
      "1296/1296 [==============================] - 3s 3ms/step - loss: 0.5056 - binary_accuracy: 0.7569\n"
     ]
    }
   ],
   "source": [
    "single_task_model = train_single_task_ranker(customer_numerical_features=['prev_items_tfidf'],\n",
    "                                             article_numerical_features=['tfidf'], n_epochs = 20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "id": "65d21f3a",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Evaluate model\n",
    "\n",
    "def evaluate(df, model, \n",
    "             customer_numerical_features=[], \n",
    "             article_numerical_features=[],\n",
    "             single_task=True):\n",
    "    \"\"\"\n",
    "    Args:\n",
    "        df (DataFrame): Train or test dataframe to evaluate.\n",
    "        model (tf.keras.Model): Trained model for inference.\n",
    "        customer_numerical_features (list): Numerical features used to describe the customer during training.\n",
    "        article_categorical_feature (str): Single, categorical feature used to describe the article during training.\n",
    "        single_task (bool): True for single-task, False for multi-task.\n",
    "\n",
    "    Returns:\n",
    "        None\n",
    "    \"\"\"\n",
    "\n",
    "    # Make predictions\n",
    "    X1= np.array([np.array(x[0]) for x in df[customer_numerical_features].values])\n",
    "    X2= np.array([np.array(x[0]) for x in df[article_numerical_features].values])\n",
    "    if single_task:\n",
    "        preds = model.predict([X1, X2], batch_size=1024*10, verbose=1)\n",
    "    else:\n",
    "        preds, _ = model.predict([X1, X2], batch_size=1024*10, verbose=1)\n",
    "    preds_rounded = (preds>=0.5).astype(int).flatten()\n",
    "\n",
    "    # Calculate MRR\n",
    "    df['prediction_score'] = preds\n",
    "    df.sort_values(by=['customer_id', 'prediction_score'], ascending=False, inplace=True)\n",
    "    reciprocal_rank = df.groupby('customer_id').apply(lambda x: 1/(np.where(x['label']==1)[0][0]+1))\n",
    "    df.drop('prediction_score', axis=1, inplace=True)\n",
    "    print(f'MRR: {reciprocal_rank.mean():.4f}')\n",
    "    print('Distribution:')\n",
    "    display(reciprocal_rank.describe())\n",
    "    \n",
    "    # Show distribution of labels and compare to model accuracy\n",
    "    print('\\nBaseline:')\n",
    "    print(df['label'].value_counts()/len(df), end='\\n\\n')\n",
    "    print(f\"Model Accuracy: {sum(preds_rounded == df['label'].values)/len(preds_rounded):.2f}\", end='\\n\\n')\n",
    "\n",
    "    # Show confusion matrix\n",
    "    print('Confusion Matrix:')\n",
    "    truth = df['label'].values\n",
    "    cm = confusion_matrix(truth, preds_rounded)\n",
    "    cmd = ConfusionMatrixDisplay(cm, display_labels=['0','1'])\n",
    "    cmd.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "id": "fd36b9d0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "519/519 [==============================] - 1s 2ms/step\n",
      "MRR: 0.8319\n",
      "Distribution:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    330712.000000\n",
       "mean          0.831893\n",
       "std           0.267609\n",
       "min           0.034483\n",
       "25%           0.500000\n",
       "50%           1.000000\n",
       "75%           1.000000\n",
       "max           1.000000\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Baseline:\n",
      "0    0.666667\n",
      "1    0.333333\n",
      "Name: label, dtype: float64\n",
      "\n",
      "Model Accuracy: 0.60\n",
      "\n",
      "Confusion Matrix:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Evaluate the training set with the single-task ranker\n",
    "evaluate(train_negative, single_task_model, \n",
    "         customer_numerical_features=['prev_items_tfidf'],\n",
    "         article_numerical_features=['tfidf'],\n",
    "         single_task=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "id": "1e80ac95",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(448893, 7)"
      ]
     },
     "execution_count": 217,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_negative.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "id": "c708ed90",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "44/44 [==============================] - 0s 2ms/step\n",
      "MRR: 0.8212\n",
      "Distribution:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    42395.000000\n",
       "mean         0.821225\n",
       "std          0.274196\n",
       "min          0.040000\n",
       "25%          0.500000\n",
       "50%          1.000000\n",
       "75%          1.000000\n",
       "max          1.000000\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Baseline:\n",
      "0    0.666667\n",
      "1    0.333333\n",
      "Name: label, dtype: float64\n",
      "\n",
      "Model Accuracy: 0.61\n",
      "\n",
      "Confusion Matrix:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Evaluate the training set with the single-task ranker\n",
    "evaluate(test_negative, single_task_model, \n",
    "         customer_numerical_features=['prev_items_tfidf'],\n",
    "         article_numerical_features=['tfidf'],\n",
    "         single_task=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "61a270b6",
   "metadata": {},
   "source": [
    "## Train Multitask Model "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 226,
   "id": "75ea903f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Garment Upper body       2008224\n",
       "Garment Lower body       1120418\n",
       "Garment Full body         699060\n",
       "Accessories               441689\n",
       "Underwear                 348659\n",
       "Swimwear                  273050\n",
       "Shoes                     191372\n",
       "Socks & Tights            121834\n",
       "Nightwear                  61771\n",
       "Unknown                    31718\n",
       "Bags                        3698\n",
       "Items                       2780\n",
       "Cosmetic                    1227\n",
       "Garment and Shoe care        677\n",
       "Stationery                   454\n",
       "Underwear/nightwear          439\n",
       "Fun                          221\n",
       "Name: product_group_name, dtype: int64"
      ]
     },
     "execution_count": 226,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_negative.product_group_name.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 229,
   "id": "4aac27f1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Distribution of purchases (label):\n",
      "0    3538194\n",
      "1    1769097\n",
      "Name: label, dtype: int64\n",
      "\n",
      "Distribution of product_group_name purchases (label2):\n",
      "1    3536157\n",
      "0    1771134\n",
      "Name: label2, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "train_negative = add_label2_to_train_negative(train_negative, selected_category='Garment Lower body')\n",
    "\n",
    "print('Distribution of purchases (label):')\n",
    "print(train_negative['label'].value_counts())\n",
    "\n",
    "print('\\nDistribution of product_group_name purchases (label2):')\n",
    "print(train_negative['label2'].value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "id": "7365ffa9",
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_multi_task_ranker(customer_numerical_features=[],\n",
    "                            article_numerical_features='',\n",
    "                            n_epochs=5,\n",
    "                            batch_size=1024*4):\n",
    "    \"\"\"\n",
    "    Args:\n",
    "        customer_numerical_features (list): Numerical features used to describe the customer.\n",
    "        article_categorical_feature (str): Single, categorical feature used to describe the article.\n",
    "            In practice, we'd use many more features to describe the article.\n",
    "        n_epochs (int): Number of epochs.\n",
    "        batch_size (int): Batch size.\n",
    "\n",
    "    Returns:\n",
    "        model (tf.Keras.Model): Trained model on train_negative.\n",
    "    \"\"\"\n",
    "    \n",
    "    # Assemble training data\n",
    "    X1= np.array([np.array(x[0]) for x in train_negative[customer_numerical_features].values])\n",
    "    X2= np.array([np.array(x[0]) for x in train_negative[article_numerical_features].values])\n",
    "    y1 = np.asarray(train_negative['label'].values)\n",
    "    y2 = np.asarray(train_negative['label2'].values)\n",
    "\n",
    "\n",
    "    # Numerical customer features\n",
    "    input1_layer = tf.keras.Input(20)\n",
    "    input1 = tf.keras.layers.Dense(20, activation='relu')(input1_layer)\n",
    "    \n",
    "    \n",
    "    # Numerical article features\n",
    "    input2_layer = tf.keras.Input(20)\n",
    "    input2 = tf.keras.layers.Dense(20, activation='relu')(input2_layer)\n",
    "\n",
    "    # Combine inputs\n",
    "    combined_layer = tf.keras.layers.concatenate([input1, input2])\n",
    "    combined_layer = tf.keras.layers.Dense(50, activation='relu')(combined_layer)\n",
    "    #combined_layer = tf.keras.layers.Dense(50, activation='relu')(combined_layer)\n",
    "    #combined_layer = tf.keras.layers.Dense(1, activation='sigmoid')(combined_layer)\n",
    "\n",
    "    \n",
    "    # Combine inputs (task 1, label)\n",
    "    combined_layer1 = tf.keras.layers.Dense(50, activation='relu', name=\"task1_layer\")(combined_layer)\n",
    "    combined_layer1 = tf.keras.layers.Dense(1, activation='sigmoid')(combined_layer1)\n",
    "\n",
    "    # Combine inputs (task 2, label2)\n",
    "    combined_layer2 = tf.keras.layers.Dense(50, activation='relu', name=\"task2_layer\")(combined_layer)\n",
    "    combined_layer2 = tf.keras.layers.Dense(1, activation='sigmoid')(combined_layer2)\n",
    "\n",
    "    # Compile model\n",
    "    bce = tf.keras.losses.BinaryCrossentropy(from_logits=False)\n",
    "    metric = tf.keras.metrics.BinaryAccuracy()\n",
    "    model = tf.keras.Model([input1_layer, input2_layer], \n",
    "                           [combined_layer1, combined_layer2])\n",
    "    model.compile(loss=bce,\n",
    "                  optimizer='adam',\n",
    "                  metrics=[metric])\n",
    "\n",
    "    # Train model\n",
    "    model.fit([X1, X2], [y1, y2], epochs=n_epochs, verbose=1, batch_size=batch_size)\n",
    "\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "id": "ec60da80",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.1649 - dense_43_loss: 0.5883 - dense_44_loss: 0.5766 - dense_43_binary_accuracy: 0.6908 - dense_44_binary_accuracy: 0.6680\n",
      "Epoch 2/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.1284 - dense_43_loss: 0.5615 - dense_44_loss: 0.5670 - dense_43_binary_accuracy: 0.7133 - dense_44_binary_accuracy: 0.6699\n",
      "Epoch 3/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.1164 - dense_43_loss: 0.5515 - dense_44_loss: 0.5650 - dense_43_binary_accuracy: 0.7217 - dense_44_binary_accuracy: 0.6704\n",
      "Epoch 4/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.1078 - dense_43_loss: 0.5437 - dense_44_loss: 0.5641 - dense_43_binary_accuracy: 0.7279 - dense_44_binary_accuracy: 0.6708\n",
      "Epoch 5/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.1014 - dense_43_loss: 0.5379 - dense_44_loss: 0.5635 - dense_43_binary_accuracy: 0.7330 - dense_44_binary_accuracy: 0.6710\n",
      "Epoch 6/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0966 - dense_43_loss: 0.5334 - dense_44_loss: 0.5632 - dense_43_binary_accuracy: 0.7369 - dense_44_binary_accuracy: 0.6712\n",
      "Epoch 7/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0924 - dense_43_loss: 0.5296 - dense_44_loss: 0.5629 - dense_43_binary_accuracy: 0.7408 - dense_44_binary_accuracy: 0.6714\n",
      "Epoch 8/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0892 - dense_43_loss: 0.5266 - dense_44_loss: 0.5626 - dense_43_binary_accuracy: 0.7431 - dense_44_binary_accuracy: 0.6715\n",
      "Epoch 9/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0864 - dense_43_loss: 0.5241 - dense_44_loss: 0.5624 - dense_43_binary_accuracy: 0.7451 - dense_44_binary_accuracy: 0.6717\n",
      "Epoch 10/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0840 - dense_43_loss: 0.5218 - dense_44_loss: 0.5621 - dense_43_binary_accuracy: 0.7468 - dense_44_binary_accuracy: 0.6718\n",
      "Epoch 11/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0822 - dense_43_loss: 0.5202 - dense_44_loss: 0.5620 - dense_43_binary_accuracy: 0.7480 - dense_44_binary_accuracy: 0.6719\n",
      "Epoch 12/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0804 - dense_43_loss: 0.5185 - dense_44_loss: 0.5619 - dense_43_binary_accuracy: 0.7491 - dense_44_binary_accuracy: 0.6720\n",
      "Epoch 13/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0788 - dense_43_loss: 0.5171 - dense_44_loss: 0.5618 - dense_43_binary_accuracy: 0.7502 - dense_44_binary_accuracy: 0.6720\n",
      "Epoch 14/20\n",
      "1296/1296 [==============================] - 6s 4ms/step - loss: 1.0774 - dense_43_loss: 0.5158 - dense_44_loss: 0.5616 - dense_43_binary_accuracy: 0.7512 - dense_44_binary_accuracy: 0.6721\n",
      "Epoch 15/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0760 - dense_43_loss: 0.5145 - dense_44_loss: 0.5615 - dense_43_binary_accuracy: 0.7518 - dense_44_binary_accuracy: 0.6721\n",
      "Epoch 16/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0746 - dense_43_loss: 0.5132 - dense_44_loss: 0.5614 - dense_43_binary_accuracy: 0.7525 - dense_44_binary_accuracy: 0.6721\n",
      "Epoch 17/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0734 - dense_43_loss: 0.5120 - dense_44_loss: 0.5613 - dense_43_binary_accuracy: 0.7534 - dense_44_binary_accuracy: 0.6721\n",
      "Epoch 18/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0722 - dense_43_loss: 0.5110 - dense_44_loss: 0.5612 - dense_43_binary_accuracy: 0.7540 - dense_44_binary_accuracy: 0.6722\n",
      "Epoch 19/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0713 - dense_43_loss: 0.5102 - dense_44_loss: 0.5611 - dense_43_binary_accuracy: 0.7545 - dense_44_binary_accuracy: 0.6722\n",
      "Epoch 20/20\n",
      "1296/1296 [==============================] - 5s 4ms/step - loss: 1.0703 - dense_43_loss: 0.5092 - dense_44_loss: 0.5611 - dense_43_binary_accuracy: 0.7553 - dense_44_binary_accuracy: 0.6723\n"
     ]
    }
   ],
   "source": [
    "multi_task_model = train_multi_task_ranker(customer_numerical_features=['prev_items_tfidf'],\n",
    "                                           article_numerical_features=['tfidf'], n_epochs=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "id": "4c631874",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "519/519 [==============================] - 1s 2ms/step\n",
      "MRR: 0.8284\n",
      "Distribution:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    330712.000000\n",
       "mean          0.828426\n",
       "std           0.269960\n",
       "min           0.027027\n",
       "25%           0.500000\n",
       "50%           1.000000\n",
       "75%           1.000000\n",
       "max           1.000000\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Baseline:\n",
      "0    0.666667\n",
      "1    0.333333\n",
      "Name: label, dtype: float64\n",
      "\n",
      "Model Accuracy: 0.74\n",
      "\n",
      "Confusion Matrix:\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfYAAAHACAYAAABQyEoMAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAA9hAAAPYQGoP6dpAAA+hUlEQVR4nO3dd3xUVf7/8fckpEEKhJICIZRAAIEEUSC4NBdFdBXWn2VZXSIC+1VBEazoooIlKiqIsqJSIq4s2EBBXUWUJlgIRAUhGnpJAhhISCBt5v7+iAyOocxkJgkz9/Xcx31s5sw9936GzeYzn3POvddiGIYhAADgE/zqOgAAAOA5JHYAAHwIiR0AAB9CYgcAwIeQ2AEA8CEkdgAAfAiJHQAAH0JiBwDAh5DYAQDwISR2AAB8CIkdAOATVq9erauvvlqxsbGyWCxasmSJy8cwDEPPPfec2rdvr6CgIDVv3lxPPvmk54OtQfXqOgAAADyhuLhYSUlJuvXWW3XttddW6xjjxo3TZ599pueee05dunRRfn6+8vPzPRxpzbLwEBgAgK+xWCxavHixhg4dam8rLS3Vww8/rP/+9786evSoOnfurGeeeUb9+/eXJG3dulVdu3bV5s2blZiYWDeBewBD8QAAUxg7dqzWr1+vhQsX6ocfftD111+vK664Qr/88oskaenSpWrTpo2WLVum1q1bq1WrVho1apTXVewkdgCAz9uzZ4/mzZund955R3369FHbtm1177336k9/+pPmzZsnSdqxY4d2796td955R/Pnz1d6eroyMjJ03XXX1XH0rmGOHQDg83788UdZrVa1b9/eob20tFSNGzeWJNlsNpWWlmr+/Pn2/ebMmaPu3bsrKyvLa4bnSewAAJ9XVFQkf39/ZWRkyN/f3+G90NBQSVJMTIzq1avnkPw7duwoqbLiJ7EDAHCe6Natm6xWqw4ePKg+ffqcdp9LLrlEFRUV2r59u9q2bStJ+vnnnyVJ8fHxtRaru1gVDwDwCUVFRcrOzpZUmchfeOEFDRgwQJGRkWrZsqVuvvlmffXVV3r++efVrVs3HTp0SCtWrFDXrl111VVXyWaz6eKLL1ZoaKimT58um82mMWPGKDw8XJ999lkdfzrnkdgBAD5h5cqVGjBgQJX21NRUpaenq7y8XE888YTmz5+v/fv3q0mTJurVq5cmT56sLl26SJIOHDigO++8U5999pkaNGigwYMH6/nnn1dkZGRtf5xqI7EDAOBDuNwNAAAfQmIHAMCHePWqeJvNpgMHDigsLEwWi6WuwwEAuMgwDB07dkyxsbHy86u5WrOkpERlZWVuHycwMFDBwcEeiKjmeHViP3DggOLi4uo6DACAm/bu3asWLVrUyLFLSkrUOj5UuQetbh8rOjpaO3fuPK+Tu1cn9rCwMEnS7o2tFB7KrAJ801/bd6nrEIAaU6FyrdXH9r/nNaGsrEy5B63andFK4WHVzxWFx2yK775LZWVlJPaacnL4PTzUz63/sYDzWT1LQF2HANSc367Lqo3p1NAwi0LDqn8em7xjyterEzsAAM6yGjZZ3bjA22rYPBdMDSKxAwBMwSZDNlU/s7vTtzYxfg0AQA145ZVX1LVrV4WHhys8PFwpKSn65JNPztrnnXfeUYcOHRQcHKwuXbro448/dvm8JHYAgCnYPPAfV7Ro0UJPP/20MjIytGHDBl166aUaMmSItmzZctr9161bp2HDhmnkyJHatGmThg4dqqFDh2rz5s0underbylbWFioiIgIHfm5DYvn4LMGxSbXdQhAjakwyrVSH6igoEDh4eE1co6TuWLvtuZur4qP67DfrVgjIyM1depUjRw5ssp7N954o4qLi7Vs2TJ7W69evZScnKxZs2Y5fQ6yIQAANcxqtWrhwoUqLi5WSkrKafdZv369Bg4c6NA2aNAgrV+/3qVzsXgOAGAKnlo8V1hY6NAeFBSkoKCg0/b58ccflZKSopKSEoWGhmrx4sXq1KnTaffNzc1VVFSUQ1tUVJRyc3NdipOKHQBgCjYZsrqxnUzscXFxioiIsG9paWlnPGdiYqIyMzP1zTff6Pbbb1dqaqp++umnGv2cVOwAALhg7969DnPsZ6rWpcp7yyckJEiSunfvru+++04vvviiXn311Sr7RkdHKy8vz6EtLy9P0dHRLsVHxQ4AMIWTQ/HubJLsl6+d3M6W2KvEYLOptLT0tO+lpKRoxYoVDm3Lly8/45z8mVCxAwBMwWoYsrpxIZirfSdOnKjBgwerZcuWOnbsmBYsWKCVK1fq008/lSQNHz5czZs3tw/ljxs3Tv369dPzzz+vq666SgsXLtSGDRv02muvuXReEjsAADXg4MGDGj58uHJychQREaGuXbvq008/1WWXXSZJ2rNnj8Ojanv37q0FCxboX//6lx566CG1a9dOS5YsUefOnV06L4kdAGAKtt82d/q7Ys6cOWd9f+XKlVXarr/+el1//fUunskRiR0AYAonV7e7098bkNgBAKZgNeTm0908F0tNYlU8AAA+hIodAGAKtT3HXldI7AAAU7DJIqssbvX3BgzFAwDgQ6jYAQCmYDMqN3f6ewMSOwDAFKxuDsW707c2MRQPAIAPoWIHAJiCWSp2EjsAwBRshkU2w41V8W70rU0MxQMA4EOo2AEApsBQPAAAPsQqP1ndGKi2ejCWmkRiBwCYguHmHLvBHDsAAKhtVOwAAFNgjh0AAB9iNfxkNdyYY/eSW8oyFA8AgA+hYgcAmIJNFtncqGdt8o6SncQOADAFs8yxMxQPAIAPoWIHAJiC+4vnGIoHAOC8UTnH7sZDYBiKBwAAtY2KHQBgCjY37xXPqngAAM4jzLEDAOBDbPIzxXXszLEDAOBDqNgBAKZgNSyyuvHoVXf61iYSOwDAFKxuLp6zMhQPAABqGxU7AMAUbIafbG6sirexKh4AgPMHQ/EAAMDrULEDAEzBJvdWtts8F0qNIrEDAEzB/RvUeMcgt3dECQAAnELFDgAwBffvFe8dtTCJHQBgCmZ5HjuJHQBgCmap2L0jSgAA4BQqdgCAKbh/gxrvqIVJ7AAAU7AZFtncuY7dS57u5h1fPwAAgFOo2AEApmBzcyjeW25QQ2IHAJiC+093847E7h1RAgAAp1CxAwBMwSqLrG7cZMadvrWJxA4AMAWG4gEAgNehYgcAmIJV7g2nWz0XSo0isQMATMEsQ/EkdgCAKfAQGAAA4HWo2AEApmC4+Tx2g8vdAAA4fzAUDwAAvA4VOwDAFMzy2FYSOwDAFKxuPt3Nnb61yTuiBADAy6Slpeniiy9WWFiYmjVrpqFDhyorK+usfdLT02WxWBy24OBgl85LYgcAmMLJoXh3NlesWrVKY8aM0ddff63ly5ervLxcl19+uYqLi8/aLzw8XDk5OfZt9+7dLp2XoXgAgCnY5CebG/Wsq33/97//ObxOT09Xs2bNlJGRob59+56xn8ViUXR0dLVilKjYAQCoFQUFBZKkyMjIs+5XVFSk+Ph4xcXFaciQIdqyZYtL56FiBwCYgtWwyOrGyvaTfQsLCx3ag4KCFBQUdNa+NptNd999ty655BJ17tz5jPslJiZq7ty56tq1qwoKCvTcc8+pd+/e2rJli1q0aOFUnFTsAABT8NQce1xcnCIiIuxbWlraOc89ZswYbd68WQsXLjzrfikpKRo+fLiSk5PVr18/vf/++2ratKleffVVpz8nFTsAwBQMN5/uZvzWd+/evQoPD7e3n6taHzt2rJYtW6bVq1c7XXWfFBAQoG7duik7O9vpPlTsAAC4IDw83GE7U2I3DENjx47V4sWL9cUXX6h169Yun8tqterHH39UTEyM032o2AEApmCVRVY3HuTiat8xY8ZowYIF+uCDDxQWFqbc3FxJUkREhEJCQiRJw4cPV/Pmze3D+VOmTFGvXr2UkJCgo0ePaurUqdq9e7dGjRrl9HlJ7AAAU7AZ7t0W1ma4tv8rr7wiSerfv79D+7x583TLLbdIkvbs2SM/v1OD50eOHNHo0aOVm5urRo0aqXv37lq3bp06derk9HlJ7AAA1ADDOPc3gZUrVzq8njZtmqZNm+bWeUnsXm7pG4310fwmytsbKEmKTyzRTeNzdfGlx067/66sYM2fGq3sH+orb1+g/m/yfl07+lCNx7l6aYTeeDZGefsC1bx1qUY+fEA9/uwY455fgjTniVj98HWorBVSfPtSTXp9p5q1KK/x+OCdbhibp5EP5Wrx600069HmimpRpvnfbj3tvk/8M15rljWUJN3++H5dcHGx4hNLtDc7SHdcluiwb0CQTXc9vU/tup5Qy3Yl+ubzcE2+ter8aNeUIv3zsQOKb1+iwwcCtODFKC1/+9Q1ym9885Oi46r+/n6Y3lgzH3JtERXcZ3Nz8Zw7fWvTeRHlzJkz1apVKwUHB6tnz5769ttv6zokr9E0ply3PnRAL/8vSy998rOSLjmmx0a01q6s099buPSERTEty3TrQwcU2cwzCfP7daEa3uPMw0RbvquvtDta6Yphv+rfn2Wp9xUFmnxra+3adirGA7sCNWFoO8UllGjqu9matSJLf787V4HBLo59wTTaJx3XVTfna8eWU79Hhw4E6G9JnRy2+VOjdLzIT999EebQ/9OFkVr9YcPTHtvPz1BZiZ8+mNNEm9aEnXafqLhSPf7mTv3wVajuuKy9Fs9uqvHP7VX3fqeucb5rcHuHWB68sY0kac3S058XNcsmi9ubN6jzin3RokWaMGGCZs2apZ49e2r69OkaNGiQsrKy1KxZs7oO77zX63LHGyWMeDBXy+Y30baM+mqVWFJl/8TkE0pMPiFJmvtU7GmPabNJb89spo//01hHDgWoeZtS3XR3rvr8paBaMS6Z3VQXDSjU9XdUjgyk3p+rjavD9MG8Jhr3zD5JUvrTMepxaaFGTcqx94ttVVat88H3Bde36oGXd2v6fS00bFyevd1ms+jIoQCHfXsPLtDqpQ1Vctzf3vbKpOaSpIjGuWrd6USV45ee8NdLEysr6k4XFys0wlpln78M/1W5ewL12pTK/x/tzQ7WBT2Kde0/DytjVeWlUAX5jn9ibxx7UAd2BuqH9Q2q87EBp9R5xf7CCy9o9OjRGjFihDp16qRZs2apfv36mjt3bl2H5nWsVmnlkoYqPe6njhed/SEDZ7PwpSh9/k6k7npmn177cpuuHX1Iz9wZX+0/RlszGqhbnyKHtu79jmlrRuXxbDbp2xXhat6mVA8Na6Mbulygu65qp3WfRFT7M8C3jX1qv75dEX7GavqkhC7HldC5RJ/+9+y38KyOjt2Pa9OaUIe2jJVh6tj99P/fqxdg06X/74g+XRgpeUnl52tO3nnOnc0b1GnFXlZWpoyMDE2cONHe5ufnp4EDB2r9+vV1GJl32bk1WHdf3U5lpX4KaWDTI3N2Kr59abWOVVZq0cIZzfT0ou3qdNFxSVJMfL62fNtAH73ZRF1TXP/CcORQPTVq4jjs36hpuY4crPz1O3q4nk4U+2vRy810ywO5GvlwjjZ8GaYpo1rp2Xezq3VO+K5+Q44oocsJ3Xllu3Pue8WwfO3+OUg/bfB8hdyoabmOHHL8YnHkUD01CLcpMNimshLHuqn3FYUKDbfqs7c9/yUDzjHLHHudJvbDhw/LarUqKirKoT0qKkrbtm2rsn9paalKS08lrD/er9esWrQt1b+XZ+n4MX+tWdZQz42L19T3f6lWcj+wK0ilJ/w18W9tHdoryi1q2/nUkOWQhC72n202i8pLLQ5tl/6/I/Zh9nMxbJX/nTKoUNf+s3K4vm3nE/ppQwN9NL96Xybgm5rGlun2KQc08W9tVF569j+ygcE2DfjrES2YHnXW/WrLoGG/6rsvw5WfF3DunQE31PkcuyvS0tI0efLkug7jvBMQaKh568r56HZdTygrs76WzG6qcc86l1h/r6S48o/l42/uUONoxyo7IPDUQrZ/L8+y/5y1qYHmPBmjZ989dcvDBmE2+8+NmlboyGHHP2ZHDgWoUbMKSVJ4pFX+9QzFt3dcExDXrkRbvmUuEqckdD2hRk0rNPPTn+1t/vWkLr2Kdc2Iw/pLq66y2SqHS/tcdVRBIYY+f6dmKuQjhwLUqGmFQ1ujphUqLvSrUq03a16mbn2K9PioVjUSC5xjk+vPVP9jf29Qp4m9SZMm8vf3V15enkN7Xl7eaZ9FO3HiRE2YMMH+urCwUHFxcTUep7cxDKm8rHpDRi3blyggyKaD+wPPWimf/CIhSYdzAuXv79j2ex27FytzTajDZXUbV5+aiwwINNQ+6bj2bXe8LeP+HUFc6gYHmWtC9c8B7R3a7pm2V3uzg/X2zKb2pC5Jg4bl6+vPwqssYPOUrRn1q1xWemHfU2tHfu/yv+Xr6OF6+ubz8CrvofYYbq5sN0js5xYYGKju3btrxYoVGjp0qKTKR9utWLFCY8eOrbK/M4/GM5u5T8Xo4ksL1bR5uU4U+enLxY30w7pQPblguyTp2btaqkl0uW59qHK1eXmZRXt+rrw8qLzcol9zArR9c4iCG1jVvHWZ6ofadN1tB/Xqo81l2KQLehSruNBfP33XQPXDrLrshiMuxzh01CHd9//a6d1ZTdXjz4Va9UEj/fJDiO6eute+z/V3HNRTt8Wrc68iJfUu0oYvw/X18ghNfdf5Bx/A950o9tfurBCHtpLjfjp2xLE9tlWpuvQq1qSbT39v7thWpQpuYFNk0woFBhtqc0HlNNOen4NUUV75pbhluxLVCzQU1siq+g2s9n12bKk8z7L5jXXNiF818l8H9NnCSCVdUqS+Vx/VpH84ntNiMXT5jfn6/J1Gslm9IzH4qt8/oa26/b1BnQ/FT5gwQampqbrooovUo0cPTZ8+XcXFxRoxYkRdh+YVjh6up6l3xSv/YD3VD7OqdccSPblgu7r3q1yFfmh/oH53t0L9mhegOy4/dTOOd2c107uzmqlrSpGmvleZRFPvz1VE4wotfClKuXsC1SDcqoQuJzTsLseRFWddcPFxPThzl954JkbpT8cotnWpHp27U606nBp6v2Rwge56ep8WvhylVya1UIs2lTen6dyT+XW4btDf8nU4J0AZq06/av7u5/Yqqfep361XllcO7Q/v0VF5+ypv9vT4f3Y43Fzm5D6DYpMkSXl7gzTpH631f5P3a+jIwzqcE6Bp98bZL3U7qVvfIkW1KNenCxt77gMCZ2ExnLnnXQ17+eWXNXXqVOXm5io5OVkzZsxQz549z9mvsLBQEREROvJzG4WHecdqRcBVg2KT6zoEoMZUGOVaqQ9UUFDg8ChUTzqZK/66fIQCGgRW+zjlxWVafNm8Go3VE+q8Ypcqn1V7uqF3AAA8xSxD8ZS5AAD4kPOiYgcAoKa5e793LncDAOA8wlA8AADwOlTsAABTMEvFTmIHAJiCWRI7Q/EAAPgQKnYAgCmYpWInsQMATMGQe5es1fltWp1EYgcAmIJZKnbm2AEA8CFU7AAAUzBLxU5iBwCYglkSO0PxAAD4ECp2AIApmKViJ7EDAEzBMCwy3EjO7vStTQzFAwDgQ6jYAQCmwPPYAQDwIWaZY2coHgAAH0LFDgAwBbMsniOxAwBMwSxD8SR2AIApmKViZ44dAAAfQsUOADAFw82heG+p2EnsAABTMCQZhnv9vQFD8QAA+BAqdgCAKdhkkYU7zwEA4BtYFQ8AALwOFTsAwBRshkUWblADAIBvMAw3V8V7ybJ4huIBAPAhVOwAAFMwy+I5EjsAwBRI7AAA+BCzLJ5jjh0AAB9CxQ4AMAWzrIonsQMATKEysbszx+7BYGoQQ/EAAPgQKnYAgCmwKh4AAB9iyL1nqnvJSDxD8QAA+BIqdgCAKTAUDwCALzHJWDyJHQBgDm5W7PKSip05dgAAfAgVOwDAFLjzHAAAPsQsi+cYigcAoAakpaXp4osvVlhYmJo1a6ahQ4cqKyvrnP3eeecddejQQcHBwerSpYs+/vhjl85LYgcAmINhcX9zwapVqzRmzBh9/fXXWr58ucrLy3X55ZeruLj4jH3WrVunYcOGaeTIkdq0aZOGDh2qoUOHavPmzU6f12IY3jJrUFVhYaEiIiJ05Oc2Cg/jOwp806DY5LoOAagxFUa5VuoDFRQUKDw8vEbOcTJXxM+eJL/6wdU+ju14iXaPerzasR46dEjNmjXTqlWr1Ldv39Puc+ONN6q4uFjLli2zt/Xq1UvJycmaNWuWU+chGwIA4ILCwkKHrbS01Kl+BQUFkqTIyMgz7rN+/XoNHDjQoW3QoEFav3690/GR2AEA5mB4YJMUFxeniIgI+5aWlnbOU9tsNt1999265JJL1Llz5zPul5ubq6ioKIe2qKgo5ebmOv0xnVoV/+GHHzp9wGuuucbpfQEAqC2eWhW/d+9eh6H4oKCgc/YdM2aMNm/erLVr11b7/M5yKrEPHTrUqYNZLBZZrVZ34gEA4LwWHh7u0hz72LFjtWzZMq1evVotWrQ4677R0dHKy8tzaMvLy1N0dLTT53NqKN5mszm1kdQBAOc1N4fhXTqVYWjs2LFavHixvvjiC7Vu3fqcfVJSUrRixQqHtuXLlyslJcXp87p1g5qSkhIFB1d/hSEAALWltm9QM2bMGC1YsEAffPCBwsLC7PPkERERCgkJkSQNHz5czZs3t8/Tjxs3Tv369dPzzz+vq666SgsXLtSGDRv02muvOX1elxfPWa1WPf7442revLlCQ0O1Y8cOSdKkSZM0Z84cVw8HAEDt8NDiOWe98sorKigoUP/+/RUTE2PfFi1aZN9nz549ysnJsb/u3bu3FixYoNdee01JSUl69913tWTJkrMuuPsjlyv2J598Um+88YaeffZZjR492t7euXNnTZ8+XSNHjnT1kAAA+BxnbhOzcuXKKm3XX3+9rr/++mqf1+WKff78+Xrttdd00003yd/f396elJSkbdu2VTsQAABqlsUD2/nP5Yp9//79SkhIqNJus9lUXl7ukaAAAPC4ai6Cc+jvBVyu2Dt16qQ1a9ZUaX/33XfVrVs3jwQFAACqx+WK/ZFHHlFqaqr2798vm82m999/X1lZWZo/f77DvW0BADivULGf3pAhQ7R06VJ9/vnnatCggR555BFt3bpVS5cu1WWXXVYTMQIA4L5afrpbXanWdex9+vTR8uXLPR0LAABwU7VvULNhwwZt3bpVUuW8e/fu3T0WFAAAnmYYlZs7/b2By4l93759GjZsmL766is1bNhQknT06FH17t1bCxcuPOd9cAEAqBPMsZ/eqFGjVF5erq1btyo/P1/5+fnaunWrbDabRo0aVRMxAgAAJ7lcsa9atUrr1q1TYmKivS0xMVEvvfSS+vTp49HgAADwGHcXwPnq4rm4uLjT3ojGarUqNjbWI0EBAOBpFqNyc6e/N3B5KH7q1Km68847tWHDBnvbhg0bNG7cOD333HMeDQ4AAI+p5YfA1BWnKvZGjRrJYjk1BFFcXKyePXuqXr3K7hUVFapXr55uvfVWDR06tEYCBQAA5+ZUYp8+fXoNhwEAQA1jjv2U1NTUmo4DAICaZZLL3ap9gxpJKikpUVlZmUNbeHi4WwEBAIDqc3nxXHFxscaOHatmzZqpQYMGatSokcMGAMB5ySSL51xO7Pfff7+++OILvfLKKwoKCtLs2bM1efJkxcbGav78+TURIwAA7jNJYnd5KH7p0qWaP3+++vfvrxEjRqhPnz5KSEhQfHy83nrrLd100001EScAAHCCyxV7fn6+2rRpI6lyPj0/P1+S9Kc//UmrV6/2bHQAAHiKSR7b6nJib9OmjXbu3ClJ6tChg95++21JlZX8yYfCAABwvjl55zl3Nm/gcmIfMWKEvv/+e0nSgw8+qJkzZyo4OFjjx4/Xfffd5/EAAQCA81yeYx8/frz954EDB2rbtm3KyMhQQkKCunbt6tHgAADwGK5jd058fLzi4+M9EQsAAHCTU4l9xowZTh/wrrvuqnYwAADUFIvcfLqbxyKpWU4l9mnTpjl1MIvFQmIHAKAOOZXYT66CP18NvXmY6tULruswgBrh162irkMAaozFWip9/0HtnIyHwAAA4ENMsnjO5cvdAADA+YuKHQBgDiap2EnsAABTcPfucT575zkAAHD+qlZiX7NmjW6++WalpKRo//79kqQ333xTa9eu9WhwAAB4jEke2+pyYn/vvfc0aNAghYSEaNOmTSotLZUkFRQU6KmnnvJ4gAAAeASJ/fSeeOIJzZo1S6+//roCAgLs7Zdccok2btzo0eAAAIBrXF48l5WVpb59+1Zpj4iI0NGjRz0REwAAHsfiuTOIjo5WdnZ2lfa1a9eqTZs2HgkKAACPO3nnOXc2L+ByYh89erTGjRunb775RhaLRQcOHNBbb72le++9V7fffntNxAgAgPtMMsfu8lD8gw8+KJvNpj//+c86fvy4+vbtq6CgIN1777268847ayJGAADgJJcTu8Vi0cMPP6z77rtP2dnZKioqUqdOnRQaGloT8QEA4BFmmWOv9p3nAgMD1alTJ0/GAgBAzeGWsqc3YMAAWSxnXkDwxRdfuBUQAACoPpcTe3JyssPr8vJyZWZmavPmzUpNTfVUXAAAeJabQ/E+W7FPmzbttO2PPfaYioqK3A4IAIAaYZKheI89BObmm2/W3LlzPXU4AABQDR57bOv69esVHBzsqcMBAOBZJqnYXU7s1157rcNrwzCUk5OjDRs2aNKkSR4LDAAAT+JytzOIiIhweO3n56fExERNmTJFl19+uccCAwAArnMpsVutVo0YMUJdunRRo0aNaiomAABQTS4tnvP399fll1/OU9wAAN7HJPeKd3lVfOfOnbVjx46aiAUAgBpzco7dnc0buJzYn3jiCd17771atmyZcnJyVFhY6LABAIC64/Qc+5QpU3TPPffoyiuvlCRdc801DreWNQxDFotFVqvV81ECAOAJXlJ1u8PpxD558mTddttt+vLLL2syHgAAagbXsTsyjMpP1K9fvxoLBgAAuMely93O9lQ3AADOZ9yg5jTat29/zuSen5/vVkAAANQIhuKrmjx5cpU7zwEAgPOHS4n9b3/7m5o1a1ZTsQAAUGMYiv8D5tcBAF7NJEPxTt+g5uSqeAAAcG6rV6/W1VdfrdjYWFksFi1ZsuSs+69cuVIWi6XKlpub69J5na7YbTabSwcGAOC8UssVe3FxsZKSknTrrbdWeeT52WRlZSk8PNz+2tUpcJcf2woAgDeq7Tn2wYMHa/DgwS6fp1mzZmrYsKHL/U5y+V7xAAB4JQ893e2Pz0gpLS31aJjJycmKiYnRZZddpq+++srl/iR2AABcEBcXp4iICPuWlpbmkePGxMRo1qxZeu+99/Tee+8pLi5O/fv318aNG106DkPxAABz8NAc+969ex3mwIOCgtwK66TExEQlJibaX/fu3Vvbt2/XtGnT9Oabbzp9HBI7AMAUPDXHHh4e7pDYa1KPHj20du1al/owFA8AwHkqMzNTMTExLvWhYgcAmEMtX+5WVFSk7Oxs++udO3cqMzNTkZGRatmypSZOnKj9+/dr/vz5kqTp06erdevWuuCCC1RSUqLZs2friy++0GeffebSeUnsAABTqO3L3TZs2KABAwbYX0+YMEGSlJqaqvT0dOXk5GjPnj3298vKynTPPfdo//79ql+/vrp27arPP//c4RjOILEDAFAD+vfvf9a7tqanpzu8vv/++3X//fe7fV4SOwDAHExyr3gSOwDAHEyS2FkVDwCAD6FiBwCYguW3zZ3+3oDEDgAwB5MMxZPYAQCmUNuXu9UV5tgBAPAhVOwAAHNgKB4AAB/jJcnZHQzFAwDgQ6jYAQCmYJbFcyR2AIA5mGSOnaF4AAB8CBU7AMAUGIoHAMCXMBQPAAC8DRU7AMAUGIoHAMCXmGQonsQOADAHkyR25tgBAPAhVOwAAFNgjh0AAF/CUDwAAPA2VOwAAFOwGIYsRvXLbnf61iYSOwDAHBiKBwAA3oaKHQBgCqyKBwDAlzAUDwAAvA0VOwDAFBiKBwDAl5hkKJ7EDgAwBbNU7MyxAwDgQ6jYAQDmwFA8AAC+xVuG093BUDwAAD6Eih0AYA6GUbm5098LkNgBAKbAqngAAOB1qNgBAObAqngAAHyHxVa5udPfG5DYfUxIcLlSh2Xqkp571DC8RNk7I/XK3Iv18/YmkqRLeu7WXy7/We3a/qrwsDLdds9ftGNXZJXjdGx/SCP+vkkd2h2W1WbRjl2NNPHxgSorq/yVSWj9q0b9Y6PaJxyWzWbR2q/jNSv9IpWUBNiP0b7tYY28eaPatf1VhmFRVnZjzZ7fXTt2nzpf3967NOzaH9U8tlAFhcH68JNEvfNB5xr+V4K38vOz6ea//6hLB+xSo0Yl+jU/RJ9/3loLFnaWZJEk3TN+vS4buNOh34aMGP3rkQGSpKhmRfr7sM1K6ppnP8YXX7bSwkUXqKLCX5IUEGDVXWO/VUJCvlrGFeqbb5tryhN9q8TTtUue/jlqo1rGF+jwofr676LOWv55G5fiBTyNxO5jxt+xTq1aHtWzM/6kX/ND9Oe+O/TMo8s16u4h+jW/voKDK7R5WzOtWtdKE+5Yf9pjdGx/SE/963MtXNxZM+f0kNVqUZtWR2TYKv8QRTY6rqcfXa5V61rp5dk9VD+kXLff+p3uG/uVHn+uvyQpOLhcT036XOu/i9NLr/eUv7+h4Tdm6qlJn+um/7tOVqufLu62Xw+OW6OZc3oo4/tYtWxeoPG3r1dpWT19+EmH2vongxe5/rqtuurKbD0/rZd2745Qu3b5mnD31youDtQHSxPt+323IUYvTO9lf11e7m//uUVcoSwWacbLPXQgJ0yt4o9q3J3fKji4QrPnXChJ8vMzVFrqrw8+TNSfLtl72liiooo05bGV+ujjdnrmud5KTsrV3Xd9o/z8YGVsjHUpXtQShuJr3urVqzV16lRlZGQoJydHixcv1tChQ+syJK8WGFihPr326NGnB+jHn6IkSW++naxeF+3T1YOylP7fblqxqq0kKapp0RmPc9uI77Tk4w5atLiLvW3fgQj7z70u2ier1U8vv95ThlGZ7F98tZdem7ZUsdGFOpAbrrjmBQoPK9P8hck69GuD32JJ0mvTliqqaZEO5Ibrz/22a923LfXRZ5V/4HLzwrTw/c66cehmffhJoqho8EedOh7S198017ffNZck5R0MVf9+u5WY+Ku09NR+5eX+OnIk5LTHyMiIVUZGrP11bm6oWjQv1FVX/WJP7KWl9fTyv3tIki7odEgNGpRXOc5VV/6i3NxQvf5bn717I3RBp0P669Ase2J3Nl7UDlbF14Li4mIlJSVp5syZdRmGz/D3M+Tvb6jsd9WJJJWW+euCDgedOkbD8BPq2P6wjhYEa9qTn2jRnLf13JRPdUGHPPs+AfWsqqjwsyd1SSorqzznBR0rz7Nvf4QKCoN0xZ9/Ub16VgUGVuiKP2dr994I5R4MrTxOgO00sdZT0ybHFdW02PV/APi8n7Y2VXJSnprHFkqSWrc+ogs6HdJ3G2Ic9uvaJU8L33pPs19dqrF3fKuwsNKzHrdBg3IdOxbkUiwdOxzWpsxoh7aMjTHq2OGwy/Gilpy8jt2dzQvUacU+ePBgDR48uC5D8CknSgK0ZVtT3XTdD9qzL0JHC4I14E+71LH9YR3IDXPqGNFRlZX8P278Xq+9cZG272qky/rt0DOPLdc/x1+jAznhytwco/+7ZYOuH7JZiz/qqOCgCo28eaMkqXHDE/ZY7nvkcj36wEr9/bofJUkHcsM08fGBstkqv09mZMbqtls26LMuOfp+c7Rio4/pumt+kiRFNjqhvEOhHv33gfd7+51Oql+/XK+/ukw2m0V+fobemJ+kL1e2tu+zISNGX62LU25uqGJijumW1O/1xOQvNf7ey+2/e78XE3NM11z9s16f082lWBo1KtHRo8EObUePBqtBg3IFBlaorKyeU/ECnuZVc+ylpaUqLT31zbuwsLAOozk/PTvjT7pnzDotnP2urFaLftkRqZVrW6ld23yn+vv5VX4j/eiz9vrsywRJ0vadjZXcNUdXXJqtuW9dqN17G2rqS5fo/27ZoFtv2iSrzaIPPu6g/CPBsv1WxQcGVmjCHev107amSpvWR35+hq4fskVPPPSFxj5wpcrK6unj5e0UE3VMj0/8QvXq2VR8PEBLPuqo4X/73lu+GKOW9e2zW5f236VnpvbW7t0N1bbNEf3fPzMqF6WtqFy0tmp1K/v+u3Y31M5djZQ+50N17XJQmd87VtiNGx/Xk1O+1Jq1LfW/TxPqJF7UHrMMxXtVYk9LS9PkyZPrOozzWk5emO59ZJCCg8pVP6Rc+Ufr66EJq5ST51z1m//bvOSefQ0d2vfsi1CzJqeGx79c20Zfrm2jhhEnVFJaTzKka/+y1X6eS/vsVFSzIo17aLB9yD5teh+9/8Yi9b54r1Z+1VqSRXP+013zFnRTo4YlKigMUrcuufbPAfzRqFsz9fY7nezJe9fuhmrWrFg3Xv/TGRNlbm6ojhYEKTbmmENij4w8rmfSVuinrU304ks9XI7lyJFgNWxY4tDWsGGJiosD7FePVCde1CCTLJ7zqjvPTZw4UQUFBfZt797Tr1aFVFIaoPyj9RXaoFQXJR/Q+u/inOqXezBUh38NUYvYAof2FjGFyjvUoMr+RwtCVFISoH6X7FJ5ub82fl+5aCgosEI2w+JQedtsla8tf/jaa7P56df8+qqo8Ff/P+3Ulm1NVVDoOMQJSFJQUIV9VOgkm80ii9+Z/+I2aXxc4WGl9i+tUmWl/uzTK5SdHakXpvdyWC/irK3bmig5Odeh7cJuudq6rYlb8QLu8qqKPSgoSEFBri1wMZvuyftlkbTvQLhio49p9PAM7d0foU+/qBxmDAstVdMmxWoceVySFPdbAj9yNERHjoZIsuidDy7Q8Bu/145dkZVz7P23K655of1SNkm6ZvA2/bStqU6UBOjCpAMaPTxDc/9zoYqPB0qSNv4Qq9HDM3Tn6G+05OMO8vOTbvzrZlltFn2/ubJqCg8rUZ+U3fphS7QCAqwadGm2+qbs1r2PDKq1fy94l2++ba6/3bhZhw7V1+7dEWrb9oj++tdt+mx5ZfUbHFyum/++WWu/itORI8GKiSnSyFs36UBOmDIyKhesNW58XM+mfa6Dhxro9TndFBFxanrv9yvpW8YVqF6ATWFhZQoJKVebNkckSTt2NJIkffRxO13zl581csQmfbq8jZKT8tS3zx498lg/p+NF7TLLULzFMM6P2UyLxeLy5W6FhYWKiIhQvx4Pq149Kjyp8oYvt960UU0aH9exoiCt/bql5i3opuO/JdzLBmTrvrHrqvR7c1FXvfl2sv31jX/9UddckaWw0DJt39VIs9+8UFu2Rdnfv+/OterZfZ+Cgyu0d3+E3v2wk/1SupMu7HpAN9/wvVq1PCqbzaLtOyM1b0E3bfulqaTKxD5l4hdq3fKoLBbpp5+bKP1376OSX0lFXYdw3ggJKdfwm39Q79571TCiVL/mh2jVqni99d/OqqjwV2BghR7912q1bXtEDRqUKz8/RBmbojX/za46erQyaV82cIfuGf/1aY9/xVV/t//8xtwPFBVV9eqM3+/TtUue/jl6o1q2LNDhw/X134WON6g5V7yQKqyl+vL7p1VQUKDw8PAaOcfJXNHryimqF1D9XFFRXqKvP36kRmP1hDpN7EVFRcrOzpYkdevWTS+88IIGDBigyMhItWzZ8pz9SewwAxI7fBmJ3fPqdCh+w4YNGjBggP31hAkTJEmpqalKT0+vo6gAAL7ILEPxdZrY+/fvr/NkJgAA4OtYFQ8AALyNV62KBwCguhiKBwDAl9iMys2d/l6AxA4AMAfm2AEAgLehYgcAmIJFbs6xeyySmkXFDgAwh1p+Hvvq1at19dVXKzY2VhaLRUuWLDlnn5UrV+rCCy9UUFCQEhISqnVPFxI7AAA1oLi4WElJSZo5c6ZT++/cuVNXXXWVBgwYoMzMTN19990aNWqUPv30U5fOy1A8AMAUavtyt8GDB2vw4MFO7z9r1iy1bt1azz//vCSpY8eOWrt2raZNm6ZBg5x/OBYVOwDAHAwPbKq89/zvt9LSUnnC+vXrNXDgQIe2QYMGaf369S4dh8QOAIAL4uLiFBERYd/S0tI8ctzc3FxFRUU5tEVFRamwsFAnTpxw+jgMxQMATMFiGLK48XySk3337t3r8HS3oKAgt2PzJBI7AMAcbL9t7vSXFB4eXiOPbY2OjlZeXp5DW15ensLDwxUSEuL0cRiKBwDgPJCSkqIVK1Y4tC1fvlwpKSkuHYfEDgAwhZND8e5srigqKlJmZqYyMzMlVV7OlpmZqT179kiSJk6cqOHDh9v3v+2227Rjxw7df//92rZtm/7973/r7bff1vjx4106L0PxAABzqOV7xW/YsEEDBgywv54wYYIkKTU1Venp6crJybEneUlq3bq1PvroI40fP14vvviiWrRoodmzZ7t0qZtEYgcAmEU17h5Xpb8L+vfvL+MsfU53V7n+/ftr06ZNrkbmgKF4AAB8CBU7AMAUavvOc3WFxA4AMIdaHoqvKwzFAwDgQ6jYAQCmYLFVbu709wYkdgCAOTAUDwAAvA0VOwDAHGr5BjV1hcQOADAFTz3d7XzHUDwAAD6Eih0AYA4mWTxHYgcAmIMh957H7h15ncQOADAH5tgBAIDXoWIHAJiDITfn2D0WSY0isQMAzMEki+cYigcAwIdQsQMAzMEmyeJmfy9AYgcAmAKr4gEAgNehYgcAmINJFs+R2AEA5mCSxM5QPAAAPoSKHQBgDiap2EnsAABz4HI3AAB8B5e7AQAAr0PFDgAwB+bYAQDwITZDsriRnG3ekdgZigcAwIdQsQMAzIGheAAAfImbiV3ekdgZigcAwIdQsQMAzIGheAAAfIjNkFvD6ayKBwAAtY2KHQBgDoatcnOnvxcgsQMAzIE5dgAAfAhz7AAAwNtQsQMAzIGheAAAfIghNxO7xyKpUQzFAwDgQ6jYAQDmwFA8AAA+xGaT5Ma16DbvuI6doXgAAHwIFTsAwBwYigcAwIeYJLEzFA8AgA+hYgcAmINJbilLYgcAmIJh2GS48YQ2d/rWJhI7AMAcDMO9qps5dgAAUNuo2AEA5mC4OcfuJRU7iR0AYA42m2RxY57cS+bYGYoHAMCHULEDAMyBoXgAAHyHYbPJcGMo3lsud2MoHgAAH0LFDgAwB5MMxVOxAwDMwWa4v1XDzJkz1apVKwUHB6tnz5769ttvz7hvenq6LBaLwxYcHOzS+UjsAADUkEWLFmnChAl69NFHtXHjRiUlJWnQoEE6ePDgGfuEh4crJyfHvu3evdulc5LYAQDmYBiV16JXe3O9Yn/hhRc0evRojRgxQp06ddKsWbNUv359zZ0794x9LBaLoqOj7VtUVJRL5ySxAwBMwbAZbm+uKCsrU0ZGhgYOHGhv8/Pz08CBA7V+/foz9isqKlJ8fLzi4uI0ZMgQbdmyxaXzktgBAObgVrVus995rrCw0GErLS097ekOHz4sq9VapeKOiopSbm7uafskJiZq7ty5+uCDD/Sf//xHNptNvXv31r59+5z+mCR2AABcEBcXp4iICPuWlpbmsWOnpKRo+PDhSk5OVr9+/fT++++radOmevXVV50+Bpe7AQBMwbAZMizVv2TN+G2Ofe/evQoPD7e3BwUFnXb/Jk2ayN/fX3l5eQ7teXl5io6OduqcAQEB6tatm7Kzs52Ok4odAGAOHhqKDw8Pd9jOlNgDAwPVvXt3rVixwt5ms9m0YsUKpaSkOBWy1WrVjz/+qJiYGKc/pldX7Ce/PVVYTz+/AfgCP6u1rkMAaszJv99GLdz8pULlbt2fpkLlLveZMGGCUlNTddFFF6lHjx6aPn26iouLNWLECEnS8OHD1bx5c/tw/pQpU9SrVy8lJCTo6NGjmjp1qnbv3q1Ro0Y5fU6vTuzHjh2TJH2V8VwdRwIAcMexY8cUERFRI8cODAxUdHS01uZ+7PaxoqOjFRgY6PT+N954ow4dOqRHHnlEubm5Sk5O1v/+9z/7gro9e/bIz+/U4PmRI0c0evRo5ebmqlGjRurevbvWrVunTp06OX1Oi1EbX5NqiM1m04EDBxQWFiaLxVLX4ZhCYWGh4uLiqswxAb6A3+/aZxiGjh07ptjYWIcE52klJSUqKytz+ziBgYEu3wmutnl1xe7n56cWLVrUdRimdHJuCfBF/H7Xrpqq1H8vODj4vE/InsLiOQAAfAiJHQAAH0Jih0uCgoL06KOPnvHyDsCb8fsNX+DVi+cAAIAjKnYAAHwIiR0AAB9CYgcAwIeQ2OG0mTNnqlWrVgoODlbPnj317bff1nVIgEesXr1aV199tWJjY2WxWLRkyZK6DgmoNhI7nLJo0SJNmDBBjz76qDZu3KikpCQNGjRIBw8erOvQALcVFxcrKSlJM2fOrOtQALexKh5O6dmzpy6++GK9/PLLkipv5xsXF6c777xTDz74YB1HB3iOxWLR4sWLNXTo0LoOBagWKnacU1lZmTIyMjRw4EB7m5+fnwYOHKj169fXYWQAgD8iseOcDh8+LKvVan8a0UlRUVHKzc2to6gAAKdDYgcAwIeQ2HFOTZo0kb+/v/Ly8hza8/LyFB0dXUdRAQBOh8SOcwoMDFT37t21YsUKe5vNZtOKFSuUkpJSh5EBAP7Iq5/HjtozYcIEpaam6qKLLlKPHj00ffp0FRcXa8SIEXUdGuC2oqIiZWdn21/v3LlTmZmZioyMVMuWLeswMsB1XO4Gp7388suaOnWqcnNzlZycrBkzZqhnz551HRbgtpUrV2rAgAFV2lNTU5Wenl77AQFuILEDAOBDmGMHAMCHkNgBAPAhJHYAAHwIiR0AAB9CYgcAwIeQ2AEA8CEkdgAAfAiJHQAAH0JiB9x0yy23aOjQofbX/fv31913313rcaxcuVIWi0VHjx494z4Wi0VLlixx+piPPfaYkpOT3Ypr165dslgsyszMdOs4AJxDYodPuuWWW2SxWGSxWBQYGKiEhARNmTJFFRUVNX7u999/X48//rhT+zqTjAHAFTwEBj7riiuu0Lx581RaWqqPP/5YY8aMUUBAgCZOnFhl37KyMgUGBnrkvJGRkR45DgBUBxU7fFZQUJCio6MVHx+v22+/XQMHDtSHH34o6dTw+ZNPPqnY2FglJiZKkvbu3asbbrhBDRs2VGRkpIYMGaJdu3bZj2m1WjVhwgQ1bNhQjRs31v33368/Pm7hj0PxpaWleuCBBxQXF6egoCAlJCRozpw52rVrl/3BI40aNZLFYtEtt9wiqfKxuGlpaWrdurVCQkKUlJSkd9991+E8H3/8sdq3b6+QkBANGDDAIU5nPfDAA2rfvr3q16+vNm3aaNKkSSovL6+y36uvvqq4uDjVr19fN9xwgwoKChzenz17tjp27Kjg4GB16NBB//73v12OBYBnkNhhGiEhISorK7O/XrFihbKysrR8+XItW7ZM5eXlGjRokMLCwrRmzRp99dVXCg0N1RVXXGHv9/zzzys9PV1z587V2rVrlZ+fr8WLF5/1vMOHD9d///tfzZgxQ1u3btWrr76q0NBQxcXF6b333pMkZWVlKScnRy+++KIkKS0tTfPnz9esWbO0ZcsWjR8/XjfffLNWrVolqfILyLXXXqurr75amZmZGjVqlB588EGX/03CwsKUnp6un376SS+++KJef/11TZs2zWGf7Oxsvf3221q6dKn+97//adOmTbrjjjvs77/11lt65JFH9OSTT2rr1q166qmnNGnSJL3xxhsuxwPAAwzAB6WmphpDhgwxDMMwbDabsXz5ciMoKMi499577e9HRUUZpaWl9j5vvvmmkZiYaNhsNntbaWmpERISYnz66aeGYRhGTEyM8eyzz9rfLy8vN1q0aGE/l2EYRr9+/Yxx48YZhmEYWVlZhiRj+fLlp43zyy+/NCQZR44csbeVlJQY9evXN9atW+ew78iRI41hw4YZhmEYEydONDp16uTw/gMPPFDlWH8kyVi8ePEZ3586darRvXt3++tHH33U8Pf3N/bt22dv++STTww/Pz8jJyfHMAzDaNu2rbFgwQKH4zz++ONGSkqKYRiGsXPnTkOSsWnTpjOeF4DnMMcOn7Vs2TKFhoaqvLxcNptNf//73/XYY4/Z3+/SpYvDvPr333+v7OxshYWFORynpKRE27dvV0FBgXJychyeQV+vXj1ddNFFVYbjT8rMzJS/v7/69evndNzZ2dk6fvy4LrvsMof2srIydevWTZK0detWhzgkKSUlxelznLRo0SLNmDFD27dvV1FRkSoqKhQeHu6wT8uWLdW8eXOH89hsNmVlZSksLEzbt2/XyJEjNXr0aPs+FRUVioiIcDkeAO4jscNnDRgwQK+88ooCAwMVGxurevUcf90bNGjg8LqoqEjdu3fXW2+9VeVYTZs2rVYMISEhLvcpKiqSJH300UcOCVWqXDfgKevXr9dNN92kyZMna9CgQYqIiNDChQv1/PPPuxzr66+/XuWLhr+/v8diBeA8Ejt8VoMGDZSQkOD0/hdeeKEWLVqkZs2aValaT4qJidE333yjvn37SqqsTDMyMnThhReedv8uXbrIZrNp1apVGjhwYJX3T44YWK1We1unTp0UFBSkPXv2nLHS79ixo30h4Elff/31uT/k76xbt07x8fF6+OGH7W27d++ust+ePXt04MABxcbG2s/j5+enxMRERUVFKTY2Vjt27NBNN93k0vkB1AwWzwG/uemmm9SkSRMNGTJEa9as0c6dO7Vy5Urddddd2rdvnyRp3Lhxevrpp7VkyRJt27ZNd9xxx1mvQW/VqpVSU1N16623asmSJfZjvv3225Kk+Ph4WSwWLVu2TIcOHVJRUZHCwsJ07733avz48XrjjTe0fft2bdy4US+99JJ9Qdptt92mX375Rffdd5+ysrK0YMECpaenu/R527Vrpz179mjhwoXavn27ZsyYcdqFgMHBwUpNTdX333+vNWvW6K677tINN9yg6OhoSdLkyZOVlpamGTNm6Oeff9aPP/6oefPm6YUXXnApHgCeQWIHflO/fn2tXr1aLVu21LXXXquOHTtq5MiRKikpsVfw99xzj/7xj38oNTVVKSkpCgsL01//+tezHveVV17RddddpzvuuEMdOnTQ6NGjVVxcLElq3ry5Jk+erAcffFBRUVEaO3asJOnxxx/XpEmTlJaWpo4dO+qKK67QRx99pNatW0uqnPd+7733tGTJEiUlJWnWrFl66qmnXPq811xzjcaPH6+xY8cqOTlZ69at06RJk6rsl5CQoGuvvVZXXnmlLr/8cnXt2tXhcrZRo0Zp9uzZmjdvnrp06aJ+/fopPT3dHiuA2mUxzrTqBwAAeB0qdgAAfAiJHQAAH0JiBwDAh5DYAQDwISR2AAB8CIkdAAAfQmIHAMCHkNgBAPAhJHYAAHwIiR0AAB9CYgcAwIeQ2AEA8CH/H5dQ5KY1L1bZAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "evaluate(train_negative, multi_task_model, \n",
    "         customer_numerical_features=['prev_items_tfidf'],\n",
    "         article_numerical_features=['tfidf'],\n",
    "         single_task=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "id": "4ae630dd",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "44/44 [==============================] - 0s 3ms/step\n",
      "MRR: 0.8142\n",
      "Distribution:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "count    42395.000000\n",
       "mean         0.814210\n",
       "std          0.278667\n",
       "min          0.040000\n",
       "25%          0.500000\n",
       "50%          1.000000\n",
       "75%          1.000000\n",
       "max          1.000000\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Baseline:\n",
      "0    0.666667\n",
      "1    0.333333\n",
      "Name: label, dtype: float64\n",
      "\n",
      "Model Accuracy: 0.73\n",
      "\n",
      "Confusion Matrix:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "evaluate(test_negative, multi_task_model, \n",
    "         customer_numerical_features=['prev_items_tfidf'],\n",
    "         article_numerical_features=['tfidf'],\n",
    "         single_task=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "oj6d21t3SyNr",
   "metadata": {
    "id": "oj6d21t3SyNr"
   },
   "source": [
    "# **Bandits for Recommendations:**"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "MAFpygeZumu1",
   "metadata": {
    "id": "MAFpygeZumu1"
   },
   "source": [
    "## **1. Introduction**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d21b257c-d1e8-42c8-8926-86b8ee5ff88b",
   "metadata": {
    "id": "d21b257c-d1e8-42c8-8926-86b8ee5ff88b"
   },
   "source": [
    "We can also use bandits to produce personalized recommendations. Compared to multi-task recommenders, bandits provide a balance between exploring new options and exploiting known information to find potential recommendations.\n",
    "  \n",
    "<br>The multi-armed bandit problem is an example of one-step reinforcement learning. To explore an example let's assume we have a gambler who is provided a slot machine with multiple arms, each with its own unknown probability distribution of payouts. The objective is to pull the arms one by one in a sequence while gathering information in order to maximize the total payout over the long run. The multi-armed bandit problem can be seen as a toy problem for reinforcement learning with one step rollout. Namely we have a game consisting of n rounds and in each round t:\n",
    "\n",
    "1. Player selects one of K actions (think of slot machines and pulling their arms, hence the name).\n",
    "2. Player gets reward of Rt . Each action i ∈ {1,2,…,K } has a fixed, but unknown to the player, reward distribution Pi with the expected reward μi .\n",
    "3. Given the history of actions and rewards, the Player updates their strategy.\n",
    "\n",
    "The goal of the player is to maximize the reward, which naturally has to be done by exploring new options (and thus learning about other machines' distributions) and exploiting known actions that proved to give high rewards so far. Mathematically the player wishes to minimize the regret.\n",
    "  \n",
    "<br>\n",
    "\n",
    "### About the Data: R6A - Yahoo! Front Page Today Module User Click Log Dataset, version 1.0 (1.1 GB)\n",
    "\n",
    "Our dataset contains a fraction of user click logs for news articles displayed in the Featured Tab of the Today Module on the [Yahoo! Front Page](http://www.yahoo.com) during the first ten days in May 2009. The articles were chosen uniformly at random from a hand-picked pool of high-quality articles, which allows one to use a recently developed method to obtain an unbiased evaluation of an arbitrary bandit algorithm.\n",
    "\n",
    "<br>The dataset contains 45,811,883 user visits to the Today Module.  For each visit, both the user and each of the candidate articles are\n",
    "associated with a feature vector of dimension 6 (including a constant feature), constructed by conjoint analysis with a bilinear model.\n",
    "\n",
    "This dataset contains 10 files, corresponding to the first 10 days in May 2009:  \n",
    "* ydata-fp-td-clicks-v1_0.20090501.gz\n",
    "* ydata-fp-td-clicks-v1_0.20090502.gz  \n",
    "...  \n",
    "* ydata-fp-td-clicks-v1_0.20090510.gz  \n",
    "\n",
    "Each line in the files corresponds to a separate user visit.  \n",
    "  \n",
    "<br>An example line is as follows:  \n",
    "1241160900 109513 0 |user 2:0.000012 3:0.000000 4:0.000006 5:0.000023 6:0.999958 1:1.000000 |109498 2:0.306008 3:0.000450 4:0.077048 5:0.230439 6:0.386055 1:1.000000 |109509 2:0.306008 3:0.000450 4:0.077048 5:0.230439 6:0.386055 1:1.000000 [[...more article features omitted...]] |109453 2:0.421669 3:0.000011 4:0.010902 5:0.309585 6:0.257833 1:1.000000  \n",
    "   \n",
    "<br>which contains the following fields delimited with spaces:  \n",
    "* `timestamp`: e.g., 1241160900\n",
    "* `displayed_article_id`: e.g., 109513\n",
    "* `user_click` (0 for no-click and 1 for click): e.g., 0\n",
    "* strings \"|user\" and \"|{article_id}\" indicate the start of user and article features\n",
    "* features are encoded as \"feature_id:feature_value\" pairs, and\n",
    "feature_id starts from 1\n",
    "  \n",
    "<br>The pool of available articles for recommendation for each user visit is the set of articles that appear in that line of data.  All user IDs (specifically, bcookies) are replaced by a common string 'user' so that no user information can be identified from this data.\n",
    "\n",
    "Regarding the 6 dimensional feature vector, Feature #1 is the constant (always 1) feature, and features #2-6 correspond to the 5 membership features constructed via conjoint analysis with a bilinear model.\n",
    "\n",
    "A unique property of this data set is that the displayed article is chosen uniformly at random from the candidate article pool.  Therefore, one can use an unbiased *offline* evaluation method to compare bandit algorithms in a reliable way.  Performance of some of the popular bandit algorithms can be found. We will cover the offline evaluation method in detail in week 4 of the course.\n",
    "  \n",
    "<br>The full dataset can be downloaded from [here](https://webscope.sandbox.yahoo.com/catalog.php?datatype=r&did=49) if you'd like, but we'll be working with a smaller sample in this course.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "mGMwRjXuDxm9",
   "metadata": {
    "id": "mGMwRjXuDxm9"
   },
   "source": [
    "## **2. Setup**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8d57d965-645f-4622-b721-7cd11500fc67",
   "metadata": {
    "id": "8d57d965-645f-4622-b721-7cd11500fc67"
   },
   "source": [
    "Now let's write a function to read in the data, and store articles / features / events in memory -- which we will treat as global variables accessible to all functions later on:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "id": "1qX4SLbYJlWm",
   "metadata": {
    "id": "1qX4SLbYJlWm"
   },
   "outputs": [],
   "source": [
    "# Imports\n",
    "import numpy as np\n",
    "import fileinput\n",
    "from IPython.display import clear_output\n",
    "import random\n",
    "import time\n",
    "\n",
    "\n",
    "# Download data\n",
    "!pip install -U --no-cache-dir gdown --pre\n",
    "!gdown --no-cookies https://drive.google.com/uc?id=1_jFxa2s7ONQ7nrhqFc1lARP6WChs2T4e\n",
    "\n",
    "clear_output()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "id": "a362ca4c-5f65-4f51-99c7-43b62fe6043f",
   "metadata": {
    "id": "a362ca4c-5f65-4f51-99c7-43b62fe6043f"
   },
   "outputs": [],
   "source": [
    "def read_data(filenames):\n",
    "    \"\"\"\n",
    "    Reads a stream of events from the list of given files.\n",
    "    \n",
    "    Parameters\n",
    "    ----------\n",
    "    filenames : list\n",
    "        List of filenames\n",
    "    \n",
    "    Stores\n",
    "    -------    \n",
    "    articles : [article_ids]\n",
    "    features : [[article_1_features] .. [article_n_features]]\n",
    "    events : [\n",
    "                 0 : displayed_article_index (relative to the pool),\n",
    "                 1 : user_click,\n",
    "                 2 : [user_features],\n",
    "                 3 : [pool_indexes]\n",
    "             ]\n",
    "    \"\"\"\n",
    "\n",
    "    global articles, features, events, n_arms, n_events\n",
    "    articles = []\n",
    "    features = []\n",
    "    events = []\n",
    "\n",
    "    skipped = 0\n",
    "\n",
    "    with fileinput.input(files=filenames) as f:\n",
    "        for line in f:\n",
    "            cols = line.split()\n",
    "            if (len(cols) - 10) % 7 != 0:\n",
    "                skipped += 1\n",
    "            else:\n",
    "                pool_idx = []\n",
    "                pool_ids = []\n",
    "                for i in range(10, len(cols) - 6, 7):\n",
    "                    id = cols[i][1:]\n",
    "                    if id not in articles:\n",
    "                        articles.append(id)\n",
    "                        features.append([float(x[2:]) for x in cols[i + 1: i + 7]])\n",
    "                    pool_idx.append(articles.index(id))\n",
    "                    pool_ids.append(id)\n",
    "\n",
    "                events.append(\n",
    "                    [\n",
    "                        pool_ids.index(cols[1]),\n",
    "                        int(cols[2]),\n",
    "                        [float(x[2:]) for x in cols[4:10]],\n",
    "                        pool_idx,\n",
    "                    ]\n",
    "                )\n",
    "    features = np.array(features)\n",
    "    n_arms = len(articles)\n",
    "    n_events = len(events)\n",
    "    print(n_events, \"events with\", n_arms, \"articles\")\n",
    "    if skipped != 0:\n",
    "        print(\"Skipped events:\", skipped)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "id": "07df3304-6701-47ff-9d0d-d5a59a5b8cb3",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "07df3304-6701-47ff-9d0d-d5a59a5b8cb3",
    "outputId": "1fbdf7bf-4b59-497d-bc70-259674cb1745"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100000 events with 21 articles\n"
     ]
    }
   ],
   "source": [
    "# Use the function to read the data in\n",
    "read_data((\"data1.txt\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "s_bLy59XFaz7",
   "metadata": {
    "id": "s_bLy59XFaz7"
   },
   "source": [
    "## **3. Implement ϵ-greedy Policy**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "95e2039e-43ba-4a09-bbc9-b2b12687e722",
   "metadata": {
    "id": "95e2039e-43ba-4a09-bbc9-b2b12687e722"
   },
   "source": [
    "Below we'll implement a simple ϵ-greedy method. Here ϵ defines the amount of exploration we wish to perform, and for the remaining 1-ϵ times the model prefers exploitation. By convention, \"epsilon\" represents the percentage of time/trials dedicated for exploration, and it is also typical to do random exploration. This introduces some form of stochasticity.\n",
    "\n",
    "Notice how the `choose_arm` method returns the best arm's index based on the ϵ-greedy policy, and the `update` method updates algorithm's parameters(matrices) for the selected arm. It looks at the selected arm, updates the number of times this arm was chosen, and then it updates the mean reward observed for the selected arm."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "id": "d74ce40f-1fda-4397-ae7d-21f6faeacad9",
   "metadata": {
    "id": "d74ce40f-1fda-4397-ae7d-21f6faeacad9"
   },
   "outputs": [],
   "source": [
    "class Egreedy:\n",
    "    \"\"\"\n",
    "    Epsilon greedy algorithm implementation\n",
    "    \"\"\"\n",
    "\n",
    "    def __init__(self, epsilon):\n",
    "        \"\"\"\n",
    "        Parameters\n",
    "        ----------\n",
    "        epsilon : number (Egreedy parameter, ideally between 0 and 1)\n",
    "        \"\"\"\n",
    "\n",
    "        self.e = round(epsilon, 1)  # epsilon parameter for Egreedy \n",
    "        self.algorithm = \"Egreedy (ε=\" + str(self.e) + \")\"\n",
    "        self.q = np.zeros(n_arms)  # average reward for each arm -- this represents the known mean reward for each arm\n",
    "        self.n = np.zeros(n_arms)  # number of times each arm was chosen\n",
    "\n",
    "    def choose_arm(self, t, user, pool_idx):\n",
    "        \"\"\"\n",
    "        Returns the best arm's index relative to the pool\n",
    "        Parameters\n",
    "        ----------\n",
    "        t : number (number of trial)\n",
    "        user : array (user features)\n",
    "        pool_idx : array of indexes (pool indexes for article identification)\n",
    "        \"\"\"\n",
    "\n",
    "        p = np.random.rand()\n",
    "        if p > self.e:\n",
    "            return np.argmax(self.q[pool_idx])\n",
    "        else:\n",
    "            return np.random.randint(low=0, high=len(pool_idx))\n",
    "\n",
    "    def update(self, displayed, reward, user, pool_idx):\n",
    "        \"\"\"\n",
    "        Updates algorithm's parameters(matrices)\n",
    "        Parameters\n",
    "        ----------\n",
    "        displayed : index (displayed article index relative to the pool)\n",
    "        reward : binary (user clicked or not)\n",
    "        user : array (user features)\n",
    "        pool_idx : array of indexes (pool indexes for article identification)\n",
    "        \"\"\"\n",
    "\n",
    "        a = pool_idx[displayed]\n",
    "        \n",
    "        # update counts pulled for chosen arm\n",
    "        self.n[a] += 1\n",
    "        \n",
    "        # update average/mean value/reward for chosen arm\n",
    "        self.q[a] += (reward - self.q[a]) / self.n[a]\n",
    "        \"\"\"\n",
    "        this can also be written as:\n",
    "        value = self.q[a]\n",
    "        new_value = ((self.n[a]-1)/float(self.n[a])) * value + (1 / float(self.n[a])) * reward\n",
    "        self.q[a] = new_value\n",
    "        \"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "NHsNS6B1JTrO",
   "metadata": {
    "id": "NHsNS6B1JTrO"
   },
   "source": [
    "## **4. Policy Evaluation**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "13ca41bc-b6b7-49c3-867b-50bc836ccae3",
   "metadata": {
    "id": "13ca41bc-b6b7-49c3-867b-50bc836ccae3"
   },
   "source": [
    "We base our evaluation on the seminal work on offline evaluation of bandits as presented in [WSDM 2011] [Unbiased Offline Evaluation of Contextual-bandit-based News Article Recommendation Algorithms](https://arxiv.org/pdf/1003.5956.pdf).\n",
    "  \n",
    "<br>Compared to machine learning in the more standard supervised learning setting, evaluation of methods in a bandit setting is frustratingly difficult. Our goal here is to measure the performance of a bandit algorithm A, that is, a rule for selecting an arm at each time step based on the preceding interactions.\n",
    "\n",
    "We suppose that there is some unknown distribution D from which tuples are drawn i.i.d. of the form (x, r1, . . . , rK), each consisting of unobserved payoffs for all arms. We also posit access to a long sequence of logged events resulting from the interaction of the uniformly random logging policy with the world. Each such event consists a selected arm a and the resulting observed payoff ra. Crucially, this logged data is partially labeled in the sense that only the payoff ra is observed for the single arm a that was chosen uniformly at random.\n",
    "\n",
    "Our goal is to use this data to evaluate a bandit algorithm A, which is a (possibly randomized) mapping for selecting the arm at time t based on the history ht−1 of t−1 preceding events. Therefore the data serves as a benchmark with which people can evaluate and compare different bandit algorithms.\n",
    "  \n",
    "The policy evaluator is shown in Algorithm 1 below. The method takes as input a bandit algorithm A and a desired\n",
    "number of \"valid\" events T on which to base the evaluation. We then step through the stream of logged events one by one. If, given the current history ht−1, it happens that the policy A chooses the same arm a as the one that was selected by the logging policy, then the event is retained (that is, added to the history), and the total payoff updated. Otherwise, if the policy A selects a different arm from the one that was taken by the logging policy, then the event is entirely ignored, and the algorithm proceeds to the next event without any change in its state.\n",
    "![Policy Evaluator](data:image/png;base64,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)\n",
    "  \n",
    "<br>We next implement this evaluation function below:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "id": "3eb3334e-253f-483c-9e39-4bdda2c3b00e",
   "metadata": {
    "id": "3eb3334e-253f-483c-9e39-4bdda2c3b00e"
   },
   "outputs": [],
   "source": [
    "def evaluate(A, size=100, learn_ratio = 0.9):\n",
    "    \"\"\"\n",
    "    Policy evaluator as described in the paper\n",
    "    Parameters\n",
    "    ----------\n",
    "    A : class (algorithm)\n",
    "    size : number (run the evaluation only on a portion of the dataset)\n",
    "    learn_ratio : number (perform learning(update parameters) only on a small portion of the traffic)\n",
    "    Returns\n",
    "    -------\n",
    "    learn : array (contains the ctr for each trial for the learning bucket)\n",
    "    deploy : array (contains the ctr for each trial for the deployment bucket)\n",
    "    \"\"\"\n",
    "    \n",
    "    start = time.time()\n",
    "    # we initialize the payoff and events parameters separately for learning phase of the events and deployment phase of events.\n",
    "    Payoff_deploy = 0 # total payoff for the deployment bucket\n",
    "    Payoff_learn = 0  # total payoff for the learning bucket\n",
    "    Events_deploy = 1 # counter of valid events for the deployment bucket\n",
    "    Events_learn = 0  # counter of valid events for the learning bucket\n",
    "\n",
    "    learn = []\n",
    "    deploy = []\n",
    "    global events\n",
    "    if size != 100:\n",
    "        k = int(n_events * size / 100)\n",
    "        events = random.sample(events, k)\n",
    "\n",
    "    \"\"\"\n",
    "    we run through the logged events, and treat each event either for learning & updating the parameters,\n",
    "    or for deployment purposes wherein we use the reward obtained as evaluation metric\n",
    "    \"\"\"\n",
    "    for t, event in enumerate(events):\n",
    "\n",
    "        displayed = event[0]\n",
    "        reward = event[1]\n",
    "        user = event[2]\n",
    "        pool_idx = event[3]\n",
    "\n",
    "        # select the arm based on the bandit policy\n",
    "        chosen = A.choose_arm(Payoff_learn + Payoff_deploy, user, pool_idx)\n",
    "        \n",
    "        \"\"\"\n",
    "        If, given the current history ht−1, it happens that the policy A chooses the same arm a\n",
    "        as the one that was selected by the logging policy, then the event is retained\n",
    "        (that is, added to the history), and the total payoff updated.\n",
    "        Otherwise, if the policy A selects a different arm from the one that was taken by the logging policy,\n",
    "        then the event is entirely ignored, and the algorithm proceeds to the next event without any change in its state.\n",
    "        \"\"\"\n",
    "        if chosen == displayed:\n",
    "            if random.random() < learn_ratio:\n",
    "                Payoff_learn += event[1]\n",
    "                Events_learn += 1\n",
    "                A.update(displayed, reward, user, pool_idx)\n",
    "                learn.append(Payoff_learn / Events_learn)\n",
    "            else:\n",
    "                Payoff_deploy += event[1]\n",
    "                Events_deploy += 1\n",
    "                deploy.append(Payoff_deploy / Events_deploy)\n",
    "\n",
    "    end = time.time()\n",
    "\n",
    "    execution_time = round(end - start, 1)\n",
    "    execution_time = (\n",
    "        str(round(execution_time / 60, 1)) + \"m\"\n",
    "        if execution_time > 60\n",
    "        else str(execution_time) + \"s\"\n",
    "    )\n",
    "    print(\n",
    "        \"{:<20}{:<10}{}\".format(\n",
    "            A.algorithm, round(Payoff_deploy / Events_deploy, 4), execution_time\n",
    "        )\n",
    "    )\n",
    "\n",
    "    return learn, deploy"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45aa4ece-69ca-4538-a88c-92e9b18a7510",
   "metadata": {
    "id": "45aa4ece-69ca-4538-a88c-92e9b18a7510"
   },
   "source": [
    "Now let's run the evaluation method on the epsilon-greedy policy for different values of epsilon, and print the corresponding reward obtained during deployment phase of evaluation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "id": "d226bd62-a6bd-4068-ab5d-4e2bfdf4bb9e",
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "d226bd62-a6bd-4068-ab5d-4e2bfdf4bb9e",
    "outputId": "4ca822ed-e696-42db-f711-dd7b533d2def"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Egreedy (ε=0.1)     0.0227    0.2s\n",
      "Egreedy (ε=0.1)     0.0247    0.2s\n",
      "Egreedy (ε=0.1)     0.0216    0.2s\n",
      "Egreedy (ε=0.2)     0.0263    0.2s\n",
      "Egreedy (ε=0.5)     0.0262    0.2s\n",
      "Egreedy (ε=0.8)     0.0267    0.2s\n"
     ]
    }
   ],
   "source": [
    "_, deploy = evaluate(Egreedy(0.1),learn_ratio=0.25)\n",
    "rnd_ctr = deploy[-1]\n",
    "\n",
    "_, deploy = evaluate(Egreedy(0.1),learn_ratio=0.5)\n",
    "rnd_ctr = deploy[-1]\n",
    "\n",
    "_, deploy = evaluate(Egreedy(0.1),learn_ratio=0.9)\n",
    "rnd_ctr = deploy[-1]\n",
    "\n",
    "_, deploy = evaluate(Egreedy(0.25),learn_ratio=0.5)\n",
    "rnd_ctr = deploy[-1]\n",
    "\n",
    "_, deploy = evaluate(Egreedy(0.5),learn_ratio=0.5)\n",
    "rnd_ctr = deploy[-1]\n",
    "\n",
    "_, deploy = evaluate(Egreedy(0.75),learn_ratio=0.5)\n",
    "rnd_ctr = deploy[-1]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aM2H5Jm4OqFb",
   "metadata": {
    "id": "aM2H5Jm4OqFb"
   },
   "source": [
    "## **5. TODO 2.3: Implement UCB1**\n",
    "[back to top](#scrollTo=wIHxLWre9THR&uniqifier=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b8bd37be-bf33-4613-b736-d674dc6d5889",
   "metadata": {
    "id": "b8bd37be-bf33-4613-b736-d674dc6d5889"
   },
   "source": [
    "Epsilon greedy performs pretty well, but it's easy to see how selecting arms at random can be inefficient. If you have one movie that 50% of users have liked, and another at 5% have liked, epsilon greedy is equally likely to pick either of these movies when exploring random arms. Upper Confidence Bound (UCB) algorithms were introduced as a class of bandit algorithm that explores more efficiently.\n",
    "\n",
    "UCB algorithms construct a confidence interval of what each arm's true performance might be, factoring in the uncertainty caused by variance in the data and the fact that we're only able to observe a limited sample of pulls for any given arm. The algorithms then optimistically assume that each arm will perform as well as its upper confidence bound, selecting the arm with the highest UCB.  \n",
    "  \n",
    "<br>\n",
    "\n",
    "### Motivation behind UCB\n",
    "Suppose an Artice A has been seen 100 times and has the best CTR. Article B has a slightly worse CTR than article A, but it hasn't been seen by as many users, so there's also more uncertainty about how well it's going to perform in the long run. For this reason, it has a larger confidence bound, giving it a slightly higher UCB score than article A. Article C was published just moments ago, so almost no users have seen it. We're extremely uncertain about how high its CTR will ultimately be, so its UCB is highest of all for now despite its initial CTR being low.\n",
    "\n",
    "Over time, more users will see articles B and C, and their confidence bounds will become more narrow and look more like that of article A. As we learn more about B and C, we'll shift from exploration toward exploitation as the articles' confidence intervals collapse toward their means. Unless the CTR of article B or C improves, the bandit will quickly start to favor article A again as the other articles' confidence bounds shrink.  \n",
    "  \n",
    "One of the most important features of the UCB is that it not only exponentially decays as the number of pulls on the given machine increases, but also increases as the timestep increases. In other words, arms that have been explored less are given a boost even if their estimated mean is low, especially if we’ve been playing for a while. In this way, the UCB1 algorithm is able to naturally define its own mix of exploration vs. exploitation without depending on a user supplied parameter like epsilon greedy.  \n",
    "  \n",
    "<br>\n",
    "\n",
    "### Mathematical Formulation\n",
    "Let $n_t(a)$ be the number of times arm a is selected in rounds 1,2,…,t and $\\mu_t(a)$ be the average reward of arm a up to time t. The upper confidence bound is defined as:\n",
    "\n",
    "$UCB_t(a) = \\mu_t(a)+ \\sqrt{\\frac{2 \\log(t)}{ N_k(t)}}$\n",
    "\n",
    "\n",
    "where $\\mu_t(a)$ can be computed as `rewards[k] / pulls[k]` and the bound can be computed as `sqrt((2 * log(t)) / pulls[k]`.\n",
    "\n",
    "The UCB1 algorithm chooses the best arm based on this optimistic estimate.\n",
    "  \n",
    "<br>Your goal is to finish the `choose_arm` and `update` methods of the `Ucb1` class below, and then evaluate the bandit:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "id": "f5881aa5-6396-469d-a582-95af04e2224a",
   "metadata": {
    "id": "f5881aa5-6396-469d-a582-95af04e2224a"
   },
   "outputs": [],
   "source": [
    "class Ucb1:\n",
    "    def __init__(self, alpha):\n",
    "        \"\"\"\n",
    "        Parameters\n",
    "        ----------\n",
    "        alpha : number (ucb parameter)\n",
    "        \"\"\"\n",
    "\n",
    "        self.alpha = round(alpha, 1)\n",
    "        self.algorithm = \"UCB1 (α=\" + str(self.alpha) + \")\"\n",
    "\n",
    "        self.q = np.zeros(n_arms)  # average reward for each arm\n",
    "        self.n = np.ones(n_arms)  # number of times each arm was chosen\n",
    "\n",
    "    def choose_arm(self, t, user, pool_idx):\n",
    "        \"\"\"\n",
    "        Returns the best arm's index relative to the pool\n",
    "        Parameters\n",
    "        ----------\n",
    "        t : number (number of trial)\n",
    "        user : array (user features)\n",
    "        pool_idx : array of indexes (pool indexes for article identification)\n",
    "        \"\"\"\n",
    "        \n",
    "        # write your code here\n",
    "        # return something here\n",
    "        ubc_term = self.alpha * np.sqrt(2 * np.log(t+1)/self.n[pool_idx])\n",
    "        ucb = self.q[pool_idx] + ubc_term\n",
    "        return np.argmax(ucb)\n",
    "\n",
    "    def update(self, displayed, reward, user, pool_idx):\n",
    "        \"\"\"\n",
    "        Updates algorithm's parameters(matrices)\n",
    "        Parameters\n",
    "        ----------\n",
    "        displayed : index (displayed article index relative to the pool)\n",
    "        reward : binary (user clicked or not)\n",
    "        user : array (user features)\n",
    "        pool_idx : array of indexes (pool indexes for article identification)\n",
    "        \"\"\"\n",
    "\n",
    "        a = pool_idx[displayed]\n",
    "        a = pool_idx[displayed]\n",
    "        \n",
    "        # update self.n[a] here\n",
    "        # update self.q[a] here\n",
    "        self.n[a] += 1\n",
    "        self.q[a] += (reward - self.q[a]) / self.n[a]\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "id": "AobbZYPvj29P",
   "metadata": {
    "id": "AobbZYPvj29P"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "UCB1 (α=0.1)        0.0302    0.6s\n"
     ]
    }
   ],
   "source": [
    "# Evaluate your implementation of UCB1\n",
    "_, deploy = evaluate(Ucb1(0.1))\n",
    "rnd_ctr = deploy[-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1a0afc01",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "collapsed_sections": [
    "NWpqDtwiKSyN",
    "ETXp_Ixa9xVH",
    "CcKGfQtEDte3",
    "8k4D0ZUPEXJ5",
    "ZOyl7LPbGeFQ",
    "ADo5s7YWHOoR",
    "MtAo2yv4H4jg",
    "NI-XuLBcJld0",
    "ibq1uMfTL2H9",
    "aMIjykLIMHZW",
    "09LGM73ePNZK",
    "NTgBnh2-PrNL",
    "oj6d21t3SyNr",
    "MAFpygeZumu1",
    "mGMwRjXuDxm9",
    "s_bLy59XFaz7",
    "NHsNS6B1JTrO",
    "aM2H5Jm4OqFb"
   ],
   "provenance": []
  },
  "environment": {
   "kernel": "python3",
   "name": "tf2-gpu.2-6.m89",
   "type": "gcloud",
   "uri": "gcr.io/deeplearning-platform-release/tf2-gpu.2-6:m89"
  },
  "gpuClass": "standard",
  "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": "3.9.15"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "02b0bbe556984c599ad1e9076b3a6fe8": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_2a353946d4874cc68c21d94b2dbc9c5f",
      "placeholder": "​",
      "style": "IPY_MODEL_1931a2f097194bb48c788ab2ecacceab",
      "value": " 266364/266364 [00:00&lt;00:00, 436200.15it/s]"
     }
    },
    "112f1c82867e4c819fd87e707f027a8c": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HBoxModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HBoxModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HBoxView",
      "box_style": "",
      "children": [
       "IPY_MODEL_87edeed6fe334333ae83fce4682c23da",
       "IPY_MODEL_968de4d2649b46d3ab2906ee4b5a51d9",
       "IPY_MODEL_02b0bbe556984c599ad1e9076b3a6fe8"
      ],
      "layout": "IPY_MODEL_2fed36e5d12f4bf58d391f3f9b41b25b"
     }
    },
    "1931a2f097194bb48c788ab2ecacceab": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "2a353946d4874cc68c21d94b2dbc9c5f": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "2c0812e3b51847e8b2cfbe926f59473f": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "2fed36e5d12f4bf58d391f3f9b41b25b": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "300449ac26444b3587071a0225b8fb5b": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_6dc688fdbe26455f9f1ab5419f8e062f",
      "placeholder": "​",
      "style": "IPY_MODEL_92b8c6b085eb41848275b673cb842b33",
      "value": " 1769097/1769097 [00:04&lt;00:00, 416129.74it/s]"
     }
    },
    "3499b6797ece417abf99198fda3fffcf": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "ProgressStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "ProgressStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "bar_color": null,
      "description_width": ""
     }
    },
    "3cc7cd57d0f24c069db2c96af4f384c7": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "3fdaac08daf847fcad219d7f6522021c": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_ad0a7b5d47da48f8af4e4587e8ce6dde",
      "placeholder": "​",
      "style": "IPY_MODEL_9a7f950631c744919489f5645752ea07",
      "value": "100%"
     }
    },
    "5778f8578b494d259037a3514b254735": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "6dc688fdbe26455f9f1ab5419f8e062f": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "8400ad754357400db79dddfaa07ba210": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "87edeed6fe334333ae83fce4682c23da": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_a5e83e39df5f4feb894b4eb88622cedd",
      "placeholder": "​",
      "style": "IPY_MODEL_8400ad754357400db79dddfaa07ba210",
      "value": "100%"
     }
    },
    "92b8c6b085eb41848275b673cb842b33": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "968de4d2649b46d3ab2906ee4b5a51d9": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "FloatProgressModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "FloatProgressModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "ProgressView",
      "bar_style": "success",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_5778f8578b494d259037a3514b254735",
      "max": 266364,
      "min": 0,
      "orientation": "horizontal",
      "style": "IPY_MODEL_3499b6797ece417abf99198fda3fffcf",
      "value": 266364
     }
    },
    "9a7f950631c744919489f5645752ea07": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "a5e83e39df5f4feb894b4eb88622cedd": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "a664a6b775b8401da8158a5c99cfca11": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "ProgressStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "ProgressStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "bar_color": null,
      "description_width": ""
     }
    },
    "ad0a7b5d47da48f8af4e4587e8ce6dde": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "c7aadd2d77644094961c0f426c8105b6": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "FloatProgressModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "FloatProgressModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "ProgressView",
      "bar_style": "success",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_3cc7cd57d0f24c069db2c96af4f384c7",
      "max": 1769097,
      "min": 0,
      "orientation": "horizontal",
      "style": "IPY_MODEL_a664a6b775b8401da8158a5c99cfca11",
      "value": 1769097
     }
    },
    "f96675512a1f4f47be2df206b1c4ada1": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HBoxModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HBoxModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HBoxView",
      "box_style": "",
      "children": [
       "IPY_MODEL_3fdaac08daf847fcad219d7f6522021c",
       "IPY_MODEL_c7aadd2d77644094961c0f426c8105b6",
       "IPY_MODEL_300449ac26444b3587071a0225b8fb5b"
      ],
      "layout": "IPY_MODEL_2c0812e3b51847e8b2cfbe926f59473f"
     }
    }
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}