travelix.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "headers loaded.. plotly plotting set..\n"
          ]
        }
      ],
      "source": [
        "# import libraries\n",
        "import os\n",
        "import json\n",
        "import re\n",
        "import sys\n",
        "import csv\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import seaborn as seab\n",
        "import pickle\n",
        "\n",
        "print('headers loaded.. plotly plotting set..')\n",
        "\n",
        "# Normalize function- to normalize values for review score and tip score//\n",
        "def normalize(x, new_min = 0, new_max = 100):\n",
        "    output = []\n",
        "    old_min, old_max = min(x), max(x)\n",
        "\n",
        "    for iter_val in x:\n",
        "        val = (new_max - new_min) / (old_max - old_min) * (iter_val - old_min) + new_min\n",
        "        output.append(val)\n",
        "\n",
        "    return output"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {},
      "outputs": [],
      "source": [
        "# Read dataframes for google colab\n",
        "# import json\n",
        "# from google.colab import drive\n",
        "\n",
        "# drive.mount('/content/drive')\n",
        "# reviews_path = \"/content/drive/MyDrive/Course Work/SEM2/Info Storage and Retrieval/project/yelp_dataset/yelp_academic_dataset_review.csv\"\n",
        "# business_path = \"/content/drive/MyDrive/Course Work/SEM2/Info Storage and Retrieval/project/yelp_dataset/yelp_academic_dataset_business.csv\"\n",
        "# user_path = \"/content/drive/MyDrive/Course Work/SEM2/Info Storage and Retrieval/project/yelp_dataset/yelp_academic_dataset_user.csv\"\n",
        "\n",
        "# reviews_df = pd.read_csv(reviews_path)\n",
        "# user_df = pd.read_csv(user_path)\n",
        "# business_df = pd.read_csv(business_path)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {},
      "outputs": [],
      "source": [
        "# Read dataframe for Local VS Code\n",
        "reviews_df = pd.read_csv('yelp_academic_dataset_review.csv')\n",
        "# user_df = pd.read_csv('yelp_academic_dataset_user.csv')\n",
        "business_df = pd.read_csv('yelp_academic_dataset_business.csv')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "['PA', 'FL', 'TN', 'IN', 'MO', 'LA', 'AZ', 'NJ', 'NV', 'AB']\n"
          ]
        }
      ],
      "source": [
        "# Get Top Ten States\n",
        "business_df = business_df.dropna(subset=['categories'])\n",
        "unique_states = business_df['state'].unique()\n",
        "state_map = dict()\n",
        "for s in unique_states:\n",
        "  state_map[s] = business_df[business_df['state'] == s].shape[0]\n",
        "# 'CA' 'MO' 'AZ' 'PA' 'TN' 'FL' 'IN' 'LA' 'AB' 'NV' 'ID' 'DE' 'IL' 'NJ' 'NC' 'CO' 'WA' 'HI' 'UT' 'TX' 'MT' 'MI' 'SD' 'XMS' 'MA' 'VI' 'VT'\n",
        "top_states = [state[0] for state in sorted(sorted(state_map.items(), key=lambda x: x[1], reverse=True), key=lambda x: x[1], reverse=True)[:10]]\n",
        "print(top_states)\n",
        "\n",
        "# create an empty dictionary to store the dataframes\n",
        "hotel_state_df_map = {}\n",
        "restaurent_state_df_map = {}\n",
        "nightlife_state_df_map = {}\n",
        "\n",
        "# Creating mask for Hotels & Travel\n",
        "hotel_mask = business_df['categories'].str.contains('Hotels & Travel')\n",
        "hotel_df = business_df[hotel_mask]\n",
        "\n",
        "# Creating mask for Restaurents\n",
        "restaurent_mask = business_df['categories'].str.contains('Restaurants')\n",
        "restaurent_df = business_df[restaurent_mask]\n",
        "\n",
        "# Creating mask for Nightlife\n",
        "nightlife_mask = business_df['categories'].str.contains('Nightlife')\n",
        "nightlife_df = business_df[nightlife_mask]\n",
        "\n",
        "for state in top_states:\n",
        "    df_name = f'business_df_{state}'\n",
        "\n",
        "    hotel_state_df = hotel_df[hotel_df['state'] == state]\n",
        "    restaurent_state_df = restaurent_df[restaurent_df['state'] == state]\n",
        "    nightlife_state_df = nightlife_df[nightlife_df['state'] == state]\n",
        "\n",
        "    exec(f\"{df_name} = hotel_state_df\")\n",
        "    # add the dataframe to the dictionary with the state abbreviation as the key\n",
        "    hotel_state_df_map[state] = hotel_state_df\n",
        "\n",
        "    exec(f\"{df_name} = restaurent_state_df\")\n",
        "    # add the dataframe to the dictionary with the state abbreviation as the key\n",
        "    restaurent_state_df_map[state] = restaurent_state_df\n",
        "\n",
        "    exec(f\"{df_name} = nightlife_state_df\")\n",
        "    # add the dataframe to the dictionary with the state abbreviation as the key\n",
        "    nightlife_state_df_map[state] = nightlife_state_df\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "# nightlife_state_df_map['LA']"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {},
      "outputs": [],
      "source": [
        "# Building Ratings Matrix\n",
        "from scipy.sparse import coo_matrix\n",
        "import numpy as np\n",
        "from collections import defaultdict\n",
        "\n",
        "class Recommendations:\n",
        "  class Business:\n",
        "    def __init__(self, name, address, city, state, postal_code, stars):\n",
        "      self.name = name\n",
        "      self.address = address\n",
        "      self.city = city\n",
        "      self.state = state\n",
        "      self.postal_code = postal_code\n",
        "      self.stars = stars\n",
        "  \n",
        "  def __init__(self, business_df, state_name, shorten=False):\n",
        "    print(f\"========Calculating For {state_name} State========\")\n",
        "    self.business_df = business_df\n",
        "    self.rating_mat = []\n",
        "    self.shorten = shorten if isinstance(shorten, bool) else False\n",
        "    self.user_num_to_user_hash_dict = dict()\n",
        "    self.user_hash_to_user_num_dict = dict()\n",
        "    self.business_num_to_business_hash_dict = dict()\n",
        "    self.business_hash_to_business_num_dict = dict()\n",
        "    self.business_recommendations = []\n",
        "    self.business_popularity = []\n",
        "    self.calculateRatingMatrix()\n",
        "    self.nonPersonalizedRecommendations()\n",
        "  \n",
        "  def calculateRatingMatrix(self):\n",
        "    print(\"Calculating rating matrix...\")\n",
        "    business_list = list(self.business_df['business_id'])\n",
        "    reviews_df_updated = reviews_df[reviews_df['business_id'].isin(business_list)]\n",
        "\n",
        "    if (self.shorten):\n",
        "          print(f\"Size Before Cutting Down: {reviews_df_updated.shape[0]}\")\n",
        "          user_counts = reviews_df_updated.groupby('user_id').size().reset_index(name='count')\n",
        "\n",
        "          # Sort the user_counts dataframe in descending order by count and select the top 100 user_ids\n",
        "          top_users = user_counts.sort_values(by='count', ascending=False).head(100)['user_id'].tolist()\n",
        "\n",
        "          # Filter the original dataframe to keep only the records that belong to the top 100 user_ids\n",
        "          reviews_df_updated = reviews_df_updated[reviews_df_updated['user_id'].isin(top_users)]\n",
        "          print(f\"Size After Cutting Down: {reviews_df_updated.shape[0]}\")\n",
        "\n",
        "\n",
        "    unique_business_id = reviews_df_updated['business_id'].unique()\n",
        "    unique_user_id = reviews_df_updated['user_id'].unique()\n",
        "\n",
        "    j = 0\n",
        "    for u in unique_user_id:\n",
        "        self.user_hash_to_user_num_dict[u] = j\n",
        "        self.user_num_to_user_hash_dict[j] = u\n",
        "        j += 1\n",
        "\n",
        "    j = 0\n",
        "    for i in unique_business_id:\n",
        "        self.business_hash_to_business_num_dict[i] = j\n",
        "        self.business_num_to_business_hash_dict[j] = i\n",
        "        j += 1\n",
        "\n",
        "    # Then, use the generated dictionaries to reindex UserID and MovieID in the data_df\n",
        "    user_list = reviews_df_updated['user_id'].values\n",
        "    movie_list = reviews_df_updated['business_id'].values\n",
        "    for j in range(len(reviews_df_updated)):\n",
        "        user_list[j] = self.user_hash_to_user_num_dict[user_list[j]]\n",
        "        movie_list[j] = self.business_hash_to_business_num_dict[movie_list[j]]\n",
        "    reviews_df_updated['user_id'] = user_list\n",
        "    reviews_df_updated['business_id'] = movie_list\n",
        "\n",
        "    num_user = len(reviews_df_updated['user_id'].unique())\n",
        "    num_movie = len(reviews_df_updated['business_id'].unique())\n",
        "\n",
        "    self.ratings_mat = coo_matrix((reviews_df_updated['stars'].values, (reviews_df_updated['user_id'].values, reviews_df_updated['business_id'].values)), shape=(num_user, num_movie)).astype(float).toarray()\n",
        "    print(f\"Size of Ratings Matrix: {self.ratings_mat.shape[0]}, {self.ratings_mat.shape[1]}\")\n",
        "  \n",
        "  def nonPersonalizedRecommendations(self):\n",
        "    print(\"Calculating NPR...\")\n",
        "    n = len(self.ratings_mat) # number of users\n",
        "    m = len(self.ratings_mat[0]) # number of movies\n",
        "\n",
        "    # Creating popularity array - size number of movies\n",
        "    self.business_popularity = np.zeros((m,))\n",
        "    self.business_popularity = self.ratings_mat.sum(axis=0) # claculating the popularity of each movie by summing the values in each column\n",
        "\n",
        "    self.business_recommendations = np.zeros((n, 50), dtype=np.int32)\n",
        "\n",
        "    for u in range(self.ratings_mat.shape[0]):\n",
        "      business_unvisited = np.where(self.ratings_mat[u] == 0)[0]\n",
        "      unwatched_popularity = self.business_popularity[business_unvisited]\n",
        "      # Sort the unwatched movies according to popularity and fetch top 50 to recommend\n",
        "      self.business_recommendations[u] = business_unvisited[np.argsort(unwatched_popularity)[::-1]][:50]\n",
        "\n",
        "    # print(\"Non personalized recommendations for first User:\")\n",
        "    # for i in range(5):\n",
        "    #   business_hash = self.getBusinessHashFromBusinessNum(self.business_recommendations[0,i])\n",
        "    #   business = self.getBusinessInfo(business_hash)\n",
        "    #   print(f\"Rank {i+1}: Business {self.business_recommendations[0,i]} - Name: {business.name} - state: {business.state} - stars: {business.stars}  - Popularity {business_popularity[self.business_recommendations[0,i]]}\")\n",
        "\n",
        "  def getNPRForuUser(self, user_num):\n",
        "    print(f\"Non personalized recommendations for User {user_num}:\")\n",
        "    for i in range(5):\n",
        "      business_hash = self.getBusinessHashFromBusinessNum(self.business_recommendations[0,i])\n",
        "      business = self.getBusinessInfo(business_hash)\n",
        "      print(f\"Rank {i+1}: Business {self.business_recommendations[0,i]} - Name: {business.name} - state: {business.state} - stars: {business.stars}  - Popularity {self.business_popularity[self.business_recommendations[0,i]]}\")\n",
        "\n",
        "\n",
        "  def getUserHashFromUserNum(self, user_num):\n",
        "    return self.user_num_to_user_hash_dict[user_num]\n",
        "\n",
        "  def getUserNumFromUserHash(self, user_hash):\n",
        "    return self.user_hash_to_user_num_dict[user_hash]\n",
        "\n",
        "  def getBusinessHashFromBusinessNum(self, business_num):\n",
        "    return self.business_num_to_business_hash_dict[business_num]\n",
        "\n",
        "  def getBusinessNumFromBusinessHash(self, business_hash):\n",
        "    return self.business_hash_to_business_num_dict[business_hash]\n",
        "  \n",
        "  def getBusinessInfo(self, business_hash):\n",
        "    bus_df = self.business_df[self.business_df['business_id'] == business_hash].iloc[0]\n",
        "    return self.Business(bus_df['name'], bus_df['address'], bus_df['city'], bus_df['state'], bus_df['postal_code'], bus_df['stars'])\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Calculating rating matrix...\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/var/folders/9r/y5cvt00j367c_9w3d96w_y800000gn/T/ipykernel_19310/2907829949.py:55: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  reviews_df_updated['user_id'] = user_list\n",
            "/var/folders/9r/y5cvt00j367c_9w3d96w_y800000gn/T/ipykernel_19310/2907829949.py:56: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  reviews_df_updated['business_id'] = movie_list\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Calculating NPR...\n"
          ]
        }
      ],
      "source": [
        "# PA_Hotel_Recommendation = Recommendations(hotel_state_df_map['PA'], 'PA')\n",
        "# FL_Hotel_Recommendation = Recommendations(hotel_state_df_map['FL'], 'FL')\n",
        "# TN_Hotel_Recommendation = Recommendations(hotel_state_df_map['TN'], 'TN')\n",
        "# IN_Hotel_Recommendation = Recommendations(hotel_state_df_map['IN'], 'IN')\n",
        "# MO_Hotel_Recommendation = Recommendations(hotel_state_df_map['MO'], 'MO')\n",
        "# LA_Hotel_Recommendation = Recommendations(hotel_state_df_map['LA'], 'LA')\n",
        "# AZ_Hotel_Recommendation = Recommendations(hotel_state_df_map['AZ'], 'AZ')\n",
        "# NJ_Hotel_Recommendation = Recommendations(hotel_state_df_map['NJ'], 'NJ')\n",
        "# NV_Hotel_Recommendation = Recommendations(hotel_state_df_map['NV'], 'NV')\n",
        "# AB_Hotel_Recommendation = Recommendations(hotel_state_df_map['AB'], 'AB')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {},
      "outputs": [],
      "source": [
        "import bz2file as bz2\n",
        "\n",
        "def compressed_pickle(title, data):\n",
        "    with bz2.BZ2File(title + \".pbz2\", \"w\") as f:\n",
        "        pickle.dump(data, f)\n",
        "\n",
        "def decompress_pickle(file):\n",
        "    data = bz2.BZ2File(file, \"rb\")\n",
        "    data = pickle.load(data)\n",
        "    return data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "# compressed_pickle(\"PA_Hotel_Recommendation\", PA_Hotel_Recommendation)\n",
        "# compressed_pickle(\"FL_Hotel_Recommendation\", FL_Hotel_Recommendation)\n",
        "# compressed_pickle(\"TN_Hotel_Recommendation\", TN_Hotel_Recommendation)\n",
        "# compressed_pickle(\"IN_Hotel_Recommendation\", IN_Hotel_Recommendation)\n",
        "# compressed_pickle(\"MO_Hotel_Recommendation\", MO_Hotel_Recommendation)\n",
        "# compressed_pickle(\"LA_Hotel_Recommendation\", LA_Hotel_Recommendation)\n",
        "# compressed_pickle(\"AZ_Hotel_Recommendation\", AZ_Hotel_Recommendation)\n",
        "# compressed_pickle(\"NJ_Hotel_Recommendation\", NJ_Hotel_Recommendation)\n",
        "# compressed_pickle(\"NV_Hotel_Recommendation\", NV_Hotel_Recommendation)\n",
        "# compressed_pickle(\"AB_Hotel_Recommendation\", AB_Hotel_Recommendation)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {},
      "outputs": [],
      "source": [
        "PA_Hotel_Recommendation = decompress_pickle(\"data/hotel/PA_Hotel_Recommendation.pbz2\")\n",
        "FL_Hotel_Recommendation = decompress_pickle(\"data/hotel/FL_Hotel_Recommendation.pbz2\")\n",
        "TN_Hotel_Recommendation = decompress_pickle(\"data/hotel/TN_Hotel_Recommendation.pbz2\")\n",
        "IN_Hotel_Recommendation = decompress_pickle(\"data/hotel/IN_Hotel_Recommendation.pbz2\")\n",
        "MO_Hotel_Recommendation = decompress_pickle(\"data/hotel/MO_Hotel_Recommendation.pbz2\")\n",
        "LA_Hotel_Recommendation = decompress_pickle(\"data/hotel/LA_Hotel_Recommendation.pbz2\")\n",
        "AZ_Hotel_Recommendation = decompress_pickle(\"data/hotel/AZ_Hotel_Recommendation.pbz2\")\n",
        "NJ_Hotel_Recommendation = decompress_pickle(\"data/hotel/NJ_Hotel_Recommendation.pbz2\")\n",
        "NV_Hotel_Recommendation = decompress_pickle(\"data/hotel/NV_Hotel_Recommendation.pbz2\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "========Calculating For PA State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 1100250\n",
            "Size After Cutting Down: 36442\n",
            "Size of Ratings Matrix: 100, 8206\n",
            "Calculating NPR...\n",
            "========Calculating For FL State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 792133\n",
            "Size After Cutting Down: 32619\n",
            "Size of Ratings Matrix: 100, 6373\n",
            "Calculating NPR...\n",
            "========Calculating For TN State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 434697\n",
            "Size After Cutting Down: 17425\n",
            "Size of Ratings Matrix: 100, 3274\n",
            "Calculating NPR...\n",
            "========Calculating For IN State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 335843\n",
            "Size After Cutting Down: 21611\n",
            "Size of Ratings Matrix: 100, 3238\n",
            "Calculating NPR...\n",
            "========Calculating For MO State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 354605\n",
            "Size After Cutting Down: 24585\n",
            "Size of Ratings Matrix: 100, 3255\n",
            "Calculating NPR...\n",
            "========Calculating For LA State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 558340\n",
            "Size After Cutting Down: 22143\n",
            "Size of Ratings Matrix: 100, 2948\n",
            "Calculating NPR...\n",
            "========Calculating For AZ State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 267036\n",
            "Size After Cutting Down: 16550\n",
            "Size of Ratings Matrix: 100, 2240\n",
            "Calculating NPR...\n",
            "========Calculating For NJ State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 171186\n",
            "Size After Cutting Down: 11793\n",
            "Size of Ratings Matrix: 100, 2512\n",
            "Calculating NPR...\n",
            "========Calculating For NV State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 243594\n",
            "Size After Cutting Down: 16982\n",
            "Size of Ratings Matrix: 100, 1519\n",
            "Calculating NPR...\n",
            "========Calculating For AB State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 68882\n",
            "Size After Cutting Down: 13310\n",
            "Size of Ratings Matrix: 100, 2190\n",
            "Calculating NPR...\n"
          ]
        }
      ],
      "source": [
        "# PA_Restaurent_Recommendation = Recommendations(restaurent_state_df_map['PA'], 'PA', True)\n",
        "# FL_Restaurent_Recommendation = Recommendations(restaurent_state_df_map['FL'], 'FL', True)\n",
        "# TN_Restaurent_Recommendation = Recommendations(restaurent_state_df_map['TN'], 'TN', True)\n",
        "# IN_Restaurent_Recommendation = Recommendations(restaurent_state_df_map['IN'], 'IN', True)\n",
        "# MO_Restaurent_Recommendation = Recommendations(restaurent_state_df_map['MO'], 'MO', True)\n",
        "# LA_Restaurent_Recommendation = Recommendations(restaurent_state_df_map['LA'], 'LA', True)\n",
        "# AZ_Restaurent_Recommendation = Recommendations(restaurent_state_df_map['AZ'], 'AZ', True)\n",
        "# NJ_Restaurent_Recommendation = Recommendations(restaurent_state_df_map['NJ'], 'NJ', True)\n",
        "# NV_Restaurent_Recommendation = Recommendations(restaurent_state_df_map['NV'], 'NV', True)\n",
        "# AB_Restaurent_Recommendation = Recommendations(restaurent_state_df_map['AB'], 'AB', True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {},
      "outputs": [],
      "source": [
        "# compressed_pickle(\"PA_Restaurent_Recommendation\", PA_Restaurent_Recommendation)\n",
        "# compressed_pickle(\"FL_Restaurent_Recommendation\", FL_Restaurent_Recommendation)\n",
        "# compressed_pickle(\"TN_Restaurent_Recommendation\", TN_Restaurent_Recommendation)\n",
        "# compressed_pickle(\"IN_Restaurent_Recommendation\", IN_Restaurent_Recommendation)\n",
        "# compressed_pickle(\"MO_Restaurent_Recommendation\", MO_Restaurent_Recommendation)\n",
        "# compressed_pickle(\"LA_Restaurent_Recommendation\", LA_Restaurent_Recommendation)\n",
        "# compressed_pickle(\"AZ_Restaurent_Recommendation\", AZ_Restaurent_Recommendation)\n",
        "# compressed_pickle(\"NJ_Restaurent_Recommendation\", NJ_Restaurent_Recommendation)\n",
        "# compressed_pickle(\"NV_Restaurent_Recommendation\", NV_Restaurent_Recommendation)\n",
        "# compressed_pickle(\"AB_Restaurent_Recommendation\", AB_Restaurent_Recommendation)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 57,
      "metadata": {},
      "outputs": [],
      "source": [
        "PA_Restaurent_Recommendation = decompress_pickle(\"data/restaurent/PA_Restaurent_Recommendation.pbz2\")\n",
        "FL_Restaurent_Recommendation = decompress_pickle(\"data/restaurent/FL_Restaurent_Recommendation.pbz2\")\n",
        "TN_Restaurent_Recommendation = decompress_pickle(\"data/restaurent/TN_Restaurent_Recommendation.pbz2\")\n",
        "IN_Restaurent_Recommendation = decompress_pickle(\"data/restaurent/IN_Restaurent_Recommendation.pbz2\")\n",
        "MO_Restaurent_Recommendation = decompress_pickle(\"data/restaurent/MO_Restaurent_Recommendation.pbz2\")\n",
        "LA_Restaurent_Recommendation = decompress_pickle(\"data/restaurent/LA_Restaurent_Recommendation.pbz2\")\n",
        "AZ_Restaurent_Recommendation = decompress_pickle(\"data/restaurent/AZ_Restaurent_Recommendation.pbz2\")\n",
        "NJ_Restaurent_Recommendation = decompress_pickle(\"data/restaurent/NJ_Restaurent_Recommendation.pbz2\")\n",
        "NV_Restaurent_Recommendation = decompress_pickle(\"data/restaurent/NV_Restaurent_Recommendation.pbz2\")\n",
        "AB_Restaurent_Recommendation = decompress_pickle(\"data/restaurent/AB_Restaurent_Recommendation.pbz2\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "========Calculating For PA State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 335684\n",
            "Size After Cutting Down: 12832\n",
            "Size of Ratings Matrix: 100, 1849\n",
            "Calculating NPR...\n",
            "========Calculating For FL State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 257816\n",
            "Size After Cutting Down: 12308\n",
            "Size of Ratings Matrix: 100, 1728\n",
            "Calculating NPR...\n",
            "========Calculating For TN State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 179480\n",
            "Size After Cutting Down: 7192\n",
            "Size of Ratings Matrix: 100, 912\n",
            "Calculating NPR...\n",
            "========Calculating For IN State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 120505\n",
            "Size After Cutting Down: 8892\n",
            "Size of Ratings Matrix: 100, 844\n",
            "Calculating NPR...\n",
            "========Calculating For MO State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 130275\n",
            "Size After Cutting Down: 10154\n",
            "Size of Ratings Matrix: 100, 1040\n",
            "Calculating NPR...\n",
            "========Calculating For LA State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 187057\n",
            "Size After Cutting Down: 8581\n",
            "Size of Ratings Matrix: 100, 967\n",
            "Calculating NPR...\n",
            "========Calculating For AZ State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 77897\n",
            "Size After Cutting Down: 5292\n",
            "Size of Ratings Matrix: 100, 472\n",
            "Calculating NPR...\n",
            "========Calculating For NJ State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 35190\n",
            "Size After Cutting Down: 2705\n",
            "Size of Ratings Matrix: 100, 357\n",
            "Calculating NPR...\n",
            "========Calculating For NV State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 67119\n",
            "Size After Cutting Down: 5202\n",
            "Size of Ratings Matrix: 100, 474\n",
            "Calculating NPR...\n",
            "========Calculating For AB State========\n",
            "Calculating rating matrix...\n",
            "Size Before Cutting Down: 18421\n",
            "Size After Cutting Down: 4124\n",
            "Size of Ratings Matrix: 100, 483\n",
            "Calculating NPR...\n"
          ]
        }
      ],
      "source": [
        "PA_Nightlife_Recommendation = Recommendations(nightlife_state_df_map['PA'], 'PA', True)\n",
        "FL_Nightlife_Recommendation = Recommendations(nightlife_state_df_map['FL'], 'FL', True)\n",
        "TN_Nightlife_Recommendation = Recommendations(nightlife_state_df_map['TN'], 'TN', True)\n",
        "IN_Nightlife_Recommendation = Recommendations(nightlife_state_df_map['IN'], 'IN', True)\n",
        "MO_Nightlife_Recommendation = Recommendations(nightlife_state_df_map['MO'], 'MO', True)\n",
        "LA_Nightlife_Recommendation = Recommendations(nightlife_state_df_map['LA'], 'LA', True)\n",
        "AZ_Nightlife_Recommendation = Recommendations(nightlife_state_df_map['AZ'], 'AZ', True)\n",
        "NJ_Nightlife_Recommendation = Recommendations(nightlife_state_df_map['NJ'], 'NJ', True)\n",
        "NV_Nightlife_Recommendation = Recommendations(nightlife_state_df_map['NV'], 'NV', True)\n",
        "AB_Nightlife_Recommendation = Recommendations(nightlife_state_df_map['AB'], 'AB', True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {},
      "outputs": [],
      "source": [
        "# compressed_pickle(\"data/nightlife/PA_Nightlife_Recommendation\", PA_Nightlife_Recommendation)\n",
        "# compressed_pickle(\"data/nightlife/FL_Nightlife_Recommendation\", FL_Nightlife_Recommendation)\n",
        "# compressed_pickle(\"data/nightlife/TN_Nightlife_Recommendation\", TN_Nightlife_Recommendation)\n",
        "# compressed_pickle(\"data/nightlife/IN_Nightlife_Recommendation\", IN_Nightlife_Recommendation)\n",
        "# compressed_pickle(\"data/nightlife/MO_Nightlife_Recommendation\", MO_Nightlife_Recommendation)\n",
        "# compressed_pickle(\"data/nightlife/LA_Nightlife_Recommendation\", LA_Nightlife_Recommendation)\n",
        "# compressed_pickle(\"data/nightlife/AZ_Nightlife_Recommendation\", AZ_Nightlife_Recommendation)\n",
        "# compressed_pickle(\"data/nightlife/NJ_Nightlife_Recommendation\", NJ_Nightlife_Recommendation)\n",
        "# compressed_pickle(\"data/nightlife/NV_Nightlife_Recommendation\", NV_Nightlife_Recommendation)\n",
        "# compressed_pickle(\"data/nightlife/AB_Nightlife_Recommendation\", AB_Nightlife_Recommendation)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {},
      "outputs": [],
      "source": [
        "PA_Nightlife_Recommendation = decompress_pickle(\"data/Nightlife/PA_Nightlife_Recommendation.pbz2\")\n",
        "FL_Nightlife_Recommendation = decompress_pickle(\"data/Nightlife/FL_Nightlife_Recommendation.pbz2\")\n",
        "TN_Nightlife_Recommendation = decompress_pickle(\"data/Nightlife/TN_Nightlife_Recommendation.pbz2\")\n",
        "IN_Nightlife_Recommendation = decompress_pickle(\"data/Nightlife/IN_Nightlife_Recommendation.pbz2\")\n",
        "MO_Nightlife_Recommendation = decompress_pickle(\"data/Nightlife/MO_Nightlife_Recommendation.pbz2\")\n",
        "LA_Nightlife_Recommendation = decompress_pickle(\"data/Nightlife/LA_Nightlife_Recommendation.pbz2\")\n",
        "AZ_Nightlife_Recommendation = decompress_pickle(\"data/Nightlife/AZ_Nightlife_Recommendation.pbz2\")\n",
        "NJ_Nightlife_Recommendation = decompress_pickle(\"data/Nightlife/NJ_Nightlife_Recommendation.pbz2\")\n",
        "NV_Nightlife_Recommendation = decompress_pickle(\"data/Nightlife/NV_Nightlife_Recommendation.pbz2\")\n",
        "AB_Nightlife_Recommendation = decompress_pickle(\"data/Nightlife/AB_Nightlife_Recommendation.pbz2\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {},
      "outputs": [],
      "source": [
        "class MF_implicit:\n",
        "    def __init__(self, train_mat, latent=5, lr=0.01, reg=0.01):\n",
        "        self.train_mat = train_mat  # the training rating matrix of size (#user, #movie)\n",
        "        \n",
        "        self.latent = latent  # the latent dimension\n",
        "        self.lr = lr  # learning rate\n",
        "        self.reg = reg  # regularization weight, i.e., the lambda in the objective function\n",
        "        \n",
        "        self.num_user, self.num_movie = train_mat.shape\n",
        "        \n",
        "        self.sample_user, self.sample_movie = self.train_mat.nonzero()  # get the user-movie paris having ratings in train_mat\n",
        "        self.num_sample = len(self.sample_user)  # the number of user-movie pairs having ratings in train_mat\n",
        "\n",
        "        # self.user_test_like = []\n",
        "        # for u in range(self.num_user):\n",
        "        #     self.user_test_like.append(np.where(self.test_mat[u, :] > 0)[0])\n",
        "\n",
        "        self.P = np.random.random((self.num_user, self.latent))  # latent factors for users, size (#user, self.latent), randomly initialized\n",
        "        self.Q = np.random.random((self.num_movie, self.latent))  # latent factors for users, size (#movie, self.latent), randomly initialized\n",
        "        \n",
        "    def negative_sampling(self):\n",
        "        negative_movie = np.random.choice(np.arange(self.num_movie), size=(len(self.sample_user)), replace=True)\n",
        "        true_negative = self.train_mat[self.sample_user, negative_movie] == 0\n",
        "        negative_user = self.sample_user[true_negative]\n",
        "        negative_movie = negative_movie[true_negative]\n",
        "        return np.concatenate([self.sample_user, negative_user]), np.concatenate([self.sample_movie, negative_movie])\n",
        "\n",
        "    def train(self, epoch=20):\n",
        "        \"\"\"\n",
        "        Goal: Write your code to train your matrix factorization model for epoch iterations in this function\n",
        "        Input: epoch -- the number of training epoch \n",
        "        \"\"\"\n",
        "        for ep in range(epoch):\n",
        "            \"\"\" \n",
        "            Write your code here to implement the training process for one epoch, \n",
        "            at the end of each epoch, run self.test() to evaluate current version of MF.\n",
        "            \"\"\"\n",
        "            # print(\"Epoch:\", ep+1)\n",
        "            s_user, s_movie_i = self.negative_sampling()\n",
        "            data = np.column_stack((s_user, s_movie_i))\n",
        "            np.random.shuffle(data)\n",
        "            for u, i  in zip(data[:, 0], data[:, 1]):\n",
        "              actual_rating = self.train_mat[u, i]\n",
        "              pu = self.P[u, :]\n",
        "              qi = self.Q[i, :]\n",
        "\n",
        "              predicted_rating = np.dot(pu, qi)\n",
        "              error = 2*(predicted_rating-actual_rating)\n",
        "\n",
        "              grad_Pu = error * qi + 2*self.reg * pu\n",
        "              grad_Qi = error * pu + 2*self.reg * qi\n",
        "\n",
        "              self.P[u, :] -= self.lr * grad_Pu\n",
        "              self.Q[i, :] -= self.lr * grad_Qi\n",
        "            self.predict()\n",
        "\n",
        "            \n",
        "    def predict(self):\n",
        "        \"\"\"\n",
        "        Write your code here to implement the prediction function, which generates the ranked lists of movies \n",
        "        by the trained MF for every user, store the result (named 'recommendation') in a numpy array of size (#user, 50), where entry (u, k) \n",
        "        represents the movie id that is ranked at position k in the recommendation list to user u. Return the 'recommendation' variable. \n",
        "        \"\"\"\n",
        "        prediction_mat = np.matmul(self.P, self.Q.T)\n",
        "        recommendation = []\n",
        "        for u in range(self.num_user):\n",
        "          scores = prediction_mat[u]\n",
        "          train_like = np.where(self.train_mat[u, :] > 0)[0]\n",
        "          scores[train_like] = -9999\n",
        "          top50_iid = np.argpartition(scores, -50)[-50:]\n",
        "          top50_iid = top50_iid[np.argsort(scores[top50_iid])[-1::-1]]\n",
        "          recommendation.append(top50_iid)\n",
        "        recommendation = np.array(recommendation)\n",
        "        return recommendation"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {},
      "outputs": [],
      "source": [
        "def calculateMF(ratings_mat):\n",
        "    mf_implicit = MF_implicit(ratings_mat, latent=5, lr=0.01, reg=0.0001)\n",
        "    mf_implicit.train(epoch=20)\n",
        "    recommendation = mf_implicit.predict()\n",
        "    return recommendation"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {},
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/var/folders/9r/y5cvt00j367c_9w3d96w_y800000gn/T/ipykernel_52000/3290410394.py:51: RuntimeWarning: overflow encountered in multiply\n",
            "  grad_Qi = error * pu + 2*self.reg * qi\n",
            "/var/folders/9r/y5cvt00j367c_9w3d96w_y800000gn/T/ipykernel_52000/3290410394.py:50: RuntimeWarning: overflow encountered in multiply\n",
            "  grad_Pu = error * qi + 2*self.reg * pu\n",
            "/var/folders/9r/y5cvt00j367c_9w3d96w_y800000gn/T/ipykernel_52000/3290410394.py:54: RuntimeWarning: invalid value encountered in subtract\n",
            "  self.Q[i, :] -= self.lr * grad_Qi\n",
            "/var/folders/9r/y5cvt00j367c_9w3d96w_y800000gn/T/ipykernel_52000/3290410394.py:53: RuntimeWarning: invalid value encountered in subtract\n",
            "  self.P[u, :] -= self.lr * grad_Pu\n"
          ]
        }
      ],
      "source": [
        "# PA_Hotel_MF = calculateMF(PA_Hotel_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"PA_Hotel_MF\", PA_Hotel_MF)\n",
        "# PA_Restaurent_MF = calculateMF(PA_Restaurent_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"PA_Restaurent_MF\", PA_Restaurent_MF)\n",
        "\n",
        "# FL_Hotel_MF = calculateMF(FL_Hotel_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"FL_Hotel_MF\", FL_Hotel_MF)\n",
        "# FL_Restaurent_MF = calculateMF(FL_Restaurent_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"FL_Restaurent_MF\", FL_Restaurent_MF)\n",
        "\n",
        "# TN_Hotel_MF = calculateMF(TN_Hotel_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"TN_Hotel_MF\", TN_Hotel_MF)\n",
        "# TN_Restaurent_MF = calculateMF(TN_Restaurent_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"TN_Restaurent_MF\", TN_Restaurent_MF)\n",
        "\n",
        "# IN_Hotel_MF = calculateMF(IN_Hotel_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"IN_Hotel_MF\", IN_Hotel_MF)\n",
        "# IN_Restaurent_MF = calculateMF(IN_Restaurent_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"IN_Restaurent_MF\", IN_Restaurent_MF)\n",
        "\n",
        "# MO_Hotel_MF = calculateMF(MO_Hotel_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"MO_Hotel_MF\", MO_Hotel_MF)\n",
        "# MO_Restaurent_MF = calculateMF(MO_Restaurent_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"MO_Restaurent_MF\", MO_Restaurent_MF)\n",
        "\n",
        "# LA_Hotel_MF = calculateMF(LA_Hotel_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"LA_Hotel_MF\", LA_Hotel_MF)\n",
        "# LA_Restaurent_MF = calculateMF(LA_Restaurent_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"LA_Restaurent_MF\", LA_Restaurent_MF)\n",
        "\n",
        "# AZ_Hotel_MF = calculateMF(AZ_Hotel_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"AZ_Hotel_MF\", AZ_Hotel_MF)\n",
        "# AZ_Restaurent_MF = calculateMF(AZ_Restaurent_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"AZ_Restaurent_MF\", AZ_Restaurent_MF)\n",
        "\n",
        "# NJ_Hotel_MF = calculateMF(NJ_Hotel_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"NJ_Hotel_MF\", NJ_Hotel_MF)\n",
        "# NJ_Restaurent_MF = calculateMF(NJ_Restaurent_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"NJ_Restaurent_MF\", NJ_Restaurent_MF)\n",
        "\n",
        "# NV_Hotel_MF = calculateMF(NV_Hotel_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"NV_Hotel_MF\", NV_Hotel_MF)\n",
        "# NV_Restaurent_MF = calculateMF(NV_Restaurent_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"NV_Restaurent_MF\", NV_Restaurent_MF)\n",
        "\n",
        "# AB_Hotel_MF = calculateMF(AB_Hotel_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"AB_Hotel_MF\", AB_Hotel_MF)\n",
        "# AB_Restaurent_MF = calculateMF(AB_Restaurent_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"AB_Restaurent_MF\", AB_Restaurent_MF)\n",
        "\n",
        "# PA_Nightlife_MF = calculateMF(PA_Nightlife_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"data/nightlife/PA_Nightlife_MF\", PA_Nightlife_MF)\n",
        "\n",
        "# FL_Nightlife_MF = calculateMF(FL_Nightlife_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"data/nightlife/FL_Nightlife_MF\", FL_Nightlife_MF)\n",
        "\n",
        "# TN_Nightlife_MF = calculateMF(TN_Nightlife_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"data/nightlife/TN_Nightlife_MF\", TN_Nightlife_MF)\n",
        "\n",
        "# IN_Nightlife_MF = calculateMF(IN_Nightlife_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"data/nightlife/IN_Nightlife_MF\", IN_Nightlife_MF)\n",
        "\n",
        "# MO_Nightlife_MF = calculateMF(MO_Nightlife_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"data/nightlife/MO_Nightlife_MF\", MO_Nightlife_MF)\n",
        "\n",
        "# LA_Nightlife_MF = calculateMF(LA_Nightlife_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"data/nightlife/LA_Nightlife_MF\", LA_Nightlife_MF)\n",
        "\n",
        "# AZ_Nightlife_MF = calculateMF(AZ_Nightlife_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"data/nightlife/AZ_Nightlife_MF\", AZ_Nightlife_MF)\n",
        "\n",
        "# NJ_Nightlife_MF = calculateMF(NJ_Nightlife_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"data/nightlife/NJ_Nightlife_MF\", NJ_Nightlife_MF)\n",
        "\n",
        "# NV_Nightlife_MF = calculateMF(NV_Nightlife_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"data/nightlife/NV_Nightlife_MF\", NV_Nightlife_MF)\n",
        "\n",
        "# AB_Nightlife_MF = calculateMF(AB_Nightlife_Recommendation.ratings_mat)\n",
        "# compressed_pickle(\"data/nightlife/AB_Nightlife_MF\", AB_Nightlife_MF)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "# Implementing Autoencoder for recommendation(basic, top-10 recommendation for each user)\n",
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "from scipy.sparse import csr_matrix\n",
        "\n",
        "def vanilla_autoencoder():\n",
        "  ratings_mat_sparse = csr_matrix(ratings_mat)\n",
        "\n",
        "  # load data\n",
        "  #ratings = np.loadtxt('ratings.csv', delimiter=',')\n",
        "  n_users, n_items = ratings_mat_sparse.shape\n",
        "\n",
        "  # split data into training and validation sets\n",
        "  split = int(0.8 * n_users)\n",
        "  train_ratings = ratings_mat_sparse[:split]\n",
        "  val_ratings = ratings_mat_sparse[split:]\n",
        "\n",
        "  # define the autoencoder model\n",
        "  input_layer = tf.keras.layers.Input(shape=(n_items,))\n",
        "  encoded_layer = tf.keras.layers.Dense(16, activation='relu')(input_layer)\n",
        "  decoded_layer = tf.keras.layers.Dense(n_items, activation='sigmoid')(encoded_layer)\n",
        "  autoencoder = tf.keras.models.Model(input_layer, decoded_layer)\n",
        "\n",
        "  # compile the model\n",
        "  autoencoder.compile(optimizer='adam', loss='binary_crossentropy')\n",
        "\n",
        "  # train the model\n",
        "  autoencoder.fit(train_ratings.toarray(), train_ratings.toarray(), epochs=20, batch_size=32, validation_data=(val_ratings, val_ratings))\n",
        "\n",
        "  # predict ratings for all users and items\n",
        "  predicted_ratings = autoencoder.predict(ratings_mat_sparse)\n",
        "\n",
        "  # print top 10 recommended items for each user\n",
        "  for i in range(n_users):\n",
        "      top_items = np.argsort(predicted_ratings[i])[::-1][:10]\n",
        "      print(f\"User {i+1}: {top_items}\")\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "from keras.layers import Input, Dense\n",
        "from keras.models import Model\n",
        "from scipy.sparse import csr_matrix\n",
        "\n",
        "def autoencoder_cf():\n",
        "  ratings_mat_sparse = csr_matrix(new_rat_mat)\n",
        "\n",
        "  # Split the data into training and testing sets\n",
        "  train_size = int(0.8 * ratings_mat_sparse.shape[0])\n",
        "  train_ratings = ratings_mat_sparse[:train_size, :]\n",
        "  test_ratings = ratings_mat_sparse[train_size:, :]\n",
        "\n",
        "  # Build the autoencoder model\n",
        "  input_layer = Input(shape=(ratings_mat_sparse.shape[1],))\n",
        "  encoded = Dense(16, activation='relu')(input_layer)\n",
        "  decoded = Dense(ratings_mat_sparse.shape[1], activation='sigmoid')(encoded)\n",
        "  autoencoder = Model(input_layer, decoded)\n",
        "\n",
        "  # Compile the model\n",
        "  autoencoder.compile(optimizer='adam', loss='binary_crossentropy')\n",
        "\n",
        "  # Train the model\n",
        "  autoencoder.fit(train_ratings.toarray(), train_ratings.toarray(), epochs=20, batch_size=32, shuffle=True, validation_data=(test_ratings.toarray(), test_ratings.toarray()))\n",
        "\n",
        "  # Extract the hidden layer output for all the users and items\n",
        "  hidden_layer = Model(input_layer, encoded)\n",
        "  batch_size = 32\n",
        "\n",
        "  def embeddings_generator(ratings_mat_sparse, batch_size):\n",
        "      num_users = ratings_mat_sparse.shape[0]\n",
        "      indices = np.arange(num_users)\n",
        "      for start_idx in range(0, num_users, batch_size):\n",
        "          end_idx = min(start_idx + batch_size, num_users)\n",
        "          batch_indices = indices[start_idx:end_idx]\n",
        "          batch_ratings = ratings_mat_sparse[batch_indices]\n",
        "          batch_embeddings = hidden_layer.predict(batch_ratings)\n",
        "          yield batch_embeddings\n",
        "\n",
        "  user_embeddings = np.concatenate(list(embeddings_generator(ratings_mat_sparse, batch_size)), axis=0)\n",
        "\n",
        "  # Compute the similarity between users and items\n",
        "  user_similarity = np.dot(user_embeddings, user_embeddings.T)\n",
        "  item_similarity = np.dot(user_embeddings.T, user_embeddings)\n",
        "\n",
        "  # Use the similarity scores to predict the ratings of the items for each user\n",
        "  user_ratings = np.dot(user_similarity, ratings_mat_sparse) / np.sum(np.abs(user_similarity), axis=1)\n",
        "  item_ratings = np.dot(item_similarity, ratings_mat_sparse.T) / np.sum(np.abs(item_similarity), axis=1)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "# Get predicted ratings for user \n",
        "def get_user_recommendation(user_id):\n",
        "  user_0_pred_ratings = user_ratings[user_id]\n",
        "  N = 10\n",
        "  # Sort predicted ratings in descending order\n",
        "  sorted_item_indices = np.argsort(user_0_pred_ratings[0].toarray().flatten())[::-1][:N]\n",
        "  print(\"Recommended items for user 0:\", sorted_item_indices)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 121,
      "metadata": {},
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "from keras.layers import Input, Dense, Dropout\n",
        "from keras.models import Model\n",
        "from scipy.sparse import csr_matrix\n",
        "from sklearn.metrics import mean_squared_error\n",
        "from math import sqrt\n",
        "from keras.callbacks import LearningRateScheduler\n",
        "\n",
        "class AE_CF:\n",
        "  def __init__(self, ratings_mat):\n",
        "    self.ratings_mat_sparse = ratings_mat\n",
        "    self.train_size = int(0.8 * self.ratings_mat_sparse.shape[0])\n",
        "    self.train_ratings = self.ratings_mat_sparse[:self.train_size, :]\n",
        "    self.test_ratings = self.ratings_mat_sparse[self.train_size:, :]\n",
        "    self.user_ratings = []\n",
        "    #self.item_ratings = []\n",
        "    self.train_cf()\n",
        "    self.rmse = self.calculate_rmse()\n",
        "\n",
        "  def embeddings_generator(self, hidden_layer, batch_size):\n",
        "    num_users = self.ratings_mat_sparse.shape[0]\n",
        "    indices = np.arange(num_users)\n",
        "    for start_idx in range(0, num_users, batch_size):\n",
        "      end_idx = min(start_idx + batch_size, num_users)\n",
        "      batch_indices = indices[start_idx:end_idx]\n",
        "      batch_ratings = self.ratings_mat_sparse[batch_indices]\n",
        "      batch_embeddings = hidden_layer.predict(batch_ratings)\n",
        "      yield batch_embeddings\n",
        "\n",
        "  def train_cf(self):\n",
        "    # Build the autoencoder model\n",
        "    input_layer = Input(shape=(self.ratings_mat_sparse.shape[1],))\n",
        "    encoded = Dense(64, activation='relu')(input_layer)\n",
        "    encoded = Dropout(0.2)(encoded)\n",
        "    encoded = Dense(32, activation='relu')(encoded)\n",
        "    encoded = Dropout(0.2)(encoded)\n",
        "    encoded = Dense(16, activation='relu')(encoded)\n",
        "    decoded = Dense(32, activation='relu')(encoded)\n",
        "    decoded = Dropout(0.2)(decoded)\n",
        "    decoded = Dense(64, activation='relu')(decoded)\n",
        "    decoded = Dropout(0.2)(decoded)\n",
        "    decoded = Dense(self.ratings_mat_sparse.shape[1], activation='sigmoid')(decoded)\n",
        "    autoencoder = Model(input_layer, decoded)\n",
        "    \n",
        "    # Define the learning rate schedule\n",
        "    def lr_schedule(epoch):\n",
        "        lr = 0.001\n",
        "        if epoch > 100:\n",
        "            lr *= 0.1\n",
        "        elif epoch > 50:\n",
        "            lr *= 0.5\n",
        "        return lr\n",
        "\n",
        "    # Compile the model with the scheduler\n",
        "    autoencoder.compile(optimizer='adam', loss='binary_crossentropy')\n",
        "    scheduler = LearningRateScheduler(lr_schedule)\n",
        "    autoencoder.fit(self.train_ratings, self.train_ratings, epochs=200, batch_size=32, shuffle=True,\n",
        "                    validation_data=(self.test_ratings, self.test_ratings), callbacks=[scheduler])\n",
        "\n",
        "    # Extract the hidden layer output for all the users and items\n",
        "    hidden_layer = Model(input_layer, encoded)\n",
        "    batch_size = 32\n",
        "    user_embeddings = np.concatenate(list(self.embeddings_generator(hidden_layer, batch_size)), axis=0)\n",
        "    #user_embeddings = hidden_layer.predict(self.ratings_mat_sparse)\n",
        "\n",
        "    # Compute the similarity between users and items\n",
        "    user_similarity = np.dot(user_embeddings, user_embeddings.T)\n",
        "    item_similarity = np.dot(user_embeddings.T, user_embeddings)\n",
        "\n",
        "    # Use the similarity scores to predict the ratings of the items for each user\n",
        "    self.user_ratings = np.dot(user_similarity, self.ratings_mat_sparse) / np.sum(np.abs(user_similarity), axis=1)[:, np.newaxis]\n",
        "  \n",
        "  def calculate_rmse(self):\n",
        "      mask = self.test_ratings != 0\n",
        "      masked_predicted_ratings = self.user_ratings[self.train_size:, :][mask]\n",
        "      masked_actual_ratings = self.test_ratings[mask]\n",
        "      rmse = sqrt(mean_squared_error(masked_actual_ratings, masked_predicted_ratings))\n",
        "      return rmse\n",
        "\n",
        "    # Get predicted ratings for user \n",
        "  def get_user_recommendation(self, user_id):\n",
        "      top_n = 12\n",
        "      # Get the predicted ratings for the user\n",
        "      user_predicted_ratings = self.user_ratings[user_id]\n",
        "      # Get the indices of the items sorted by predicted ratings\n",
        "      sorted_item_indices = np.argsort(user_predicted_ratings)\n",
        "      # Get the indices of the top-n items\n",
        "      top_n_item_indices = sorted_item_indices[-top_n:]\n",
        "      return top_n_item_indices\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 122,
      "metadata": {},
      "outputs": [],
      "source": [
        "def calculateAECF(ratings_mat):\n",
        "    ratings_mat = ratings_mat[:100,:500]\n",
        "    return AE_CF(ratings_mat)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "aecf_recommendations = decompress_pickle(\"LA_Hotel_AECF.pbz2\")\n",
        "recommendations_class = LA_Hotel_Recommendation"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 105,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "true\n",
            "true\n",
            "true\n",
            "true\n",
            "true\n",
            "true\n",
            "true\n",
            "true\n"
          ]
        }
      ],
      "source": [
        "#np.set_printoptions(threshold=np.inf)\n",
        "for x in LA_Restaurent_Recommendation.ratings_mat[0]:\n",
        "    if x > 5:\n",
        "        print(\"true\")\n",
        "# with open(\"abc.txt\", \"a\") as f:\n",
        "#   print(LA_Restaurent_Recommendation.ratings_mat[0], file=f)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 123,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch 1/200\n",
            "3/3 [==============================] - 2s 419ms/step - loss: 0.6961 - val_loss: 0.7164 - lr: 0.0010\n",
            "Epoch 2/200\n",
            "3/3 [==============================] - 0s 31ms/step - loss: 0.6855 - val_loss: 0.7112 - lr: 0.0010\n",
            "Epoch 3/200\n",
            "3/3 [==============================] - 0s 61ms/step - loss: 0.6748 - val_loss: 0.7065 - lr: 0.0010\n",
            "Epoch 4/200\n",
            "3/3 [==============================] - 0s 57ms/step - loss: 0.6603 - val_loss: 0.7016 - lr: 0.0010\n",
            "Epoch 5/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: 0.6392 - val_loss: 0.6955 - lr: 0.0010\n",
            "Epoch 6/200\n",
            "3/3 [==============================] - 0s 36ms/step - loss: 0.6148 - val_loss: 0.6881 - lr: 0.0010\n",
            "Epoch 7/200\n",
            "3/3 [==============================] - 0s 40ms/step - loss: 0.5684 - val_loss: 0.6806 - lr: 0.0010\n",
            "Epoch 8/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: 0.5209 - val_loss: 0.6732 - lr: 0.0010\n",
            "Epoch 9/200\n",
            "3/3 [==============================] - 0s 53ms/step - loss: 0.4425 - val_loss: 0.6690 - lr: 0.0010\n",
            "Epoch 10/200\n",
            "3/3 [==============================] - 0s 38ms/step - loss: 0.3748 - val_loss: 0.6791 - lr: 0.0010\n",
            "Epoch 11/200\n",
            "3/3 [==============================] - 0s 34ms/step - loss: 0.3672 - val_loss: 0.6981 - lr: 0.0010\n",
            "Epoch 12/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: 0.2718 - val_loss: 0.7118 - lr: 0.0010\n",
            "Epoch 13/200\n",
            "3/3 [==============================] - 0s 80ms/step - loss: 0.2377 - val_loss: 0.7203 - lr: 0.0010\n",
            "Epoch 14/200\n",
            "3/3 [==============================] - 0s 62ms/step - loss: 0.1731 - val_loss: 0.7247 - lr: 0.0010\n",
            "Epoch 15/200\n",
            "3/3 [==============================] - 0s 38ms/step - loss: 0.0530 - val_loss: 0.7230 - lr: 0.0010\n",
            "Epoch 16/200\n",
            "3/3 [==============================] - 0s 55ms/step - loss: -0.2039 - val_loss: 0.7257 - lr: 0.0010\n",
            "Epoch 17/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -0.3658 - val_loss: 0.7345 - lr: 0.0010\n",
            "Epoch 18/200\n",
            "3/3 [==============================] - 0s 30ms/step - loss: -0.4171 - val_loss: 0.7544 - lr: 0.0010\n",
            "Epoch 19/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -0.3331 - val_loss: 0.7548 - lr: 0.0010\n",
            "Epoch 20/200\n",
            "3/3 [==============================] - 0s 53ms/step - loss: -1.3291 - val_loss: 0.7428 - lr: 0.0010\n",
            "Epoch 21/200\n",
            "3/3 [==============================] - 0s 56ms/step - loss: -1.3048 - val_loss: 0.7294 - lr: 0.0010\n",
            "Epoch 22/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -2.7748 - val_loss: 0.7726 - lr: 0.0010\n",
            "Epoch 23/200\n",
            "3/3 [==============================] - 0s 30ms/step - loss: -2.4295 - val_loss: 0.8075 - lr: 0.0010\n",
            "Epoch 24/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -4.5211 - val_loss: 0.8635 - lr: 0.0010\n",
            "Epoch 25/200\n",
            "3/3 [==============================] - 0s 37ms/step - loss: -6.8151 - val_loss: 0.8796 - lr: 0.0010\n",
            "Epoch 26/200\n",
            "3/3 [==============================] - 0s 61ms/step - loss: -8.9754 - val_loss: 0.8902 - lr: 0.0010\n",
            "Epoch 27/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -12.7152 - val_loss: 0.8681 - lr: 0.0010\n",
            "Epoch 28/200\n",
            "3/3 [==============================] - 0s 30ms/step - loss: -15.5934 - val_loss: 0.7126 - lr: 0.0010\n",
            "Epoch 29/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -21.3550 - val_loss: 0.3353 - lr: 0.0010\n",
            "Epoch 30/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -35.4601 - val_loss: -0.0138 - lr: 0.0010\n",
            "Epoch 31/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -40.6537 - val_loss: -0.6160 - lr: 0.0010\n",
            "Epoch 32/200\n",
            "3/3 [==============================] - 0s 54ms/step - loss: -59.3825 - val_loss: -1.6522 - lr: 0.0010\n",
            "Epoch 33/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -83.5705 - val_loss: -3.0749 - lr: 0.0010\n",
            "Epoch 34/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -115.7696 - val_loss: -4.2346 - lr: 0.0010\n",
            "Epoch 35/200\n",
            "3/3 [==============================] - 0s 35ms/step - loss: -135.1998 - val_loss: -6.5537 - lr: 0.0010\n",
            "Epoch 36/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -197.2944 - val_loss: -9.6499 - lr: 0.0010\n",
            "Epoch 37/200\n",
            "3/3 [==============================] - 0s 32ms/step - loss: -223.1198 - val_loss: -14.3994 - lr: 0.0010\n",
            "Epoch 38/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -356.8474 - val_loss: -18.9532 - lr: 0.0010\n",
            "Epoch 39/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -379.6292 - val_loss: -27.3593 - lr: 0.0010\n",
            "Epoch 40/200\n",
            "3/3 [==============================] - 0s 38ms/step - loss: -474.3350 - val_loss: -39.8557 - lr: 0.0010\n",
            "Epoch 41/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -747.5480 - val_loss: -55.0859 - lr: 0.0010\n",
            "Epoch 42/200\n",
            "3/3 [==============================] - 0s 30ms/step - loss: -1043.0281 - val_loss: -69.5541 - lr: 0.0010\n",
            "Epoch 43/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -1239.4928 - val_loss: -91.1965 - lr: 0.0010\n",
            "Epoch 44/200\n",
            "3/3 [==============================] - 0s 44ms/step - loss: -1450.0225 - val_loss: -124.5943 - lr: 0.0010\n",
            "Epoch 45/200\n",
            "3/3 [==============================] - 0s 56ms/step - loss: -1690.2272 - val_loss: -162.5067 - lr: 0.0010\n",
            "Epoch 46/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -2608.7964 - val_loss: -211.8946 - lr: 0.0010\n",
            "Epoch 47/200\n",
            "3/3 [==============================] - 0s 32ms/step - loss: -2800.6880 - val_loss: -261.6283 - lr: 0.0010\n",
            "Epoch 48/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -3565.1375 - val_loss: -315.2165 - lr: 0.0010\n",
            "Epoch 49/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -4772.0718 - val_loss: -382.8686 - lr: 0.0010\n",
            "Epoch 50/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -5561.5576 - val_loss: -487.0943 - lr: 0.0010\n",
            "Epoch 51/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -6773.9463 - val_loss: -589.8781 - lr: 0.0010\n",
            "Epoch 52/200\n",
            "3/3 [==============================] - 0s 38ms/step - loss: -6576.5430 - val_loss: -652.2501 - lr: 5.0000e-04\n",
            "Epoch 53/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -9750.6611 - val_loss: -734.4285 - lr: 5.0000e-04\n",
            "Epoch 54/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -10076.4512 - val_loss: -815.9948 - lr: 5.0000e-04\n",
            "Epoch 55/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -10930.2754 - val_loss: -909.7531 - lr: 5.0000e-04\n",
            "Epoch 56/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -13156.9062 - val_loss: -998.5041 - lr: 5.0000e-04\n",
            "Epoch 57/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -12010.0732 - val_loss: -1096.4895 - lr: 5.0000e-04\n",
            "Epoch 58/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -12125.4561 - val_loss: -1197.2703 - lr: 5.0000e-04\n",
            "Epoch 59/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -16967.2852 - val_loss: -1310.1044 - lr: 5.0000e-04\n",
            "Epoch 60/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -16042.5439 - val_loss: -1431.9092 - lr: 5.0000e-04\n",
            "Epoch 61/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -15452.8281 - val_loss: -1552.3912 - lr: 5.0000e-04\n",
            "Epoch 62/200\n",
            "3/3 [==============================] - 0s 25ms/step - loss: -19857.6250 - val_loss: -1681.0388 - lr: 5.0000e-04\n",
            "Epoch 63/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -22254.0703 - val_loss: -1820.2848 - lr: 5.0000e-04\n",
            "Epoch 64/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -22527.7012 - val_loss: -1982.3201 - lr: 5.0000e-04\n",
            "Epoch 65/200\n",
            "3/3 [==============================] - 0s 31ms/step - loss: -19618.9961 - val_loss: -2134.9048 - lr: 5.0000e-04\n",
            "Epoch 66/200\n",
            "3/3 [==============================] - 0s 49ms/step - loss: -24476.7617 - val_loss: -2303.5034 - lr: 5.0000e-04\n",
            "Epoch 67/200\n",
            "3/3 [==============================] - 0s 32ms/step - loss: -23203.5723 - val_loss: -2456.6890 - lr: 5.0000e-04\n",
            "Epoch 68/200\n",
            "3/3 [==============================] - 0s 39ms/step - loss: -28325.8633 - val_loss: -2609.3911 - lr: 5.0000e-04\n",
            "Epoch 69/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -34649.5781 - val_loss: -2817.3521 - lr: 5.0000e-04\n",
            "Epoch 70/200\n",
            "3/3 [==============================] - 0s 48ms/step - loss: -31846.4746 - val_loss: -3045.3835 - lr: 5.0000e-04\n",
            "Epoch 71/200\n",
            "3/3 [==============================] - 0s 33ms/step - loss: -39337.8633 - val_loss: -3255.4199 - lr: 5.0000e-04\n",
            "Epoch 72/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -39840.9609 - val_loss: -3449.5630 - lr: 5.0000e-04\n",
            "Epoch 73/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -41258.8750 - val_loss: -3744.5103 - lr: 5.0000e-04\n",
            "Epoch 74/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -42169.5703 - val_loss: -4004.3242 - lr: 5.0000e-04\n",
            "Epoch 75/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -45470.8008 - val_loss: -4309.3755 - lr: 5.0000e-04\n",
            "Epoch 76/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -46814.2461 - val_loss: -4565.5640 - lr: 5.0000e-04\n",
            "Epoch 77/200\n",
            "3/3 [==============================] - 0s 25ms/step - loss: -57649.0820 - val_loss: -4875.7583 - lr: 5.0000e-04\n",
            "Epoch 78/200\n",
            "3/3 [==============================] - 0s 24ms/step - loss: -53252.2383 - val_loss: -5234.7168 - lr: 5.0000e-04\n",
            "Epoch 79/200\n",
            "3/3 [==============================] - 0s 24ms/step - loss: -61315.6484 - val_loss: -5628.7510 - lr: 5.0000e-04\n",
            "Epoch 80/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -62174.6367 - val_loss: -6039.2812 - lr: 5.0000e-04\n",
            "Epoch 81/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -66112.9141 - val_loss: -6424.7603 - lr: 5.0000e-04\n",
            "Epoch 82/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -82120.7969 - val_loss: -6840.4141 - lr: 5.0000e-04\n",
            "Epoch 83/200\n",
            "3/3 [==============================] - 0s 30ms/step - loss: -83452.2969 - val_loss: -7252.0913 - lr: 5.0000e-04\n",
            "Epoch 84/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -80057.6797 - val_loss: -7729.8818 - lr: 5.0000e-04\n",
            "Epoch 85/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -82496.7656 - val_loss: -8264.1406 - lr: 5.0000e-04\n",
            "Epoch 86/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -98916.5234 - val_loss: -8811.1533 - lr: 5.0000e-04\n",
            "Epoch 87/200\n",
            "3/3 [==============================] - 0s 30ms/step - loss: -98349.6094 - val_loss: -9365.1836 - lr: 5.0000e-04\n",
            "Epoch 88/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -96291.0469 - val_loss: -9968.7627 - lr: 5.0000e-04\n",
            "Epoch 89/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -103690.8984 - val_loss: -10610.1348 - lr: 5.0000e-04\n",
            "Epoch 90/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -108334.2109 - val_loss: -11199.4092 - lr: 5.0000e-04\n",
            "Epoch 91/200\n",
            "3/3 [==============================] - 0s 30ms/step - loss: -116866.1719 - val_loss: -11898.4053 - lr: 5.0000e-04\n",
            "Epoch 92/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -120454.3750 - val_loss: -12686.6309 - lr: 5.0000e-04\n",
            "Epoch 93/200\n",
            "3/3 [==============================] - 0s 66ms/step - loss: -134196.2188 - val_loss: -13456.5830 - lr: 5.0000e-04\n",
            "Epoch 94/200\n",
            "3/3 [==============================] - 0s 30ms/step - loss: -154044.6875 - val_loss: -14201.7686 - lr: 5.0000e-04\n",
            "Epoch 95/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -149672.0312 - val_loss: -15078.2246 - lr: 5.0000e-04\n",
            "Epoch 96/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -157739.3594 - val_loss: -16022.5215 - lr: 5.0000e-04\n",
            "Epoch 97/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -161362.9219 - val_loss: -16909.3320 - lr: 5.0000e-04\n",
            "Epoch 98/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -156171.7031 - val_loss: -17854.5996 - lr: 5.0000e-04\n",
            "Epoch 99/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -174665.0781 - val_loss: -19002.9590 - lr: 5.0000e-04\n",
            "Epoch 100/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -195809.3906 - val_loss: -20125.7930 - lr: 5.0000e-04\n",
            "Epoch 101/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -196391.5469 - val_loss: -21185.6445 - lr: 5.0000e-04\n",
            "Epoch 102/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -217075.2969 - val_loss: -21423.8340 - lr: 1.0000e-04\n",
            "Epoch 103/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -234044.0938 - val_loss: -21685.6367 - lr: 1.0000e-04\n",
            "Epoch 104/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -211244.7812 - val_loss: -21948.5723 - lr: 1.0000e-04\n",
            "Epoch 105/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -199049.4219 - val_loss: -22194.4961 - lr: 1.0000e-04\n",
            "Epoch 106/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -221012.9688 - val_loss: -22464.3906 - lr: 1.0000e-04\n",
            "Epoch 107/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -263261.9375 - val_loss: -22726.9199 - lr: 1.0000e-04\n",
            "Epoch 108/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -239847.5000 - val_loss: -22947.7539 - lr: 1.0000e-04\n",
            "Epoch 109/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -244875.7031 - val_loss: -23120.6914 - lr: 1.0000e-04\n",
            "Epoch 110/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -232706.9688 - val_loss: -23326.3477 - lr: 1.0000e-04\n",
            "Epoch 111/200\n",
            "3/3 [==============================] - 0s 30ms/step - loss: -220026.4688 - val_loss: -23561.6348 - lr: 1.0000e-04\n",
            "Epoch 112/200\n",
            "3/3 [==============================] - 0s 25ms/step - loss: -250320.4531 - val_loss: -23795.2383 - lr: 1.0000e-04\n",
            "Epoch 113/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -227575.0469 - val_loss: -24006.7227 - lr: 1.0000e-04\n",
            "Epoch 114/200\n",
            "3/3 [==============================] - 0s 25ms/step - loss: -256996.2812 - val_loss: -24238.8477 - lr: 1.0000e-04\n",
            "Epoch 115/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -271344.2500 - val_loss: -24479.8359 - lr: 1.0000e-04\n",
            "Epoch 116/200\n",
            "3/3 [==============================] - 0s 45ms/step - loss: -285455.3438 - val_loss: -24726.5898 - lr: 1.0000e-04\n",
            "Epoch 117/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -249235.2969 - val_loss: -24960.0312 - lr: 1.0000e-04\n",
            "Epoch 118/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -278717.8750 - val_loss: -25175.9746 - lr: 1.0000e-04\n",
            "Epoch 119/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -273900.1875 - val_loss: -25414.9102 - lr: 1.0000e-04\n",
            "Epoch 120/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -276570.6562 - val_loss: -25640.6445 - lr: 1.0000e-04\n",
            "Epoch 121/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -299139.9062 - val_loss: -25886.7012 - lr: 1.0000e-04\n",
            "Epoch 122/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -257986.0469 - val_loss: -26096.0684 - lr: 1.0000e-04\n",
            "Epoch 123/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -264641.5000 - val_loss: -26281.5430 - lr: 1.0000e-04\n",
            "Epoch 124/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -251488.0469 - val_loss: -26468.3633 - lr: 1.0000e-04\n",
            "Epoch 125/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -294544.4688 - val_loss: -26694.5781 - lr: 1.0000e-04\n",
            "Epoch 126/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -241306.8438 - val_loss: -26921.2695 - lr: 1.0000e-04\n",
            "Epoch 127/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -296135.8750 - val_loss: -27177.9414 - lr: 1.0000e-04\n",
            "Epoch 128/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -270397.5000 - val_loss: -27432.0781 - lr: 1.0000e-04\n",
            "Epoch 129/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -256028.1719 - val_loss: -27703.1875 - lr: 1.0000e-04\n",
            "Epoch 130/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -303675.2812 - val_loss: -27948.8066 - lr: 1.0000e-04\n",
            "Epoch 131/200\n",
            "3/3 [==============================] - 0s 32ms/step - loss: -307311.0000 - val_loss: -28170.1406 - lr: 1.0000e-04\n",
            "Epoch 132/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -333978.6562 - val_loss: -28360.8945 - lr: 1.0000e-04\n",
            "Epoch 133/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -299419.9375 - val_loss: -28561.8379 - lr: 1.0000e-04\n",
            "Epoch 134/200\n",
            "3/3 [==============================] - 0s 26ms/step - loss: -307555.8750 - val_loss: -28782.4570 - lr: 1.0000e-04\n",
            "Epoch 135/200\n",
            "3/3 [==============================] - 0s 25ms/step - loss: -289736.0625 - val_loss: -29007.0195 - lr: 1.0000e-04\n",
            "Epoch 136/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -293428.3438 - val_loss: -29230.8867 - lr: 1.0000e-04\n",
            "Epoch 137/200\n",
            "3/3 [==============================] - 0s 53ms/step - loss: -322393.8125 - val_loss: -29454.2695 - lr: 1.0000e-04\n",
            "Epoch 138/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -275860.6250 - val_loss: -29685.3398 - lr: 1.0000e-04\n",
            "Epoch 139/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -284695.1875 - val_loss: -29924.8008 - lr: 1.0000e-04\n",
            "Epoch 140/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -317083.1250 - val_loss: -30158.2246 - lr: 1.0000e-04\n",
            "Epoch 141/200\n",
            "3/3 [==============================] - 0s 30ms/step - loss: -337732.9375 - val_loss: -30379.3750 - lr: 1.0000e-04\n",
            "Epoch 142/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -290605.8438 - val_loss: -30591.3066 - lr: 1.0000e-04\n",
            "Epoch 143/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -299846.3438 - val_loss: -30813.0840 - lr: 1.0000e-04\n",
            "Epoch 144/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -266462.6875 - val_loss: -31046.4531 - lr: 1.0000e-04\n",
            "Epoch 145/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -380484.9062 - val_loss: -31277.6191 - lr: 1.0000e-04\n",
            "Epoch 146/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -328212.0938 - val_loss: -31488.1445 - lr: 1.0000e-04\n",
            "Epoch 147/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -374192.9062 - val_loss: -31699.2871 - lr: 1.0000e-04\n",
            "Epoch 148/200\n",
            "3/3 [==============================] - 0s 34ms/step - loss: -340887.3125 - val_loss: -31903.5527 - lr: 1.0000e-04\n",
            "Epoch 149/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -342059.0625 - val_loss: -32143.6836 - lr: 1.0000e-04\n",
            "Epoch 150/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -334512.6875 - val_loss: -32356.9688 - lr: 1.0000e-04\n",
            "Epoch 151/200\n",
            "3/3 [==============================] - 0s 32ms/step - loss: -292570.7812 - val_loss: -32593.8340 - lr: 1.0000e-04\n",
            "Epoch 152/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -329538.7188 - val_loss: -32833.2266 - lr: 1.0000e-04\n",
            "Epoch 153/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -335226.0312 - val_loss: -33063.0938 - lr: 1.0000e-04\n",
            "Epoch 154/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -310882.5000 - val_loss: -33291.9219 - lr: 1.0000e-04\n",
            "Epoch 155/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -331719.8125 - val_loss: -33505.0859 - lr: 1.0000e-04\n",
            "Epoch 156/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -398107.4062 - val_loss: -33734.7891 - lr: 1.0000e-04\n",
            "Epoch 157/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -399497.2500 - val_loss: -33972.1562 - lr: 1.0000e-04\n",
            "Epoch 158/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -329307.9375 - val_loss: -34221.9180 - lr: 1.0000e-04\n",
            "Epoch 159/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -360580.3750 - val_loss: -34483.5938 - lr: 1.0000e-04\n",
            "Epoch 160/200\n",
            "3/3 [==============================] - 0s 49ms/step - loss: -386343.1875 - val_loss: -34695.3633 - lr: 1.0000e-04\n",
            "Epoch 161/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -362028.3438 - val_loss: -34936.5781 - lr: 1.0000e-04\n",
            "Epoch 162/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -365387.9688 - val_loss: -35191.6797 - lr: 1.0000e-04\n",
            "Epoch 163/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -335796.8438 - val_loss: -35455.8398 - lr: 1.0000e-04\n",
            "Epoch 164/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -384838.8125 - val_loss: -35702.7109 - lr: 1.0000e-04\n",
            "Epoch 165/200\n",
            "3/3 [==============================] - 0s 32ms/step - loss: -373222.3750 - val_loss: -35945.4375 - lr: 1.0000e-04\n",
            "Epoch 166/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -367469.2188 - val_loss: -36163.7422 - lr: 1.0000e-04\n",
            "Epoch 167/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -381218.5625 - val_loss: -36397.0273 - lr: 1.0000e-04\n",
            "Epoch 168/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -433430.6875 - val_loss: -36621.2734 - lr: 1.0000e-04\n",
            "Epoch 169/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -413127.3125 - val_loss: -36844.2422 - lr: 1.0000e-04\n",
            "Epoch 170/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -400222.5938 - val_loss: -37094.3047 - lr: 1.0000e-04\n",
            "Epoch 171/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -387406.7812 - val_loss: -37349.9766 - lr: 1.0000e-04\n",
            "Epoch 172/200\n",
            "3/3 [==============================] - 0s 32ms/step - loss: -478763.4062 - val_loss: -37611.9570 - lr: 1.0000e-04\n",
            "Epoch 173/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -336177.0625 - val_loss: -37874.4805 - lr: 1.0000e-04\n",
            "Epoch 174/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -466066.6562 - val_loss: -38123.9531 - lr: 1.0000e-04\n",
            "Epoch 175/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -373685.4062 - val_loss: -38343.5938 - lr: 1.0000e-04\n",
            "Epoch 176/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -415830.5000 - val_loss: -38583.6328 - lr: 1.0000e-04\n",
            "Epoch 177/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -413201.8125 - val_loss: -38825.1914 - lr: 1.0000e-04\n",
            "Epoch 178/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -353458.0625 - val_loss: -39056.3633 - lr: 1.0000e-04\n",
            "Epoch 179/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -474762.4062 - val_loss: -39281.5312 - lr: 1.0000e-04\n",
            "Epoch 180/200\n",
            "3/3 [==============================] - 0s 54ms/step - loss: -362270.4375 - val_loss: -39475.2070 - lr: 1.0000e-04\n",
            "Epoch 181/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -450308.6875 - val_loss: -39715.1914 - lr: 1.0000e-04\n",
            "Epoch 182/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -379256.3438 - val_loss: -39989.8516 - lr: 1.0000e-04\n",
            "Epoch 183/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -391307.6562 - val_loss: -40298.5469 - lr: 1.0000e-04\n",
            "Epoch 184/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -448234.6875 - val_loss: -40609.4883 - lr: 1.0000e-04\n",
            "Epoch 185/200\n",
            "3/3 [==============================] - 0s 27ms/step - loss: -418644.4688 - val_loss: -40898.6016 - lr: 1.0000e-04\n",
            "Epoch 186/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -395202.4375 - val_loss: -41176.6875 - lr: 1.0000e-04\n",
            "Epoch 187/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -461015.5938 - val_loss: -41458.8867 - lr: 1.0000e-04\n",
            "Epoch 188/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -465801.6875 - val_loss: -41720.7031 - lr: 1.0000e-04\n",
            "Epoch 189/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -397374.3125 - val_loss: -41976.2109 - lr: 1.0000e-04\n",
            "Epoch 190/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -448902.8438 - val_loss: -42228.0508 - lr: 1.0000e-04\n",
            "Epoch 191/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -376045.9062 - val_loss: -42432.7305 - lr: 1.0000e-04\n",
            "Epoch 192/200\n",
            "3/3 [==============================] - 0s 32ms/step - loss: -423370.9062 - val_loss: -42683.0508 - lr: 1.0000e-04\n",
            "Epoch 193/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -494273.9375 - val_loss: -42925.1406 - lr: 1.0000e-04\n",
            "Epoch 194/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -420166.5000 - val_loss: -43196.4883 - lr: 1.0000e-04\n",
            "Epoch 195/200\n",
            "3/3 [==============================] - 0s 29ms/step - loss: -496702.7500 - val_loss: -43495.8633 - lr: 1.0000e-04\n",
            "Epoch 196/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -483281.0938 - val_loss: -43818.6406 - lr: 1.0000e-04\n",
            "Epoch 197/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -393259.7500 - val_loss: -44131.7266 - lr: 1.0000e-04\n",
            "Epoch 198/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -504587.1875 - val_loss: -44442.6172 - lr: 1.0000e-04\n",
            "Epoch 199/200\n",
            "3/3 [==============================] - 0s 28ms/step - loss: -497329.5625 - val_loss: -44755.8750 - lr: 1.0000e-04\n",
            "Epoch 200/200\n",
            "3/3 [==============================] - 0s 47ms/step - loss: -428126.4375 - val_loss: -45075.1758 - lr: 1.0000e-04\n",
            "1/1 [==============================] - 0s 109ms/step\n",
            "1/1 [==============================] - 0s 28ms/step\n",
            "1/1 [==============================] - 0s 15ms/step\n",
            "1/1 [==============================] - 0s 78ms/step\n",
            "3.92793973073431\n"
          ]
        }
      ],
      "source": [
        "LA_Restaurent_AECF = calculateAECF(LA_Restaurent_Recommendation.ratings_mat)\n",
        "print(LA_Restaurent_AECF.rmse)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 103,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "true\n",
            "true\n",
            "true\n"
          ]
        }
      ],
      "source": [
        "for x in LA_Restaurent_AECF.test_ratings[0]:\n",
        "    if x > 5:\n",
        "        print(\"true\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 79,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "(100, 100)\n"
          ]
        }
      ],
      "source": [
        "business_ids = aecf_recommendations.user_ratings\n",
        "#print(business_ids[0].toarray().flatten())\n",
        "print(business_ids.shape)\n",
        "# recommendations_class.ratings_mat.shape\n",
        "\n",
        "# business_list = []\n",
        "# for i in range(12):\n",
        "#     business_hash = recommendations_class.getBusinessHashFromBusinessNum(\n",
        "#         business_ids[i])\n",
        "#     print(len(recommendations_class.business_num_to_business_hash_dict))\n",
        "#     print(business_hash)\n",
        "#     business = recommendations_class.getBusinessInfo(business_hash)\n",
        "#     business_list.append(business)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "PA_Restaurent_AECF = calculateAECF(PA_Restaurent_Recommendation.ratings_mat)\n",
        "FL_Restaurent_AECF = calculateAECF(FL_Restaurent_Recommendation.ratings_mat)\n",
        "TN_Restaurent_AECF = calculateAECF(TN_Restaurent_Recommendation.ratings_mat)\n",
        "IN_Restaurent_AECF = calculateAECF(IN_Restaurent_Recommendation.ratings_mat)\n",
        "MO_Restaurent_AECF = calculateAECF(MO_Restaurent_Recommendation.ratings_mat)\n",
        "LA_Restaurent_AECF = calculateAECF(LA_Restaurent_Recommendation.ratings_mat)\n",
        "AZ_Restaurent_AECF = calculateAECF(AZ_Restaurent_Recommendation.ratings_mat)\n",
        "NJ_Restaurent_AECF = calculateAECF(NJ_Restaurent_Recommendation.ratings_mat)\n",
        "NV_Restaurent_AECF = calculateAECF(NV_Restaurent_Recommendation.ratings_mat)\n",
        "AB_Restaurent_AECF = calculateAECF(AB_Restaurent_Recommendation.ratings_mat)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {},
      "outputs": [],
      "source": [
        "compressed_pickle(\"data/reataurent/PA_Restaurent_AECF\", PA_Restaurent_AECF)\n",
        "compressed_pickle(\"data/reataurent/FL_Restaurent_AECF\", FL_Restaurent_AECF)\n",
        "compressed_pickle(\"data/reataurent/TN_Restaurent_AECF\", TN_Restaurent_AECF)\n",
        "compressed_pickle(\"data/reataurent/IN_Restaurent_AECF\", IN_Restaurent_AECF)\n",
        "compressed_pickle(\"data/reataurent/MO_Restaurent_AECF\", MO_Restaurent_AECF)\n",
        "compressed_pickle(\"data/reataurent/LA_Restaurent_AECF\", LA_Restaurent_AECF)\n",
        "compressed_pickle(\"data/reataurent/AZ_Restaurent_AECF\", AZ_Restaurent_AECF)\n",
        "compressed_pickle(\"data/reataurent/NJ_Restaurent_AECF\", NJ_Restaurent_AECF)\n",
        "compressed_pickle(\"data/reataurent/NV_Restaurent_AECF\", NV_Restaurent_AECF)\n",
        "compressed_pickle(\"data/reataurent/AB_Restaurent_AECF\", AB_Restaurent_AECF)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "PA_Hotel_AECF = calculateAECF(PA_Hotel_Recommendation.ratings_mat)\n",
        "FL_Hotel_AECF = calculateAECF(FL_Hotel_Recommendation.ratings_mat)\n",
        "TN_Hotel_AECF = calculateAECF(TN_Hotel_Recommendation.ratings_mat)\n",
        "IN_Hotel_AECF = calculateAECF(IN_Hotel_Recommendation.ratings_mat)\n",
        "MO_Hotel_AECF = calculateAECF(MO_Hotel_Recommendation.ratings_mat)\n",
        "LA_Hotel_AECF = calculateAECF(LA_Hotel_Recommendation.ratings_mat)\n",
        "AZ_Hotel_AECF = calculateAECF(AZ_Hotel_Recommendation.ratings_mat)\n",
        "NJ_Hotel_AECF = calculateAECF(NJ_Hotel_Recommendation.ratings_mat)\n",
        "NV_Hotel_AECF = calculateAECF(NV_Hotel_Recommendation.ratings_mat)\n",
        "AB_Hotel_AECF = calculateAECF(AB_Hotel_Recommendation.ratings_mat)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {},
      "outputs": [],
      "source": [
        "compressed_pickle(\"data/hotel/PA_Hotel_AECF\", PA_Hotel_AECF)\n",
        "compressed_pickle(\"data/hotel/FL_Hotel_AECF\", FL_Hotel_AECF)\n",
        "compressed_pickle(\"data/hotel/TN_Hotel_AECF\", TN_Hotel_AECF)\n",
        "compressed_pickle(\"data/hotel/IN_Hotel_AECF\", IN_Hotel_AECF)\n",
        "compressed_pickle(\"data/hotel/MO_Hotel_AECF\", MO_Hotel_AECF)\n",
        "compressed_pickle(\"data/hotel/LA_Hotel_AECF\", LA_Hotel_AECF)\n",
        "compressed_pickle(\"data/hotel/AZ_Hotel_AECF\", AZ_Hotel_AECF)\n",
        "compressed_pickle(\"data/hotel/NJ_Hotel_AECF\", NJ_Hotel_AECF)\n",
        "compressed_pickle(\"data/hotel/NV_Hotel_AECF\", NV_Hotel_AECF)\n",
        "compressed_pickle(\"data/hotel/AB_Hotel_AECF\", AB_Hotel_AECF)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "PA_Nightlife_AECF = calculateAECF(PA_Nightlife_Recommendation.ratings_mat)\n",
        "FL_Nightlife_AECF = calculateAECF(FL_Nightlife_Recommendation.ratings_mat)\n",
        "TN_Nightlife_AECF = calculateAECF(TN_Nightlife_Recommendation.ratings_mat)\n",
        "IN_Nightlife_AECF = calculateAECF(IN_Nightlife_Recommendation.ratings_mat)\n",
        "MO_Nightlife_AECF = calculateAECF(MO_Nightlife_Recommendation.ratings_mat)\n",
        "LA_Nightlife_AECF = calculateAECF(LA_Nightlife_Recommendation.ratings_mat)\n",
        "AZ_Nightlife_AECF = calculateAECF(AZ_Nightlife_Recommendation.ratings_mat)\n",
        "NJ_Nightlife_AECF = calculateAECF(NJ_Nightlife_Recommendation.ratings_mat)\n",
        "NV_Nightlife_AECF = calculateAECF(NV_Nightlife_Recommendation.ratings_mat)\n",
        "AB_Nightlife_AECF = calculateAECF(AB_Nightlife_Recommendation.ratings_mat)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {},
      "outputs": [],
      "source": [
        "compressed_pickle(\"data/Nightlife/PA_Nightlife_AECF\", PA_Nightlife_AECF)\n",
        "compressed_pickle(\"data/Nightlife/FL_Nightlife_AECF\", FL_Nightlife_AECF)\n",
        "compressed_pickle(\"data/Nightlife/TN_Nightlife_AECF\", TN_Nightlife_AECF)\n",
        "compressed_pickle(\"data/Nightlife/IN_Nightlife_AECF\", IN_Nightlife_AECF)\n",
        "compressed_pickle(\"data/Nightlife/MO_Nightlife_AECF\", MO_Nightlife_AECF)\n",
        "compressed_pickle(\"data/Nightlife/LA_Nightlife_AECF\", LA_Nightlife_AECF)\n",
        "compressed_pickle(\"data/Nightlife/AZ_Nightlife_AECF\", AZ_Nightlife_AECF)\n",
        "compressed_pickle(\"data/Nightlife/NJ_Nightlife_AECF\", NJ_Nightlife_AECF)\n",
        "compressed_pickle(\"data/Nightlife/NV_Nightlife_AECF\", NV_Nightlife_AECF)\n",
        "compressed_pickle(\"data/Nightlife/AB_Nightlife_AECF\", AB_Nightlife_AECF)"
      ]
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3.10.6 ('tf')",
      "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": "3.10.10"
    },
    "orig_nbformat": 4,
    "vscode": {
      "interpreter": {
        "hash": "8a48a9194367956c9607f494d6809ea58a846b01fefe7fdc109363ac5be64d1f"
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 2
}