Recommendation System.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import warnings\n",
    "from tqdm import tqdm\n",
    "warnings.filterwarnings('ignore')\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "import matplotlib.pyplot as plt\n",
    "from scipy.sparse import csr_matrix\n",
    "from sklearn.neighbors import KNeighborsClassifier,NearestNeighbors\n",
    "import pathlib\n",
    "import seaborn as sns\n",
    "import time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "class RecommendationSystem:\n",
    "    \"\"\"\n",
    "    item-based collaborative filtering for movie-lens dataset using sklearn's nearest neighbors\n",
    "    \n",
    "    \"\"\"\n",
    "    def __init__(self,movie_th,user_th,movie_path,ratings_path,users_path):\n",
    "        \n",
    "        \"\"\"\n",
    "        Description:\n",
    "            1. Initialization of threshold values\n",
    "            2. Initialization of file paths\n",
    "            3. Initialization of models\n",
    "        \n",
    "        \"\"\"\n",
    "        self.movie_th = movie_th\n",
    "        self.user_th = user_th\n",
    "        self.movies_path = movie_path\n",
    "        self.ratings_path = ratings_path\n",
    "        self.path = pathlib.Path().absolute() / 'ml-1m'\n",
    "        self.users_path = users_path\n",
    "        self.model = NearestNeighbors()\n",
    "        self.data_matrix = None \n",
    "        self.dic_movie_name = None\n",
    "        self.dic_movie_id = None\n",
    "    def define_model_parameters(self,n_neighbors,algorithm,metric,jobs=1):\n",
    "        \n",
    "        #print(n_neighbors,algorithm,metric,jobs)\n",
    "        \"\"\"\n",
    "        Args:\n",
    "            n_neighbors : number of neighbors used for calculating similarity\n",
    "            algorithm : type of algorithm used i.e brute force, KDTRee, BallTree for building model\n",
    "            metric : how similarity is being measured i.e cosine, l1,l2 norm \n",
    "            jobs :  number of processors to be used in parallel\n",
    "        \"\"\"\n",
    "        self.model.set_params(**{\n",
    "            'n_neighbors':n_neighbors,\n",
    "            'algorithm': algorithm,\n",
    "            'metric': metric,\n",
    "            'n_jobs' : jobs\n",
    "        })\n",
    "            \n",
    "        \n",
    "        \n",
    "    def read_clean_prepare(self):\n",
    "        \n",
    "        \n",
    "        \n",
    "        self.ratings = pd.read_csv(\n",
    "            self.path/self.ratings_path,\n",
    "            sep='::',\n",
    "            names=['userId','movieId','rating','timestamp'],\n",
    "            usecols = ['userId','movieId','rating'],\n",
    "            dtype= {'userId':np.int32,'movieId':np.int32,'rating':np.float32}\n",
    "        )\n",
    "        \n",
    "        self.movies = pd.read_csv(\n",
    "            self.path/self.movies_path,\n",
    "            sep='::',\n",
    "            names=['movieId','title','genres'],\n",
    "            usecols = ['movieId','title'],\n",
    "            dtype= {'movieId':np.int32,'title':str}\n",
    "        )\n",
    "        \n",
    "        self.movies.title = self.movies.title.str.lower()\n",
    "        self.movies['name'] = self.movies.title.str.replace(r'\\(\\d{4}\\)',\"\")\n",
    "        self.movies.name = self.movies.name.str.rstrip()\n",
    "        self.dic_movie_name = dict([(row.movieId,row.title) for row in self.movies.itertuples()])\n",
    "        self.dic_movie_id = dict([(row.name,row.movieId) for row in self.movies.itertuples()])\n",
    "        temp = self.ratings.groupby('movieId').agg({'userId':len})\n",
    "        self.pop_movies = temp[temp['userId']>=recommender.movie_th].index\n",
    "        \n",
    "        \n",
    "        self.data_matrix = self.ratings.pivot(index = 'movieId',columns = 'userId', values = 'rating').fillna(0)\n",
    "        \n",
    "        self.data_matrix = csr_matrix(self.data_matrix)\n",
    "        \n",
    "        \n",
    "        time.sleep(3)\n",
    "        self.define_model_parameters(20,'brute','cosine')\n",
    "        \n",
    "        \n",
    "        self.model.fit(self.data_matrix)\n",
    "        \n",
    "    def plot_rating_freq(self):\n",
    "        temp = self.ratings.groupby('movieId').agg({'userId':len})\n",
    "\n",
    "        #temp = temp.reset_index()\n",
    "        temp = temp.sort_values(by=['userId'],ascending=False)\n",
    "        temp.index = pd.RangeIndex(0,temp.shape[0])\n",
    "        plt.plot(temp.index,temp['userId'])\n",
    "        plt.title('movie rating frequency')\n",
    "        plt.xlabel(\"movies\")\n",
    "        plt.ylabel(\"number of ratings\")\n",
    "        \n",
    "        \n",
    "    def recommend(self,fav,how_many):\n",
    "        \n",
    "        \n",
    "        \"\"\"\n",
    "        Args:\n",
    "        \n",
    "            fav: name of favorite movie\n",
    "            how_many : number of similar movies to be recommended\n",
    "            \n",
    "        Returns:\n",
    "            \n",
    "            Exits if the movie not in database else returns the name of the movies.\n",
    "        \n",
    "        \"\"\"\n",
    "        if fav not in self.dic_movie_id:\n",
    "            print(\"Movie not found in the database\\n\")\n",
    "            exit() \n",
    "        \n",
    "        fav_id = self.dic_movie_id[fav]\n",
    "        _,recommendations = self.model.kneighbors(self.data_matrix[fav_id],n_neighbors=how_many+1)\n",
    "        \n",
    "        recommendations = [self.dic_movie_name[x] for x in recommendations[0][1:]]\n",
    "        #print(recommendations)\n",
    "        print(\"Top {} recommendations are:\\n\".format(how_many))\n",
    "        \n",
    "        for i in range(how_many):\n",
    "            \n",
    "            print(\"{}. {}\".format(i+1,recommendations[i]))\n",
    "        return "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [],
   "source": [
    "recommender = RecommendationSystem(50,50,'movies.dat','ratings.dat','users.dat')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "20 brute cosine 1\n"
     ]
    }
   ],
   "source": [
    "recommender.read_clean_prepare()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "recommender.plot_rating_freq()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3706, 6040)"
      ]
     },
     "execution_count": 143,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "recommender.data_matrix.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Top 10 recommendations are:\n",
      "\n",
      "1. patriot games (1992)\n",
      "2. land and freedom (tierra y libertad) (1995)\n",
      "3. postino, il (the postman) (1994)\n",
      "4. mis�rables, les (1998)\n",
      "5. friday the 13th part vii: the new blood (1988)\n",
      "6. three colors: white (1994)\n",
      "7. four weddings and a funeral (1994)\n",
      "8. extreme measures (1996)\n",
      "9. back to the future part iii (1990)\n",
      "10. superweib, das (1996)\n"
     ]
    }
   ],
   "source": [
    "recommender.recommend('toy story',10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}