{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "df = pd.read_csv('../data/ratings.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#df['rating'] = df['rating']-3\n",
    "#df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[4.  5.  3.  2.  1.  4.5 3.5 2.5 0.5 1.5]\n",
      "['userId' 'movieId' 'rating' 'timestamp']\n",
      "[     1      3      6 ... 160836 163937 163981]\n"
     ]
    }
   ],
   "source": [
    "rating = df['rating'].unique()\n",
    "print(rating)\n",
    "print(df.columns.values)\n",
    "\n",
    "#df['userId'] = 'u'+df['userId']\n",
    "#df['movieId'] = 'm'+df['movieId'].unique()\n",
    "\n",
    "movie = df['movieId'].unique()\n",
    "print(movie)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# garantir que os ids dof filmes e usuários sejam diferentes\n",
    "df['movieId'] = df['movieId'] + 610"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import networkx as nx\n",
    "from networkx.algorithms import bipartite\n",
    "\n",
    "#df = df.query(expr='rating>3.5')\n",
    "user_id_list = df['userId'].unique()\n",
    "movie_id_list = df['movieId'].unique()\n",
    "df = df.filter(['userId','movieId','rating'])\n",
    "df_b  = df.filter(['userId','movieId'])\n",
    "df.filter(['userId','movieId'])\n",
    "\n",
    "B = nx.Graph()\n",
    "# Add nodes with the node attribute \"bipartite\"\n",
    "B.add_nodes_from(np.array(user_id_list), bipartite=0)\n",
    "B.add_nodes_from(np.array(movie_id_list), bipartite=1)\n",
    "# Add edges only between nodes of opposite node sets\n",
    "B.add_edges_from(np.array(df_b))\n",
    "nx.write_gexf(B, 'graph_b.gexf')\n",
    "\n",
    "df_r45 = df.query(expr='rating>=4').filter(['userId','movieId']) #gostou\n",
    "B_r45 = nx.Graph()\n",
    "# Add nodes with the node attribute \"bipartite\"\n",
    "B_r45.add_nodes_from(np.array(user_id_list), bipartite=0)\n",
    "B_r45.add_nodes_from(np.array(movie_id_list), bipartite=1)\n",
    "# Add edges only between nodes of opposite node sets\n",
    "B_r45.add_edges_from(np.array(df_r45))\n",
    "nx.write_gexf(B_r45, 'graph_b_r45.gexf')\n",
    "\n",
    "df_indiferente = df.query(expr='rating<4 & rating>=2').filter(['userId','movieId']) #indiferente\n",
    "B_indiferente = nx.Graph()\n",
    "# Add nodes with the node attribute \"bipartite\"\n",
    "B_indiferente.add_nodes_from(np.array(user_id_list), bipartite=0)\n",
    "B_indiferente.add_nodes_from(np.array(movie_id_list), bipartite=1)\n",
    "# Add edges only between nodes of opposite node sets\n",
    "B_indiferente.add_edges_from(np.array(df_indiferente))\n",
    "nx.write_gexf(B_indiferente, 'graph_b_indiferente.gexf')\n",
    "\n",
    "df_r01 = df.query(expr='rating<2').filter(['userId','movieId']) #não gostou\n",
    "B_r01 = nx.Graph()\n",
    "# Add nodes with the node attribute \"bipartite\"\n",
    "B_r01.add_nodes_from(np.array(user_id_list), bipartite=0)\n",
    "B_r01.add_nodes_from(np.array(movie_id_list), bipartite=1)\n",
    "# Add edges only between nodes of opposite node sets\n",
    "B_r01.add_edges_from(np.array(df_r01))\n",
    "nx.write_gexf(B_r01, 'graph_b_r01.gexf')\n",
    "\n",
    "# rating centralizado\n",
    "df['rating'] = df['rating'] -2.75\n",
    "# B= bipartido W=weight/rating\n",
    "BW = nx.Graph()\n",
    "# Add nodes with the node attribute \"bipartite\"\n",
    "BW.add_nodes_from(np.array(user_id_list), bipartite=0)\n",
    "BW.add_nodes_from(np.array(movie_id_list), bipartite=1)\n",
    "# Add edges only between nodes of opposite node sets\n",
    "BW.add_weighted_edges_from(np.array(df))\n",
    "nx.write_gexf(BW, 'graph_bw.gexf')\n",
    "\n",
    "am = bipartite.biadjacency_matrix(B, row_order=user_id_list)\n",
    "G = bipartite.from_biadjacency_matrix(am)\n",
    "nx.write_gexf(G, 'graph_g.gexf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "am_user = am.dot(am.T)\n",
    "df = pd.DataFrame(am_user)\n",
    "G = nx.Graph(am_user)\n",
    "nx.write_gexf(G, 'graph_am_user.gexf')\n",
    "\n",
    "#muito pesado\n",
    "#am_movie = am.T.dot(am)\n",
    "#df = pd.DataFrame(am_movie)\n",
    "#G = nx.Graph(am_movie)\n",
    "#nx.write_gexf(G, 'graph_am_movie.gexf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10334\n",
      "9724\n"
     ]
    }
   ],
   "source": [
    "print(len(movie_id_list)+len(user_id_list))\n",
    "print(len(movie_id_list))"
   ]
  }
 ],
 "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": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}