Ejercicio_Sistemas_Recomendacion.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Recommendation Engine / System\n",
    "## Proyecto: Recomendador de Repositorios de Github\n",
    "### Usaremos Collaborative Filtering\n",
    "Artículo completo en www.aprendemachinelearning.com"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.006503Z",
     "start_time": "2019-08-24T11:17:09.176829Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.neighbors import NearestNeighbors\n",
    "import matplotlib.pyplot as plt\n",
    "import sklearn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.022820Z",
     "start_time": "2019-08-24T11:17:20.009253Z"
    }
   },
   "outputs": [],
   "source": [
    "df_users = pd.read_csv(\"users.csv\")\n",
    "df_repos = pd.read_csv(\"repos.csv\")\n",
    "df_ratings = pd.read_csv(\"ratings.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.039228Z",
     "start_time": "2019-08-24T11:17:20.026564Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userId</th>\n",
       "      <th>username</th>\n",
       "      <th>name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>iris9112</td>\n",
       "      <td>Isabel Ruiz Buriticá</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>dianaclarke</td>\n",
       "      <td>Diana</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>nateprewitt</td>\n",
       "      <td>Nate Prewitt</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>oldani</td>\n",
       "      <td>Ordanis Sanchez</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>waflessnet</td>\n",
       "      <td>waflessnet</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   userId     username                  name\n",
       "0       1     iris9112  Isabel Ruiz Buriticá\n",
       "1       2  dianaclarke                 Diana\n",
       "2       3  nateprewitt          Nate Prewitt\n",
       "3       4       oldani       Ordanis Sanchez\n",
       "4       5   waflessnet            waflessnet"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_users.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.051589Z",
     "start_time": "2019-08-24T11:17:20.041546Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>repoId</th>\n",
       "      <th>title</th>\n",
       "      <th>categories</th>\n",
       "      <th>stars</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>airbnb / javascript</td>\n",
       "      <td>completar</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>kamranahmedse / developer-roadmap</td>\n",
       "      <td>Roadmap to becoming a web developer in 2019</td>\n",
       "      <td>85800.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>microsoft / vscode</td>\n",
       "      <td>Visual Studio Code</td>\n",
       "      <td>80855.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>torvalds / linux</td>\n",
       "      <td>Linux kernel source tree</td>\n",
       "      <td>78761.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>ytdl-org / youtube-dl</td>\n",
       "      <td>Command-line program to download videos from Y...</td>\n",
       "      <td>53909.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   repoId                              title  \\\n",
       "0       1                airbnb / javascript   \n",
       "1       2  kamranahmedse / developer-roadmap   \n",
       "2       3                 microsoft / vscode   \n",
       "3       4                   torvalds / linux   \n",
       "4       5              ytdl-org / youtube-dl   \n",
       "\n",
       "                                          categories    stars  \n",
       "0                                          completar      NaN  \n",
       "1        Roadmap to becoming a web developer in 2019  85800.0  \n",
       "2                                 Visual Studio Code  80855.0  \n",
       "3                           Linux kernel source tree  78761.0  \n",
       "4  Command-line program to download videos from Y...  53909.0  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_repos.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.063652Z",
     "start_time": "2019-08-24T11:17:20.054058Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>userId</th>\n",
       "      <th>repoId</th>\n",
       "      <th>rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   userId  repoId  rating\n",
       "0       1       1       2\n",
       "1       1       2       3\n",
       "2       1       3       4\n",
       "3       1       4       5\n",
       "4       1       5       3"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_ratings.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.071557Z",
     "start_time": "2019-08-24T11:17:20.065997Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "30 users\n",
      "167 items\n"
     ]
    }
   ],
   "source": [
    "n_users = df_ratings.userId.unique().shape[0]\n",
    "n_items = df_ratings.repoId.unique().shape[0]\n",
    "print (str(n_users) + ' users')\n",
    "print (str(n_items) + ' items')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.297814Z",
     "start_time": "2019-08-24T11:17:20.074222Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([94., 62., 66., 28., 40., 12., 14.,  8.]),\n",
       " array([1.   , 1.875, 2.75 , 3.625, 4.5  , 5.375, 6.25 , 7.125, 8.   ]),\n",
       " <a list of 8 Patch objects>)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(df_ratings.rating,bins=8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.309486Z",
     "start_time": "2019-08-24T11:17:20.300762Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "rating\n",
       "1    94\n",
       "2    62\n",
       "3    66\n",
       "4    28\n",
       "5    40\n",
       "6    12\n",
       "7    14\n",
       "8     8\n",
       "Name: userId, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_ratings.groupby([\"rating\"])[\"userId\"].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.476113Z",
     "start_time": "2019-08-24T11:17:20.312575Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([94., 31., 22.,  7.,  8.,  2.,  2.,  1.]),\n",
       " array([1.   , 1.875, 2.75 , 3.625, 4.5  , 5.375, 6.25 , 7.125, 8.   ]),\n",
       " <a list of 8 Patch objects>)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(df_ratings.groupby([\"repoId\"])[\"repoId\"].count(),bins=8)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Método 1: User-Based"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.546851Z",
     "start_time": "2019-08-24T11:17:20.479056Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>repoId</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>...</th>\n",
       "      <th>158</th>\n",
       "      <th>159</th>\n",
       "      <th>160</th>\n",
       "      <th>161</th>\n",
       "      <th>162</th>\n",
       "      <th>163</th>\n",
       "      <th>164</th>\n",
       "      <th>165</th>\n",
       "      <th>166</th>\n",
       "      <th>167</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>userId</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>30 rows × 167 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "repoId  1    2    3    4    5    6    7    8    9    10  ...   158  159  160  \\\n",
       "userId                                                   ...                   \n",
       "1       2.0  3.0  4.0  5.0  3.0  1.0  5.0  1.0  0.0  4.0 ...   0.0  0.0  0.0   \n",
       "2       0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "3       0.0  0.0  0.0  0.0  0.0  0.0  5.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "4       0.0  0.0  0.0  0.0  0.0  0.0  5.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "5       0.0  0.0  0.0  5.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "6       0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "7       0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "8       0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "9       0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "10      0.0  0.0  0.0  5.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "11      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "12      0.0  0.0  0.0  0.0  3.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "13      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "14      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "15      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "16      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "17      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "18      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "19      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "20      0.0  3.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "21      2.0  0.0  4.0  5.0  0.0  0.0  5.0  0.0  1.0  0.0 ...   0.0  0.0  0.0   \n",
       "22      0.0  3.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "23      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   1.0  0.0  0.0   \n",
       "24      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  4.0 ...   0.0  1.0  3.0   \n",
       "25      0.0  0.0  0.0  0.0  3.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  3.0   \n",
       "26      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "27      0.0  0.0  4.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  0.0   \n",
       "28      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0 ...   0.0  0.0  3.0   \n",
       "29      0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  0.0  4.0 ...   0.0  0.0  0.0   \n",
       "30      0.0  0.0  4.0  5.0  0.0  0.0  5.0  0.0  0.0  4.0 ...   0.0  0.0  0.0   \n",
       "\n",
       "repoId  161  162  163  164  165  166  167  \n",
       "userId                                     \n",
       "1       0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "2       0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "3       0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "4       0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "5       0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "6       0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "7       0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "8       0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "9       0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "10      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "11      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "12      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "13      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "14      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "15      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "16      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "17      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "18      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "19      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "20      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "21      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "22      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "23      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "24      1.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "25      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "26      0.0  2.0  1.0  0.0  0.0  0.0  0.0  \n",
       "27      0.0  0.0  0.0  0.0  0.0  0.0  0.0  \n",
       "28      0.0  0.0  0.0  1.0  0.0  0.0  0.0  \n",
       "29      0.0  0.0  0.0  0.0  1.0  1.0  0.0  \n",
       "30      0.0  2.0  0.0  0.0  0.0  0.0  1.0  \n",
       "\n",
       "[30 rows x 167 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_matrix = pd.pivot_table(df_ratings, values='rating', index='userId', columns='repoId').fillna(0)\n",
    "df_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.556202Z",
     "start_time": "2019-08-24T11:17:20.549811Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sparsity: 6.43%\n"
     ]
    }
   ],
   "source": [
    "ratings = df_matrix.values\n",
    "sparsity = float(len(ratings.nonzero()[0]))\n",
    "sparsity /= (ratings.shape[0] * ratings.shape[1])\n",
    "sparsity *= 100\n",
    "print('Sparsity: {:4.2f}%'.format(sparsity))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.563450Z",
     "start_time": "2019-08-24T11:17:20.559194Z"
    }
   },
   "outputs": [],
   "source": [
    "ratings_train, ratings_test = train_test_split(ratings, test_size = 0.2, shuffle=False, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.572537Z",
     "start_time": "2019-08-24T11:17:20.566581Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24, 167)"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ratings_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.581343Z",
     "start_time": "2019-08-24T11:17:20.575533Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(6, 167)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ratings_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<ul>\n",
    "<li>Matriz de similitud entre los usuarios (distancia del coseno -vectores-).</li>\n",
    "<li>Predecir la valoración desconocida de un ítem i para un usuario activo u basandonos en la suma ponderada de todas las valoraciones del resto de usuarios para dicho ítem.</li>\n",
    "<li>Recomendaremos los nuevos ítems a los usuarios según lo establecido en los pasos anteriores.</li>\n",
    "</ul>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.590966Z",
     "start_time": "2019-08-24T11:17:20.584226Z"
    }
   },
   "outputs": [],
   "source": [
    "sim_matrix = 1 - sklearn.metrics.pairwise.cosine_distances(ratings)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.606259Z",
     "start_time": "2019-08-24T11:17:20.594409Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(30, 30)"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sim_matrix.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.619099Z",
     "start_time": "2019-08-24T11:17:20.610504Z"
    }
   },
   "outputs": [],
   "source": [
    "# sim_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.961669Z",
     "start_time": "2019-08-24T11:17:20.626832Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams['figure.figsize'] = (20.0, 5.0)\n",
    "plt.imshow(sim_matrix);\n",
    "plt.colorbar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.973503Z",
     "start_time": "2019-08-24T11:17:20.967091Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(24, 24)\n",
      "(6, 6)\n"
     ]
    }
   ],
   "source": [
    "#separar las filas y columnas de train y test\n",
    "sim_matrix_train = sim_matrix[0:24,0:24]\n",
    "sim_matrix_test = sim_matrix[24:30,24:30]\n",
    "print(sim_matrix_train.shape)\n",
    "print(sim_matrix_test.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predicciones (las recomendaciones!)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.984488Z",
     "start_time": "2019-08-24T11:17:20.976334Z"
    }
   },
   "outputs": [],
   "source": [
    "users_predictions = sim_matrix_train.dot(ratings_train) / np.array([np.abs(sim_matrix_train).sum(axis=1)]).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:20.995708Z",
     "start_time": "2019-08-24T11:17:20.988423Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24, 167)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "users_predictions.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:21.552006Z",
     "start_time": "2019-08-24T11:17:21.001562Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.rcParams['figure.figsize'] = (20.0, 5.0)\n",
    "plt.imshow(users_predictions);\n",
    "plt.colorbar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Veamos un ejemplo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:21.573447Z",
     "start_time": "2019-08-24T11:17:21.555761Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4    ytdl-org / youtube-dl\n",
      "Name: title, dtype: object puntaje: 2.061791042203265\n",
      "84    dipanjanS / practical-machine-learning-with-py...\n",
      "Name: title, dtype: object puntaje: 2.4496166855518746\n",
      "99    abhat222 / Data-Science--Cheat-Sheet\n",
      "Name: title, dtype: object puntaje: 3.362878201910944\n"
     ]
    }
   ],
   "source": [
    "USUARIO_EJEMPLO = 'jbagnato' # debe existir en nuestro dataset de train!\n",
    "data = df_users[df_users['username'] == USUARIO_EJEMPLO]\n",
    "usuario_ver = data.iloc[0]['userId'] -1 # resta 1 para obtener el index de pandas\n",
    "user0=users_predictions.argsort()[usuario_ver]\n",
    "\n",
    "# Veamos los tres recomendados con mayor puntaje en la predic para este usuario\n",
    "for i, aRepo in enumerate(user0[-3:]):\n",
    "    selRepo = df_repos[df_repos['repoId']==(aRepo+1)]\n",
    "    print(selRepo['title'] , 'puntaje:', users_predictions[usuario_ver][aRepo])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Medir el Error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:21.586593Z",
     "start_time": "2019-08-24T11:17:21.576948Z"
    }
   },
   "outputs": [],
   "source": [
    "def get_mse(preds, actuals):\n",
    "    if preds.shape[1] != actuals.shape[1]:\n",
    "        actuals = actuals.T\n",
    "    preds = preds[actuals.nonzero()].flatten()\n",
    "    actuals = actuals[actuals.nonzero()].flatten()\n",
    "    return mean_squared_error(preds, actuals)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:21.598997Z",
     "start_time": "2019-08-24T11:17:21.590908Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3.3906497514974663"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_mse(users_predictions, ratings_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:21.617859Z",
     "start_time": "2019-08-24T11:17:21.602589Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4.722040854666231"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Realizo las predicciones para el test set\n",
    "users_predictions_test = sim_matrix.dot(ratings) / np.array([np.abs(sim_matrix).sum(axis=1)]).T\n",
    "users_predictions_test = users_predictions_test[24:30,:]\n",
    "\n",
    "get_mse(users_predictions_test, ratings_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Método 2: Usando Nearest Neighbors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:21.636356Z",
     "start_time": "2019-08-24T11:17:21.621408Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NearestNeighbors(algorithm='auto', leaf_size=30, metric='minkowski',\n",
       "                 metric_params=None, n_jobs=None, n_neighbors=8, p=2,\n",
       "                 radius='cosine')"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "k = 8\n",
    "neighbors = NearestNeighbors(k, 'cosine')\n",
    "neighbors.fit(ratings_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:21.653412Z",
     "start_time": "2019-08-24T11:17:21.639928Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24, 8)"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_k_distances, top_k_users = neighbors.kneighbors(ratings_train, return_distance=True)\n",
    "top_k_distances.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:21.664383Z",
     "start_time": "2019-08-24T11:17:21.656625Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.        ,  7.87400787,  8.66025404,  8.83176087,  9.2736185 ,\n",
       "       10.24695077, 10.63014581, 10.95445115])"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_k_distances[usuario_ver]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:21.678598Z",
     "start_time": "2019-08-24T11:17:21.668182Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24, 8)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_k_users.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:21.689376Z",
     "start_time": "2019-08-24T11:17:21.681953Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([11,  8, 15, 12,  4,  5, 19,  1])"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res = top_k_users[usuario_ver]\n",
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:18:54.878119Z",
     "start_time": "2019-08-24T11:18:54.846499Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    userId  username                  name\n",
      "11      12  jbagnato  Juan Ignacio Bagnato\n",
      "   userId      username         name\n",
      "8       9  camiloyate09  Camilo Yate\n",
      "    userId         username           name\n",
      "15      16  Armando115Tapia  Armando Tapia\n",
      "    userId         username                                name\n",
      "12      13  otromaximiliano  Lindolfo Maximiliano Ayala Aguirre\n",
      "   userId    username        name\n",
      "4       5  waflessnet  waflessnet\n",
      "   userId   username                name\n",
      "5       6  luisbalru  Luis Balderas Ruiz\n",
      "    userId    username               name\n",
      "19      20  diazandr3s  Andres Diaz-Pinto\n",
      "   userId     username   name\n",
      "1       2  dianaclarke  Diana\n"
     ]
    }
   ],
   "source": [
    "for a_index in res:\n",
    "    print(df_users[df_users['userId'] == (a_index+1)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:21.775667Z",
     "start_time": "2019-08-24T11:17:21.766733Z"
    }
   },
   "outputs": [],
   "source": [
    "users_predicts_k = np.zeros(ratings_train.shape)\n",
    "for i in range(ratings_train.shape[0]):# para cada usuario del conjunto de entrenamiento\n",
    "    users_predicts_k[i,:] = top_k_distances[i].T.dot(ratings_train[top_k_users][i]) / np.array([np.abs(top_k_distances[i].T).sum(axis=0)]).T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:21.789460Z",
     "start_time": "2019-08-24T11:17:21.782373Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24, 167)"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "users_predicts_k.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.186864Z",
     "start_time": "2019-08-24T11:17:21.792814Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(users_predicts_k);\n",
    "plt.colorbar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.217999Z",
     "start_time": "2019-08-24T11:17:22.190216Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3    torvalds / linux\n",
      "Name: title, dtype: object puntaje: 0.6975667679985449\n",
      "31    freeCodeCamp / freeCodeCamp\n",
      "Name: title, dtype: object puntaje: 0.9595266134630276\n",
      "61    jbagnato / machine-learning\n",
      "Name: title, dtype: object puntaje: 1.0790939117855096\n",
      "13    python / cpython\n",
      "Name: title, dtype: object puntaje: 1.3183999040179721\n"
     ]
    }
   ],
   "source": [
    "user0=users_predicts_k.argsort()[usuario_ver][-4:]\n",
    "# los tres con mayor puntaje en la predic para este usuario\n",
    "# agrego 4 puesto que uno es mi propio repositorio\n",
    "for aRepo in user0:\n",
    "    selRepo = df_repos[df_repos['repoId'] == (aRepo+1)]\n",
    "    print(selRepo['title'], 'puntaje:', users_predicts_k[usuario_ver][aRepo])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.231642Z",
     "start_time": "2019-08-24T11:17:22.221631Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8.124686270208018"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_mse(users_predicts_k, ratings_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.247970Z",
     "start_time": "2019-08-24T11:17:22.237464Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "13.755121956634824"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_mse(users_predicts_k, ratings_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Método 3: Collaborative Filtering Basado en Items"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.261199Z",
     "start_time": "2019-08-24T11:17:22.254134Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "167"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#basado en items\n",
    "n_repos = ratings_train.shape[1]\n",
    "n_repos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.275236Z",
     "start_time": "2019-08-24T11:17:22.265225Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "NearestNeighbors(algorithm='auto', leaf_size=30, metric='minkowski',\n",
       "                 metric_params=None, n_jobs=None, n_neighbors=167, p=2,\n",
       "                 radius='cosine')"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "neighbors = NearestNeighbors(n_repos, 'cosine')\n",
    "neighbors.fit(ratings_train.T)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.293004Z",
     "start_time": "2019-08-24T11:17:22.278397Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(167, 167)"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_k_distances, top_k_items = neighbors.kneighbors(ratings_train.T, return_distance=True)\n",
    "top_k_distances.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.304552Z",
     "start_time": "2019-08-24T11:17:22.296780Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[  0,  15,  18, ...,  62,  13,  61],\n",
       "       [  1, 144, 151, ...,  62,  13,  61],\n",
       "       [  2,   0,  18, ...,  62,  13,  61],\n",
       "       ...,\n",
       "       [164, 163, 162, ...,  62,  61,  13],\n",
       "       [164, 163, 162, ...,  62,  61,  13],\n",
       "       [164, 163, 162, ...,  62,  61,  13]])"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_k_items"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.319143Z",
     "start_time": "2019-08-24T11:17:22.308189Z"
    }
   },
   "outputs": [],
   "source": [
    "data = df_repos[df_repos['title'] == 'jbagnato / machine-learning']\n",
    "repo_ver = data.iloc[0]['repoId'] -1\n",
    "#print(repo_ver)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.341692Z",
     "start_time": "2019-08-24T11:17:22.322893Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>repoId</th>\n",
       "      <th>title</th>\n",
       "      <th>categories</th>\n",
       "      <th>stars</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>62</td>\n",
       "      <td>jbagnato / machine-learning</td>\n",
       "      <td>Código Python, Jupyter Notebooks, archivos csv...</td>\n",
       "      <td>48.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    repoId                        title  \\\n",
       "61      62  jbagnato / machine-learning   \n",
       "\n",
       "                                           categories  stars  \n",
       "61  Código Python, Jupyter Notebooks, archivos csv...   48.0  "
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_repos[df_repos['repoId'].isin([repo_ver+1])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.350637Z",
     "start_time": "2019-08-24T11:17:22.345468Z"
    }
   },
   "outputs": [],
   "source": [
    "item_preds = ratings_train.dot(top_k_distances) / np.array([np.abs(top_k_distances).sum(axis=1)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.860116Z",
     "start_time": "2019-08-24T11:17:22.353627Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(item_preds);\n",
    "plt.colorbar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.886744Z",
     "start_time": "2019-08-24T11:17:22.868897Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "164    nodejs / node\n",
      "Name: title, dtype: object puntaje: 0.7915561710178032\n",
      "165    OWASP / CheatSheetSeries\n",
      "Name: title, dtype: object puntaje: 0.836463821866165\n",
      "166    explosion / spaCy\n",
      "Name: title, dtype: object puntaje: 0.8728575365246009\n"
     ]
    }
   ],
   "source": [
    "user0=item_preds.argsort()[usuario_ver]\n",
    "user0\n",
    "\n",
    "# los tres con mayor puntaje en la predic para este usuario\n",
    "for aRepo in user0[-3:]:\n",
    "    selRepo = df_repos[df_repos['repoId']==(aRepo+1)]\n",
    "    print(selRepo['title'], 'puntaje:', item_preds[usuario_ver][aRepo])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.903769Z",
     "start_time": "2019-08-24T11:17:22.893219Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10.03834337791184"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_mse(item_preds, ratings_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:22.914196Z",
     "start_time": "2019-08-24T11:17:22.906636Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "13.482774653471974"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_mse(item_preds, ratings_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Método 4: recomendacion por Correlacion "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:23.018846Z",
     "start_time": "2019-08-24T11:17:22.917365Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>pearsonR</th>\n",
       "      <th>rating</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>repoId</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>0.341515</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>0.341515</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>95</th>\n",
       "      <td>0.341515</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>0.336601</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>138</th>\n",
       "      <td>0.336601</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>137</th>\n",
       "      <td>0.336601</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>140</th>\n",
       "      <td>0.336601</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>88</th>\n",
       "      <td>0.336601</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>0.336601</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        pearsonR  rating\n",
       "repoId                  \n",
       "62      1.000000       7\n",
       "87      0.341515       3\n",
       "92      0.341515       3\n",
       "95      0.341515       3\n",
       "139     0.336601       1\n",
       "138     0.336601       1\n",
       "137     0.336601       1\n",
       "140     0.336601       1\n",
       "88      0.336601       1\n",
       "67      0.336601       1"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "average_rating = pd.DataFrame(df_ratings.groupby('repoId')['rating'].mean())\n",
    "average_rating['ratingCount'] = pd.DataFrame(df_ratings.groupby('repoId')['rating'].count())\n",
    "average_rating.sort_values('ratingCount', ascending=False).head()\n",
    "\n",
    "mi_repo_ratings = df_matrix[62]\n",
    "similar_to_mine = df_matrix.corrwith(mi_repo_ratings)\n",
    "corr_mine = pd.DataFrame(similar_to_mine, columns=['pearsonR'])\n",
    "corr_mine.dropna(inplace=True)\n",
    "corr_summary = corr_mine.join(average_rating['rating'])\n",
    "corr_summary[corr_summary['rating']>=1].sort_values('pearsonR', ascending=False).head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:23.049417Z",
     "start_time": "2019-08-24T11:17:23.026199Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>repoId</th>\n",
       "      <th>title</th>\n",
       "      <th>categories</th>\n",
       "      <th>stars</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>92</td>\n",
       "      <td>joanby / python-ml-course</td>\n",
       "      <td>Curso de Introducción a Machine Learning con P...</td>\n",
       "      <td>156.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    repoId                      title  \\\n",
       "91      92  joanby / python-ml-course   \n",
       "\n",
       "                                           categories  stars  \n",
       "91  Curso de Introducción a Machine Learning con P...  156.0  "
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Veamos uno de los recomendados\n",
    "df_repos[df_repos['repoId'] == 92]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Método 5: Repo mas popular -sin collaborative filtering-"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Esta es la manera básica y sin uso de machine learning de ofrecer recomendaciones. <br>\n",
    "estas no serán personalizadas, serán iguales para cualquier usuario"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-24T11:17:23.407502Z",
     "start_time": "2019-08-24T11:17:23.052803Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "popular_repo = df_repos[['title','stars']].groupby('stars').sum().reset_index()\n",
    "popular_repo_top_20 = popular_repo.sort_values('stars', ascending=False).head(n=10)\n",
    "\n",
    "plt.rcdefaults()\n",
    " \n",
    "objects = (list(popular_repo_top_20['title']))\n",
    "y_pos = np.arange(len(objects))\n",
    "performance = list(popular_repo_top_20['stars'])\n",
    " \n",
    "plt.bar(y_pos, performance, align='center', alpha=0.5)\n",
    "plt.xticks(y_pos, objects, rotation='vertical')\n",
    "plt.ylabel('Stars count')\n",
    "plt.title('Most popular repo')\n",
    " \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}