{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Classical Music Recommendation Playground\n",
    "\n",
    "This notebook will show how to implement simple recommender system follwing two different approaches: **Collaborative Filtering** (user based) and **Content Based** recommendation.\n",
    "\n",
    "> DISCLAIMER:\n",
    "> The used dataset is NOT a real dataset, but it has been artificially generated for the Tutorial purposes.\n",
    "> It absolutely should NOT be used as training data for any application."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import scipy.spatial.distance as distance"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data import\n",
    "\n",
    "### data.csv\n",
    "\n",
    "User listening experience dataset. 100 **users** -- numeric identifiers from 0 to 99 -- interact (or not) with 100 **items** (classical composers).\n",
    "\n",
    "1 = interaction, 0 = no interaction (implicit feedback)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Wolfgang Amadeus Mozart</th>\n",
       "      <th>Franz Liszt</th>\n",
       "      <th>Joseph Haydn</th>\n",
       "      <th>Johannes Brahms</th>\n",
       "      <th>Robert Schumann</th>\n",
       "      <th>Antonio Vivaldi</th>\n",
       "      <th>Roland de Lassus</th>\n",
       "      <th>Frédéric Chopin</th>\n",
       "      <th>Franz Schubert</th>\n",
       "      <th>Domenico Scarlatti</th>\n",
       "      <th>...</th>\n",
       "      <th>Arnold Schoenberg</th>\n",
       "      <th>Bruno Mantovani</th>\n",
       "      <th>Antonín Dvořák</th>\n",
       "      <th>Piotr Ilitch Tchaïkovski</th>\n",
       "      <th>Johann Christian Bach</th>\n",
       "      <th>Aaron Copland</th>\n",
       "      <th>Ferruccio Busoni</th>\n",
       "      <th>Ralph Vaughan Williams</th>\n",
       "      <th>Zoltán Kodály</th>\n",
       "      <th>Leonard Bernstein</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 100 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Wolfgang Amadeus Mozart  Franz Liszt  Joseph Haydn  Johannes Brahms  \\\n",
       "0                        1            1             0                0   \n",
       "1                        1            1             1                1   \n",
       "2                        0            1             0                0   \n",
       "3                        0            0             0                0   \n",
       "4                        0            1             0                1   \n",
       "\n",
       "   Robert Schumann  Antonio Vivaldi  Roland de Lassus  Frédéric Chopin  \\\n",
       "0                1                1                 0                0   \n",
       "1                1                1                 1                1   \n",
       "2                0                0                 0                1   \n",
       "3                0                0                 0                1   \n",
       "4                0                1                 0                0   \n",
       "\n",
       "   Franz Schubert  Domenico Scarlatti        ...          Arnold Schoenberg  \\\n",
       "0               0                   1        ...                          0   \n",
       "1               1                   1        ...                          1   \n",
       "2               1                   0        ...                          0   \n",
       "3               0                   0        ...                          0   \n",
       "4               0                   1        ...                          0   \n",
       "\n",
       "   Bruno Mantovani  Antonín Dvořák  Piotr Ilitch Tchaïkovski  \\\n",
       "0                0               0                         0   \n",
       "1                0               1                         1   \n",
       "2                0               0                         0   \n",
       "3                0               1                         0   \n",
       "4                0               1                         0   \n",
       "\n",
       "   Johann Christian Bach  Aaron Copland  Ferruccio Busoni  \\\n",
       "0                      1              0                 0   \n",
       "1                      1              1                 0   \n",
       "2                      0              0                 0   \n",
       "3                      0              0                 0   \n",
       "4                      0              0                 0   \n",
       "\n",
       "   Ralph Vaughan Williams  Zoltán Kodály  Leonard Bernstein  \n",
       "0                       0              0                  0  \n",
       "1                       0              1                  1  \n",
       "2                       0              0                  1  \n",
       "3                       0              0                  0  \n",
       "4                       0              0                  1  \n",
       "\n",
       "[5 rows x 100 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv('data.csv', index_col=0)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here a (quite-predictable) list of the top 10 most popular composer."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Wolfgang Amadeus Mozart    89\n",
       "Ludwig van Beethoven       83\n",
       "Johann Strauss             80\n",
       "Antonio Vivaldi            79\n",
       "Johann Sebastian Bach      76\n",
       "Joseph Haydn               73\n",
       "Georg Friedrich Haendel    71\n",
       "Franz Liszt                71\n",
       "Maurice Ravel              71\n",
       "Giuseppe Verdi             70\n",
       "dtype: int64"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.sum().sort_values(ascending=False)[0:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### artists.csv\n",
    "\n",
    "The 100 involved artists with label, uris and 17 embedding dimension coming from the [music embeddings repo](https://github.com/DOREMUS-ANR/music-embeddings)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>label</th>\n",
       "      <th>0</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>11</th>\n",
       "      <th>12</th>\n",
       "      <th>13</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "      <th>16</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>uri</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>http://data.doremus.org/artist/4802a043-23bb-3b8d-a443-4a3bd22ccc63</th>\n",
       "      <td>Wolfgang Amadeus Mozart</td>\n",
       "      <td>-0.049424</td>\n",
       "      <td>0.012972</td>\n",
       "      <td>0.030435</td>\n",
       "      <td>0.672381</td>\n",
       "      <td>0.705714</td>\n",
       "      <td>-0.003292</td>\n",
       "      <td>0.040397</td>\n",
       "      <td>0.033374</td>\n",
       "      <td>0.023954</td>\n",
       "      <td>-0.085231</td>\n",
       "      <td>0.158234</td>\n",
       "      <td>0.044664</td>\n",
       "      <td>0.018166</td>\n",
       "      <td>0.010425</td>\n",
       "      <td>-0.111956</td>\n",
       "      <td>0.142959</td>\n",
       "      <td>-0.030154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>http://data.doremus.org/artist/aabcd2ee-ac9b-30f2-8096-e9de8b3c7a81</th>\n",
       "      <td>Franz Liszt</td>\n",
       "      <td>0.000628</td>\n",
       "      <td>-0.008797</td>\n",
       "      <td>-0.007513</td>\n",
       "      <td>0.724762</td>\n",
       "      <td>0.796191</td>\n",
       "      <td>0.010984</td>\n",
       "      <td>0.017313</td>\n",
       "      <td>0.072120</td>\n",
       "      <td>0.025766</td>\n",
       "      <td>-0.085152</td>\n",
       "      <td>0.156424</td>\n",
       "      <td>0.043058</td>\n",
       "      <td>0.020398</td>\n",
       "      <td>0.032555</td>\n",
       "      <td>-0.048986</td>\n",
       "      <td>0.033035</td>\n",
       "      <td>-0.001050</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>http://data.doremus.org/artist/12fa21ff-cfa4-31d6-87d9-a22315193b04</th>\n",
       "      <td>Joseph Haydn</td>\n",
       "      <td>-2.000000</td>\n",
       "      <td>-2.000000</td>\n",
       "      <td>-2.000000</td>\n",
       "      <td>0.649524</td>\n",
       "      <td>0.722857</td>\n",
       "      <td>-0.001422</td>\n",
       "      <td>0.035583</td>\n",
       "      <td>0.029520</td>\n",
       "      <td>0.023965</td>\n",
       "      <td>-0.085292</td>\n",
       "      <td>0.158400</td>\n",
       "      <td>0.042192</td>\n",
       "      <td>0.008899</td>\n",
       "      <td>0.007677</td>\n",
       "      <td>-0.091435</td>\n",
       "      <td>0.147538</td>\n",
       "      <td>0.000403</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>http://data.doremus.org/artist/f9a2ac39-a62d-3be2-8abb-e564de0ec96d</th>\n",
       "      <td>Johannes Brahms</td>\n",
       "      <td>-0.010495</td>\n",
       "      <td>-0.003960</td>\n",
       "      <td>0.000920</td>\n",
       "      <td>0.745714</td>\n",
       "      <td>0.806667</td>\n",
       "      <td>0.003321</td>\n",
       "      <td>0.019724</td>\n",
       "      <td>0.059037</td>\n",
       "      <td>0.024170</td>\n",
       "      <td>-0.085409</td>\n",
       "      <td>0.158257</td>\n",
       "      <td>0.044797</td>\n",
       "      <td>0.028943</td>\n",
       "      <td>0.022953</td>\n",
       "      <td>-0.063101</td>\n",
       "      <td>0.111707</td>\n",
       "      <td>-0.025850</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>http://data.doremus.org/artist/f753314d-87a7-32a9-9218-da98ae4f9812</th>\n",
       "      <td>Robert Schumann</td>\n",
       "      <td>0.000628</td>\n",
       "      <td>-0.008797</td>\n",
       "      <td>-0.007513</td>\n",
       "      <td>0.723810</td>\n",
       "      <td>0.767619</td>\n",
       "      <td>0.003386</td>\n",
       "      <td>0.021836</td>\n",
       "      <td>0.059332</td>\n",
       "      <td>0.023935</td>\n",
       "      <td>-0.084914</td>\n",
       "      <td>0.158104</td>\n",
       "      <td>0.045242</td>\n",
       "      <td>0.030560</td>\n",
       "      <td>0.023502</td>\n",
       "      <td>-0.129961</td>\n",
       "      <td>0.110820</td>\n",
       "      <td>-0.069069</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                      label  \\\n",
       "uri                                                                           \n",
       "http://data.doremus.org/artist/4802a043-23bb-3b...  Wolfgang Amadeus Mozart   \n",
       "http://data.doremus.org/artist/aabcd2ee-ac9b-30...              Franz Liszt   \n",
       "http://data.doremus.org/artist/12fa21ff-cfa4-31...             Joseph Haydn   \n",
       "http://data.doremus.org/artist/f9a2ac39-a62d-3b...          Johannes Brahms   \n",
       "http://data.doremus.org/artist/f753314d-87a7-32...          Robert Schumann   \n",
       "\n",
       "                                                           0         1  \\\n",
       "uri                                                                      \n",
       "http://data.doremus.org/artist/4802a043-23bb-3b... -0.049424  0.012972   \n",
       "http://data.doremus.org/artist/aabcd2ee-ac9b-30...  0.000628 -0.008797   \n",
       "http://data.doremus.org/artist/12fa21ff-cfa4-31... -2.000000 -2.000000   \n",
       "http://data.doremus.org/artist/f9a2ac39-a62d-3b... -0.010495 -0.003960   \n",
       "http://data.doremus.org/artist/f753314d-87a7-32...  0.000628 -0.008797   \n",
       "\n",
       "                                                           2         3  \\\n",
       "uri                                                                      \n",
       "http://data.doremus.org/artist/4802a043-23bb-3b...  0.030435  0.672381   \n",
       "http://data.doremus.org/artist/aabcd2ee-ac9b-30... -0.007513  0.724762   \n",
       "http://data.doremus.org/artist/12fa21ff-cfa4-31... -2.000000  0.649524   \n",
       "http://data.doremus.org/artist/f9a2ac39-a62d-3b...  0.000920  0.745714   \n",
       "http://data.doremus.org/artist/f753314d-87a7-32... -0.007513  0.723810   \n",
       "\n",
       "                                                           4         5  \\\n",
       "uri                                                                      \n",
       "http://data.doremus.org/artist/4802a043-23bb-3b...  0.705714 -0.003292   \n",
       "http://data.doremus.org/artist/aabcd2ee-ac9b-30...  0.796191  0.010984   \n",
       "http://data.doremus.org/artist/12fa21ff-cfa4-31...  0.722857 -0.001422   \n",
       "http://data.doremus.org/artist/f9a2ac39-a62d-3b...  0.806667  0.003321   \n",
       "http://data.doremus.org/artist/f753314d-87a7-32...  0.767619  0.003386   \n",
       "\n",
       "                                                           6         7  \\\n",
       "uri                                                                      \n",
       "http://data.doremus.org/artist/4802a043-23bb-3b...  0.040397  0.033374   \n",
       "http://data.doremus.org/artist/aabcd2ee-ac9b-30...  0.017313  0.072120   \n",
       "http://data.doremus.org/artist/12fa21ff-cfa4-31...  0.035583  0.029520   \n",
       "http://data.doremus.org/artist/f9a2ac39-a62d-3b...  0.019724  0.059037   \n",
       "http://data.doremus.org/artist/f753314d-87a7-32...  0.021836  0.059332   \n",
       "\n",
       "                                                           8         9  \\\n",
       "uri                                                                      \n",
       "http://data.doremus.org/artist/4802a043-23bb-3b...  0.023954 -0.085231   \n",
       "http://data.doremus.org/artist/aabcd2ee-ac9b-30...  0.025766 -0.085152   \n",
       "http://data.doremus.org/artist/12fa21ff-cfa4-31...  0.023965 -0.085292   \n",
       "http://data.doremus.org/artist/f9a2ac39-a62d-3b...  0.024170 -0.085409   \n",
       "http://data.doremus.org/artist/f753314d-87a7-32...  0.023935 -0.084914   \n",
       "\n",
       "                                                          10        11  \\\n",
       "uri                                                                      \n",
       "http://data.doremus.org/artist/4802a043-23bb-3b...  0.158234  0.044664   \n",
       "http://data.doremus.org/artist/aabcd2ee-ac9b-30...  0.156424  0.043058   \n",
       "http://data.doremus.org/artist/12fa21ff-cfa4-31...  0.158400  0.042192   \n",
       "http://data.doremus.org/artist/f9a2ac39-a62d-3b...  0.158257  0.044797   \n",
       "http://data.doremus.org/artist/f753314d-87a7-32...  0.158104  0.045242   \n",
       "\n",
       "                                                          12        13  \\\n",
       "uri                                                                      \n",
       "http://data.doremus.org/artist/4802a043-23bb-3b...  0.018166  0.010425   \n",
       "http://data.doremus.org/artist/aabcd2ee-ac9b-30...  0.020398  0.032555   \n",
       "http://data.doremus.org/artist/12fa21ff-cfa4-31...  0.008899  0.007677   \n",
       "http://data.doremus.org/artist/f9a2ac39-a62d-3b...  0.028943  0.022953   \n",
       "http://data.doremus.org/artist/f753314d-87a7-32...  0.030560  0.023502   \n",
       "\n",
       "                                                          14        15  \\\n",
       "uri                                                                      \n",
       "http://data.doremus.org/artist/4802a043-23bb-3b... -0.111956  0.142959   \n",
       "http://data.doremus.org/artist/aabcd2ee-ac9b-30... -0.048986  0.033035   \n",
       "http://data.doremus.org/artist/12fa21ff-cfa4-31... -0.091435  0.147538   \n",
       "http://data.doremus.org/artist/f9a2ac39-a62d-3b... -0.063101  0.111707   \n",
       "http://data.doremus.org/artist/f753314d-87a7-32... -0.129961  0.110820   \n",
       "\n",
       "                                                          16  \n",
       "uri                                                           \n",
       "http://data.doremus.org/artist/4802a043-23bb-3b... -0.030154  \n",
       "http://data.doremus.org/artist/aabcd2ee-ac9b-30... -0.001050  \n",
       "http://data.doremus.org/artist/12fa21ff-cfa4-31...  0.000403  \n",
       "http://data.doremus.org/artist/f9a2ac39-a62d-3b... -0.025850  \n",
       "http://data.doremus.org/artist/f753314d-87a7-32... -0.069069  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "artists = pd.read_csv('artists.csv', index_col=0)\n",
    "artists.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Utils"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Wolfgang Amadeus Mozart',\n",
       " 'Joseph Haydn',\n",
       " 'Robert Schumann',\n",
       " 'Antonio Vivaldi',\n",
       " 'Franz Schubert',\n",
       " 'Georg Philipp Telemann',\n",
       " 'Ludwig van Beethoven',\n",
       " 'Alessandro Scarlatti',\n",
       " 'Benjamin Britten',\n",
       " 'Johann Sebastian Bach',\n",
       " 'Richard Wagner',\n",
       " 'Luigi Cherubini',\n",
       " 'Giuseppe Verdi',\n",
       " 'Johann Strauss',\n",
       " 'Niccolò Paganini',\n",
       " 'Gaetano Donizetti',\n",
       " 'Jean-Baptiste Lully',\n",
       " 'George Gershwin',\n",
       " 'Piotr Ilitch Tchaïkovski']"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# returns the list of items which a given user has interacted to\n",
    "def get_items(id):\n",
    "    user = data.loc[id]\n",
    "    return user[user[:] == 1].axes[0].tolist()\n",
    "\n",
    "# example user 7\n",
    "get_items(7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "masked_array(data=[-0.049423877149820335, 0.012972225435078144,\n",
       "                   0.030434519052505493, 0.6723809242248535,\n",
       "                   0.7057142853736877, -0.003291688393801451,\n",
       "                   0.04039749875664711, 0.03337432071566582,\n",
       "                   0.023954134434461597, -0.08523057401180267,\n",
       "                   0.1582336723804474, 0.04466380551457405,\n",
       "                   0.018166353926062584, 0.01042507402598858,\n",
       "                   -0.11195577681064606, 0.14295919239521027,\n",
       "                   -0.030153987929224968],\n",
       "             mask=[False, False, False, False, False, False, False, False,\n",
       "                   False, False, False, False, False, False, False, False,\n",
       "                   False],\n",
       "       fill_value=1e+20)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# retrieve the embeddings for an artists given a label\n",
    "def get_emb(label):\n",
    "    a = artists.loc[artists['label'] == label]\n",
    "    embs =  a.drop('label', axis=1).values[0]\n",
    "    \n",
    "    return np.ma.array(embs, mask=embs == -2.)\n",
    "\n",
    "    \n",
    "# example\n",
    "get_emb('Wolfgang Amadeus Mozart')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Collaborative filtering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Find most similar users to the given one\n",
    "def most_similar_users(user, k):\n",
    "    user_vec = user.values # user listening\n",
    "    \n",
    "     # search among all the other users\n",
    "    pool = data.drop(user.name)\n",
    "    \n",
    "    # apply the cosine distance to each element of the pool, and sort accordingly\n",
    "    pool['distance'] = data.apply(lambda u: distance.cosine(user_vec, u.values), axis=1)\n",
    "    pool = pool.sort_values('distance').drop('distance', 1)\n",
    "    \n",
    "    # return the first k users\n",
    "    return pool[:k]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Select the most popular artists among a subset of users\n",
    "def most_popular_among(user_subset, k=10):\n",
    "    return user_subset.sum().sort_values(ascending=False).index.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Recommend artists by looking at similar users\n",
    "def collaborative_filtering(user, k=10):\n",
    "    _user = data.loc[user]\n",
    "    # find k most similar users\n",
    "    similar_users = most_similar_users(_user, k)\n",
    "    # get k closest items\n",
    "    most_popular = most_popular_among(similar_users)\n",
    "    # remove the ones already in the list\n",
    "    prediction = [x for x in most_popular if x not in get_items(user)]\n",
    "    return prediction[:k]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Which artist would be recommended to our user? He already listened these ones:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Wolfgang Amadeus Mozart',\n",
       " 'Frédéric Chopin',\n",
       " 'Franz Schubert',\n",
       " 'Ludwig van Beethoven',\n",
       " 'Carl Philipp Emanuel Bach',\n",
       " 'Richard Strauss',\n",
       " 'Francis Poulenc',\n",
       " 'Maurice Ravel',\n",
       " 'Felix Mendelssohn Bartholdy',\n",
       " 'Giuseppe Verdi',\n",
       " 'Igor Stravinsky',\n",
       " 'Georg Friedrich Haendel']"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "user_example = 8\n",
    "get_items(user_example)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The recommendation proposes other Germans composers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Johann Strauss',\n",
       " 'Franz Liszt',\n",
       " 'Johannes Brahms',\n",
       " 'Richard Wagner',\n",
       " 'Claude Debussy',\n",
       " 'Johann Sebastian Bach',\n",
       " 'Gustav Mahler',\n",
       " 'Luigi Cherubini',\n",
       " 'Jean-Baptiste Lully',\n",
       " 'Antonio Vivaldi']"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "collaborative_filtering(user_example)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Content based recommendation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Define similarity metric"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def compute_similarity(seed, target, w=1):\n",
    "    b1 = np.where(seed.mask==True)[0]\n",
    "    b2 = np.where(target.mask==True)[0]\n",
    "    bad_pos = np.unique(np.concatenate([b1, b2]))\n",
    "\n",
    "    _seed = np.delete(seed, bad_pos, axis=0)\n",
    "    _target = np.delete(target, bad_pos, axis=0)\n",
    "    _w = np.delete(w, bad_pos, axis=0)\n",
    "\n",
    "    if len(_seed) == 0:\n",
    "        return 0\n",
    "\n",
    "    # distance\n",
    "    d = weighted_l2(_seed, _target, _w)\n",
    "\n",
    "    # how much info I am not finding\n",
    "    penalty = len([x for x in b2 if x not in b1]) / len(seed)\n",
    "\n",
    "    # score\n",
    "    s = (max_distance - d) / max_distance\n",
    "    return s * (1 - penalty)\n",
    "\n",
    "\n",
    "def weighted_l2(a, b, w=1):\n",
    "    q = a - b\n",
    "    return np.sqrt((w * q * q).sum())\n",
    "\n",
    "\n",
    "_ones = np.ones(17)\n",
    "max_distance = weighted_l2(_ones,-_ones, _ones)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compute all the similarity scores between couple of artists and put them in a Data Frame."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Library/Frameworks/Python.framework/Versions/3.6/lib/python3.6/site-packages/ipykernel_launcher.py:8: DeprecationWarning: in the future the special handling of scalars will be removed from delete and raise an error\n",
      "  \n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>label</th>\n",
       "      <th>Wolfgang Amadeus Mozart</th>\n",
       "      <th>Franz Liszt</th>\n",
       "      <th>Joseph Haydn</th>\n",
       "      <th>Johannes Brahms</th>\n",
       "      <th>Robert Schumann</th>\n",
       "      <th>Antonio Vivaldi</th>\n",
       "      <th>Roland de Lassus</th>\n",
       "      <th>Frédéric Chopin</th>\n",
       "      <th>Franz Schubert</th>\n",
       "      <th>Domenico Scarlatti</th>\n",
       "      <th>...</th>\n",
       "      <th>Arnold Schoenberg</th>\n",
       "      <th>Bruno Mantovani</th>\n",
       "      <th>Antonín Dvořák</th>\n",
       "      <th>Piotr Ilitch Tchaïkovski</th>\n",
       "      <th>Johann Christian Bach</th>\n",
       "      <th>Aaron Copland</th>\n",
       "      <th>Ferruccio Busoni</th>\n",
       "      <th>Ralph Vaughan Williams</th>\n",
       "      <th>Zoltán Kodály</th>\n",
       "      <th>Leonard Bernstein</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>label</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>Wolfgang Amadeus Mozart</th>\n",
       "      <td>1</td>\n",
       "      <td>0.977317</td>\n",
       "      <td>0.818705</td>\n",
       "      <td>0.981598</td>\n",
       "      <td>0.985071</td>\n",
       "      <td>0.814204</td>\n",
       "      <td>0.624089</td>\n",
       "      <td>0.976958</td>\n",
       "      <td>0.987308</td>\n",
       "      <td>0.813783</td>\n",
       "      <td>...</td>\n",
       "      <td>0.803884</td>\n",
       "      <td>0.573075</td>\n",
       "      <td>0.808275</td>\n",
       "      <td>0.808963</td>\n",
       "      <td>0.818815</td>\n",
       "      <td>0.628535</td>\n",
       "      <td>0.805652</td>\n",
       "      <td>0.80302</td>\n",
       "      <td>0.797534</td>\n",
       "      <td>0.627795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Franz Liszt</th>\n",
       "      <td>0.977317</td>\n",
       "      <td>1</td>\n",
       "      <td>0.806518</td>\n",
       "      <td>0.989032</td>\n",
       "      <td>0.983511</td>\n",
       "      <td>0.798843</td>\n",
       "      <td>0.616185</td>\n",
       "      <td>0.98243</td>\n",
       "      <td>0.979911</td>\n",
       "      <td>0.800846</td>\n",
       "      <td>...</td>\n",
       "      <td>0.807869</td>\n",
       "      <td>0.576931</td>\n",
       "      <td>0.813884</td>\n",
       "      <td>0.813808</td>\n",
       "      <td>0.805256</td>\n",
       "      <td>0.6364</td>\n",
       "      <td>0.807672</td>\n",
       "      <td>0.80757</td>\n",
       "      <td>0.798923</td>\n",
       "      <td>0.635377</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Joseph Haydn</th>\n",
       "      <td>0.994142</td>\n",
       "      <td>0.979343</td>\n",
       "      <td>1</td>\n",
       "      <td>0.982494</td>\n",
       "      <td>0.984184</td>\n",
       "      <td>0.988096</td>\n",
       "      <td>0.794793</td>\n",
       "      <td>0.977726</td>\n",
       "      <td>0.988863</td>\n",
       "      <td>0.990187</td>\n",
       "      <td>...</td>\n",
       "      <td>0.975743</td>\n",
       "      <td>0.742932</td>\n",
       "      <td>0.981111</td>\n",
       "      <td>0.982625</td>\n",
       "      <td>0.993293</td>\n",
       "      <td>0.799882</td>\n",
       "      <td>0.976397</td>\n",
       "      <td>0.974162</td>\n",
       "      <td>0.970276</td>\n",
       "      <td>0.798628</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Johannes Brahms</th>\n",
       "      <td>0.981598</td>\n",
       "      <td>0.989032</td>\n",
       "      <td>0.809113</td>\n",
       "      <td>1</td>\n",
       "      <td>0.988774</td>\n",
       "      <td>0.800753</td>\n",
       "      <td>0.614646</td>\n",
       "      <td>0.979436</td>\n",
       "      <td>0.982709</td>\n",
       "      <td>0.802281</td>\n",
       "      <td>...</td>\n",
       "      <td>0.815153</td>\n",
       "      <td>0.578398</td>\n",
       "      <td>0.818782</td>\n",
       "      <td>0.817388</td>\n",
       "      <td>0.807047</td>\n",
       "      <td>0.638088</td>\n",
       "      <td>0.81323</td>\n",
       "      <td>0.811394</td>\n",
       "      <td>0.803853</td>\n",
       "      <td>0.637382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Robert Schumann</th>\n",
       "      <td>0.985071</td>\n",
       "      <td>0.983511</td>\n",
       "      <td>0.810504</td>\n",
       "      <td>0.988774</td>\n",
       "      <td>1</td>\n",
       "      <td>0.804982</td>\n",
       "      <td>0.617933</td>\n",
       "      <td>0.976734</td>\n",
       "      <td>0.981611</td>\n",
       "      <td>0.804426</td>\n",
       "      <td>...</td>\n",
       "      <td>0.810159</td>\n",
       "      <td>0.576856</td>\n",
       "      <td>0.813493</td>\n",
       "      <td>0.812816</td>\n",
       "      <td>0.810574</td>\n",
       "      <td>0.634682</td>\n",
       "      <td>0.814013</td>\n",
       "      <td>0.809981</td>\n",
       "      <td>0.798908</td>\n",
       "      <td>0.634023</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 100 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "label                   Wolfgang Amadeus Mozart Franz Liszt Joseph Haydn  \\\n",
       "label                                                                      \n",
       "Wolfgang Amadeus Mozart                       1    0.977317     0.818705   \n",
       "Franz Liszt                            0.977317           1     0.806518   \n",
       "Joseph Haydn                           0.994142    0.979343            1   \n",
       "Johannes Brahms                        0.981598    0.989032     0.809113   \n",
       "Robert Schumann                        0.985071    0.983511     0.810504   \n",
       "\n",
       "label                   Johannes Brahms Robert Schumann Antonio Vivaldi  \\\n",
       "label                                                                     \n",
       "Wolfgang Amadeus Mozart        0.981598        0.985071        0.814204   \n",
       "Franz Liszt                    0.989032        0.983511        0.798843   \n",
       "Joseph Haydn                   0.982494        0.984184        0.988096   \n",
       "Johannes Brahms                       1        0.988774        0.800753   \n",
       "Robert Schumann                0.988774               1        0.804982   \n",
       "\n",
       "label                   Roland de Lassus Frédéric Chopin Franz Schubert  \\\n",
       "label                                                                     \n",
       "Wolfgang Amadeus Mozart         0.624089        0.976958       0.987308   \n",
       "Franz Liszt                     0.616185         0.98243       0.979911   \n",
       "Joseph Haydn                    0.794793        0.977726       0.988863   \n",
       "Johannes Brahms                 0.614646        0.979436       0.982709   \n",
       "Robert Schumann                 0.617933        0.976734       0.981611   \n",
       "\n",
       "label                   Domenico Scarlatti        ...         \\\n",
       "label                                             ...          \n",
       "Wolfgang Amadeus Mozart           0.813783        ...          \n",
       "Franz Liszt                       0.800846        ...          \n",
       "Joseph Haydn                      0.990187        ...          \n",
       "Johannes Brahms                   0.802281        ...          \n",
       "Robert Schumann                   0.804426        ...          \n",
       "\n",
       "label                   Arnold Schoenberg Bruno Mantovani Antonín Dvořák  \\\n",
       "label                                                                      \n",
       "Wolfgang Amadeus Mozart          0.803884        0.573075       0.808275   \n",
       "Franz Liszt                      0.807869        0.576931       0.813884   \n",
       "Joseph Haydn                     0.975743        0.742932       0.981111   \n",
       "Johannes Brahms                  0.815153        0.578398       0.818782   \n",
       "Robert Schumann                  0.810159        0.576856       0.813493   \n",
       "\n",
       "label                   Piotr Ilitch Tchaïkovski Johann Christian Bach  \\\n",
       "label                                                                    \n",
       "Wolfgang Amadeus Mozart                 0.808963              0.818815   \n",
       "Franz Liszt                             0.813808              0.805256   \n",
       "Joseph Haydn                            0.982625              0.993293   \n",
       "Johannes Brahms                         0.817388              0.807047   \n",
       "Robert Schumann                         0.812816              0.810574   \n",
       "\n",
       "label                   Aaron Copland Ferruccio Busoni Ralph Vaughan Williams  \\\n",
       "label                                                                           \n",
       "Wolfgang Amadeus Mozart      0.628535         0.805652                0.80302   \n",
       "Franz Liszt                    0.6364         0.807672                0.80757   \n",
       "Joseph Haydn                 0.799882         0.976397               0.974162   \n",
       "Johannes Brahms              0.638088          0.81323               0.811394   \n",
       "Robert Schumann              0.634682         0.814013               0.809981   \n",
       "\n",
       "label                   Zoltán Kodály Leonard Bernstein  \n",
       "label                                                    \n",
       "Wolfgang Amadeus Mozart      0.797534          0.627795  \n",
       "Franz Liszt                  0.798923          0.635377  \n",
       "Joseph Haydn                 0.970276          0.798628  \n",
       "Johannes Brahms              0.803853          0.637382  \n",
       "Robert Schumann              0.798908          0.634023  \n",
       "\n",
       "[5 rows x 100 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "similarity_matrix = pd.DataFrame(index=artists['label'], columns=artists['label'])\n",
    "for i in np.arange(len(similarity_matrix)):\n",
    "    seed = artists.iloc[i]['label']\n",
    "    for j in np.arange(len(similarity_matrix)):\n",
    "        \n",
    "        if i == j:\n",
    "            similarity_matrix.iloc[i][j] = 1\n",
    "            continue\n",
    "        \n",
    "        target = artists.iloc[j]['label']\n",
    "        similarity_matrix.iloc[i][j] = compute_similarity(get_emb(seed), get_emb(target))\n",
    "        \n",
    "similarity_matrix.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def content_based(user, k=10):\n",
    "    _items = get_items(user)\n",
    "    \n",
    "    # remove the items already in the list\n",
    "    candidates = similarity_matrix.drop(labels=_items, axis=1)\n",
    "    \n",
    "    # choose the artists that maximise the similarity among all the items\n",
    "    candidates = candidates.loc[_items]\n",
    "    return candidates.sum().sort_values(ascending=False).index.tolist()[0:k]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Wolfgang Amadeus Mozart',\n",
       " 'Frédéric Chopin',\n",
       " 'Franz Schubert',\n",
       " 'Ludwig van Beethoven',\n",
       " 'Carl Philipp Emanuel Bach',\n",
       " 'Richard Strauss',\n",
       " 'Francis Poulenc',\n",
       " 'Maurice Ravel',\n",
       " 'Felix Mendelssohn Bartholdy',\n",
       " 'Giuseppe Verdi',\n",
       " 'Igor Stravinsky',\n",
       " 'Georg Friedrich Haendel']"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_items(user_example)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Johannes Brahms',\n",
       " 'Claude Debussy',\n",
       " 'Robert Schumann',\n",
       " 'Bedřich Smetana',\n",
       " 'César Franck',\n",
       " 'Jean Sibelius',\n",
       " 'Carl Maria von Weber',\n",
       " 'Gabriel Fauré',\n",
       " 'Edward Elgar',\n",
       " 'Franz Liszt']"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "content_based(user_example)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bonus \n",
    "\n",
    "What happens with a user that appreciate just a particular composer?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Wolfgang Amadeus Mozart</th>\n",
       "      <th>Franz Liszt</th>\n",
       "      <th>Joseph Haydn</th>\n",
       "      <th>Johannes Brahms</th>\n",
       "      <th>Robert Schumann</th>\n",
       "      <th>Antonio Vivaldi</th>\n",
       "      <th>Roland de Lassus</th>\n",
       "      <th>Frédéric Chopin</th>\n",
       "      <th>Franz Schubert</th>\n",
       "      <th>Domenico Scarlatti</th>\n",
       "      <th>...</th>\n",
       "      <th>Arnold Schoenberg</th>\n",
       "      <th>Bruno Mantovani</th>\n",
       "      <th>Antonín Dvořák</th>\n",
       "      <th>Piotr Ilitch Tchaïkovski</th>\n",
       "      <th>Johann Christian Bach</th>\n",
       "      <th>Aaron Copland</th>\n",
       "      <th>Ferruccio Busoni</th>\n",
       "      <th>Ralph Vaughan Williams</th>\n",
       "      <th>Zoltán Kodály</th>\n",
       "      <th>Leonard Bernstein</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>100</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>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>...</td>\n",
       "      <td>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",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 100 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Wolfgang Amadeus Mozart  Franz Liszt  Joseph Haydn  Johannes Brahms  \\\n",
       "100                      0.0          0.0           0.0              0.0   \n",
       "\n",
       "     Robert Schumann  Antonio Vivaldi  Roland de Lassus  Frédéric Chopin  \\\n",
       "100              0.0              1.0               0.0              0.0   \n",
       "\n",
       "     Franz Schubert  Domenico Scarlatti        ...          Arnold Schoenberg  \\\n",
       "100             0.0                 0.0        ...                        0.0   \n",
       "\n",
       "     Bruno Mantovani  Antonín Dvořák  Piotr Ilitch Tchaïkovski  \\\n",
       "100              0.0             0.0                       0.0   \n",
       "\n",
       "     Johann Christian Bach  Aaron Copland  Ferruccio Busoni  \\\n",
       "100                    0.0            0.0               0.0   \n",
       "\n",
       "     Ralph Vaughan Williams  Zoltán Kodály  Leonard Bernstein  \n",
       "100                     0.0            0.0                0.0  \n",
       "\n",
       "[1 rows x 100 columns]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "new_user = np.zeros(len(data.loc[0]))\n",
    "new_user_id = len(data)\n",
    "data.loc[new_user_id] = new_user\n",
    "data.loc[new_user_id]['Antonio Vivaldi'] = 1\n",
    "data.loc[[new_user_id]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Wolfgang Amadeus Mozart',\n",
       " 'Johann Strauss',\n",
       " 'Claudio Monteverdi',\n",
       " 'Joseph Haydn',\n",
       " 'Piotr Ilitch Tchaïkovski',\n",
       " 'Georg Philipp Telemann',\n",
       " 'Johannes Brahms',\n",
       " 'Johann Sebastian Bach',\n",
       " 'Niccolò Paganini',\n",
       " 'Franz Liszt']"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "collaborative_filtering(new_user_id)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Alessandro Scarlatti',\n",
       " 'Johann Sebastian Bach',\n",
       " 'Georg Friedrich Haendel',\n",
       " 'François Couperin',\n",
       " 'Henry Purcell',\n",
       " 'Carl Philipp Emanuel Bach',\n",
       " 'Domenico Scarlatti',\n",
       " 'Georg Philipp Telemann',\n",
       " 'Baldassare Galuppi',\n",
       " 'André Campra']"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "content_based(new_user_id)"
   ]
  },
  {
   "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}