Song_recommendations_with_genre1.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Song recommendations with genre.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyMzObjGP2Cxt9O8dbZgAQTT",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/sonichandni/Song-Recommendations/blob/master/Song_recommendations_with_genre1.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jX0NqcbRmFlH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "%tensorflow_version 2.x"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pPief-NzmaMR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import random\n",
        "from tqdm import tqdm\n",
        "from gensim.models import Word2Vec \n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "\n",
        "import warnings;\n",
        "warnings.filterwarnings('ignore')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "npKOsaSomkAd",
        "colab_type": "code",
        "outputId": "6fd4f357-a07b-4067-fea3-eb77087f907e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 393
        }
      },
      "source": [
        "!wget https://raw.githubusercontent.com/sonichandni/Song-Recommendations/master/songs.csv\n",
        "df = pd.read_csv('songs.csv',  encoding = \"ISO-8859-1\")\n",
        "df.head()"
      ],
      "execution_count": 138,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2020-03-16 10:45:49--  https://raw.githubusercontent.com/sonichandni/Song-Recommendations/master/songs.csv\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 202933 (198K) [text/plain]\n",
            "Saving to: ‘songs.csv.2’\n",
            "\n",
            "\rsongs.csv.2           0%[                    ]       0  --.-KB/s               \rsongs.csv.2         100%[===================>] 198.18K  --.-KB/s    in 0.005s  \n",
            "\n",
            "2020-03-16 10:45:49 (41.3 MB/s) - ‘songs.csv.2’ saved [202933/202933]\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "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>Song-Name</th>\n",
              "      <th>Singer/Artists</th>\n",
              "      <th>Genre</th>\n",
              "      <th>Album/Movie</th>\n",
              "      <th>User-Rating</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Aankh Marey</td>\n",
              "      <td>Kumar Sanu, Mika Singh, Neha Kakkar</td>\n",
              "      <td>BollywoodDance</td>\n",
              "      <td>Simmba</td>\n",
              "      <td>8.8/10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Coca Cola</td>\n",
              "      <td>Neha Kakkar, Tony Kakkar</td>\n",
              "      <td>BollywoodDanceRomantic</td>\n",
              "      <td>Luka Chuppi</td>\n",
              "      <td>9.0/10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Apna Time Aayega</td>\n",
              "      <td>Ranveer Singh</td>\n",
              "      <td>BollywoodDance</td>\n",
              "      <td>Gully Boy</td>\n",
              "      <td>9.7/10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Mungda</td>\n",
              "      <td>Jyotica Tangri, Shaan, Subhro Ganguly</td>\n",
              "      <td>BollywoodDance</td>\n",
              "      <td>Total Dhamaal</td>\n",
              "      <td>9.1/10</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Tere Bin</td>\n",
              "      <td>Asees Kaur, Rahat Fateh Ali Khan, Tanishk Bagchi</td>\n",
              "      <td>BollywoodRomantic</td>\n",
              "      <td>Simmba</td>\n",
              "      <td>9.2/10</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          Song-Name  ... User-Rating\n",
              "1       Aankh Marey  ...      8.8/10\n",
              "2         Coca Cola  ...      9.0/10\n",
              "3  Apna Time Aayega  ...      9.7/10\n",
              "4            Mungda  ...      9.1/10\n",
              "5          Tere Bin  ...      9.2/10\n",
              "\n",
              "[5 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 138
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oJ18kjpnnAFP",
        "colab_type": "code",
        "outputId": "ebe65033-fd5f-4534-d474-23fb07f46d5a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "df.shape"
      ],
      "execution_count": 139,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(2420, 5)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 139
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2dcGkCcsnFoc",
        "colab_type": "code",
        "outputId": "2b94ff3d-52f8-44fb-e6b4-cfd1874a727c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 123
        }
      },
      "source": [
        "df.isnull().sum()"
      ],
      "execution_count": 140,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Song-Name          0\n",
              "Singer/Artists    10\n",
              "Genre              0\n",
              "Album/Movie        3\n",
              "User-Rating        0\n",
              "dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 140
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ywb3ZcednIb5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df['SongCode'] = [\"SB\" + str(i) for i in range(df.shape[0])]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sM5v7kKSopQJ",
        "colab_type": "code",
        "outputId": "1f3fd954-0d3e-4598-8f15-81de7e527987",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "df.head()"
      ],
      "execution_count": 142,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>Song-Name</th>\n",
              "      <th>Singer/Artists</th>\n",
              "      <th>Genre</th>\n",
              "      <th>Album/Movie</th>\n",
              "      <th>User-Rating</th>\n",
              "      <th>SongCode</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>Aankh Marey</td>\n",
              "      <td>Kumar Sanu, Mika Singh, Neha Kakkar</td>\n",
              "      <td>BollywoodDance</td>\n",
              "      <td>Simmba</td>\n",
              "      <td>8.8/10</td>\n",
              "      <td>SB0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Coca Cola</td>\n",
              "      <td>Neha Kakkar, Tony Kakkar</td>\n",
              "      <td>BollywoodDanceRomantic</td>\n",
              "      <td>Luka Chuppi</td>\n",
              "      <td>9.0/10</td>\n",
              "      <td>SB1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Apna Time Aayega</td>\n",
              "      <td>Ranveer Singh</td>\n",
              "      <td>BollywoodDance</td>\n",
              "      <td>Gully Boy</td>\n",
              "      <td>9.7/10</td>\n",
              "      <td>SB2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Mungda</td>\n",
              "      <td>Jyotica Tangri, Shaan, Subhro Ganguly</td>\n",
              "      <td>BollywoodDance</td>\n",
              "      <td>Total Dhamaal</td>\n",
              "      <td>9.1/10</td>\n",
              "      <td>SB3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>Tere Bin</td>\n",
              "      <td>Asees Kaur, Rahat Fateh Ali Khan, Tanishk Bagchi</td>\n",
              "      <td>BollywoodRomantic</td>\n",
              "      <td>Simmba</td>\n",
              "      <td>9.2/10</td>\n",
              "      <td>SB4</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          Song-Name  ... SongCode\n",
              "1       Aankh Marey  ...      SB0\n",
              "2         Coca Cola  ...      SB1\n",
              "3  Apna Time Aayega  ...      SB2\n",
              "4            Mungda  ...      SB3\n",
              "5          Tere Bin  ...      SB4\n",
              "\n",
              "[5 rows x 6 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 142
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BiITqiZpowYP",
        "colab_type": "code",
        "outputId": "d45895e3-6bc1-4502-f623-89bd78617eb3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "songs = df[\"Genre\"].unique().tolist()\n",
        "len(songs)"
      ],
      "execution_count": 143,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "21"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 143
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DJjxrE71o4tC",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# shuffle song ID's\n",
        "random.shuffle(songs)\n",
        "\n",
        "# extract 90% of song ID's\n",
        "song_train = [songs[i] for i in range(round(0.9*len(songs)))]\n",
        "\n",
        "# split data into train and validation set\n",
        "train_df = df[df['Genre'].isin(song_train)]\n",
        "validation_df = df[~df['Genre'].isin(song_train)]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Gc7zCxkhpPyv",
        "colab_type": "code",
        "outputId": "e5a42f5c-4eb2-44d4-c2d9-d087dea7bb03",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# list to capture the genre\n",
        "like_train = []\n",
        "\n",
        "# populate the list with the product codes\n",
        "for i in tqdm(song_train):\n",
        "    temp = train_df[train_df[\"Genre\"] == i][\"SongCode\"].tolist()\n",
        "    like_train.append(temp)"
      ],
      "execution_count": 145,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "100%|██████████| 19/19 [00:00<00:00, 950.65it/s]\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RibMuIdWpg-3",
        "colab_type": "code",
        "outputId": "bfef3b50-5556-4a79-d868-b5d36a81958d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# list to capture the genre\n",
        "like_val = []\n",
        "\n",
        "# populate the list with the song codes\n",
        "for i in tqdm(validation_df['Genre'].unique()):\n",
        "    temp = validation_df[validation_df[\"Genre\"] == i][\"SongCode\"].tolist()\n",
        "    like_val.append(temp)"
      ],
      "execution_count": 146,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "100%|██████████| 2/2 [00:00<00:00, 510.88it/s]\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pAIFd-lWpjjJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# train word2vec model\n",
        "model = Word2Vec(window = 5, sg = 1, hs = 0,\n",
        "                 negative = 10, # for negative sampling\n",
        "                 alpha=0.03, min_alpha=0.0007,\n",
        "                 seed = 14, min_count=1)\n",
        "\n",
        "model.build_vocab(like_train, progress_per=200)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lHG2xC4_sBrR",
        "colab_type": "code",
        "outputId": "edbabc26-fc39-4368-cd8d-be7b8d4e22bb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "model.train(like_train, total_examples = model.corpus_count, \n",
        "            epochs=10, report_delay=1)"
      ],
      "execution_count": 148,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(23810, 23810)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 148
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jqn3PmvAU8F_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.init_sims(replace=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2jhId_ZJU-QJ",
        "colab_type": "code",
        "outputId": "0e1fb535-7f58-4ba8-86c5-9d4d4b3d21de",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "print(model)"
      ],
      "execution_count": 150,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Word2Vec(vocab=2381, size=100, alpha=0.03)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3lKwHzaHU_1L",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# # extract all vectors\n",
        "# X = model[model.wv.vocab]\n",
        "\n",
        "# X.shape\n",
        "# test = model.wv.most_similar('SB1057')\n",
        "# # for t in test:\n",
        "# #   for d in df:\n",
        "#     # print(d.type)\n",
        "#     # # if(d['SongCode'] == t[0]):\n",
        "#     #   print(d['Song-Name'])\n",
        "#     #   print(t[0])\n",
        "# # df.head()\n",
        "# test\n",
        "# # df.loc[df['SongCode'].isin(test[0])]\n",
        "# # df.loc[df['SongCode'] == 'SB1']\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Q6nsDvo9rVMH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# columns = list(df)\n",
        "# data = []\n",
        "# df1 = pd.DataFrame(columns = df.columns)\n",
        "# for t in test:\n",
        "#   c = df.loc[lambda df: df['SongCode'] == t[0]]\n",
        "#   print(c.values)\n",
        "#   # zipped = zip(columns, c.values)\n",
        "#   # a_dictionary = dict(zipped)\n",
        "#   # print(a_dictionary)\n",
        "#   data.append(c.values)\n",
        "\n",
        "# # \n",
        "# # df1 = df1.append(data, True)\n",
        "# # # df1 = pd.DataFrame(data, columns=columns)\n",
        "# # df1 = pd.DataFrame(data=data,columns=df.columns)\n",
        "# # df1"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z9aXwk-WIy_1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# columns = list(df)\n",
        "# data = []\n",
        "# df1 = pd.DataFrame(columns = df.columns)\n",
        "# for t in test:\n",
        "#   df1 = df.loc[df['SongCode'] == t[0]]\n",
        "#   df1 = df.loc[lambda df: df['SongCode'] == t[0]]\n",
        "  \n",
        "# df1"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "F7m8PujPVBk-",
        "colab_type": "code",
        "outputId": "013aa235-7767-4428-d8b7-50b59dcd705e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 554
        }
      },
      "source": [
        "import umap\n",
        "\n",
        "cluster_embedding = umap.UMAP(n_neighbors=30, min_dist=0.0,\n",
        "                              n_components=2, random_state=42).fit_transform(X)\n",
        "\n",
        "plt.figure(figsize=(10,9))\n",
        "plt.scatter(cluster_embedding[:, 0], cluster_embedding[:, 1], s=3, cmap='Spectral')"
      ],
      "execution_count": 153,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.collections.PathCollection at 0x7f9903571e10>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 153
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlIAAAIICAYAAAClygDiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGlmkJEmSGlmkJEmSGlmkJEmSGn0H4+i97nq6yW8AAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 720x648 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UW-yalAFVEac",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "sng = train_df[[\"SongCode\", \"Song-Name\",\"Genre\"]]\n",
        "# sng\n",
        "# remove duplicates\n",
        "# sng.drop_duplicates(inplace=True, subset='SongCode', keep=\"last\")\n",
        "\n",
        "# # create song-ID and Song-Name dictionary\n",
        "sng_dict = sng.groupby('SongCode')['Song-Name'].apply(list).to_dict()\n",
        "# sng_dict"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qHzbvGH2VGgB",
        "colab_type": "code",
        "outputId": "98ce07d4-5d78-4588-a44d-de43a13d58a2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "# test the dictionary\n",
        "sng_dict['SB649']"
      ],
      "execution_count": 155,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['Subhanallah']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 155
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ueW0v5_NVIJy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def similar_songs(v, n = 6):\n",
        "    \n",
        "    # extract most similar song for the input vector\n",
        "    ms = model.similar_by_vector(v, topn= n+1)[1:]\n",
        "    # print(ms)\n",
        "    # print(sng_dict)\n",
        "    # extract name and similarity score of the similar song\n",
        "    new_ms = []\n",
        "    for j in ms:\n",
        "        pair = (sng_dict[j[0]][0], j[1])\n",
        "        new_ms.append(pair)\n",
        "        \n",
        "    return new_ms    "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ztjWr42SVK4b",
        "colab_type": "code",
        "outputId": "d57ac242-285d-4a88-b49a-57b7542d3f7e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 123
        }
      },
      "source": [
        "similar_songs(model['SB649'])"
      ],
      "execution_count": 157,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[('Musafir', 0.6709380745887756),\n",
              " ('Tu Hi Meri Shab Hai', 0.6442582607269287),\n",
              " ('Koi Mil Gaya', 0.6425020694732666),\n",
              " ('Aap Ki Nazron Ne Samjha', 0.635926365852356),\n",
              " ('Meri Khamoshi Hai', 0.6357775926589966),\n",
              " ('Dooriyan', 0.6345308423042297)]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 157
        }
      ]
    }
  ]
}