Music_Genre_Classification.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Music_Genre_Classification.ipynb",
      "provenance": [],
      "collapsed_sections": [
        "EX9D0gMATq42",
        "901QWODvU0O1",
        "AVPA8lcBV2M7",
        "DXsE6jl8k6xh",
        "sms8mrx3Wbgj",
        "_kUwZQMGXZi4",
        "7dmHHk-iWPuW",
        "oKGk6dB_cL8S"
      ],
      "mount_file_id": "1bfguLC5dirxjR6gxz8lVLSTGQKmN9xkw",
      "authorship_tag": "ABX9TyPBn52ZmRuTV9/b8ibNCxA5",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/sultanbst123/MachineHack_Genre-Music-Classification/blob/main/Music_Genre_Classification.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "794SpT8iTnH9"
      },
      "source": [
        "#Data modelling with Python"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EX9D0gMATq42"
      },
      "source": [
        "##Library & File"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wwk-e_4KTJf2",
        "outputId": "432467a0-4aa4-4376-fce4-a61130b3af8f"
      },
      "source": [
        "!pip install --upgrade scikit-learn"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Requirement already satisfied: scikit-learn in /usr/local/lib/python3.7/dist-packages (0.22.2.post1)\n",
            "Collecting scikit-learn\n",
            "  Downloading scikit_learn-1.0.1-cp37-cp37m-manylinux_2_12_x86_64.manylinux2010_x86_64.whl (23.2 MB)\n",
            "\u001b[K     |████████████████████████████████| 23.2 MB 1.5 MB/s \n",
            "\u001b[?25hRequirement already satisfied: scipy>=1.1.0 in /usr/local/lib/python3.7/dist-packages (from scikit-learn) (1.4.1)\n",
            "Requirement already satisfied: joblib>=0.11 in /usr/local/lib/python3.7/dist-packages (from scikit-learn) (1.0.1)\n",
            "Collecting threadpoolctl>=2.0.0\n",
            "  Downloading threadpoolctl-3.0.0-py3-none-any.whl (14 kB)\n",
            "Requirement already satisfied: numpy>=1.14.6 in /usr/local/lib/python3.7/dist-packages (from scikit-learn) (1.19.5)\n",
            "Installing collected packages: threadpoolctl, scikit-learn\n",
            "  Attempting uninstall: scikit-learn\n",
            "    Found existing installation: scikit-learn 0.22.2.post1\n",
            "    Uninstalling scikit-learn-0.22.2.post1:\n",
            "      Successfully uninstalled scikit-learn-0.22.2.post1\n",
            "Successfully installed scikit-learn-1.0.1 threadpoolctl-3.0.0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hx22UW7hTfyK",
        "outputId": "1256eaf7-ca4d-4bf9-f3a3-b3e65e3fdd4e"
      },
      "source": [
        "\n",
        "!pip install catboost\n",
        "!pip install optuna\n",
        "!pip install borutashap"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Collecting catboost\n",
            "  Downloading catboost-1.0.0-cp37-none-manylinux1_x86_64.whl (76.4 MB)\n",
            "\u001b[K     |████████████████████████████████| 76.4 MB 38 kB/s \n",
            "\u001b[?25hRequirement already satisfied: plotly in /usr/local/lib/python3.7/dist-packages (from catboost) (4.4.1)\n",
            "Requirement already satisfied: scipy in /usr/local/lib/python3.7/dist-packages (from catboost) (1.4.1)\n",
            "Requirement already satisfied: pandas>=0.24.0 in /usr/local/lib/python3.7/dist-packages (from catboost) (1.1.5)\n",
            "Requirement already satisfied: six in /usr/local/lib/python3.7/dist-packages (from catboost) (1.15.0)\n",
            "Requirement already satisfied: matplotlib in /usr/local/lib/python3.7/dist-packages (from catboost) (3.2.2)\n",
            "Requirement already satisfied: numpy>=1.16.0 in /usr/local/lib/python3.7/dist-packages (from catboost) (1.19.5)\n",
            "Requirement already satisfied: graphviz in /usr/local/lib/python3.7/dist-packages (from catboost) (0.10.1)\n",
            "Requirement already satisfied: python-dateutil>=2.7.3 in /usr/local/lib/python3.7/dist-packages (from pandas>=0.24.0->catboost) (2.8.2)\n",
            "Requirement already satisfied: pytz>=2017.2 in /usr/local/lib/python3.7/dist-packages (from pandas>=0.24.0->catboost) (2018.9)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->catboost) (1.3.2)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.7/dist-packages (from matplotlib->catboost) (0.10.0)\n",
            "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->catboost) (2.4.7)\n",
            "Requirement already satisfied: retrying>=1.3.3 in /usr/local/lib/python3.7/dist-packages (from plotly->catboost) (1.3.3)\n",
            "Installing collected packages: catboost\n",
            "Successfully installed catboost-1.0.0\n",
            "Collecting optuna\n",
            "  Downloading optuna-2.10.0-py3-none-any.whl (308 kB)\n",
            "\u001b[K     |████████████████████████████████| 308 kB 5.1 MB/s \n",
            "\u001b[?25hRequirement already satisfied: numpy in /usr/local/lib/python3.7/dist-packages (from optuna) (1.19.5)\n",
            "Requirement already satisfied: scipy!=1.4.0 in /usr/local/lib/python3.7/dist-packages (from optuna) (1.4.1)\n",
            "Requirement already satisfied: sqlalchemy>=1.1.0 in /usr/local/lib/python3.7/dist-packages (from optuna) (1.4.25)\n",
            "Requirement already satisfied: PyYAML in /usr/local/lib/python3.7/dist-packages (from optuna) (3.13)\n",
            "Collecting cliff\n",
            "  Downloading cliff-3.9.0-py3-none-any.whl (80 kB)\n",
            "\u001b[K     |████████████████████████████████| 80 kB 8.8 MB/s \n",
            "\u001b[?25hRequirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.7/dist-packages (from optuna) (21.0)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.7/dist-packages (from optuna) (4.62.3)\n",
            "Collecting colorlog\n",
            "  Downloading colorlog-6.5.0-py2.py3-none-any.whl (11 kB)\n",
            "Collecting alembic\n",
            "  Downloading alembic-1.7.4-py3-none-any.whl (209 kB)\n",
            "\u001b[K     |████████████████████████████████| 209 kB 49.3 MB/s \n",
            "\u001b[?25hCollecting cmaes>=0.8.2\n",
            "  Downloading cmaes-0.8.2-py3-none-any.whl (15 kB)\n",
            "Requirement already satisfied: pyparsing>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from packaging>=20.0->optuna) (2.4.7)\n",
            "Requirement already satisfied: greenlet!=0.4.17 in /usr/local/lib/python3.7/dist-packages (from sqlalchemy>=1.1.0->optuna) (1.1.2)\n",
            "Requirement already satisfied: importlib-metadata in /usr/local/lib/python3.7/dist-packages (from sqlalchemy>=1.1.0->optuna) (4.8.1)\n",
            "Requirement already satisfied: importlib-resources in /usr/local/lib/python3.7/dist-packages (from alembic->optuna) (5.2.2)\n",
            "Collecting Mako\n",
            "  Downloading Mako-1.1.5-py2.py3-none-any.whl (75 kB)\n",
            "\u001b[K     |████████████████████████████████| 75 kB 4.1 MB/s \n",
            "\u001b[?25hCollecting stevedore>=2.0.1\n",
            "  Downloading stevedore-3.5.0-py3-none-any.whl (49 kB)\n",
            "\u001b[K     |████████████████████████████████| 49 kB 5.4 MB/s \n",
            "\u001b[?25hCollecting cmd2>=1.0.0\n",
            "  Downloading cmd2-2.2.0-py3-none-any.whl (144 kB)\n",
            "\u001b[K     |████████████████████████████████| 144 kB 66.9 MB/s \n",
            "\u001b[?25hRequirement already satisfied: PrettyTable>=0.7.2 in /usr/local/lib/python3.7/dist-packages (from cliff->optuna) (2.2.1)\n",
            "Collecting autopage>=0.4.0\n",
            "  Downloading autopage-0.4.0-py3-none-any.whl (20 kB)\n",
            "Collecting pbr!=2.1.0,>=2.0.0\n",
            "  Downloading pbr-5.6.0-py2.py3-none-any.whl (111 kB)\n",
            "\u001b[K     |████████████████████████████████| 111 kB 72.9 MB/s \n",
            "\u001b[?25hRequirement already satisfied: typing-extensions in /usr/local/lib/python3.7/dist-packages (from cmd2>=1.0.0->cliff->optuna) (3.7.4.3)\n",
            "Collecting colorama>=0.3.7\n",
            "  Downloading colorama-0.4.4-py2.py3-none-any.whl (16 kB)\n",
            "Requirement already satisfied: wcwidth>=0.1.7 in /usr/local/lib/python3.7/dist-packages (from cmd2>=1.0.0->cliff->optuna) (0.2.5)\n",
            "Collecting pyperclip>=1.6\n",
            "  Downloading pyperclip-1.8.2.tar.gz (20 kB)\n",
            "Requirement already satisfied: attrs>=16.3.0 in /usr/local/lib/python3.7/dist-packages (from cmd2>=1.0.0->cliff->optuna) (21.2.0)\n",
            "Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.7/dist-packages (from importlib-metadata->sqlalchemy>=1.1.0->optuna) (3.6.0)\n",
            "Requirement already satisfied: MarkupSafe>=0.9.2 in /usr/local/lib/python3.7/dist-packages (from Mako->alembic->optuna) (2.0.1)\n",
            "Building wheels for collected packages: pyperclip\n",
            "  Building wheel for pyperclip (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for pyperclip: filename=pyperclip-1.8.2-py3-none-any.whl size=11136 sha256=8ce8aeb2846e77f398f59c3be7e7a76a180cbc3bb1d20a0629a0d3d5621d1b33\n",
            "  Stored in directory: /root/.cache/pip/wheels/9f/18/84/8f69f8b08169c7bae2dde6bd7daf0c19fca8c8e500ee620a28\n",
            "Successfully built pyperclip\n",
            "Installing collected packages: pyperclip, pbr, colorama, stevedore, Mako, cmd2, autopage, colorlog, cmaes, cliff, alembic, optuna\n",
            "Successfully installed Mako-1.1.5 alembic-1.7.4 autopage-0.4.0 cliff-3.9.0 cmaes-0.8.2 cmd2-2.2.0 colorama-0.4.4 colorlog-6.5.0 optuna-2.10.0 pbr-5.6.0 pyperclip-1.8.2 stevedore-3.5.0\n",
            "Collecting borutashap\n",
            "  Downloading BorutaShap-1.0.16-py3-none-any.whl (13 kB)\n",
            "Requirement already satisfied: matplotlib in /usr/local/lib/python3.7/dist-packages (from borutashap) (3.2.2)\n",
            "Requirement already satisfied: scikit-learn in /usr/local/lib/python3.7/dist-packages (from borutashap) (1.0.1)\n",
            "Requirement already satisfied: tqdm in /usr/local/lib/python3.7/dist-packages (from borutashap) (4.62.3)\n",
            "Collecting shap>=0.34.0\n",
            "  Downloading shap-0.40.0.tar.gz (371 kB)\n",
            "\u001b[K     |████████████████████████████████| 371 kB 5.1 MB/s \n",
            "\u001b[?25h  Installing build dependencies ... \u001b[?25l\u001b[?25hdone\n",
            "  Getting requirements to build wheel ... \u001b[?25l\u001b[?25hdone\n",
            "    Preparing wheel metadata ... \u001b[?25l\u001b[?25hdone\n",
            "Requirement already satisfied: scipy in /usr/local/lib/python3.7/dist-packages (from borutashap) (1.4.1)\n",
            "Requirement already satisfied: statsmodels in /usr/local/lib/python3.7/dist-packages (from borutashap) (0.10.2)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.7/dist-packages (from borutashap) (1.19.5)\n",
            "Requirement already satisfied: seaborn in /usr/local/lib/python3.7/dist-packages (from borutashap) (0.11.2)\n",
            "Requirement already satisfied: pandas in /usr/local/lib/python3.7/dist-packages (from borutashap) (1.1.5)\n",
            "Collecting slicer==0.0.7\n",
            "  Downloading slicer-0.0.7-py3-none-any.whl (14 kB)\n",
            "Requirement already satisfied: packaging>20.9 in /usr/local/lib/python3.7/dist-packages (from shap>=0.34.0->borutashap) (21.0)\n",
            "Requirement already satisfied: numba in /usr/local/lib/python3.7/dist-packages (from shap>=0.34.0->borutashap) (0.51.2)\n",
            "Requirement already satisfied: cloudpickle in /usr/local/lib/python3.7/dist-packages (from shap>=0.34.0->borutashap) (1.3.0)\n",
            "Requirement already satisfied: pyparsing>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from packaging>20.9->shap>=0.34.0->borutashap) (2.4.7)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.7/dist-packages (from matplotlib->borutashap) (0.10.0)\n",
            "Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->borutashap) (2.8.2)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->borutashap) (1.3.2)\n",
            "Requirement already satisfied: six in /usr/local/lib/python3.7/dist-packages (from cycler>=0.10->matplotlib->borutashap) (1.15.0)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.7/dist-packages (from numba->shap>=0.34.0->borutashap) (57.4.0)\n",
            "Requirement already satisfied: llvmlite<0.35,>=0.34.0.dev0 in /usr/local/lib/python3.7/dist-packages (from numba->shap>=0.34.0->borutashap) (0.34.0)\n",
            "Requirement already satisfied: pytz>=2017.2 in /usr/local/lib/python3.7/dist-packages (from pandas->borutashap) (2018.9)\n",
            "Requirement already satisfied: joblib>=0.11 in /usr/local/lib/python3.7/dist-packages (from scikit-learn->borutashap) (1.0.1)\n",
            "Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.7/dist-packages (from scikit-learn->borutashap) (3.0.0)\n",
            "Requirement already satisfied: patsy>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from statsmodels->borutashap) (0.5.2)\n",
            "Building wheels for collected packages: shap\n",
            "  Building wheel for shap (PEP 517) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for shap: filename=shap-0.40.0-cp37-cp37m-linux_x86_64.whl size=509259 sha256=3a46ce958e222753bf569a83d851fca90e4d38683e360bd72252419db0d5b306\n",
            "  Stored in directory: /root/.cache/pip/wheels/ec/35/84/e304841ac4b910bc95fe9a6e5302eb2507b4333728851dcbfb\n",
            "Successfully built shap\n",
            "Installing collected packages: slicer, shap, borutashap\n",
            "Successfully installed borutashap-1.0.16 shap-0.40.0 slicer-0.0.7\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VwQkHKwVTlzc"
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "import optuna\n",
        "from BorutaShap import BorutaShap\n",
        " \n",
        "# classification\n",
        "import xgboost, lightgbm, catboost\n",
        "from xgboost import XGBClassifier\n",
        "from catboost import CatBoostClassifier\n",
        "from lightgbm import LGBMClassifier\n",
        "from sklearn.tree import DecisionTreeClassifier, ExtraTreeClassifier\n",
        "from sklearn.ensemble import (RandomForestClassifier, StackingClassifier, GradientBoostingClassifier,\n",
        "                             AdaBoostClassifier, HistGradientBoostingClassifier) \n",
        "from sklearn.svm import SVC\n",
        "from sklearn.linear_model import LogisticRegression, SGDClassifier\n",
        "from sklearn.neighbors import KNeighborsClassifier\n",
        " \n",
        "# regression\n",
        "from xgboost import XGBRegressor\n",
        "from catboost import CatBoostRegressor\n",
        "from lightgbm import LGBMRegressor\n",
        "from sklearn.tree import DecisionTreeRegressor, ExtraTreeRegressor\n",
        "from sklearn.ensemble import (RandomForestRegressor, GradientBoostingRegressor,\n",
        "                             AdaBoostRegressor, HistGradientBoostingRegressor) \n",
        "from sklearn.svm import SVR\n",
        "from sklearn.linear_model import LinearRegression, Ridge, Lasso, ElasticNet, SGDOneClassSVM\n",
        "from sklearn.neighbors import KNeighborsRegressor\n",
        "\n",
        "#preprocessing\n",
        "from sklearn.model_selection import train_test_split, KFold, cross_val_score, StratifiedKFold, GridSearchCV\n",
        "from sklearn.metrics import (accuracy_score, classification_report, confusion_matrix, \n",
        "                            log_loss, roc_auc_score, plot_roc_curve,plot_confusion_matrix) \n",
        "from sklearn.metrics import mean_absolute_error, mean_squared_error, mean_squared_log_error\n",
        "from sklearn.preprocessing import LabelEncoder,OneHotEncoder, StandardScaler, MinMaxScaler, RobustScaler\n",
        "from sklearn.calibration import CalibrationDisplay\n",
        "\n",
        "\n",
        "import warnings\n",
        "warnings.filterwarnings(\"ignore\")\n",
        " \n",
        "%matplotlib inline"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LGtnZC89TxFC"
      },
      "source": [
        "train = pd.read_csv(\"/content/drive/MyDrive/Genre music classification/train_genre_music(clean).csv\")\n",
        "test = pd.read_csv(\"/content/drive/MyDrive/Genre music classification/test_genre_music(clean).csv\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "U_N8Bd6oUndx",
        "outputId": "9eeeb78a-860f-4e50-e718-33f2c0c8bfe6"
      },
      "source": [
        "train.info()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "<class 'pandas.core.frame.DataFrame'>\n",
            "RangeIndex: 17996 entries, 0 to 17995\n",
            "Data columns (total 23 columns):\n",
            " #   Column                 Non-Null Count  Dtype  \n",
            "---  ------                 --------------  -----  \n",
            " 0   Popularity             17996 non-null  float64\n",
            " 1   danceability           17996 non-null  float64\n",
            " 2   energy                 17996 non-null  float64\n",
            " 3   key                    17996 non-null  int64  \n",
            " 4   loudness               17996 non-null  float64\n",
            " 5   mode                   17996 non-null  int64  \n",
            " 6   speechiness            17996 non-null  float64\n",
            " 7   acousticness           17996 non-null  float64\n",
            " 8   instrumentalness       17996 non-null  float64\n",
            " 9   liveness               17996 non-null  float64\n",
            " 10  valence                17996 non-null  float64\n",
            " 11  tempo                  17996 non-null  float64\n",
            " 12  duration_in min/ms     17996 non-null  float64\n",
            " 13  time_signature         17996 non-null  int64  \n",
            " 14  Class_Genre            17996 non-null  object \n",
            " 15  length_of_artist_name  17996 non-null  int64  \n",
            " 16  length_of_track_name   17996 non-null  int64  \n",
            " 17  length_words_artist    17996 non-null  int64  \n",
            " 18  length_words_track     17996 non-null  int64  \n",
            " 19  most_popularity        17996 non-null  float64\n",
            " 20  most_energy            17996 non-null  float64\n",
            " 21  new                    17996 non-null  float64\n",
            " 22  new2                   17996 non-null  float64\n",
            "dtypes: float64(15), int64(7), object(1)\n",
            "memory usage: 3.2+ MB\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ctGXcnwDUpmb",
        "outputId": "c9d4c72a-6d2d-43d2-a1bd-3f2c42a4c183"
      },
      "source": [
        "train.head()"
      ],
      "execution_count": null,
      "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>Popularity</th>\n",
              "      <th>danceability</th>\n",
              "      <th>energy</th>\n",
              "      <th>key</th>\n",
              "      <th>loudness</th>\n",
              "      <th>mode</th>\n",
              "      <th>speechiness</th>\n",
              "      <th>acousticness</th>\n",
              "      <th>instrumentalness</th>\n",
              "      <th>liveness</th>\n",
              "      <th>valence</th>\n",
              "      <th>tempo</th>\n",
              "      <th>duration_in min/ms</th>\n",
              "      <th>time_signature</th>\n",
              "      <th>Class_Genre</th>\n",
              "      <th>length_of_artist_name</th>\n",
              "      <th>length_of_track_name</th>\n",
              "      <th>length_words_artist</th>\n",
              "      <th>length_words_track</th>\n",
              "      <th>most_popularity</th>\n",
              "      <th>most_energy</th>\n",
              "      <th>new</th>\n",
              "      <th>new2</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>60.0</td>\n",
              "      <td>0.854</td>\n",
              "      <td>0.564</td>\n",
              "      <td>1</td>\n",
              "      <td>-4.964</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0485</td>\n",
              "      <td>0.017100</td>\n",
              "      <td>0.003910</td>\n",
              "      <td>0.0849</td>\n",
              "      <td>0.8990</td>\n",
              "      <td>134.071</td>\n",
              "      <td>234596.0</td>\n",
              "      <td>4</td>\n",
              "      <td>HipHop</td>\n",
              "      <td>10</td>\n",
              "      <td>37</td>\n",
              "      <td>2</td>\n",
              "      <td>7</td>\n",
              "      <td>82.0</td>\n",
              "      <td>0.843</td>\n",
              "      <td>0.709</td>\n",
              "      <td>134.071</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>54.0</td>\n",
              "      <td>0.382</td>\n",
              "      <td>0.814</td>\n",
              "      <td>3</td>\n",
              "      <td>-7.230</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0406</td>\n",
              "      <td>0.001100</td>\n",
              "      <td>0.004010</td>\n",
              "      <td>0.1010</td>\n",
              "      <td>0.5690</td>\n",
              "      <td>116.454</td>\n",
              "      <td>251733.0</td>\n",
              "      <td>4</td>\n",
              "      <td>Rock</td>\n",
              "      <td>6</td>\n",
              "      <td>12</td>\n",
              "      <td>1</td>\n",
              "      <td>3</td>\n",
              "      <td>79.0</td>\n",
              "      <td>0.814</td>\n",
              "      <td>0.598</td>\n",
              "      <td>116.454</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>35.0</td>\n",
              "      <td>0.434</td>\n",
              "      <td>0.614</td>\n",
              "      <td>6</td>\n",
              "      <td>-8.334</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0525</td>\n",
              "      <td>0.486000</td>\n",
              "      <td>0.000196</td>\n",
              "      <td>0.3940</td>\n",
              "      <td>0.7870</td>\n",
              "      <td>147.681</td>\n",
              "      <td>109667.0</td>\n",
              "      <td>4</td>\n",
              "      <td>Indie Alt</td>\n",
              "      <td>13</td>\n",
              "      <td>18</td>\n",
              "      <td>2</td>\n",
              "      <td>5</td>\n",
              "      <td>50.0</td>\n",
              "      <td>0.854</td>\n",
              "      <td>0.524</td>\n",
              "      <td>147.681</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>66.0</td>\n",
              "      <td>0.853</td>\n",
              "      <td>0.597</td>\n",
              "      <td>10</td>\n",
              "      <td>-6.528</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0555</td>\n",
              "      <td>0.021200</td>\n",
              "      <td>0.003910</td>\n",
              "      <td>0.1220</td>\n",
              "      <td>0.5690</td>\n",
              "      <td>107.033</td>\n",
              "      <td>173968.0</td>\n",
              "      <td>4</td>\n",
              "      <td>HipHop</td>\n",
              "      <td>4</td>\n",
              "      <td>26</td>\n",
              "      <td>1</td>\n",
              "      <td>5</td>\n",
              "      <td>66.0</td>\n",
              "      <td>0.713</td>\n",
              "      <td>0.725</td>\n",
              "      <td>107.033</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>53.0</td>\n",
              "      <td>0.167</td>\n",
              "      <td>0.975</td>\n",
              "      <td>2</td>\n",
              "      <td>-4.279</td>\n",
              "      <td>1</td>\n",
              "      <td>0.2160</td>\n",
              "      <td>0.000169</td>\n",
              "      <td>0.016100</td>\n",
              "      <td>0.1720</td>\n",
              "      <td>0.0918</td>\n",
              "      <td>199.060</td>\n",
              "      <td>229960.0</td>\n",
              "      <td>4</td>\n",
              "      <td>Rock</td>\n",
              "      <td>21</td>\n",
              "      <td>33</td>\n",
              "      <td>4</td>\n",
              "      <td>6</td>\n",
              "      <td>81.0</td>\n",
              "      <td>0.983</td>\n",
              "      <td>0.571</td>\n",
              "      <td>199.060</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Popularity  danceability  energy  ...  most_energy    new     new2\n",
              "0        60.0         0.854   0.564  ...        0.843  0.709  134.071\n",
              "1        54.0         0.382   0.814  ...        0.814  0.598  116.454\n",
              "2        35.0         0.434   0.614  ...        0.854  0.524  147.681\n",
              "3        66.0         0.853   0.597  ...        0.713  0.725  107.033\n",
              "4        53.0         0.167   0.975  ...        0.983  0.571  199.060\n",
              "\n",
              "[5 rows x 23 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ne5K6qK8Uyf9"
      },
      "source": [
        "def save_ram(df, verbose=True):\n",
        "    numerics = ['int16', 'int32', 'int64', 'float16', 'float32', 'float64']\n",
        "    start_mem = df.memory_usage().sum() / 1024**2\n",
        "    for col in df.columns:\n",
        "        col_type = df[col].dtypes\n",
        "        if col_type in numerics:\n",
        "            c_min = df[col].min()\n",
        "            c_max = df[col].max()\n",
        "            if str(col_type)[:3] == 'int':\n",
        "                if c_min > np.iinfo(np.int8).min and c_max < np.iinfo(np.int8).max:\n",
        "                    df[col] = df[col].astype(np.int8)\n",
        "                elif c_min > np.iinfo(np.int16).min and c_max < np.iinfo(np.int16).max:\n",
        "                    df[col] = df[col].astype(np.int16)\n",
        "                elif c_min > np.iinfo(np.int32).min and c_max < np.iinfo(np.int32).max:\n",
        "                    df[col] = df[col].astype(np.int32)\n",
        "                elif c_min > np.iinfo(np.int64).min and c_max < np.iinfo(np.int64).max:\n",
        "                    df[col] = df[col].astype(np.int64)\n",
        "            else:\n",
        "                if c_min > np.finfo(np.float16).min and c_max < np.finfo(np.float16).max:\n",
        "                    df[col] = df[col].astype(np.float16)\n",
        "                elif c_min > np.finfo(np.float32).min and c_max < np.finfo(np.float32).max:\n",
        "                    df[col] = df[col].astype(np.float32)\n",
        "                else:\n",
        "                    df[col] = df[col].astype(np.float64)\n",
        "\n",
        "    end_mem = df.memory_usage().sum() / 1024**2\n",
        "\n",
        "    print(f\"After Diet : {round(end_mem, 2)}MB\")\n",
        "    print(f\"Reduced: {round(100*(start_mem - end_mem)/(start_mem), 2)}%\")\n",
        "\n",
        "    return df"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "miBE41YKi-la",
        "outputId": "a35e76ab-5f69-4a69-9310-7c46c061672a"
      },
      "source": [
        "train = save_ram(train)\n",
        "test = save_ram(test)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "After Diet : 0.84MB\n",
            "Reduced: 73.37%\n",
            "After Diet : 0.29MB\n",
            "Reduced: 77.27%\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "901QWODvU0O1"
      },
      "source": [
        "## Split data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ItMRD_JbU2-4"
      },
      "source": [
        "tester = test.copy() \n",
        "\n",
        "x = train.drop(\"Class_Genre\", axis=1)\n",
        "y = train.Class_Genre"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5HhugWoiVKbO"
      },
      "source": [
        "cater = x.select_dtypes(exclude=\"number\").columns"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DI9UPWpUVVpF"
      },
      "source": [
        "x[cater] = x[cater].astype(\"category\")\n",
        "tester[cater] = tester[cater].astype(\"category\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9e_D6KgPVIX9"
      },
      "source": [
        "x_train,x_val,y_train,y_val = train_test_split(x, y, test_size=0.2, random_state=1618)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jbvy2nJUVx2s",
        "outputId": "18cf0e97-7255-4140-d8c1-c99d2b4acbc0"
      },
      "source": [
        "x_train.head()"
      ],
      "execution_count": null,
      "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>Popularity</th>\n",
              "      <th>danceability</th>\n",
              "      <th>energy</th>\n",
              "      <th>key</th>\n",
              "      <th>loudness</th>\n",
              "      <th>mode</th>\n",
              "      <th>speechiness</th>\n",
              "      <th>acousticness</th>\n",
              "      <th>instrumentalness</th>\n",
              "      <th>liveness</th>\n",
              "      <th>valence</th>\n",
              "      <th>tempo</th>\n",
              "      <th>duration_in min/ms</th>\n",
              "      <th>time_signature</th>\n",
              "      <th>length_of_artist_name</th>\n",
              "      <th>length_of_track_name</th>\n",
              "      <th>length_words_artist</th>\n",
              "      <th>length_words_track</th>\n",
              "      <th>most_popularity</th>\n",
              "      <th>most_energy</th>\n",
              "      <th>new</th>\n",
              "      <th>new2</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>13297</th>\n",
              "      <td>53.0</td>\n",
              "      <td>0.343994</td>\n",
              "      <td>0.911133</td>\n",
              "      <td>10</td>\n",
              "      <td>-5.996094</td>\n",
              "      <td>1</td>\n",
              "      <td>0.053986</td>\n",
              "      <td>0.000050</td>\n",
              "      <td>0.001660</td>\n",
              "      <td>0.064209</td>\n",
              "      <td>0.509766</td>\n",
              "      <td>135.2500</td>\n",
              "      <td>320333.000000</td>\n",
              "      <td>4</td>\n",
              "      <td>23</td>\n",
              "      <td>9</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>55.0</td>\n",
              "      <td>0.937988</td>\n",
              "      <td>0.627441</td>\n",
              "      <td>135.250000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17216</th>\n",
              "      <td>71.0</td>\n",
              "      <td>0.717773</td>\n",
              "      <td>0.690918</td>\n",
              "      <td>4</td>\n",
              "      <td>-6.441406</td>\n",
              "      <td>1</td>\n",
              "      <td>0.038696</td>\n",
              "      <td>0.015900</td>\n",
              "      <td>0.003910</td>\n",
              "      <td>0.062683</td>\n",
              "      <td>0.551758</td>\n",
              "      <td>100.0000</td>\n",
              "      <td>213107.000000</td>\n",
              "      <td>4</td>\n",
              "      <td>24</td>\n",
              "      <td>14</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>71.0</td>\n",
              "      <td>0.716797</td>\n",
              "      <td>0.704590</td>\n",
              "      <td>100.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5014</th>\n",
              "      <td>20.0</td>\n",
              "      <td>0.637207</td>\n",
              "      <td>0.612793</td>\n",
              "      <td>8</td>\n",
              "      <td>-7.343750</td>\n",
              "      <td>1</td>\n",
              "      <td>0.074829</td>\n",
              "      <td>0.318115</td>\n",
              "      <td>0.000003</td>\n",
              "      <td>0.111023</td>\n",
              "      <td>0.570801</td>\n",
              "      <td>90.0625</td>\n",
              "      <td>4.400000</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>35</td>\n",
              "      <td>1</td>\n",
              "      <td>5</td>\n",
              "      <td>20.0</td>\n",
              "      <td>0.629883</td>\n",
              "      <td>0.625000</td>\n",
              "      <td>2.060547</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>385</th>\n",
              "      <td>20.0</td>\n",
              "      <td>0.269043</td>\n",
              "      <td>0.154053</td>\n",
              "      <td>9</td>\n",
              "      <td>-18.078125</td>\n",
              "      <td>1</td>\n",
              "      <td>0.038696</td>\n",
              "      <td>0.930176</td>\n",
              "      <td>0.750977</td>\n",
              "      <td>0.109009</td>\n",
              "      <td>0.104004</td>\n",
              "      <td>134.8750</td>\n",
              "      <td>6.858217</td>\n",
              "      <td>4</td>\n",
              "      <td>76</td>\n",
              "      <td>59</td>\n",
              "      <td>12</td>\n",
              "      <td>11</td>\n",
              "      <td>20.0</td>\n",
              "      <td>0.154053</td>\n",
              "      <td>0.211548</td>\n",
              "      <td>4.511719</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16199</th>\n",
              "      <td>73.0</td>\n",
              "      <td>0.404053</td>\n",
              "      <td>0.720215</td>\n",
              "      <td>10</td>\n",
              "      <td>-9.179688</td>\n",
              "      <td>0</td>\n",
              "      <td>0.062286</td>\n",
              "      <td>0.000056</td>\n",
              "      <td>0.835938</td>\n",
              "      <td>0.046600</td>\n",
              "      <td>0.475098</td>\n",
              "      <td>121.8750</td>\n",
              "      <td>258467.000000</td>\n",
              "      <td>4</td>\n",
              "      <td>21</td>\n",
              "      <td>27</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>75.0</td>\n",
              "      <td>0.875977</td>\n",
              "      <td>0.562012</td>\n",
              "      <td>121.875000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       Popularity  danceability    energy  ...  most_energy       new        new2\n",
              "13297        53.0      0.343994  0.911133  ...     0.937988  0.627441  135.250000\n",
              "17216        71.0      0.717773  0.690918  ...     0.716797  0.704590  100.000000\n",
              "5014         20.0      0.637207  0.612793  ...     0.629883  0.625000    2.060547\n",
              "385          20.0      0.269043  0.154053  ...     0.154053  0.211548    4.511719\n",
              "16199        73.0      0.404053  0.720215  ...     0.875977  0.562012  121.875000\n",
              "\n",
              "[5 rows x 22 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1U113meMV1ku",
        "outputId": "43643b2d-2c80-4007-9c85-2b63114b790d"
      },
      "source": [
        "y.value_counts()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Rock             4949\n",
              "Indie Alt        2587\n",
              "Pop              2524\n",
              "Metal            1854\n",
              "HipHop           1447\n",
              "Alt Music        1373\n",
              "Blues            1272\n",
              "Acoustic/Folk     625\n",
              "Instrumental      576\n",
              "Bollywood         402\n",
              "Country           387\n",
              "Name: Class_Genre, dtype: int64"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AVPA8lcBV2M7"
      },
      "source": [
        "##Test data with LGBM"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CdOhKFpGg0Dw"
      },
      "source": [
        "Test with LGBM"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BsOpI9FSV9FA"
      },
      "source": [
        "LG = LGBMClassifier(random_state=18)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zgfkMoRHWO5l",
        "outputId": "2aae143d-e598-4ea4-9f2c-467883869377"
      },
      "source": [
        "LG.fit(x_train, y_train)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "LGBMClassifier(random_state=18)"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gtAgqjqvWTYr",
        "outputId": "b8105e6b-a48f-44a9-9cdd-3426cdd4448e"
      },
      "source": [
        "print(f\"accuracy train: {LG.score(x_train, y_train)}\")\n",
        "print(f\"accuracy val: {LG.score(x_val, y_val)}\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "accuracy train: 0.777229786051681\n",
            "accuracy val: 0.5861111111111111\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xXkP4eteYV1D"
      },
      "source": [
        "pred = LG.predict_proba(x_val)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hbhHk1M8Yjmj",
        "outputId": "700ff267-48d6-4793-e55d-8448e04a2afc"
      },
      "source": [
        "log_loss(y_val, pred)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1.0935778049172022"
            ]
          },
          "metadata": {},
          "execution_count": 78
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 238
        },
        "id": "uxFpwpfSY57f",
        "outputId": "7e6c9d13-e31c-48e1-fc2a-91bb79d08bd5"
      },
      "source": [
        "Label = {}\n",
        "train_label_ohe = pd.get_dummies(y_train)\n",
        "val_label_ohe = pd.get_dummies(y_val) \n",
        "\n",
        "for i in range(11): # all label\n",
        "    mdl = LGBMClassifier(random_state=19)\n",
        "    mdl.fit(x_train,train_label_ohe.iloc[:,i]) \n",
        "    #plot \n",
        "    ax = plt.gca()\n",
        "    Label[i] = plot_roc_curve(mdl, x_val, val_label_ohe.iloc[:,i], ax=ax,name=train_label_ohe.iloc[:,i].name)\n",
        "\n",
        "plt.legend(title=\"Music Genre\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f5590bfd050>"
            ]
          },
          "metadata": {},
          "execution_count": 83
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 278
        },
        "id": "27LHvPZ6kNWO",
        "outputId": "04a84411-0d9e-4774-f7cf-2cdbd96ab891"
      },
      "source": [
        "fig, ax = plt.subplots(1, 1,figsize=(20,15))\n",
        "plot_confusion_matrix(LG, x_val, y_val,ax=ax)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x7f09a39d5f90>"
            ]
          },
          "metadata": {},
          "execution_count": 22
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x1080 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 210
        },
        "id": "mFROc7aZj3hN",
        "outputId": "46029287-a310-418e-f3d2-bea43f759fc2"
      },
      "source": [
        "lightgbm.plot_importance(LG)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f09a3ba3d10>"
            ]
          },
          "metadata": {},
          "execution_count": 18
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qAglkuA9k5-2"
      },
      "source": [
        "loss = []\n",
        "acc_train = []\n",
        "acc_val = []"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DXsE6jl8k6xh"
      },
      "source": [
        "## Model-1"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Imc_V-XRk8kf"
      },
      "source": [
        "x_train_1 = x_train.copy()\n",
        "x_val_1 = x_val.copy()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wni9CmkwlMXl",
        "outputId": "e436ff42-72e9-40fe-b9c0-a78788b7cc29"
      },
      "source": [
        "x_train_1.dtypes"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Popularity               float16\n",
              "danceability             float16\n",
              "energy                   float16\n",
              "key                         int8\n",
              "loudness                 float16\n",
              "mode                        int8\n",
              "speechiness              float16\n",
              "acousticness             float16\n",
              "instrumentalness         float16\n",
              "liveness                 float16\n",
              "valence                  float16\n",
              "tempo                    float16\n",
              "duration_in min/ms       float32\n",
              "time_signature              int8\n",
              "length_of_artist_name      int16\n",
              "length_of_track_name       int16\n",
              "length_words_artist         int8\n",
              "length_words_track          int8\n",
              "most_popularity          float16\n",
              "most_energy              float16\n",
              "new                      float16\n",
              "new2                     float16\n",
              "dtype: object"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yZhlITXElQ0-"
      },
      "source": [
        "XGB = XGBClassifier(random_state=10)\n",
        "GBT = GradientBoostingClassifier(random_state=20)\n",
        "HGBT = HistGradientBoostingClassifier(random_state=30)\n",
        "ADA = AdaBoostClassifier(random_state=40)\n",
        "EXT = ExtraTreeClassifier(random_state=50)\n",
        "RF = RandomForestClassifier(random_state=60)\n",
        "DT = DecisionTreeClassifier(random_state=70)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a7Juw9UDlYww"
      },
      "source": [
        "def score_1(model,x_train, y_train, x_val, y_val):\n",
        "\n",
        "    #fit model\n",
        "    model.fit(x_train,y_train)\n",
        "\n",
        "    #predict proba\n",
        "    prob = model.predict_proba(x_val)\n",
        "    loss.append(log_loss(y_val,prob)) \n",
        "\n",
        "    # accuracy train & val\n",
        "    acc_train.append(model.score(x_train,y_train)) \n",
        "    acc_val.append(model.score(x_val,y_val))\n",
        "    \n",
        "    return model"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4mfXMJH3laj-",
        "outputId": "c52b811e-a52a-4804-e35d-7f8895577a57"
      },
      "source": [
        "for i in [XGB,ADA,GBT,HGBT,DT,RF,EXT]:\n",
        "    print(30*\"-\")\n",
        "    print(i)\n",
        "    score_1(i, x_train_1, y_train, x_val_1, y_val)\n",
        "    print(10*\"->\" + \" DONE\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "------------------------------\n",
            "XGBClassifier(random_state=10)\n",
            "->->->->->->->->->-> DONE\n",
            "------------------------------\n",
            "AdaBoostClassifier(random_state=40)\n",
            "->->->->->->->->->-> DONE\n",
            "------------------------------\n",
            "GradientBoostingClassifier(random_state=20)\n",
            "->->->->->->->->->-> DONE\n",
            "------------------------------\n",
            "HistGradientBoostingClassifier(random_state=30)\n",
            "->->->->->->->->->-> DONE\n",
            "------------------------------\n",
            "DecisionTreeClassifier(random_state=70)\n",
            "->->->->->->->->->-> DONE\n",
            "------------------------------\n",
            "RandomForestClassifier(random_state=60)\n",
            "->->->->->->->->->-> DONE\n",
            "------------------------------\n",
            "ExtraTreeClassifier(random_state=50)\n",
            "->->->->->->->->->-> DONE\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sms8mrx3Wbgj"
      },
      "source": [
        "##Model-2"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jXXRt4UyWfgl"
      },
      "source": [
        "x_train_2 = x_train.copy() \n",
        "x_val_2 = x_val.copy()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GQuBLSUEWuLg",
        "outputId": "b0353b7f-84f3-410f-995f-386fb38258b6"
      },
      "source": [
        "x_train_2.head()"
      ],
      "execution_count": null,
      "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>Popularity</th>\n",
              "      <th>danceability</th>\n",
              "      <th>energy</th>\n",
              "      <th>key</th>\n",
              "      <th>loudness</th>\n",
              "      <th>mode</th>\n",
              "      <th>speechiness</th>\n",
              "      <th>acousticness</th>\n",
              "      <th>instrumentalness</th>\n",
              "      <th>liveness</th>\n",
              "      <th>valence</th>\n",
              "      <th>tempo</th>\n",
              "      <th>duration_in min/ms</th>\n",
              "      <th>time_signature</th>\n",
              "      <th>length_of_artist_name</th>\n",
              "      <th>length_of_track_name</th>\n",
              "      <th>length_words_artist</th>\n",
              "      <th>length_words_track</th>\n",
              "      <th>most_popularity</th>\n",
              "      <th>most_energy</th>\n",
              "      <th>new</th>\n",
              "      <th>new2</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>13297</th>\n",
              "      <td>53.0</td>\n",
              "      <td>0.343994</td>\n",
              "      <td>0.911133</td>\n",
              "      <td>10</td>\n",
              "      <td>-5.996094</td>\n",
              "      <td>1</td>\n",
              "      <td>0.053986</td>\n",
              "      <td>0.000050</td>\n",
              "      <td>0.001660</td>\n",
              "      <td>0.064209</td>\n",
              "      <td>0.509766</td>\n",
              "      <td>135.2500</td>\n",
              "      <td>320333.000000</td>\n",
              "      <td>4</td>\n",
              "      <td>23</td>\n",
              "      <td>9</td>\n",
              "      <td>3</td>\n",
              "      <td>1</td>\n",
              "      <td>55.0</td>\n",
              "      <td>0.937988</td>\n",
              "      <td>0.627441</td>\n",
              "      <td>135.250000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>17216</th>\n",
              "      <td>71.0</td>\n",
              "      <td>0.717773</td>\n",
              "      <td>0.690918</td>\n",
              "      <td>4</td>\n",
              "      <td>-6.441406</td>\n",
              "      <td>1</td>\n",
              "      <td>0.038696</td>\n",
              "      <td>0.015900</td>\n",
              "      <td>0.003910</td>\n",
              "      <td>0.062683</td>\n",
              "      <td>0.551758</td>\n",
              "      <td>100.0000</td>\n",
              "      <td>213107.000000</td>\n",
              "      <td>4</td>\n",
              "      <td>24</td>\n",
              "      <td>14</td>\n",
              "      <td>3</td>\n",
              "      <td>3</td>\n",
              "      <td>71.0</td>\n",
              "      <td>0.716797</td>\n",
              "      <td>0.704590</td>\n",
              "      <td>100.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5014</th>\n",
              "      <td>20.0</td>\n",
              "      <td>0.637207</td>\n",
              "      <td>0.612793</td>\n",
              "      <td>8</td>\n",
              "      <td>-7.343750</td>\n",
              "      <td>1</td>\n",
              "      <td>0.074829</td>\n",
              "      <td>0.318115</td>\n",
              "      <td>0.000003</td>\n",
              "      <td>0.111023</td>\n",
              "      <td>0.570801</td>\n",
              "      <td>90.0625</td>\n",
              "      <td>4.400000</td>\n",
              "      <td>4</td>\n",
              "      <td>9</td>\n",
              "      <td>35</td>\n",
              "      <td>1</td>\n",
              "      <td>5</td>\n",
              "      <td>20.0</td>\n",
              "      <td>0.629883</td>\n",
              "      <td>0.625000</td>\n",
              "      <td>2.060547</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>385</th>\n",
              "      <td>20.0</td>\n",
              "      <td>0.269043</td>\n",
              "      <td>0.154053</td>\n",
              "      <td>9</td>\n",
              "      <td>-18.078125</td>\n",
              "      <td>1</td>\n",
              "      <td>0.038696</td>\n",
              "      <td>0.930176</td>\n",
              "      <td>0.750977</td>\n",
              "      <td>0.109009</td>\n",
              "      <td>0.104004</td>\n",
              "      <td>134.8750</td>\n",
              "      <td>6.858217</td>\n",
              "      <td>4</td>\n",
              "      <td>76</td>\n",
              "      <td>59</td>\n",
              "      <td>12</td>\n",
              "      <td>11</td>\n",
              "      <td>20.0</td>\n",
              "      <td>0.154053</td>\n",
              "      <td>0.211548</td>\n",
              "      <td>4.511719</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>16199</th>\n",
              "      <td>73.0</td>\n",
              "      <td>0.404053</td>\n",
              "      <td>0.720215</td>\n",
              "      <td>10</td>\n",
              "      <td>-9.179688</td>\n",
              "      <td>0</td>\n",
              "      <td>0.062286</td>\n",
              "      <td>0.000056</td>\n",
              "      <td>0.835938</td>\n",
              "      <td>0.046600</td>\n",
              "      <td>0.475098</td>\n",
              "      <td>121.8750</td>\n",
              "      <td>258467.000000</td>\n",
              "      <td>4</td>\n",
              "      <td>21</td>\n",
              "      <td>27</td>\n",
              "      <td>3</td>\n",
              "      <td>4</td>\n",
              "      <td>75.0</td>\n",
              "      <td>0.875977</td>\n",
              "      <td>0.562012</td>\n",
              "      <td>121.875000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       Popularity  danceability    energy  ...  most_energy       new        new2\n",
              "13297        53.0      0.343994  0.911133  ...     0.937988  0.627441  135.250000\n",
              "17216        71.0      0.717773  0.690918  ...     0.716797  0.704590  100.000000\n",
              "5014         20.0      0.637207  0.612793  ...     0.629883  0.625000    2.060547\n",
              "385          20.0      0.269043  0.154053  ...     0.154053  0.211548    4.511719\n",
              "16199        73.0      0.404053  0.720215  ...     0.875977  0.562012  121.875000\n",
              "\n",
              "[5 rows x 22 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zgFWxHYlWyI3"
      },
      "source": [
        "CB = CatBoostClassifier(random_state=929)\n",
        "LGBM = LGBMClassifier(random_state=299)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RR_zHq8JWdn4"
      },
      "source": [
        "\n",
        "def score_2(model,x_train, y_train, x_val, y_val):\n",
        "       \n",
        "    #lgbm\n",
        "    if model == LGBM:\n",
        "       model.fit(x_train,y_train,verbose=False)\n",
        "\n",
        "       #predict proba\n",
        "       prob = model.predict_proba(x_val)\n",
        "       loss.append(log_loss(y_val,prob))\n",
        "\n",
        "       # accuracy train & val\n",
        "       acc_train.append(model.score(x_train,y_train))\n",
        "       acc_val.append(model.score(x_val,y_val))\n",
        "    \n",
        "   #catboost\n",
        "    else:         \n",
        "       model.fit(x_train,y_train,verbose=False)\n",
        "\n",
        "       #predict proba\n",
        "       prob = model.predict_proba(x_val)\n",
        "       loss.append(log_loss(y_val,prob))\n",
        "\n",
        "       # accuracy train & val\n",
        "       acc_train.append(model.score(x_train,y_train))\n",
        "       acc_val.append(model.score(x_val,y_val))\n",
        "    \n",
        "\n",
        "    return model"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sZmzQgE2W9LF",
        "outputId": "57d66ae3-95a1-4217-ac24-04fbd114d4dc"
      },
      "source": [
        "for i in [CB,LGBM]:\n",
        "    print(30*\"-\")\n",
        "    print(i)\n",
        "    score_2(i, x_train_2, y_train, x_val_2, y_val)\n",
        "    print(10*\"->\" + \" DONE\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "------------------------------\n",
            "<catboost.core.CatBoostClassifier object at 0x7f10d3540990>\n",
            "->->->->->->->->->-> DONE\n",
            "------------------------------\n",
            "LGBMClassifier(random_state=299)\n",
            "->->->->->->->->->-> DONE\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_kUwZQMGXZi4"
      },
      "source": [
        "## Model-3"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HovMO-d9XcF7"
      },
      "source": [
        "x_train_3 = x_train.copy()\n",
        "x_val_3 = x_val.copy()"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8zJzHf1pXiqo"
      },
      "source": [
        "sc = StandardScaler()\n",
        "x_train_3 = sc.fit_transform(x_train_3)\n",
        "x_val_3 = sc.transform(x_val_3)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MRTzkU2hXwIP"
      },
      "source": [
        "LR = LogisticRegression()\n",
        "#KNN = KNeighborsRegressor()\n",
        "#SGDC = SGDClassifier(random_state=2828)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MVzkuor_YCSj"
      },
      "source": [
        "def score_3(model,x_train, y_train, x_val, y_val):\n",
        "\n",
        "    model.fit(x_train,y_train) \n",
        "\n",
        "    # predict proba\n",
        "    prob = model.predict_proba(x_val)\n",
        "    loss.append(log_loss(y_val,prob))\n",
        "\n",
        "    # accuracy train & val\n",
        "    acc_train.append(model.score(x_train,y_train))\n",
        "    acc_val.append(model.score(x_val,y_val))\n",
        "    \n",
        "    return model"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "F41WSBPgYVsD",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a8f02624-1108-47b4-e0c2-1236575f50e9"
      },
      "source": [
        "for i in [LR]:\n",
        "    print(30*\"-\")\n",
        "    print(i)\n",
        "    score_3(i, x_train_3, y_train, x_val_3, y_val)\n",
        "    print(10*\"->\" + \" DONE\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "------------------------------\n",
            "LogisticRegression()\n",
            "->->->->->->->->->-> DONE\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7dmHHk-iWPuW"
      },
      "source": [
        "## Model-selection"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GG3SUvslnIu-"
      },
      "source": [
        "all_models = pd.DataFrame({\"models\": [\"XGB\", \"ADA\", \"GBT\", \"HGBT\", \"DT\", \"RF\", \"EXT\",\"CB\",\"LGBM\",\"LR\"], \n",
        "                          \"Log_loss\": loss, \n",
        "                          \"Train_accuracy\": acc_train,\n",
        "                          \"Val_accuracy\": acc_val})\n",
        "#all_models = all_models.set_index(\"models\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jCE_VQVHnVXo",
        "outputId": "ed98599d-2319-4ca4-8559-9c5f125db884"
      },
      "source": [
        "all_models"
      ],
      "execution_count": null,
      "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>models</th>\n",
              "      <th>Log_loss</th>\n",
              "      <th>Train_accuracy</th>\n",
              "      <th>Val_accuracy</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>XGB</td>\n",
              "      <td>1.156597</td>\n",
              "      <td>0.615796</td>\n",
              "      <td>0.565278</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>ADA</td>\n",
              "      <td>1.870463</td>\n",
              "      <td>0.402126</td>\n",
              "      <td>0.403056</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>GBT</td>\n",
              "      <td>1.133772</td>\n",
              "      <td>0.657335</td>\n",
              "      <td>0.574444</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>HGBT</td>\n",
              "      <td>1.111064</td>\n",
              "      <td>0.769936</td>\n",
              "      <td>0.581667</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>DT</td>\n",
              "      <td>19.292372</td>\n",
              "      <td>0.921784</td>\n",
              "      <td>0.418889</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>5</th>\n",
              "      <td>RF</td>\n",
              "      <td>1.456237</td>\n",
              "      <td>0.921784</td>\n",
              "      <td>0.540000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>6</th>\n",
              "      <td>EXT</td>\n",
              "      <td>21.853486</td>\n",
              "      <td>0.921784</td>\n",
              "      <td>0.341667</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7</th>\n",
              "      <td>CB</td>\n",
              "      <td>1.104484</td>\n",
              "      <td>0.823979</td>\n",
              "      <td>0.574444</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>8</th>\n",
              "      <td>LGBM</td>\n",
              "      <td>1.094712</td>\n",
              "      <td>0.777716</td>\n",
              "      <td>0.583333</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>9</th>\n",
              "      <td>LR</td>\n",
              "      <td>1.279637</td>\n",
              "      <td>0.534593</td>\n",
              "      <td>0.513611</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  models   Log_loss  Train_accuracy  Val_accuracy\n",
              "0    XGB   1.156597        0.615796      0.565278\n",
              "1    ADA   1.870463        0.402126      0.403056\n",
              "2    GBT   1.133772        0.657335      0.574444\n",
              "3   HGBT   1.111064        0.769936      0.581667\n",
              "4     DT  19.292372        0.921784      0.418889\n",
              "5     RF   1.456237        0.921784      0.540000\n",
              "6    EXT  21.853486        0.921784      0.341667\n",
              "7     CB   1.104484        0.823979      0.574444\n",
              "8   LGBM   1.094712        0.777716      0.583333\n",
              "9     LR   1.279637        0.534593      0.513611"
            ]
          },
          "metadata": {},
          "execution_count": 51
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "274QRqloa9vZ",
        "outputId": "f6d22a3f-aceb-47cb-b10c-b49e830749f1"
      },
      "source": [
        "# plot log_loss\n",
        "\n",
        "plt.figure(figsize=(17, 8)) \n",
        "sns.barplot(x = \"models\", y=\"Log_loss\", data=all_models.sort_values(\"Log_loss\", ascending=True))\n",
        "plt.title(\"Log_loss\")\n",
        "plt.xlabel(\"Models\")\n",
        "plt.ylabel(\"Log_loss\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0, 0.5, 'Log_loss')"
            ]
          },
          "metadata": {},
          "execution_count": 33
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1224x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KGstheoRcLll"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oKGk6dB_cL8S"
      },
      "source": [
        "##Hyperparameters - tuning"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GrG_MbYmcUVX"
      },
      "source": [
        "#lgbm\n",
        "def objective(trial):\n",
        "    params = {\n",
        "        #\"device_type\": trial.suggest_categorical(\"device_type\", ['gpu']),\n",
        "        #'objective': trial.suggest_categorical('objective',['multiclass']), \n",
        "        \"n_estimators\": trial.suggest_categorical(\"n_estimators\", [10000]),\n",
        "        \"learning_rate\": trial.suggest_float(\"learning_rate\", 0.01, 0.3),\n",
        "        \"num_leaves\": trial.suggest_int(\"num_leaves\", 20, 3000, step=20),\n",
        "        \"max_depth\": trial.suggest_int(\"max_depth\", 3, 12),\n",
        "        \"min_data_in_leaf\": trial.suggest_int(\"min_data_in_leaf\", 200, 10000, step=100),\n",
        "        \"lambda_l1\": trial.suggest_int(\"lambda_l1\", 0, 100, step=5),\n",
        "        \"lambda_l2\": trial.suggest_int(\"lambda_l2\", 0, 100, step=5),\n",
        "        \"min_gain_to_split\": trial.suggest_float(\"min_gain_to_split\", 0, 15),\n",
        "        \"bagging_fraction\": trial.suggest_float(\"bagging_fraction\", 0.2, 0.95, step=0.1),\n",
        "        \"bagging_freq\": trial.suggest_categorical(\"bagging_freq\", [1]),\n",
        "        \"feature_fraction\": trial.suggest_float(\"feature_fraction\", 0.2, 0.95, step=0.1),\n",
        "    }\n",
        "    model = LGBMClassifier(**params)  \n",
        "    model.fit(x_train,y_train,eval_set=[(x_val,y_val)],verbose=False,early_stopping_rounds=100)\n",
        "\n",
        "    y_preds = model.predict_proba(x_val)  \n",
        "    roc = log_loss(y_val, y_preds)\n",
        "\n",
        "    return roc"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4KisF1ShcfMs",
        "outputId": "3d8cd4da-988f-46ce-c47c-c0ffc803242d"
      },
      "source": [
        "study = optuna.create_study(direction='minimize') #for loss\n",
        "study.optimize(objective, n_trials=75)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\u001b[32m[I 2021-10-26 07:51:25,004]\u001b[0m A new study created in memory with name: no-name-4238c3b0-019b-4e28-8ff5-378169b1c5dc\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:25,817]\u001b[0m Trial 0 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.01450885251916211, 'num_leaves': 1200, 'max_depth': 8, 'min_data_in_leaf': 3900, 'lambda_l1': 15, 'lambda_l2': 40, 'min_gain_to_split': 12.584108810423245, 'bagging_fraction': 0.30000000000000004, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 0 with value: 2.136604827384357.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:26,657]\u001b[0m Trial 1 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.07384125442201207, 'num_leaves': 1620, 'max_depth': 4, 'min_data_in_leaf': 6300, 'lambda_l1': 5, 'lambda_l2': 45, 'min_gain_to_split': 3.644010353296009, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 0 with value: 2.136604827384357.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:27,434]\u001b[0m Trial 2 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.14716561893270638, 'num_leaves': 920, 'max_depth': 7, 'min_data_in_leaf': 3100, 'lambda_l1': 55, 'lambda_l2': 20, 'min_gain_to_split': 12.21216738915156, 'bagging_fraction': 0.2, 'bagging_freq': 1, 'feature_fraction': 0.30000000000000004}. Best is trial 0 with value: 2.136604827384357.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:28,116]\u001b[0m Trial 3 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.09053707624139422, 'num_leaves': 2820, 'max_depth': 7, 'min_data_in_leaf': 9600, 'lambda_l1': 40, 'lambda_l2': 50, 'min_gain_to_split': 6.056556008643773, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 0 with value: 2.136604827384357.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:30,539]\u001b[0m Trial 4 finished with value: 1.58715341401846 and parameters: {'n_estimators': 10000, 'learning_rate': 0.11846356760961273, 'num_leaves': 2140, 'max_depth': 12, 'min_data_in_leaf': 2500, 'lambda_l1': 30, 'lambda_l2': 15, 'min_gain_to_split': 14.391187573355184, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 4 with value: 1.58715341401846.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:32,299]\u001b[0m Trial 5 finished with value: 2.073845303289843 and parameters: {'n_estimators': 10000, 'learning_rate': 0.05873998386069488, 'num_leaves': 2740, 'max_depth': 9, 'min_data_in_leaf': 600, 'lambda_l1': 100, 'lambda_l2': 55, 'min_gain_to_split': 6.531740158722722, 'bagging_fraction': 0.2, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 4 with value: 1.58715341401846.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:35,429]\u001b[0m Trial 6 finished with value: 2.160115201228966 and parameters: {'n_estimators': 10000, 'learning_rate': 0.07682035894111704, 'num_leaves': 2240, 'max_depth': 7, 'min_data_in_leaf': 500, 'lambda_l1': 95, 'lambda_l2': 90, 'min_gain_to_split': 6.599761862522039, 'bagging_fraction': 0.4, 'bagging_freq': 1, 'feature_fraction': 0.30000000000000004}. Best is trial 4 with value: 1.58715341401846.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:36,119]\u001b[0m Trial 7 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.06909153699997972, 'num_leaves': 120, 'max_depth': 8, 'min_data_in_leaf': 9700, 'lambda_l1': 80, 'lambda_l2': 95, 'min_gain_to_split': 4.404869563303725, 'bagging_fraction': 0.5, 'bagging_freq': 1, 'feature_fraction': 0.2}. Best is trial 4 with value: 1.58715341401846.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:36,945]\u001b[0m Trial 8 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.24185975367978454, 'num_leaves': 2400, 'max_depth': 3, 'min_data_in_leaf': 3900, 'lambda_l1': 100, 'lambda_l2': 35, 'min_gain_to_split': 14.96497184954366, 'bagging_fraction': 0.2, 'bagging_freq': 1, 'feature_fraction': 0.2}. Best is trial 4 with value: 1.58715341401846.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:37,803]\u001b[0m Trial 9 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.23823312938899682, 'num_leaves': 2760, 'max_depth': 6, 'min_data_in_leaf': 6000, 'lambda_l1': 20, 'lambda_l2': 90, 'min_gain_to_split': 4.168403281578556, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.30000000000000004}. Best is trial 4 with value: 1.58715341401846.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:44,012]\u001b[0m Trial 10 finished with value: 1.329410594263911 and parameters: {'n_estimators': 10000, 'learning_rate': 0.17657312509603862, 'num_leaves': 1820, 'max_depth': 12, 'min_data_in_leaf': 2200, 'lambda_l1': 45, 'lambda_l2': 5, 'min_gain_to_split': 0.0884891640639891, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 10 with value: 1.329410594263911.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:48,257]\u001b[0m Trial 11 finished with value: 1.3475950538491912 and parameters: {'n_estimators': 10000, 'learning_rate': 0.16894804150825168, 'num_leaves': 1840, 'max_depth': 12, 'min_data_in_leaf': 2000, 'lambda_l1': 45, 'lambda_l2': 0, 'min_gain_to_split': 0.6993258978551253, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 10 with value: 1.329410594263911.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:51,021]\u001b[0m Trial 12 finished with value: 1.363148872032055 and parameters: {'n_estimators': 10000, 'learning_rate': 0.19452125604952392, 'num_leaves': 1740, 'max_depth': 12, 'min_data_in_leaf': 1900, 'lambda_l1': 60, 'lambda_l2': 5, 'min_gain_to_split': 0.5690646896261485, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 10 with value: 1.329410594263911.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:51:57,546]\u001b[0m Trial 13 finished with value: 1.6296999187120194 and parameters: {'n_estimators': 10000, 'learning_rate': 0.19026687391450076, 'num_leaves': 880, 'max_depth': 10, 'min_data_in_leaf': 4900, 'lambda_l1': 70, 'lambda_l2': 0, 'min_gain_to_split': 0.07669266052884482, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 10 with value: 1.329410594263911.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:00,664]\u001b[0m Trial 14 finished with value: 1.357437512595016 and parameters: {'n_estimators': 10000, 'learning_rate': 0.18422270577618416, 'num_leaves': 1860, 'max_depth': 11, 'min_data_in_leaf': 2200, 'lambda_l1': 35, 'lambda_l2': 20, 'min_gain_to_split': 2.011316974697424, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.4}. Best is trial 10 with value: 1.329410594263911.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:02,584]\u001b[0m Trial 15 finished with value: 1.4774616797566695 and parameters: {'n_estimators': 10000, 'learning_rate': 0.297354435467148, 'num_leaves': 1280, 'max_depth': 10, 'min_data_in_leaf': 1300, 'lambda_l1': 50, 'lambda_l2': 75, 'min_gain_to_split': 9.275197757457484, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 10 with value: 1.329410594263911.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:03,259]\u001b[0m Trial 16 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.1417139714086964, 'num_leaves': 400, 'max_depth': 11, 'min_data_in_leaf': 7600, 'lambda_l1': 45, 'lambda_l2': 0, 'min_gain_to_split': 2.0332965339302547, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.4}. Best is trial 10 with value: 1.329410594263911.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:06,144]\u001b[0m Trial 17 finished with value: 1.9250554236308897 and parameters: {'n_estimators': 10000, 'learning_rate': 0.23152329420523934, 'num_leaves': 2000, 'max_depth': 12, 'min_data_in_leaf': 3400, 'lambda_l1': 70, 'lambda_l2': 30, 'min_gain_to_split': 2.112347571747755, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 10 with value: 1.329410594263911.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:07,949]\u001b[0m Trial 18 finished with value: 1.735976536601773 and parameters: {'n_estimators': 10000, 'learning_rate': 0.28273759394171044, 'num_leaves': 1340, 'max_depth': 10, 'min_data_in_leaf': 5000, 'lambda_l1': 25, 'lambda_l2': 65, 'min_gain_to_split': 8.995216435550475, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.4}. Best is trial 10 with value: 1.329410594263911.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:15,996]\u001b[0m Trial 19 finished with value: 1.1380643594384816 and parameters: {'n_estimators': 10000, 'learning_rate': 0.16815294025998284, 'num_leaves': 880, 'max_depth': 5, 'min_data_in_leaf': 1700, 'lambda_l1': 0, 'lambda_l2': 10, 'min_gain_to_split': 0.9611821778050054, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:20,327]\u001b[0m Trial 20 finished with value: 1.1930754550359755 and parameters: {'n_estimators': 10000, 'learning_rate': 0.11395286504586583, 'num_leaves': 580, 'max_depth': 4, 'min_data_in_leaf': 200, 'lambda_l1': 10, 'lambda_l2': 15, 'min_gain_to_split': 2.5778152377693493, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:24,831]\u001b[0m Trial 21 finished with value: 1.1417579906376043 and parameters: {'n_estimators': 10000, 'learning_rate': 0.11914684539528111, 'num_leaves': 660, 'max_depth': 5, 'min_data_in_leaf': 200, 'lambda_l1': 0, 'lambda_l2': 15, 'min_gain_to_split': 3.065901776588544, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:30,152]\u001b[0m Trial 22 finished with value: 1.14466023926356 and parameters: {'n_estimators': 10000, 'learning_rate': 0.11539284328057053, 'num_leaves': 560, 'max_depth': 5, 'min_data_in_leaf': 200, 'lambda_l1': 0, 'lambda_l2': 25, 'min_gain_to_split': 3.0901314839902905, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:34,094]\u001b[0m Trial 23 finished with value: 1.2130283677150937 and parameters: {'n_estimators': 10000, 'learning_rate': 0.1262736220648553, 'num_leaves': 680, 'max_depth': 5, 'min_data_in_leaf': 1200, 'lambda_l1': 0, 'lambda_l2': 30, 'min_gain_to_split': 5.076423094781566, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:39,877]\u001b[0m Trial 24 finished with value: 1.1735202857756888 and parameters: {'n_estimators': 10000, 'learning_rate': 0.10599964144960929, 'num_leaves': 180, 'max_depth': 5, 'min_data_in_leaf': 1000, 'lambda_l1': 0, 'lambda_l2': 25, 'min_gain_to_split': 3.159511113950482, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:45,184]\u001b[0m Trial 25 finished with value: 1.2791497414005057 and parameters: {'n_estimators': 10000, 'learning_rate': 0.036743087604136, 'num_leaves': 1040, 'max_depth': 5, 'min_data_in_leaf': 200, 'lambda_l1': 10, 'lambda_l2': 10, 'min_gain_to_split': 8.136947283040458, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:47,446]\u001b[0m Trial 26 finished with value: 1.2990505118080287 and parameters: {'n_estimators': 10000, 'learning_rate': 0.21001813187556032, 'num_leaves': 420, 'max_depth': 3, 'min_data_in_leaf': 1400, 'lambda_l1': 20, 'lambda_l2': 25, 'min_gain_to_split': 5.322473564623111, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:52:58,538]\u001b[0m Trial 27 finished with value: 1.2191826746172632 and parameters: {'n_estimators': 10000, 'learning_rate': 0.15766003641215592, 'num_leaves': 680, 'max_depth': 6, 'min_data_in_leaf': 2800, 'lambda_l1': 0, 'lambda_l2': 10, 'min_gain_to_split': 1.3394377237038293, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:02,720]\u001b[0m Trial 28 finished with value: 1.2348954111907677 and parameters: {'n_estimators': 10000, 'learning_rate': 0.1398803017432503, 'num_leaves': 20, 'max_depth': 4, 'min_data_in_leaf': 1400, 'lambda_l1': 10, 'lambda_l2': 35, 'min_gain_to_split': 3.063554043020852, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:03,478]\u001b[0m Trial 29 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.01307432540138767, 'num_leaves': 1120, 'max_depth': 6, 'min_data_in_leaf': 3800, 'lambda_l1': 15, 'lambda_l2': 40, 'min_gain_to_split': 1.3153026744054475, 'bagging_fraction': 0.4, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:05,831]\u001b[0m Trial 30 finished with value: 1.4151647581980902 and parameters: {'n_estimators': 10000, 'learning_rate': 0.0908510185983247, 'num_leaves': 380, 'max_depth': 5, 'min_data_in_leaf': 800, 'lambda_l1': 20, 'lambda_l2': 15, 'min_gain_to_split': 10.486632988961484, 'bagging_fraction': 0.5, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:10,446]\u001b[0m Trial 31 finished with value: 1.1816612350430757 and parameters: {'n_estimators': 10000, 'learning_rate': 0.10365278173979622, 'num_leaves': 200, 'max_depth': 5, 'min_data_in_leaf': 900, 'lambda_l1': 0, 'lambda_l2': 20, 'min_gain_to_split': 3.7544912130207515, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:15,498]\u001b[0m Trial 32 finished with value: 1.182459040951064 and parameters: {'n_estimators': 10000, 'learning_rate': 0.10620324635916464, 'num_leaves': 820, 'max_depth': 6, 'min_data_in_leaf': 200, 'lambda_l1': 5, 'lambda_l2': 25, 'min_gain_to_split': 3.229915951124986, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:26,988]\u001b[0m Trial 33 finished with value: 1.2388351466129355 and parameters: {'n_estimators': 10000, 'learning_rate': 0.04045096956498076, 'num_leaves': 1460, 'max_depth': 4, 'min_data_in_leaf': 1700, 'lambda_l1': 0, 'lambda_l2': 45, 'min_gain_to_split': 5.2041054447561255, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:27,684]\u001b[0m Trial 34 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.12831523310773443, 'num_leaves': 300, 'max_depth': 3, 'min_data_in_leaf': 7400, 'lambda_l1': 10, 'lambda_l2': 25, 'min_gain_to_split': 1.5146759873379894, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:31,356]\u001b[0m Trial 35 finished with value: 1.197881230402782 and parameters: {'n_estimators': 10000, 'learning_rate': 0.158611820564874, 'num_leaves': 540, 'max_depth': 4, 'min_data_in_leaf': 900, 'lambda_l1': 5, 'lambda_l2': 10, 'min_gain_to_split': 3.211842607090641, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:35,481]\u001b[0m Trial 36 finished with value: 1.4728519894203218 and parameters: {'n_estimators': 10000, 'learning_rate': 0.093279213631023, 'num_leaves': 980, 'max_depth': 7, 'min_data_in_leaf': 3000, 'lambda_l1': 30, 'lambda_l2': 55, 'min_gain_to_split': 4.701256531691218, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:36,168]\u001b[0m Trial 37 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.05451086607060071, 'num_leaves': 760, 'max_depth': 5, 'min_data_in_leaf': 8900, 'lambda_l1': 15, 'lambda_l2': 40, 'min_gain_to_split': 7.16944528985592, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:39,514]\u001b[0m Trial 38 finished with value: 1.3182622363692271 and parameters: {'n_estimators': 10000, 'learning_rate': 0.13470818517973276, 'num_leaves': 520, 'max_depth': 8, 'min_data_in_leaf': 2600, 'lambda_l1': 5, 'lambda_l2': 20, 'min_gain_to_split': 5.94757255006294, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:40,223]\u001b[0m Trial 39 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.07057378689115089, 'num_leaves': 200, 'max_depth': 6, 'min_data_in_leaf': 4500, 'lambda_l1': 30, 'lambda_l2': 30, 'min_gain_to_split': 12.574155489658537, 'bagging_fraction': 0.30000000000000004, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:43,042]\u001b[0m Trial 40 finished with value: 1.2577742351188181 and parameters: {'n_estimators': 10000, 'learning_rate': 0.20719454793805786, 'num_leaves': 80, 'max_depth': 7, 'min_data_in_leaf': 700, 'lambda_l1': 15, 'lambda_l2': 15, 'min_gain_to_split': 2.593408972778103, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.9}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:47,871]\u001b[0m Trial 41 finished with value: 1.1875224960731774 and parameters: {'n_estimators': 10000, 'learning_rate': 0.10838407007958206, 'num_leaves': 200, 'max_depth': 5, 'min_data_in_leaf': 900, 'lambda_l1': 0, 'lambda_l2': 20, 'min_gain_to_split': 4.207327160177095, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:53,617]\u001b[0m Trial 42 finished with value: 1.1919220201155656 and parameters: {'n_estimators': 10000, 'learning_rate': 0.09549000633941936, 'num_leaves': 260, 'max_depth': 5, 'min_data_in_leaf': 1600, 'lambda_l1': 0, 'lambda_l2': 5, 'min_gain_to_split': 3.75858499155507, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:53:59,799]\u001b[0m Trial 43 finished with value: 1.2107952119487815 and parameters: {'n_estimators': 10000, 'learning_rate': 0.08074355951840151, 'num_leaves': 2980, 'max_depth': 6, 'min_data_in_leaf': 1000, 'lambda_l1': 5, 'lambda_l2': 20, 'min_gain_to_split': 3.803933024413221, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:54:03,569]\u001b[0m Trial 44 finished with value: 1.193120259679639 and parameters: {'n_estimators': 10000, 'learning_rate': 0.15241837241775447, 'num_leaves': 620, 'max_depth': 4, 'min_data_in_leaf': 500, 'lambda_l1': 5, 'lambda_l2': 35, 'min_gain_to_split': 2.871719903704611, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.7}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:54:12,708]\u001b[0m Trial 45 finished with value: 1.211273071857456 and parameters: {'n_estimators': 10000, 'learning_rate': 0.12192110568225153, 'num_leaves': 480, 'max_depth': 3, 'min_data_in_leaf': 2200, 'lambda_l1': 10, 'lambda_l2': 25, 'min_gain_to_split': 0.8039085181966228, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:54:13,447]\u001b[0m Trial 46 finished with value: 2.136604827384357 and parameters: {'n_estimators': 10000, 'learning_rate': 0.10628242299616185, 'num_leaves': 300, 'max_depth': 5, 'min_data_in_leaf': 5900, 'lambda_l1': 25, 'lambda_l2': 10, 'min_gain_to_split': 6.026936929519299, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:54:16,105]\u001b[0m Trial 47 finished with value: 1.4510174494670016 and parameters: {'n_estimators': 10000, 'learning_rate': 0.16460475575778183, 'num_leaves': 800, 'max_depth': 4, 'min_data_in_leaf': 1800, 'lambda_l1': 85, 'lambda_l2': 15, 'min_gain_to_split': 1.752607842419257, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.8}. Best is trial 19 with value: 1.1380643594384816.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:54:31,088]\u001b[0m Trial 48 finished with value: 1.0934061861873998 and parameters: {'n_estimators': 10000, 'learning_rate': 0.08179834985372536, 'num_leaves': 160, 'max_depth': 7, 'min_data_in_leaf': 700, 'lambda_l1': 0, 'lambda_l2': 5, 'min_gain_to_split': 0.8351071746419583, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:54:43,038]\u001b[0m Trial 49 finished with value: 1.1838632624883754 and parameters: {'n_estimators': 10000, 'learning_rate': 0.05735688242518877, 'num_leaves': 1180, 'max_depth': 8, 'min_data_in_leaf': 400, 'lambda_l1': 15, 'lambda_l2': 5, 'min_gain_to_split': 0.9856643473776541, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.4}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:55:28,723]\u001b[0m Trial 50 finished with value: 1.1869925306340763 and parameters: {'n_estimators': 10000, 'learning_rate': 0.07847357297951606, 'num_leaves': 20, 'max_depth': 9, 'min_data_in_leaf': 2400, 'lambda_l1': 5, 'lambda_l2': 0, 'min_gain_to_split': 0.03634005114182148, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:55:35,911]\u001b[0m Trial 51 finished with value: 1.1530083103250137 and parameters: {'n_estimators': 10000, 'learning_rate': 0.10207635932722109, 'num_leaves': 140, 'max_depth': 6, 'min_data_in_leaf': 1100, 'lambda_l1': 0, 'lambda_l2': 10, 'min_gain_to_split': 2.43134382860853, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:55:40,989]\u001b[0m Trial 52 finished with value: 1.154576834590636 and parameters: {'n_estimators': 10000, 'learning_rate': 0.1443360410655525, 'num_leaves': 140, 'max_depth': 7, 'min_data_in_leaf': 1200, 'lambda_l1': 0, 'lambda_l2': 10, 'min_gain_to_split': 2.352123760904561, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:55:44,850]\u001b[0m Trial 53 finished with value: 1.2142779611165617 and parameters: {'n_estimators': 10000, 'learning_rate': 0.17714836373062914, 'num_leaves': 380, 'max_depth': 7, 'min_data_in_leaf': 1500, 'lambda_l1': 10, 'lambda_l2': 5, 'min_gain_to_split': 2.254730256452063, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.4}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:55:51,014]\u001b[0m Trial 54 finished with value: 1.20243663882268 and parameters: {'n_estimators': 10000, 'learning_rate': 0.1430768842976036, 'num_leaves': 920, 'max_depth': 7, 'min_data_in_leaf': 600, 'lambda_l1': 20, 'lambda_l2': 10, 'min_gain_to_split': 0.6848782608032966, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:55:56,859]\u001b[0m Trial 55 finished with value: 1.2129725682005743 and parameters: {'n_estimators': 10000, 'learning_rate': 0.11802603940217628, 'num_leaves': 640, 'max_depth': 6, 'min_data_in_leaf': 2000, 'lambda_l1': 5, 'lambda_l2': 0, 'min_gain_to_split': 2.3822423875048626, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:56:02,810]\u001b[0m Trial 56 finished with value: 1.1135805689753329 and parameters: {'n_estimators': 10000, 'learning_rate': 0.15186291022351978, 'num_leaves': 120, 'max_depth': 7, 'min_data_in_leaf': 500, 'lambda_l1': 0, 'lambda_l2': 15, 'min_gain_to_split': 1.565813545948548, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:56:05,165]\u001b[0m Trial 57 finished with value: 1.621758159671639 and parameters: {'n_estimators': 10000, 'learning_rate': 0.1298447102660791, 'num_leaves': 440, 'max_depth': 9, 'min_data_in_leaf': 500, 'lambda_l1': 10, 'lambda_l2': 15, 'min_gain_to_split': 13.74435744948202, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:56:11,664]\u001b[0m Trial 58 finished with value: 1.1568541676578672 and parameters: {'n_estimators': 10000, 'learning_rate': 0.17044143988496574, 'num_leaves': 300, 'max_depth': 8, 'min_data_in_leaf': 300, 'lambda_l1': 5, 'lambda_l2': 75, 'min_gain_to_split': 1.2979645702279594, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.30000000000000004}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:56:17,014]\u001b[0m Trial 59 finished with value: 1.410814659239801 and parameters: {'n_estimators': 10000, 'learning_rate': 0.08290356594842878, 'num_leaves': 100, 'max_depth': 6, 'min_data_in_leaf': 1200, 'lambda_l1': 60, 'lambda_l2': 5, 'min_gain_to_split': 0.3284018724674569, 'bagging_fraction': 0.6000000000000001, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:56:47,165]\u001b[0m Trial 60 finished with value: 1.2890077956551225 and parameters: {'n_estimators': 10000, 'learning_rate': 0.04386355935611176, 'num_leaves': 1580, 'max_depth': 6, 'min_data_in_leaf': 3500, 'lambda_l1': 0, 'lambda_l2': 0, 'min_gain_to_split': 1.914756173479754, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.4}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:56:57,212]\u001b[0m Trial 61 finished with value: 1.1178821314933267 and parameters: {'n_estimators': 10000, 'learning_rate': 0.1484629658676244, 'num_leaves': 120, 'max_depth': 7, 'min_data_in_leaf': 1200, 'lambda_l1': 0, 'lambda_l2': 10, 'min_gain_to_split': 1.0013170701603606, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:57:05,779]\u001b[0m Trial 62 finished with value: 1.0983054659557892 and parameters: {'n_estimators': 10000, 'learning_rate': 0.1492253339714773, 'num_leaves': 740, 'max_depth': 7, 'min_data_in_leaf': 600, 'lambda_l1': 0, 'lambda_l2': 15, 'min_gain_to_split': 0.8418693909846142, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:57:09,526]\u001b[0m Trial 63 finished with value: 1.1771283514486237 and parameters: {'n_estimators': 10000, 'learning_rate': 0.20137622793603374, 'num_leaves': 1060, 'max_depth': 7, 'min_data_in_leaf': 200, 'lambda_l1': 15, 'lambda_l2': 15, 'min_gain_to_split': 0.9388601144010922, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:57:18,278]\u001b[0m Trial 64 finished with value: 1.111012142663804 and parameters: {'n_estimators': 10000, 'learning_rate': 0.18572616935544586, 'num_leaves': 760, 'max_depth': 8, 'min_data_in_leaf': 700, 'lambda_l1': 5, 'lambda_l2': 20, 'min_gain_to_split': 0.3227479228696942, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:57:25,019]\u001b[0m Trial 65 finished with value: 1.1047918717121976 and parameters: {'n_estimators': 10000, 'learning_rate': 0.22152414402461307, 'num_leaves': 760, 'max_depth': 8, 'min_data_in_leaf': 700, 'lambda_l1': 5, 'lambda_l2': 5, 'min_gain_to_split': 0.2895400919892792, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:57:35,583]\u001b[0m Trial 66 finished with value: 1.180726178268044 and parameters: {'n_estimators': 10000, 'learning_rate': 0.23065593658303213, 'num_leaves': 1340, 'max_depth': 8, 'min_data_in_leaf': 1800, 'lambda_l1': 10, 'lambda_l2': 5, 'min_gain_to_split': 0.5341546633913232, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:57:41,216]\u001b[0m Trial 67 finished with value: 1.2120675148054367 and parameters: {'n_estimators': 10000, 'learning_rate': 0.2608660644900196, 'num_leaves': 2440, 'max_depth': 9, 'min_data_in_leaf': 1500, 'lambda_l1': 25, 'lambda_l2': 5, 'min_gain_to_split': 0.14123190987008416, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:57:50,151]\u001b[0m Trial 68 finished with value: 1.1604059263026552 and parameters: {'n_estimators': 10000, 'learning_rate': 0.22152945988714512, 'num_leaves': 900, 'max_depth': 8, 'min_data_in_leaf': 700, 'lambda_l1': 5, 'lambda_l2': 100, 'min_gain_to_split': 1.1037787201339389, 'bagging_fraction': 0.7, 'bagging_freq': 1, 'feature_fraction': 0.4}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:57:55,956]\u001b[0m Trial 69 finished with value: 1.3483568157829697 and parameters: {'n_estimators': 10000, 'learning_rate': 0.18444397139057683, 'num_leaves': 760, 'max_depth': 9, 'min_data_in_leaf': 2200, 'lambda_l1': 10, 'lambda_l2': 20, 'min_gain_to_split': 1.8558907736282233, 'bagging_fraction': 0.5, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:57:58,230]\u001b[0m Trial 70 finished with value: 1.3078220813340684 and parameters: {'n_estimators': 10000, 'learning_rate': 0.2534031104275999, 'num_leaves': 720, 'max_depth': 8, 'min_data_in_leaf': 600, 'lambda_l1': 40, 'lambda_l2': 10, 'min_gain_to_split': 1.5499126951889455, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.5}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:58:06,190]\u001b[0m Trial 71 finished with value: 1.1010690010119117 and parameters: {'n_estimators': 10000, 'learning_rate': 0.19086478454278022, 'num_leaves': 880, 'max_depth': 7, 'min_data_in_leaf': 1100, 'lambda_l1': 0, 'lambda_l2': 15, 'min_gain_to_split': 0.4901926242197584, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:58:16,234]\u001b[0m Trial 72 finished with value: 1.1028073473114228 and parameters: {'n_estimators': 10000, 'learning_rate': 0.19645390420770312, 'num_leaves': 1020, 'max_depth': 8, 'min_data_in_leaf': 1400, 'lambda_l1': 0, 'lambda_l2': 15, 'min_gain_to_split': 0.35595574521603224, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:58:26,948]\u001b[0m Trial 73 finished with value: 1.107497980307151 and parameters: {'n_estimators': 10000, 'learning_rate': 0.1949179874792965, 'num_leaves': 1280, 'max_depth': 7, 'min_data_in_leaf': 1300, 'lambda_l1': 5, 'lambda_l2': 30, 'min_gain_to_split': 0.06909999096966912, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n",
            "\u001b[32m[I 2021-10-26 07:58:33,728]\u001b[0m Trial 74 finished with value: 1.1186067885686084 and parameters: {'n_estimators': 10000, 'learning_rate': 0.19481705189503193, 'num_leaves': 1320, 'max_depth': 8, 'min_data_in_leaf': 800, 'lambda_l1': 5, 'lambda_l2': 15, 'min_gain_to_split': 0.5594884407511969, 'bagging_fraction': 0.9, 'bagging_freq': 1, 'feature_fraction': 0.6000000000000001}. Best is trial 48 with value: 1.0934061861873998.\u001b[0m\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pneHs7VqckHS",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7e24c746-2411-4d0f-faec-7249ae7451a8"
      },
      "source": [
        "print('Number of finished trials:', len(study.trials))\n",
        "print('Best trial:', study.best_trial.params)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of finished trials: 75\n",
            "Best trial: {'n_estimators': 10000, 'learning_rate': 0.08179834985372536, 'num_leaves': 160, 'max_depth': 7, 'min_data_in_leaf': 700, 'lambda_l1': 0, 'lambda_l2': 5, 'min_gain_to_split': 0.8351071746419583, 'bagging_fraction': 0.8, 'bagging_freq': 1, 'feature_fraction': 0.5}\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zNQnci0Wm-1h",
        "outputId": "f193ff48-1546-4127-bdac-940786a56381"
      },
      "source": [
        "lg_params = study.best_trial.params\n",
        "lg_params"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'bagging_fraction': 0.8,\n",
              " 'bagging_freq': 1,\n",
              " 'feature_fraction': 0.5,\n",
              " 'lambda_l1': 0,\n",
              " 'lambda_l2': 5,\n",
              " 'learning_rate': 0.08179834985372536,\n",
              " 'max_depth': 7,\n",
              " 'min_data_in_leaf': 700,\n",
              " 'min_gain_to_split': 0.8351071746419583,\n",
              " 'n_estimators': 10000,\n",
              " 'num_leaves': 160}"
            ]
          },
          "metadata": {},
          "execution_count": 39
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Yo4cWrK1oDUf"
      },
      "source": [
        "##Cross-validation"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bHUXQnWMoGvU"
      },
      "source": [
        "def cross_val(X, y, model, params, folds=7):\n",
        "\n",
        "    loss = []\n",
        "    \n",
        "    skf = StratifiedKFold(n_splits=folds, shuffle=True, random_state=21)\n",
        "\n",
        "    for fold, (train_idx, test_idx) in enumerate(skf.split(X, y)):\n",
        "        print(f\"Fold: {fold+1}\")\n",
        "        x_train, y_train = X.iloc[train_idx], y.iloc[train_idx]\n",
        "        x_test, y_test = X.iloc[test_idx], y.iloc[test_idx]\n",
        " \n",
        "        alg = model(**params,random_state=2021)\n",
        "        alg.fit(x_train,y_train,eval_set=[(x_test,y_test)],verbose=False,early_stopping_rounds=100)\n",
        "\n",
        "        pred = alg.predict_proba(x_test)\n",
        "        log = log_loss(y_test, pred)\n",
        "        loss.append(log)\n",
        "     \n",
        "        print(f\" Log_loss: {log}\")\n",
        "        print(\"-\"*50)\n",
        "     \n",
        "    print(\"\\nmean log_loss:\",np.array(loss).mean())\n",
        "\n",
        "    return alg"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kZACbwZ7ovSH",
        "outputId": "31bbe2d4-b710-4075-86b2-2d1bb5c437c4"
      },
      "source": [
        "score_lg = cross_val(x, y, LGBMClassifier, lg_params)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Fold: 1\n",
            " Log_loss: 1.1100669655045408\n",
            "--------------------------------------------------\n",
            "Fold: 2\n",
            " Log_loss: 1.1031464262581097\n",
            "--------------------------------------------------\n",
            "Fold: 3\n",
            " Log_loss: 1.079215286855158\n",
            "--------------------------------------------------\n",
            "Fold: 4\n",
            " Log_loss: 1.056136593874037\n",
            "--------------------------------------------------\n",
            "Fold: 5\n",
            " Log_loss: 1.0953080341978378\n",
            "--------------------------------------------------\n",
            "Fold: 6\n",
            " Log_loss: 1.1043463714204387\n",
            "--------------------------------------------------\n",
            "Fold: 7\n",
            " Log_loss: 1.1072018227546203\n",
            "--------------------------------------------------\n",
            "\n",
            "mean log_loss: 1.0936316429806774\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "w4uKDpNbp56Y"
      },
      "source": [
        "# ngetest \n",
        "\n",
        "prob = score_lg.predict_proba(x_val)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "H98V7rIFqEUI",
        "outputId": "9a34c9de-09bb-4fbd-b097-0d784d0b4c31"
      },
      "source": [
        "log_loss(y_val,prob)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9298632781470977"
            ]
          },
          "metadata": {},
          "execution_count": 58
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AMWaMF01rP2S",
        "outputId": "db95f6f2-7bdf-4461-9b13-0556a167b472"
      },
      "source": [
        "print(f\"accuracy train: {score_lg.score(x_train, y_train)}\")\n",
        "print(f\"accuracy val: {score_lg.score(x_val, y_val)}\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "accuracy train: 0.650805779383162\n",
            "accuracy val: 0.66\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 278
        },
        "id": "GvU3y7VGqVjv",
        "outputId": "3b928074-3e95-4be1-caaa-ddcd1c495969"
      },
      "source": [
        "\n",
        "fig, ax = plt.subplots(1, 1,figsize=(20,15)) \n",
        "\n",
        "plot_confusion_matrix(score_lg, x_val, y_val,ax=ax)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x7f10c9b17690>"
            ]
          },
          "metadata": {},
          "execution_count": 63
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x1080 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 443
        },
        "id": "vLuB3USlq-8Z",
        "outputId": "aaac68f9-655d-411c-fde5-46bdbe3dba89"
      },
      "source": [
        "classification_report(y_val, score_lg.predict(x_val))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'               precision    recall  f1-score   support\\n\\nAcoustic/Folk       0.87      0.93      0.90       128\\n    Alt Music       0.58      0.05      0.10       280\\n        Blues       0.70      0.69      0.69       266\\n    Bollywood       0.92      0.91      0.91        88\\n      Country       0.81      0.93      0.87        70\\n       HipHop       0.77      0.80      0.78       301\\n    Indie Alt       0.54      0.49      0.51       525\\n Instrumental       0.98      0.98      0.98       106\\n        Metal       0.76      0.67      0.71       391\\n          Pop       0.67      0.65      0.66       498\\n         Rock       0.57      0.77      0.65       947\\n\\n     accuracy                           0.66      3600\\n    macro avg       0.74      0.72      0.71      3600\\n weighted avg       0.66      0.66      0.64      3600\\n'"
            ]
          },
          "metadata": {},
          "execution_count": 65
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 370
        },
        "id": "tTJMxh-zt9Yg",
        "outputId": "fdc1598f-48ba-43d7-cab5-1ae6205714dc"
      },
      "source": [
        "# plot_feature_importance\n",
        "fig, ax = plt.subplots(1, 1,figsize=(15, 20))\n",
        "\n",
        "lightgbm.plot_importance(score_lg, ax=ax)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7f10c990a0d0>"
            ]
          },
          "metadata": {},
          "execution_count": 91
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x1440 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LuKQxUn7rcIM"
      },
      "source": [
        " make a Subsmisism"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Mu3PGIqbrfr9"
      },
      "source": [
        "sub = pd.read_csv(\"/content/drive/MyDrive/Genre music classification/submission.csv\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 456
        },
        "id": "bozDd5ntrpXn",
        "outputId": "98be89d7-7edc-4969-c025-cff013202650"
      },
      "source": [
        "sub"
      ],
      "execution_count": null,
      "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>Acoustic/Folk_0</th>\n",
              "      <th>Alt_Music_1</th>\n",
              "      <th>Blues_2</th>\n",
              "      <th>Bollywood_3</th>\n",
              "      <th>Country_4</th>\n",
              "      <th>HipHop_5</th>\n",
              "      <th>Indie Alt_6</th>\n",
              "      <th>Instrumental_7</th>\n",
              "      <th>Metal_8</th>\n",
              "      <th>Pop_9</th>\n",
              "      <th>Rock_10</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\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",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\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",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\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",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\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",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7708</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7709</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7710</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7711</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7712</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>7713 rows × 11 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      Acoustic/Folk_0  Alt_Music_1  Blues_2  ...  Metal_8  Pop_9  Rock_10\n",
              "0                   1            0        0  ...        0      0        0\n",
              "1                   0            1        0  ...        0      0        0\n",
              "2                   0            0        1  ...        0      0        0\n",
              "3                   0            0        0  ...        0      0        0\n",
              "4                   0            0        0  ...        0      0        0\n",
              "...               ...          ...      ...  ...      ...    ...      ...\n",
              "7708                0            0        0  ...        0      0        0\n",
              "7709                0            0        0  ...        0      0        0\n",
              "7710                0            0        0  ...        0      0        0\n",
              "7711                0            0        0  ...        0      0        0\n",
              "7712                0            0        0  ...        0      0        0\n",
              "\n",
              "[7713 rows x 11 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 68
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j2FLH_RHrrxA"
      },
      "source": [
        "# predict test \n",
        "\n",
        "y_test = score_lg.predict(tester)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 456
        },
        "id": "MCS4KN9gr1bF",
        "outputId": "9ae19eb2-b642-4e3d-8207-bf80b214ea6e"
      },
      "source": [
        "new_sub = pd.get_dummies(y_test)\n",
        "new_sub"
      ],
      "execution_count": null,
      "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>Acoustic/Folk</th>\n",
              "      <th>Alt Music</th>\n",
              "      <th>Blues</th>\n",
              "      <th>Bollywood</th>\n",
              "      <th>Country</th>\n",
              "      <th>HipHop</th>\n",
              "      <th>Indie Alt</th>\n",
              "      <th>Instrumental</th>\n",
              "      <th>Metal</th>\n",
              "      <th>Pop</th>\n",
              "      <th>Rock</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7708</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7709</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7710</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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7711</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>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>7712</th>\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>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>7713 rows × 11 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "      Acoustic/Folk  Alt Music  Blues  ...  Metal  Pop  Rock\n",
              "0                 0          0      0  ...      0    0     1\n",
              "1                 0          0      0  ...      0    0     1\n",
              "2                 0          0      0  ...      0    1     0\n",
              "3                 0          0      0  ...      0    0     1\n",
              "4                 0          0      0  ...      0    1     0\n",
              "...             ...        ...    ...  ...    ...  ...   ...\n",
              "7708              0          0      0  ...      0    1     0\n",
              "7709              0          0      0  ...      0    0     1\n",
              "7710              0          0      0  ...      0    1     0\n",
              "7711              0          0      0  ...      0    0     0\n",
              "7712              0          0      0  ...      0    0     0\n",
              "\n",
              "[7713 rows x 11 columns]"
            ]
          },
          "metadata": {},
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iN6RMAJFsBX_"
      },
      "source": [
        "new_sub.to_csv(\"my_subs.csv\", index=False)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-KBFpgd0ssBv"
      },
      "source": [
        ""
      ],
      "execution_count": null,
      "outputs": []
    }
  ]
}