{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "%matplotlib inline"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "\n# regression\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "from ai4water.datasets import busan_beach\nfrom skopt.plots import plot_objective\nfrom autotab import OptimizePipeline\n\ndata = busan_beach()\n\npl = OptimizePipeline(\n    inputs_to_transform=data.columns.tolist()[0:-1],\n    outputs_to_transform=data.columns.tolist()[-1:],\n    parent_iterations=30,\n    child_iterations=0,  # don't optimize hyperparamters only for demonstration\n    parent_algorithm='bayes',\n    child_algorithm='random',\n    eval_metric='mse',\n    monitor=['r2', 'r2_score'],\n    models=[ \"LinearRegression\",\n            \"LassoLars\",\n            \"Lasso\",\n            \"RandomForestRegressor\",\n            \"HistGradientBoostingRegressor\",\n             \"CatBoostRegressor\",\n             \"XGBRegressor\",\n             \"LGBMRegressor\",\n             \"GradientBoostingRegressor\",\n             \"ExtraTreeRegressor\",\n             \"ExtraTreesRegressor\"\n             ],\n\n    input_features=data.columns.tolist()[0:-1],\n    output_features=data.columns.tolist()[-1:],\n    split_random=True,\n)\n\nresults = pl.fit(data=data, process_results=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.optimizer_._plot_convergence(save=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.optimizer_._plot_parallel_coords(figsize=(16, 8), save=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.optimizer_._plot_distributions(save=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.optimizer_.plot_importance(save=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "_ = plot_objective(results)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.optimizer_._plot_evaluations(save=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.optimizer_._plot_edf(save=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.bfe_all_best_models(data=data)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.dumbbell_plot(data=data, save=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.dumbbell_plot(data=data, metric_name='r2', save=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.taylor_plot(data=data, save=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.compare_models()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.compare_models(plot_type=\"bar_chart\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.compare_models(\"r2\", plot_type=\"bar_chart\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "print(f\"all results are save in {pl.path} folder\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "collapsed": false
      },
      "outputs": [],
      "source": [
        "pl.cleanup()"
      ]
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.9"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}