fyr91/pdp-notebook-2.ipynb

## pdp-notebook-2.ipynb
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import pickle\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "from matplotlib import pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Load model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[18:42:11] WARNING: src/objective/regression_obj.cu:152: reg:linear is now deprecated in favor of reg:squarederror.\n"
     ]
    }
   ],
   "source": [
    "filename = 'model.sav'\n",
    "model = pickle.load(open(filename, 'rb'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Load data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"2017.csv\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### PDP"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Country', 'Happiness.Rank', 'Happiness.Score', 'Whisker.high',\n",
       "       'Whisker.low', 'Economy..GDP.per.Capita.', 'Family',\n",
       "       'Health..Life.Expectancy.', 'Freedom', 'Generosity',\n",
       "       'Trust..Government.Corruption.', 'Dystopia.Residual'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def pdp(feature, df, x_labels, y_label):\n",
    "    df_copy = df.copy()\n",
    "    unique_vals = np.unique(df_copy[feature].values)\n",
    "    y = []\n",
    "    for val in unique_vals:\n",
    "        df_copy[feature] = val\n",
    "        X = df_copy[x_labels]\n",
    "        y.append(np.average(model.predict(X)))\n",
    "    plt.ylim(2, 8)\n",
    "    g = sns.lineplot(x=unique_vals, y=y)\n",
    "    g.set(xticks=unique_vals)\n",
    "    g.set(xticklabels=[])\n",
    "    g.set(xlabel = feature)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for v in ['Economy..GDP.per.Capita.', 'Family',\n",
    "       'Health..Life.Expectancy.', 'Freedom', 'Generosity',\n",
    "       'Trust..Government.Corruption.', 'Dystopia.Residual']:\n",
    "    pdp(v, df, ['Economy..GDP.per.Capita.', 'Family',\n",
    "       'Health..Life.Expectancy.', 'Freedom', 'Generosity',\n",
    "       'Trust..Government.Corruption.', 'Dystopia.Residual'], 'Happiness.Score')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "analysis flow",
   "language": "python",
   "name": "analysis"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}