{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn.datasets import load_iris, load_breast_cancer\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.inspection import permutation_importance as skpermutation_importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def accuracy_score(y_true, y_pred):\n",
    "    return np.mean(y_true == y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def permutation_importance(estimator, X, y, n_repeats=5, random_state=0):\n",
    "    baseline_score = accuracy_score(y, estimator.predict(X))\n",
    "    importances = np.zeros((X.shape[1], n_repeats))\n",
    "    rng = np.random.RandomState(random_state)\n",
    "    for i in range(X.shape[1]):\n",
    "        original_features = X[:, i].copy()\n",
    "        for j in range(n_repeats):\n",
    "            rng.shuffle(X[:, i])\n",
    "            importances[i, j] = accuracy_score(y, estimator.predict(X))\n",
    "        X[:, i] = original_features\n",
    "    importances = baseline_score - importances\n",
    "    return importances, importances.mean(axis=1), importances.std(axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "X, y = load_iris(return_X_y=True)\n",
    "clf = RandomForestClassifier(random_state=0).fit(X, y)\n",
    "importance1, mean1, std1 = permutation_importance(clf, X, y, random_state=0)\n",
    "importance2 = skpermutation_importance(clf, X, y, random_state=0)\n",
    "assert np.allclose(importance1, importance2.importances)\n",
    "assert np.allclose(mean1, importance2.importances_mean)\n",
    "assert np.allclose(std1, importance2.importances_std)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "X, y = load_breast_cancer(return_X_y=True)\n",
    "clf = RandomForestClassifier(random_state=0).fit(X, y)\n",
    "importance1, mean1, std1 = permutation_importance(clf, X, y, random_state=0)\n",
    "importance2 = skpermutation_importance(clf, X, y, random_state=0)\n",
    "assert np.allclose(importance1, importance2.importances)\n",
    "assert np.allclose(mean1, importance2.importances_mean)\n",
    "assert np.allclose(std1, importance2.importances_std)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "dev",
   "language": "python",
   "name": "dev"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}