diff --git a/python-package/examples/FastRGF/FastRGF_regressor_on_boston_dataset.py b/python-package/examples/FastRGF/FastRGF_regressor_on_boston_dataset.py index d0e14d08..0c43bdbc 100644 --- a/python-package/examples/FastRGF/FastRGF_regressor_on_boston_dataset.py +++ b/python-package/examples/FastRGF/FastRGF_regressor_on_boston_dataset.py @@ -1,20 +1,20 @@ import time -from sklearn.datasets import load_boston +from sklearn.datasets import load_diabetes from sklearn.utils.validation import check_random_state from sklearn.ensemble import RandomForestRegressor from rgf.sklearn import FastRGFRegressor, RGFRegressor -boston = load_boston() +diabetes = load_diabetes() rng = check_random_state(42) -perm = rng.permutation(boston.target.size) -boston.data = boston.data[perm] -boston.target = boston.target[perm] +perm = rng.permutation(diabetes.target.size) +diabetes.data = diabetes.data[perm] +diabetes.target = diabetes.target[perm] -train_x = boston.data[:300] -test_x = boston.data[300:] -train_y = boston.target[:300] -test_y = boston.target[300:] +train_x = diabetes.data[:300] +test_x = diabetes.data[300:] +train_y = diabetes.target[:300] +test_y = diabetes.target[300:] start = time.time() reg = RGFRegressor() diff --git a/python-package/examples/RGF/comparison_RGF_and_RF_regressors_on_boston_dataset.py b/python-package/examples/RGF/comparison_RGF_and_RF_regressors_on_diabetes_dataset.py similarity index 64% rename from python-package/examples/RGF/comparison_RGF_and_RF_regressors_on_boston_dataset.py rename to python-package/examples/RGF/comparison_RGF_and_RF_regressors_on_diabetes_dataset.py index a7bf7197..9f460e79 100644 --- a/python-package/examples/RGF/comparison_RGF_and_RF_regressors_on_boston_dataset.py +++ b/python-package/examples/RGF/comparison_RGF_and_RF_regressors_on_diabetes_dataset.py @@ -1,18 +1,20 @@ -from sklearn.datasets import load_boston +from sklearn.datasets import load_diabetes from sklearn.utils.validation import check_random_state from sklearn.model_selection import cross_val_score from sklearn.metrics import make_scorer, mean_squared_error from sklearn.ensemble import RandomForestRegressor from rgf.sklearn import RGFRegressor -boston = load_boston() +diabetes = load_diabetes() rng = check_random_state(42) -perm = rng.permutation(boston.target.size) -boston.data = boston.data[perm] -boston.target = boston.target[perm] +perm = rng.permutation(diabetes.target.size) +diabetes.data = diabetes.data[perm] +diabetes.target = diabetes.target[perm] -rgf = RGFRegressor(max_leaf=300, - algorithm="RGF_Sib", +rgf = RGFRegressor(max_leaf=30, + n_iter=5, + learning_rate=0.2, + algorithm="RGF", test_interval=100, loss="LS", verbose=False) @@ -24,20 +26,20 @@ n_folds = 3 rgf_scores = cross_val_score(rgf, - boston.data, - boston.target, + diabetes.data, + diabetes.target, scoring=make_scorer(mean_squared_error), cv=n_folds) rf_scores = cross_val_score(rf, - boston.data, - boston.target, + diabetes.data, + diabetes.target, scoring=make_scorer(mean_squared_error), cv=n_folds) rgf_score = sum(rgf_scores)/n_folds print('RGF Regressor MSE: {0:.5f}'.format(rgf_score)) -# >>>RGF Regressor MSE: 11.79409 +# >>> RGF Regressor MSE: 3377.46076 rf_score = sum(rf_scores)/n_folds print('Random Forest Regressor MSE: {0:.5f}'.format(rf_score)) -# >>>Random Forest Regressor MSE: 13.58614 +# >>> Random Forest Regressor MSE: 3441.01988 diff --git a/python-package/examples/RGF/regression_on_boston_dataset.ipynb b/python-package/examples/RGF/regression_on_boston_dataset.ipynb deleted file mode 100644 index 459d8232..00000000 --- a/python-package/examples/RGF/regression_on_boston_dataset.ipynb +++ /dev/null @@ -1,174 +0,0 @@ -{ - "cells": [ - { - "cell_type": "code", - "execution_count": 1, - "metadata": { - "collapsed": false - }, - "outputs": [], - "source": [ - "from sklearn.datasets import load_boston\n", - "from sklearn.model_selection import cross_val_score, train_test_split\n", - "from sklearn.metrics import make_scorer, mean_squared_error\n", - "from sklearn.ensemble import RandomForestRegressor\n", - "from rgf.sklearn import RGFRegressor\n", - "import matplotlib.pyplot as plt" - ] - }, - { - "cell_type": "code", - "execution_count": 2, - "metadata": { - "collapsed": false - }, - "outputs": [], - "source": [ - "boston = load_boston()\n", - "X_train, X_test, y_train, y_test = train_test_split(boston.data,\n", - " boston.target,\n", - " test_size=0.1,\n", - " random_state=42)" - ] - }, - { - "cell_type": "code", - "execution_count": 3, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "rgf = RGFRegressor(max_leaf=300,\n", - " algorithm=\"RGF_Sib\",\n", - " test_interval=100,\n", - " loss=\"LS\",\n", - " verbose=False)\n", - "rf = RandomForestRegressor(n_estimators=600,\n", - " min_samples_leaf=3,\n", - " max_depth=10,\n", - " random_state=42)\n", - "n_folds = 3" - ] - }, - { - "cell_type": "code", - "execution_count": 4, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "rgf_scores = cross_val_score(rgf,\n", - " X_train,\n", - " y_train,\n", - " scoring=make_scorer(mean_squared_error),\n", - " cv=n_folds)\n", - "rf_scores = cross_val_score(rf,\n", - " X_train,\n", - " y_train,\n", - " scoring=make_scorer(mean_squared_error),\n", - " cv=n_folds)" - ] - }, - { - "cell_type": "code", - "execution_count": 5, - "metadata": { - "collapsed": false - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "RGF Regressor MSE: 12.59373\n", - "Random Forest Regressor MSE: 13.80435\n" - ] - } - ], - "source": [ - "rgf_score = sum(rgf_scores)/n_folds\n", - "print('RGF Regressor MSE: {0:.5f}'.format(rgf_score))\n", - "rf_score = sum(rf_scores)/n_folds\n", - "print('Random Forest Regressor MSE: {0:.5f}'.format(rf_score))" - ] - }, - { - "cell_type": "code", - "execution_count": 6, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [ - "y_pred_rgf = rgf.fit(X_train, y_train).predict(X_test)\n", - "y_pred_rf = rf.fit(X_train, y_train).predict(X_test)" - ] - }, - { - "cell_type": "code", - "execution_count": 7, - "metadata": { - "collapsed": false - }, - "outputs": [ - { - "data": { - "image/png": "iVBORw0KGgoAAAANSUhEUgAAAysAAAHHCAYAAAC/TseMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRCQr/X+Xz9oj1mBSlQAAAABJRU5ErkJggg==\n", - "text/plain": [ - "" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "%matplotlib inline\n", - "\n", - "plt.figure(figsize=(9.5,5))\n", - "origin_plot = plt.scatter(X_test[:, 5], y_test, s=45, color=\"black\")\n", - "rgf_plot = plt.scatter(X_test[:, 5], y_pred_rgf, s=45, color=\"red\", alpha=0.6)\n", - "rf_plot = plt.scatter(X_test[:, 5], y_pred_rf, s=45, color=\"green\", alpha=0.6)\n", - "plt.xlabel(\"Average number of rooms per dwelling\")\n", - "plt.ylabel(\"Median value of owner-occupied homes in $1000's\")\n", - "plt.legend([origin_plot, rgf_plot, rf_plot],\n", - " [\"Ground Truth\", \"RGF\", \"Random Forest\"],\n", - " loc=\"upper left\")\n", - "plt.show()" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "metadata": { - "collapsed": true - }, - "outputs": [], - "source": [] - } - ], - "metadata": { - "anaconda-cloud": {}, - "kernelspec": { - "display_name": "Python [default]", - "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": "3.5.2" - } - }, - "nbformat": 4, - "nbformat_minor": 1 -} diff --git a/python-package/examples/RGF/regression_on_diabetes_dataset.ipynb b/python-package/examples/RGF/regression_on_diabetes_dataset.ipynb new file mode 100644 index 00000000..3dcc719e --- /dev/null +++ b/python-package/examples/RGF/regression_on_diabetes_dataset.ipynb @@ -0,0 +1,162 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "metadata": {}, + "outputs": [], + "source": [ + "from sklearn.datasets import load_diabetes\n", + "from sklearn.model_selection import cross_val_score, train_test_split\n", + "from sklearn.metrics import make_scorer, mean_squared_error\n", + "from sklearn.ensemble import RandomForestRegressor\n", + "from rgf.sklearn import RGFRegressor\n", + "import matplotlib.pyplot as plt" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": {}, + "outputs": [], + "source": [ + "diabetes = load_diabetes()\n", + "X_train, X_test, y_train, y_test = train_test_split(diabetes.data,\n", + " diabetes.target,\n", + " test_size=0.1,\n", + " random_state=42)" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": {}, + "outputs": [], + "source": [ + "rgf = RGFRegressor(max_leaf=30,\n", + " n_iter=5,\n", + " learning_rate=0.2,\n", + " algorithm=\"RGF\",\n", + " test_interval=100,\n", + " loss=\"LS\",\n", + " verbose=False)\n", + "rf = RandomForestRegressor(n_estimators=600,\n", + " min_samples_leaf=3,\n", + " max_depth=10,\n", + " random_state=42)\n", + "n_folds = 3" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": {}, + "outputs": [], + "source": [ + "rgf_scores = cross_val_score(rgf,\n", + " X_train,\n", + " y_train,\n", + " scoring=make_scorer(mean_squared_error),\n", + " cv=n_folds)\n", + "rf_scores = cross_val_score(rf,\n", + " X_train,\n", + " y_train,\n", + " scoring=make_scorer(mean_squared_error),\n", + " cv=n_folds)" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "RGF Regressor MSE: 3310.10304\n", + "Random Forest Regressor MSE: 3361.16454\n" + ] + } + ], + "source": [ + "rgf_score = sum(rgf_scores)/n_folds\n", + "print('RGF Regressor MSE: {0:.5f}'.format(rgf_score))\n", + "rf_score = sum(rf_scores)/n_folds\n", + "print('Random Forest Regressor MSE: {0:.5f}'.format(rf_score))" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": {}, + "outputs": [], + "source": [ + "y_pred_rgf = rgf.fit(X_train, y_train).predict(X_test)\n", + "y_pred_rf = rf.fit(X_train, y_train).predict(X_test)" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": {}, + "outputs": [ + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "%matplotlib inline\n", + "\n", + "plt.figure(figsize=(9.5,5))\n", + "origin_plot = plt.scatter(X_test[:, 2], y_test, s=45, color=\"black\")\n", + "rgf_plot = plt.scatter(X_test[:, 2], y_pred_rgf, s=45, color=\"red\", alpha=0.6)\n", + "rf_plot = plt.scatter(X_test[:, 2], y_pred_rf, s=45, color=\"green\", alpha=0.6)\n", + "plt.xlabel(\"Body mass index (scaled)\")\n", + "plt.ylabel(\"Quantitative measure of disease progression one year after baseline\")\n", + "plt.legend([origin_plot, rgf_plot, rf_plot],\n", + " [\"Ground Truth\", \"RGF\", \"Random Forest\"],\n", + " loc=\"upper left\")\n", + "plt.show()" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": {}, + "outputs": [], + "source": [] + } + ], + "metadata": { + "anaconda-cloud": {}, + "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": "3.8.8" + } + }, + "nbformat": 4, + "nbformat_minor": 1 +}