sklearn_linear_regression_diabetes_dataset_with_matplotlib_plot.py

1import matplotlib.pyplot as plt
2import numpy as np
3
4from sklearn import datasets, linear_model
5from sklearn.metrics import mean_squared_error, r2_score
6
7# Load the diabetes dataset
8diabetes_X, diabetes_y = datasets.load_diabetes(return_X_y=True)
9
10# Use only one feature
11diabetes_X = diabetes_X[:, np.newaxis, 2]
12
13# Split the data into training/testing sets
14diabetes_X_train = diabetes_X[:-20]
15diabetes_X_test = diabetes_X[-20:]
16
17# Split the targets into training/testing sets
18diabetes_y_train = diabetes_y[:-20]
19diabetes_y_test = diabetes_y[-20:]
20
21# Create linear regression object
22regr = linear_model.LinearRegression()
23
24# Train the model using the training sets
25regr.fit(diabetes_X_train, diabetes_y_train)
26
27# Make predictions using the testing set
28diabetes_y_pred = regr.predict(diabetes_X_test)
29
30# The coefficients
31print("Coefficients: \n", regr.coef_)
32# The mean squared error
33print("Mean squared error: %.2f" % mean_squared_error(diabetes_y_test, diabetes_y_pred))
34# The coefficient of determination: 1 is perfect prediction
35print("Coefficient of determination: %.2f" % r2_score(diabetes_y_test, diabetes_y_pred))
36
37# Plot outputs
38plt.scatter(diabetes_X_test, diabetes_y_test, color="black")
39plt.plot(diabetes_X_test, diabetes_y_pred, color="blue", linewidth=3)
40
41plt.xticks(())
42plt.yticks(())
43
44plt.show()