Plotting Cross-Validated Predictions

This example shows how to use cross_val_predict to visualize prediction errors.

../_images/sphx_glr_plot_cv_predict_001.png

from sklearn import datasets
from sklearn.model_selection import cross_val_predict
from sklearn import linear_model
import matplotlib.pyplot as plt
 
lr = linear_model.LinearRegression()
boston = datasets.load_boston()
y = boston.target
 
# cross_val_predict returns an array of the same size as `y` where each entry
# is a prediction obtained by cross validation:
predicted = cross_val_predict(lr, boston.data, y, cv=10)
 
fig, ax = plt.subplots()
ax.scatter(y, predicted)
ax.plot([y.min(), y.max()], [y.min(), y.max()], 'k--', lw=4)
ax.set_xlabel('Measured')
ax.set_ylabel('Predicted')
plt.show()

Total running time of the script: (0 minutes 0.114 seconds)

Download Python source code: plot_cv_predict.py

Download IPython notebook: plot_cv_predict.ipynb

Links:

http://scikit-learn.org/stable/auto_examples/plot_cv_predict.html

doc_scikit_learn

2025-01-10 15:47:30

Comments