PredictionErrorDisplay

classsklearn.metrics.PredictionErrorDisplay(*, y_true, y_pred)[source]

Visualization of the prediction error of a regression model.

This tool can display “residuals vs predicted” or “actual vs predicted” using scatter plots to qualitatively assess the behavior of a regressor, preferably on held-out data points.

See the details in the docstrings of from_estimator or from_predictions to create a visualizer. All parameters are stored as attributes.

For general information regarding scikit-learn visualization tools, read more in the Visualization Guide. For details regarding interpreting these plots, refer to the Model Evaluation Guide.

Added in version 1.2.

Parameters:

y_truendarray of shape (n_samples,)

True values.

y_predndarray of shape (n_samples,)

Prediction values.

Attributes:

line_matplotlib Artist

Optimal line representing y_true == y_pred. Therefore, it is a diagonal line for kind="predictions" and a horizontal line for kind="residuals".

errors_lines_matplotlib Artist or None

Residual lines. If with_errors=False, then it is set to None.

scatter_matplotlib Artist

Scatter data points.

ax_matplotlib Axes

Axes with the different matplotlib axis.

figure_matplotlib Figure

Figure containing the scatter and lines.

See also

PredictionErrorDisplay.from_estimator

Prediction error visualization given an estimator and some data.

PredictionErrorDisplay.from_predictions

Prediction error visualization given the true and predicted targets.

Examples

>>> import matplotlib.pyplot as plt
>>> from sklearn.datasets import load_diabetes
>>> from sklearn.linear_model import Ridge
>>> from sklearn.metrics import PredictionErrorDisplay
>>> X, y = load_diabetes(return_X_y=True)
>>> ridge = Ridge().fit(X, y)
>>> y_pred = ridge.predict(X)
>>> display = PredictionErrorDisplay(y_true=y, y_pred=y_pred)
>>> display.plot()
<...>
>>> plt.show()

classmethodfrom_estimator(estimator, X, y, *, kind='residual_vs_predicted', subsample=1000, random_state=None, ax=None, scatter_kwargs=None, line_kwargs=None)[source]

Plot the prediction error given a regressor and some data.

For general information regarding scikit-learn visualization tools, read more in the Visualization Guide. For details regarding interpreting these plots, refer to the Model Evaluation Guide.

Added in version 1.2.

Parameters:

estimatorestimator instance

Fitted regressor or a fitted Pipeline in which the last estimator is a regressor.

X{array-like, sparse matrix} of shape (n_samples, n_features)

Input values.

yarray-like of shape (n_samples,)

Target values.

kind{“actual_vs_predicted”, “residual_vs_predicted”}, default=”residual_vs_predicted”

The type of plot to draw:

• “actual_vs_predicted” draws the observed values (y-axis) vs. the predicted values (x-axis).
• “residual_vs_predicted” draws the residuals, i.e. difference between observed and predicted values, (y-axis) vs. the predicted values (x-axis).

subsamplefloat, int or None, default=1_000

Sampling the samples to be shown on the scatter plot. If float, it should be between 0 and 1 and represents the proportion of the original dataset. If int, it represents the number of samples display on the scatter plot. If None, no subsampling will be applied. by default, 1000 samples or less will be displayed.

random_stateint or RandomState, default=None

Controls the randomness when subsample is not None. See Glossary for details.

axmatplotlib axes, default=None

Axes object to plot on. If None, a new figure and axes is created.

scatter_kwargsdict, default=None

Dictionary with keywords passed to the matplotlib.pyplot.scatter call.

line_kwargsdict, default=None

Dictionary with keyword passed to the matplotlib.pyplot.plot call to draw the optimal line.

Returns:

displayPredictionErrorDisplay

Object that stores the computed values.

See also

PredictionErrorDisplay

Prediction error visualization for regression.

PredictionErrorDisplay.from_predictions

Prediction error visualization given the true and predicted targets.

Examples

>>> import matplotlib.pyplot as plt
>>> from sklearn.datasets import load_diabetes
>>> from sklearn.linear_model import Ridge
>>> from sklearn.metrics import PredictionErrorDisplay
>>> X, y = load_diabetes(return_X_y=True)
>>> ridge = Ridge().fit(X, y)
>>> disp = PredictionErrorDisplay.from_estimator(ridge, X, y)
>>> plt.show()

classmethodfrom_predictions(y_true, y_pred, *, kind='residual_vs_predicted', subsample=1000, random_state=None, ax=None, scatter_kwargs=None, line_kwargs=None)[source]

Plot the prediction error given the true and predicted targets.

For general information regarding scikit-learn visualization tools, read more in the Visualization Guide. For details regarding interpreting these plots, refer to the Model Evaluation Guide.

Added in version 1.2.

Parameters:

y_truearray-like of shape (n_samples,)

True target values.

y_predarray-like of shape (n_samples,)

Predicted target values.

kind{“actual_vs_predicted”, “residual_vs_predicted”}, default=”residual_vs_predicted”

The type of plot to draw:

• “actual_vs_predicted” draws the observed values (y-axis) vs. the predicted values (x-axis).
• “residual_vs_predicted” draws the residuals, i.e. difference between observed and predicted values, (y-axis) vs. the predicted values (x-axis).

subsamplefloat, int or None, default=1_000

Sampling the samples to be shown on the scatter plot. If float, it should be between 0 and 1 and represents the proportion of the original dataset. If int, it represents the number of samples display on the scatter plot. If None, no subsampling will be applied. by default, 1000 samples or less will be displayed.

random_stateint or RandomState, default=None

Controls the randomness when subsample is not None. See Glossary for details.

axmatplotlib axes, default=None

Axes object to plot on. If None, a new figure and axes is created.

scatter_kwargsdict, default=None

Dictionary with keywords passed to the matplotlib.pyplot.scatter call.

line_kwargsdict, default=None

Dictionary with keyword passed to the matplotlib.pyplot.plot call to draw the optimal line.

Returns:

displayPredictionErrorDisplay

Object that stores the computed values.

See also

PredictionErrorDisplay

Prediction error visualization for regression.

PredictionErrorDisplay.from_estimator

Prediction error visualization given an estimator and some data.

Examples

>>> import matplotlib.pyplot as plt
>>> from sklearn.datasets import load_diabetes
>>> from sklearn.linear_model import Ridge
>>> from sklearn.metrics import PredictionErrorDisplay
>>> X, y = load_diabetes(return_X_y=True)
>>> ridge = Ridge().fit(X, y)
>>> y_pred = ridge.predict(X)
>>> disp = PredictionErrorDisplay.from_predictions(y_true=y, y_pred=y_pred)
>>> plt.show()

plot(ax=None, *, kind='residual_vs_predicted', scatter_kwargs=None, line_kwargs=None)[source]

Plot visualization.

Extra keyword arguments will be passed to matplotlib’s plot.

Parameters:

axmatplotlib axes, default=None

Axes object to plot on. If None, a new figure and axes is created.

kind{“actual_vs_predicted”, “residual_vs_predicted”}, default=”residual_vs_predicted”

The type of plot to draw:

• “actual_vs_predicted” draws the observed values (y-axis) vs. the predicted values (x-axis).
• “residual_vs_predicted” draws the residuals, i.e. difference between observed and predicted values, (y-axis) vs. the predicted values (x-axis).

scatter_kwargsdict, default=None

Dictionary with keywords passed to the matplotlib.pyplot.scatter call.

line_kwargsdict, default=None

Dictionary with keyword passed to the matplotlib.pyplot.plot call to draw the optimal line.

Returns:

displayPredictionErrorDisplay

Object that stores computed values.

Gallery examples

Plotting Cross-Validated Predictions

Release Highlights for scikit-learn 1.2

Combine predictors using stacking

Lagged features for time series forecasting

Time-related feature engineering

Common pitfalls in the interpretation of coefficients of linear models

Plotting Cross-Validated Predictions

Effect of transforming the targets in regression model