roc#

ComparisonReport.metrics.roc(*, data_source='test')[source]#

Plot the ROC curve.

Parameters:

data_source{“test”, “train”, “both”}, default=”test”

The data source to use.

“test” : use the test set provided when creating the report.
“train” : use the train set provided when creating the report.
“both” : use both the train and test sets to compute the metrics.

Returns:

RocCurveDisplay: The ROC curve display.

See also

RocCurveDisplay: Display class for ROC curve plots.

Notes

To keep the stored display lightweight, the ROC curve is downsampled to at most 500 points per class and per child report. Sampling is performed by picking evenly-spaced indices on the sorted thresholds.

Examples

>>> from sklearn.datasets import load_breast_cancer
>>> from sklearn.linear_model import LogisticRegression
>>> from skore import evaluate
>>> X, y = load_breast_cancer(return_X_y=True)
>>> estimator_1 = LogisticRegression(max_iter=10000, random_state=42)
>>> estimator_2 = LogisticRegression(max_iter=10000, random_state=43)
>>> comparison_report = evaluate([estimator_1, estimator_2], X, y, splitter=0.2)
>>> display = comparison_report.metrics.roc()
>>> display.plot()