CalibrationDisplay

class skore.CalibrationDisplay(*, calibration_report, report_type, report_pos_label=None)

Calibration curve visualization.

Parameters:

calibration_reportpandas.DataFrame

Calibration report of the estimator. Its columns are estimator, data_source, split, label, predicted_probability, and fraction_of_positives.

report_type{“estimator”, “cross-validation”, “comparison-estimator”, “comparison-cross-validation”}

Report type from which the display is created.

report_pos_labelint, float, bool, str or None, default=None

Positive class label configured on the report.

Examples

>>> from sklearn.datasets import make_classification
>>> from sklearn.linear_model import LogisticRegression
>>> from skore import evaluate
>>> X, y = make_classification(
...     n_samples=100_000, n_features=20, n_informative=2, n_redundant=10,
...     random_state=42)
>>> report = evaluate(LogisticRegression(), X, y, splitter=0.2)
>>> display = report.inspection.calibration_curve(n_bins="auto", strategy="uniform")
>>> display.frame().head(5)
    predicted_probability  fraction_of_positives data_source  label
0               0.013323               0.018338        test      0
1               0.052453               0.049888        test      0
2               0.088494               0.093787        test      0
3               0.124554               0.113264        test      0
4               0.160065               0.145329        test      0

frame(*, label=<DEFAULT>)

Return the data used for the display as a DataFrame.

Parameters:

labelint, float, bool, str or None, default=report pos_label

The class whose curve to select. Use None to show all classes.

Returns:

DataFrame

The returned frame always contains:

• predicted_probability

• fraction_of_positives

• data_source

It also contains label when label=None (one-vs-rest view).

When a specific label is selected, the label column is dropped from

the output to provide a single-curve table.

Depending on the report type, estimator and/or split metadata may

be removed from the returned frame.

help()

Display display help using rich or HTML.

property labels

Return the list of labels available in the calibration data.

plot(*, label=<DEFAULT>)

Plot the calibration curve.

Parameters:

labelint, float, bool, str or None, default=report pos_label

The class whose curve to plot. Use None to show all classes.

Returns:

matplotlib.figure.Figure

Examples

>>> from sklearn.datasets import make_classification
>>> from sklearn.linear_model import LogisticRegression
>>> from skore import evaluate
>>> X, y = make_classification(
...     n_samples=100_000, n_features=20, n_informative=2, n_redundant=10,
...     random_state=42)
>>> report = evaluate(LogisticRegression(), X, y, splitter=0.2)
>>> display = report.inspection.calibration_curve()
>>> display.plot()

set_style(*, policy='update', line_kwargs=None, ax_set_kwargs=None)

Set the style parameters for the display.

Parameters:

policy{“override”, “update”}, default=”update”

Policy to use when setting the style parameters. If “override”, existing settings are set to the provided values. If “update”, existing settings are not changed; only settings that were previously unset are changed.

line_kwargsdict, default=None

Keyword arguments passed to seaborn.lineplot().

ax_set_kwargsdict, default=None

Keyword arguments passed to matplotlib.axes.Axes.set(). Useful keys include xlabel, ylabel, xlim, and ylim.

Returns:

None

static style_plot(plot_func)

Apply consistent style to skore displays.

This decorator: 1. Applies default style settings 2. Runs plot_func under plt.ioff() so figures are not shown until returned 3. Calls Figure.tight_layout() on the returned figure when applicable 4. Restores the original style settings

Parameters:

plot_funccallable

The plot function to be decorated.

Returns:

callable

The decorated plot function.