add example

examples/diagnose/plot_cross_validation.py

@@ -0,0 +1,56 @@
+"""
+=======================
+Cross-validation report
+=======================
+
+This example shows how the :class:`skore.CrossValidationReport` class can be used to
+get insights from any scikit-learn estimator.
+"""
+
+# %%
+from skrub.datasets import fetch_employee_salaries
+
+dataset = fetch_employee_salaries()
+df = dataset.X
+y = dataset.y
+
+# %%
+from skrub import TableReport
+
+TableReport(df)
+
+# %%
+TableReport(y.to_frame())
+
+# %%
+from skrub import tabular_learner
+
+estimator = tabular_learner("regressor")
+estimator
+
+# %%
+from skore import CrossValidationReport
+from sklearn.model_selection import RepeatedKFold
+
+cv = RepeatedKFold(n_splits=10, n_repeats=3, random_state=42)
+reporter = CrossValidationReport(estimator, df, y, cv=cv, n_jobs=4)
+reporter
+
+# %%
+results = reporter.metrics.r2()
+results
+
+# %%
+results.aggregate(["mean", "std"], axis=0)
+
+# %%
+ax = results.plot.kde()
+ax.set_xlim(0, 1)
+_ = ax.set_title("R2 score distribution")
+
+# %%
+results = reporter.metrics.report_metrics()
+results
+
+# %%
+results.aggregate(["mean", "std"], axis=0)