feat(skore): Add data_source="all" option to EstimatorReport.metrics.summarize

skore/src/skore/_sklearn/_estimator/metrics_accessor.py

@@ -63,7 +63,7 @@ def __init__(self, parent: EstimatorReport) -> None:
     def summarize(
         self,
         *,
-        data_source: DataSource = "test",
+        data_source: Literal["test", "train", "X_y", "all"] = "test",
         X: ArrayLike | None = None,
         y: ArrayLike | None = None,
         scoring: Scoring | list[Scoring] | None = None,
@@ -77,12 +77,13 @@ def summarize(
 
         Parameters
         ----------
-        data_source : {"test", "train", "X_y"}, default="test"
+        data_source : {"test", "train", "X_y", "all"}, 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.
             - "X_y" : use the provided `X` and `y` to compute the metric.
+            - "all" : use both the train and test sets to compute the metrics.
 
         X : array-like of shape (n_samples, n_features), default=None
             New data on which to compute the metric. By default, we use the validation
@@ -160,7 +161,66 @@ class is set to the one provided when creating the report. If `None`,
                                   LogisticRegression Favorability
         Metric   Label / Average
         F1 Score               1             0.96...          (↗︎)
+        >>> report.metrics.summarize(
+        ...    indicator_favorability=True,
+        ...    data_source="all"
+        ... ).frame().drop(["Fit time (s)", "Predict time (s)"])
+                          LogisticRegression (train)  ...  Favorability
+        Metric                                        ...
+        Precision                           0.962963  ...          (↗︎)
+        Recall                              0.973783  ...          (↗︎)
+        ROC AUC                             0.994300  ...          (↗︎)
+        Brier score                         0.027365  ...          (↘︎)
+        >>> # Using scikit-learn metrics
+        >>> report.metrics.summarize(
+        ...     scoring=["f1"],
+        ...     indicator_favorability=True,
+        ... ).frame()
+                                  LogisticRegression Favorability
+        Metric   Label / Average
+        F1 Score               1             0.96...          (↗︎)
         """
+        if data_source == "all":
+            train_summary = self.summarize(
+                data_source="train",
+                X=X,
+                y=y,
+                scoring=scoring,
+                scoring_names=scoring_names,
+                scoring_kwargs=scoring_kwargs,
+                pos_label=pos_label,
+                indicator_favorability=indicator_favorability,
+                flat_index=flat_index,
+            )
+            test_summary = self.summarize(
+                data_source="test",
+                X=X,
+                y=y,
+                scoring=scoring,
+                scoring_names=scoring_names,
+                scoring_kwargs=scoring_kwargs,
+                pos_label=pos_label,
+                indicator_favorability=indicator_favorability,
+                flat_index=flat_index,
+            )
+            # Add suffix to the dataframes to distinguish train and test.
+            train_df = train_summary.frame().add_suffix(" (train)")
+            test_df = test_summary.frame().add_suffix(" (test)")
+            combined = pd.concat(
+                [
+                    train_df,
+                    test_df,
+                ],
+                axis=1,
+            )
+            if indicator_favorability:
+                combined["Favorability"] = train_df["Favorability (train)"]
+                combined.drop(
+                    columns=["Favorability (train)", "Favorability (test)"],
+                    inplace=True,
+                )
+            return MetricsSummaryDisplay(summarize_data=combined)
+
         if pos_label is _DEFAULT:
             pos_label = self._parent.pos_label
 
@@ -429,6 +489,18 @@ class is set to the one provided when creating the report. If `None`,
                 )
         return MetricsSummaryDisplay(summarize_data=results)
 
+        return self._summarize(
+            data_source=data_source,
+            X=X,
+            y=y,
+            scoring=scoring,
+            scoring_names=scoring_names,
+            scoring_kwargs=scoring_kwargs,
+            pos_label=pos_label,
+            indicator_favorability=indicator_favorability,
+            flat_index=flat_index,
+        )
+
     def _compute_metric_scores(
         self,
         metric_fn: Callable,