feat: Add dedicated Display classes to ComparisonReport (#1309)

Duplicate the plot classes for `ComparisonReport` use-cases:
- `RocCurveDisplay`,
- `PredictionRecallCurveDisplay`,
- `PredictionErrorDisplay`.

I've chosen to simplify a few things from the original classes, in
particular by removing the user's ability to customize the plot.
The plot tests strictly compare pixel by pixel the matplotlib result
with an expected image.

The next iteration should be to factorize these duplicated classes with
the one in `skore.sklearn._plot`.

---

**Roc curves**

![binary-classification](https://github.com/user-attachments/assets/b3e4fd8a-e201-4f3a-83be-01165c8a9c9d)

![multiclass-classification](https://github.com/user-attachments/assets/6ec2f490-7c55-483d-ada2-5af67a0ff1bd)

**PR curves**

![pr-binary-classification](https://github.com/user-attachments/assets/79a07266-5ddb-42e4-99cb-e6f3fefdfc6d)

![pr-multiclass-classification](https://github.com/user-attachments/assets/80b200cd-20af-4a51-8fe6-401f56e9b534)

**Prediction error curves**

![Figure_1](https://github.com/user-attachments/assets/d64de1de-84de-4f25-9fa9-7ee7e3e17ad9)

![Figure_2](https://github.com/user-attachments/assets/b2d8889e-6573-4dd7-8170-c5974fd767e0)

examples/getting_started/plot_skore_getting_started.py

@@ -162,25 +162,34 @@
 # %%
 from skore import ComparisonReport
 
-comparator = ComparisonReport(reports=[log_reg_report, rf_report])
+comparison_report = ComparisonReport(reports=[log_reg_report, rf_report])
 
 # %%
 # As for the :class:`~skore.EstimatorReport` and the
 # :class:`~skore.CrossValidationReport`, we have a helper:
 
 # %%
-comparator.help()
+comparison_report.help()
 
 # %%
 # Let us display the result of our benchmark:
 
 # %%
-benchmark_metrics = comparator.metrics.report_metrics()
+benchmark_metrics = comparison_report.metrics.report_metrics()
 benchmark_metrics
 
 # %%
 # We have the result of our benchmark.
 
+# %%
+# We display the ROC curve for the two estimator reports we want to compare, by
+# superimposing them on the same figure:
+
+# %%
+comparison_report.metrics.roc().plot()
+plt.tight_layout()
+
+
 # %%
 # Train-test split with skore
 # ^^^^^^^^^^^^^^^^^^^^^^^^^^^

skore/src/skore/sklearn/__init__.py

@@ -11,11 +11,11 @@
 from skore.sklearn.train_test_split.train_test_split import train_test_split
 
 __all__ = [
-    "train_test_split",
+    "ComparisonReport",
     "CrossValidationReport",
     "EstimatorReport",
-    "ComparisonReport",
-    "RocCurveDisplay",
     "PrecisionRecallCurveDisplay",
     "PredictionErrorDisplay",
+    "RocCurveDisplay",
+    "train_test_split",
 ]

skore/src/skore/sklearn/_comparison/metrics_accessor.py

@@ -5,7 +5,12 @@
 from sklearn.utils.metaestimators import available_if
 
 from skore.externals._pandas_accessors import DirNamesMixin
-from skore.sklearn._base import _BaseAccessor
+from skore.sklearn._base import _BaseAccessor, _get_cached_response_values
+from skore.sklearn._comparison.precision_recall_curve_display import (
+    PrecisionRecallCurveDisplay,
+)
+from skore.sklearn._comparison.prediction_error_display import PredictionErrorDisplay
+from skore.sklearn._comparison.roc_curve_display import RocCurveDisplay
 from skore.utils._accessor import _check_supported_ml_task
 from skore.utils._index import flatten_multi_index
 from skore.utils._progress_bar import progress_decorator
@@ -1108,3 +1113,328 @@ def __repr__(self):
             class_name="skore.ComparisonReport.metrics",
             help_method_name="report.metrics.help()",
         )
+
+    @progress_decorator(description="Computing predictions for display")
+    def _get_display(
+        self,
+        *,
+        X,
+        y,
+        data_source,
+        response_method,
+        display_class,
+        display_kwargs,
+    ):
+        """Get the display from the cache or compute it.
+
+        Parameters
+        ----------
+        X : array-like of shape (n_samples, n_features)
+            The data.
+
+        y : array-like of shape (n_samples,)
+            The target.
+
+        data_source : {"test", "train", "X_y"}, 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.
+
+        response_method : str
+            The response method.
+
+        display_class : class
+            The display class.
+
+        display_kwargs : dict
+            The display kwargs used by `display_class._from_predictions`.
+
+        Returns
+        -------
+        display : display_class
+            The display.
+        """
+        cache_key = (self._parent._hash, display_class.__name__)
+        cache_key += tuple(display_kwargs.values())
+        cache_key += (data_source,)
+
+        progress = self._progress_info["current_progress"]
+        main_task = self._progress_info["current_task"]
+        total_estimators = len(self._parent.estimator_reports_)
+        progress.update(main_task, total=total_estimators)
+
+        if cache_key in self._parent._cache:
+            display = self._parent._cache[cache_key]
+        else:
+            y_true, y_pred = [], []
+
+            for report in self._parent.estimator_reports_:
+                report_X, report_y, _ = report.metrics._get_X_y_and_data_source_hash(
+                    data_source=data_source,
+                    X=X,
+                    y=y,
+                )
+
+                y_true.append(report_y)
+                y_pred.append(
+                    _get_cached_response_values(
+                        cache=report._cache,
+                        estimator_hash=report._hash,
+                        estimator=report._estimator,
+                        X=report_X,
+                        response_method=response_method,
+                        data_source=data_source,
+                        data_source_hash=None,
+                        pos_label=display_kwargs.get("pos_label", None),
+                    )
+                )
+                progress.update(main_task, advance=1, refresh=True)
+
+            display = display_class._from_predictions(
+                y_true,
+                y_pred,
+                estimators=[r.estimator_ for r in self._parent.estimator_reports_],
+                estimator_names=self._parent.report_names_,
+                ml_task=self._parent._ml_task,
+                data_source=data_source,
+                **display_kwargs,
+            )
+            self._parent._cache[cache_key] = display
+
+        return display
+
+    @available_if(
+        _check_supported_ml_task(
+            supported_ml_tasks=["binary-classification", "multiclass-classification"]
+        )
+    )
+    def roc(self, *, data_source="test", X=None, y=None, pos_label=None, ax=None):
+        """Plot the ROC curve.
+
+        Parameters
+        ----------
+        data_source : {"test", "train", "X_y"}, 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.
+
+        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
+            set provided when creating the report.
+
+        y : array-like of shape (n_samples,), default=None
+            New target on which to compute the metric. By default, we use the target
+            provided when creating the report.
+
+        pos_label : int, float, bool or str, default=None
+            The positive class.
+
+        Returns
+        -------
+        RocCurveDisplay
+            The ROC curve display.
+
+        Examples
+        --------
+        >>> from sklearn.datasets import load_breast_cancer
+        >>> from sklearn.linear_model import LogisticRegression
+        >>> from sklearn.model_selection import train_test_split
+        >>> from skore import ComparisonReport, EstimatorReport
+        >>> X, y = load_breast_cancer(return_X_y=True)
+        >>> X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)
+        >>> estimator_1 = LogisticRegression(max_iter=10000, random_state=42)
+        >>> estimator_report_1 = EstimatorReport(
+        ...     estimator_1,
+        ...     X_train=X_train,
+        ...     y_train=y_train,
+        ...     X_test=X_test,
+        ...     y_test=y_test,
+        ... )
+        >>> estimator_2 = LogisticRegression(max_iter=10000, random_state=43)
+        >>> estimator_report_2 = EstimatorReport(
+        ...     estimator_2,
+        ...     X_train=X_train,
+        ...     y_train=y_train,
+        ...     X_test=X_test,
+        ...     y_test=y_test,
+        ... )
+        >>> comparison_report = ComparisonReport(
+        ...     [estimator_report_1, estimator_report_2]
+        ... )
+        >>> display = comparison_report.metrics.roc()
+        >>> display.plot()
+        """
+        response_method = ("predict_proba", "decision_function")
+        display_kwargs = {"pos_label": pos_label}
+        return self._get_display(
+            X=X,
+            y=y,
+            data_source=data_source,
+            response_method=response_method,
+            display_class=RocCurveDisplay,
+            display_kwargs=display_kwargs,
+        )
+
+    @available_if(
+        _check_supported_ml_task(
+            supported_ml_tasks=["binary-classification", "multiclass-classification"]
+        )
+    )
+    def precision_recall(self, *, data_source="test", X=None, y=None, pos_label=None):
+        """Plot the precision-recall curve.
+
+        Parameters
+        ----------
+        data_source : {"test", "train", "X_y"}, 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.
+
+        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
+            set provided when creating the report.
+
+        y : array-like of shape (n_samples,), default=None
+            New target on which to compute the metric. By default, we use the target
+            provided when creating the report.
+
+        pos_label : int, float, bool or str, default=None
+            The positive class.
+
+        Returns
+        -------
+        PrecisionRecallCurveDisplay
+            The precision-recall curve display.
+
+        Examples
+        --------
+        >>> from sklearn.datasets import load_breast_cancer
+        >>> from sklearn.linear_model import LogisticRegression
+        >>> from sklearn.model_selection import train_test_split
+        >>> from skore import ComparisonReport, EstimatorReport
+        >>> X, y = load_breast_cancer(return_X_y=True)
+        >>> X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)
+        >>> estimator_1 = LogisticRegression(max_iter=10000, random_state=42)
+        >>> estimator_report_1 = EstimatorReport(
+        ...     estimator_1,
+        ...     X_train=X_train,
+        ...     y_train=y_train,
+        ...     X_test=X_test,
+        ...     y_test=y_test,
+        ... )
+        >>> estimator_2 = LogisticRegression(max_iter=10000, random_state=43)
+        >>> estimator_report_2 = EstimatorReport(
+        ...     estimator_2,
+        ...     X_train=X_train,
+        ...     y_train=y_train,
+        ...     X_test=X_test,
+        ...     y_test=y_test,
+        ... )
+        >>> comparison_report = ComparisonReport(
+        ...     [estimator_report_1, estimator_report_2]
+        ... )
+        >>> display = comparison_report.metrics.precision_recall()
+        >>> display.plot()
+        """
+        response_method = ("predict_proba", "decision_function")
+        display_kwargs = {"pos_label": pos_label}
+        return self._get_display(
+            X=X,
+            y=y,
+            data_source=data_source,
+            response_method=response_method,
+            display_class=PrecisionRecallCurveDisplay,
+            display_kwargs=display_kwargs,
+        )
+
+    @available_if(_check_supported_ml_task(supported_ml_tasks=["regression"]))
+    def prediction_error(
+        self,
+        *,
+        data_source="test",
+        X=None,
+        y=None,
+        subsample=1_000,
+        random_state=None,
+    ):
+        """Plot the prediction error of a regression model.
+
+        Extra keyword arguments will be passed to matplotlib's `plot`.
+
+        Parameters
+        ----------
+        data_source : {"test", "train", "X_y"}, 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.
+
+        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
+            set provided when creating the report.
+
+        y : array-like of shape (n_samples,), default=None
+            New target on which to compute the metric. By default, we use the target
+            provided when creating the report.
+
+        subsample : float, 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, 1,000 samples or less will be displayed.
+
+        random_state : int, default=None
+            The random state to use for the subsampling.
+
+        Returns
+        -------
+        PredictionErrorDisplay
+            The prediction error display.
+
+        Examples
+        --------
+        >>> from sklearn.datasets import load_diabetes
+        >>> from sklearn.linear_model import Ridge
+        >>> from sklearn.model_selection import train_test_split
+        >>> from skore import ComparisonReport, EstimatorReport
+        >>> X, y = load_diabetes(return_X_y=True)
+        >>> X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)
+        >>> estimator_1 = Ridge(random_state=42)
+        >>> estimator_report_1 = EstimatorReport(
+        ...     estimator_1,
+        ...     X_train=X_train,
+        ...     y_train=y_train,
+        ...     X_test=X_test,
+        ...     y_test=y_test,
+        ... )
+        >>> estimator_2 = Ridge(random_state=43)
+        >>> estimator_report_2 = EstimatorReport(
+        ...     estimator_2,
+        ...     X_train=X_train,
+        ...     y_train=y_train,
+        ...     X_test=X_test,
+        ...     y_test=y_test,
+        ... )
+        >>> comparison_report = ComparisonReport(
+        ...     [estimator_report_1, estimator_report_2]
+        ... )
+        >>> display = comparison_report.metrics.prediction_error()
+        >>> display.plot(kind="actual_vs_predicted")
+        """
+        display_kwargs = {"subsample": subsample, "random_state": random_state}
+        return self._get_display(
+            X=X,
+            y=y,
+            data_source=data_source,
+            response_method="predict",
+            display_class=PredictionErrorDisplay,
+            display_kwargs=display_kwargs,
+        )