wip

Author: jaffourt

sklearn/metrics/_ranking.py

@@ -29,6 +29,7 @@
     column_or_1d,
 )
 from sklearn.utils._array_api import (
+    _average,
     _max_precision_float_dtype,
     get_namespace_and_device,
     size,
@@ -1412,6 +1413,7 @@ def coverage_error(y_true, y_score, *, sample_weight=None):
     >>> coverage_error(y_true, y_score)
     1.5
     """
+    xp, _, device_ = get_namespace_and_device(y_true, y_score, sample_weight)
     y_true = check_array(y_true, ensure_2d=True)
     y_score = check_array(y_score, ensure_2d=True)
     check_consistent_length(y_true, y_score, sample_weight)
@@ -1423,12 +1425,12 @@ def coverage_error(y_true, y_score, *, sample_weight=None):
     if y_true.shape != y_score.shape:
         raise ValueError("y_true and y_score have different shape")
 
-    y_score_mask = np.ma.masked_array(y_score, mask=np.logical_not(y_true))
-    y_min_relevant = y_score_mask.min(axis=1).reshape((-1, 1))
-    coverage = (y_score >= y_min_relevant).sum(axis=1)
-    coverage = coverage.filled(0)
-
-    return float(np.average(coverage, weights=sample_weight))
+    y_true_logical_not = xp.astype(y_true, xp.bool, device=device_, copy=False)
+    inf_val = xp.asarray(xp.inf, dtype=y_score.dtype, device=device_)
+    y_score_relevant_only = xp.where(y_true_logical_not, y_score, inf_val)
+    y_min_relevant = xp.reshape(xp.min(y_score_relevant_only, axis=1), (-1, 1))
+    coverage = xp.count_nonzero(y_score >= y_min_relevant, axis=1)
+    return _average(coverage, weights=sample_weight)
 
 
 @validate_params(

sklearn/metrics/tests/test_common.py

@@ -2125,6 +2125,35 @@ def check_array_api_multilabel_classification_metric(
         )
 
 
+def check_array_api_continuous_classification_metric(
+    metric, array_namespace, device, dtype_name
+):
+    y_true_np = np.array([[0, 1, 0], [1, 1, 0]])
+    y_score_np = np.array([[0.1, 10.0, -3], [0, 1, 3]])
+
+    check_array_api_metric(
+        metric,
+        array_namespace,
+        device,
+        dtype_name,
+        a_np=y_true_np,
+        b_np=y_score_np,
+        sample_weight=None,
+    )
+
+    sample_weight = np.array([0.1, 2.0], dtype=dtype_name)
+
+    check_array_api_metric(
+        metric,
+        array_namespace,
+        device,
+        dtype_name,
+        a_np=y_true_np,
+        b_np=y_score_np,
+        sample_weight=sample_weight,
+    )
+
+
 def check_array_api_regression_metric(metric, array_namespace, device, dtype_name):
     func_name = metric.func.__name__ if isinstance(metric, partial) else metric.__name__
     if func_name == "mean_poisson_deviance" and sp_version < parse_version("1.14.0"):