Fix bootstrapper not working with kwargs arguments

CHANGELOG.md

@@ -36,6 +36,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Fixed axis names with Precision-Recall curve ([#2462](https://github.com/Lightning-AI/torchmetrics/pull/2462))
 
 
+- Fixed `BootStrapper` wrapper not working with `kwargs` provided argument ([#2503](https://github.com/Lightning-AI/torchmetrics/pull/2503))
+
+
 - Fixed warnings being suppressed in `MeanAveragePrecision` when requested ([#2501](https://github.com/Lightning-AI/torchmetrics/pull/2501))
 
 

src/torchmetrics/wrappers/bootstrapping.py

@@ -129,11 +129,11 @@ def update(self, *args: Any, **kwargs: Any) -> None:
 
         """
         args_sizes = apply_to_collection(args, Tensor, len)
-        kwargs_sizes = list(apply_to_collection(kwargs, Tensor, len))
+        kwargs_sizes = apply_to_collection(kwargs, Tensor, len)
         if len(args_sizes) > 0:
             size = args_sizes[0]
         elif len(kwargs_sizes) > 0:
-            size = kwargs_sizes[0]
+            size = next(iter(kwargs_sizes.values()))
         else:
             raise ValueError("None of the input contained tensors, so could not determine the sampling size")
 

tests/unittests/wrappers/test_bootstrapping.py

@@ -22,7 +22,7 @@
 from sklearn.metrics import mean_squared_error, precision_score, recall_score
 from torch import Tensor
 from torchmetrics.classification import MulticlassF1Score, MulticlassPrecision, MulticlassRecall
-from torchmetrics.regression import MeanSquaredError
+from torchmetrics.regression import MeanAbsoluteError, MeanSquaredError
 from torchmetrics.wrappers.bootstrapping import BootStrapper, _bootstrap_sampler
 
 from unittests._helpers import seed_all
@@ -140,3 +140,23 @@ def test_low_sample_amount(sampling_strategy):
         MulticlassF1Score(num_classes=3, average=None), num_bootstraps=20, sampling_strategy=sampling_strategy
     )
     assert bootstrap_f1(preds, target)  # does not work
+
+
+def test_args_and_kwargs_works():
+    """Test that metric works with both args and kwargs and mix.
+
+    See issue: https://github.com/Lightning-AI/torchmetrics/issues/2450
+
+    """
+    x = torch.rand(100)
+    y = x + torch.randn_like(x)
+    ae = MeanAbsoluteError()
+    assert ae(x, y) == ae(preds=x, target=y)
+
+    bootstrapped_ae = BootStrapper(ae)
+    res1 = bootstrapped_ae(x, y)
+    res2 = bootstrapped_ae(x, target=y)
+    res3 = bootstrapped_ae(preds=x, target=y)
+
+    assert (res1["mean"].shape == res2["mean"].shape) & (res2["mean"].shape == res3["mean"].shape)
+    assert (res1["std"].shape == res2["std"].shape) & (res2["mean"].shape == res3["std"].shape)