style: apply black formatting

Signed-off-by: Rusheel Sharma <rusheelhere@gmail.com>

Author: Rusheel86

monai/losses/focal_loss.py

@@ -246,9 +246,7 @@ def forward(self, input: torch.Tensor, target: torch.Tensor) -> torch.Tensor:
             pass
 
         else:
-            raise ValueError(
-                f"Unsupported reduction: {self.reduction}, available options are [\"mean\", \"sum\", \"none\"]."
-            )
+            raise ValueError(f'Unsupported reduction: {self.reduction}, available options are ["mean", "sum", "none"].')
 
         return loss
 

monai/losses/unified_focal_loss.py

@@ -69,9 +69,7 @@ def forward(self, y_pred: torch.Tensor, y_true: torch.Tensor) -> torch.Tensor:
             if n_pred_ch == 1:
                 warnings.warn("single channel prediction, `to_onehot_y=True` ignored.")
             else:
-                if self.ignore_index is not None and (
-                    self.ignore_index < 0 or self.ignore_index >= n_pred_ch
-                ):
+                if self.ignore_index is not None and (self.ignore_index < 0 or self.ignore_index >= n_pred_ch):
                     # Replace sentinel ignore_index with a valid class before one_hot
                     y_true = torch.where(y_true == self.ignore_index, torch.tensor(0, device=y_true.device), y_true)
                 y_true = one_hot(y_true, num_classes=n_pred_ch)
@@ -148,9 +146,7 @@ def forward(self, y_pred: torch.Tensor, y_true: torch.Tensor) -> torch.Tensor:
             if n_pred_ch == 1:
                 warnings.warn("single channel prediction, `to_onehot_y=True` ignored.")
             else:
-                if self.ignore_index is not None and (
-                    self.ignore_index < 0 or self.ignore_index >= n_pred_ch
-                ):
+                if self.ignore_index is not None and (self.ignore_index < 0 or self.ignore_index >= n_pred_ch):
                     # Replace sentinel ignore_index with a valid class before one_hot
                     y_true = torch.where(y_true == self.ignore_index, torch.tensor(0, device=y_true.device), y_true)
                 y_true = one_hot(y_true, num_classes=n_pred_ch)
@@ -278,14 +274,10 @@ def forward(self, y_pred: torch.Tensor, y_true: torch.Tensor) -> torch.Tensor:
         if not (self.to_onehot_y and orig_pred_ch != 1):
             if not (orig_pred_ch == 1 and y_true.shape[1] == 1):
                 if y_true.shape != y_pred.shape:
-                    raise ValueError(
-                        f"ground truth has different shape ({y_true.shape}) from input ({y_pred.shape})"
-                    )
+                    raise ValueError(f"ground truth has different shape ({y_true.shape}) from input ({y_pred.shape})")
 
         if self.to_onehot_y and orig_pred_ch != 1:
-            if self.ignore_index is not None and (
-                self.ignore_index < 0 or self.ignore_index >= self.num_classes
-            ):
+            if self.ignore_index is not None and (self.ignore_index < 0 or self.ignore_index >= self.num_classes):
                 # Replace sentinel ignore_index with a valid class before one_hot
                 y_true = torch.where(y_true == self.ignore_index, torch.tensor(0, device=y_true.device), y_true)
             y_true = one_hot(y_true, num_classes=self.num_classes)
@@ -296,9 +288,7 @@ def forward(self, y_pred: torch.Tensor, y_true: torch.Tensor) -> torch.Tensor:
 
         mask = create_ignore_mask(original_y_true, self.ignore_index)
 
-        use_mask = self.ignore_index is not None and (
-            self.ignore_index < 0 or self.ignore_index >= self.num_classes
-        )
+        use_mask = self.ignore_index is not None and (self.ignore_index < 0 or self.ignore_index >= self.num_classes)
         if use_mask:
             y_pred_masked, y_true_masked = mask_loss_inputs(y_pred, y_true, self.ignore_index, mask=mask)
         else:

monai/losses/utils.py

@@ -71,10 +71,7 @@ def compute_tp_fp_fn(
 
 
 def mask_loss_inputs(
-    input: torch.Tensor,
-    target: torch.Tensor,
-    ignore_index: int | None,
-    mask: torch.Tensor | None = None,
+    input: torch.Tensor, target: torch.Tensor, ignore_index: int | None, mask: torch.Tensor | None = None
 ) -> tuple[torch.Tensor, torch.Tensor]:
     """Apply ignore_index masking to loss inputs."""
     if mask is None and ignore_index is not None:

monai/metrics/confusion_matrix.py

@@ -184,7 +184,12 @@ class range, ignored regions are inferred from spatial locations where all label
         mask = None
     else:
         mask = create_ignore_mask(original_y if ignore_index is not None else y, ignore_index)
-        if mask is not None and not include_background and mask.shape[1] != y_pred.shape[1] and mask.shape[1] > y_pred.shape[1]:
+        if (
+            mask is not None
+            and not include_background
+            and mask.shape[1] != y_pred.shape[1]
+            and mask.shape[1] > y_pred.shape[1]
+        ):
             mask = mask[:, 1:]
 
     # convert to [BNS], where S is the number of pixels for one sample.

monai/metrics/meaniou.py

@@ -162,7 +162,12 @@ def compute_iou(
         mask = None
     else:
         mask = create_ignore_mask(original_y if ignore_index is not None else y, ignore_index)
-        if mask is not None and not include_background and mask.shape[1] != y_pred.shape[1] and mask.shape[1] > y_pred.shape[1]:
+        if (
+            mask is not None
+            and not include_background
+            and mask.shape[1] != y_pred.shape[1]
+            and mask.shape[1] > y_pred.shape[1]
+        ):
             mask = mask[:, 1:]
     if mask is not None:
         if mask.shape != y_pred.shape:

tests/losses/test_ignore_index_losses.py

@@ -76,7 +76,9 @@ def test_no_ignore_behavior(self, loss_class, kwargs):
         output = loss_func(input_data, target)
         self.assertFalse(torch.isnan(output))
 
-    @parameterized.expand([(loss_class, kwargs, ignore_index) for loss_class, kwargs in CLASS_INDEX_TEST_CASES for ignore_index in (0, 1)])
+    @parameterized.expand(
+        [(loss_class, kwargs, ignore_index) for loss_class, kwargs in CLASS_INDEX_TEST_CASES for ignore_index in (0, 1)]
+    )
     def test_loss_ignore_class_index(self, loss_class, kwargs, ignore_index):
         loss_func = loss_class(ignore_index=ignore_index, **kwargs)
 

tests/metrics/test_ignore_index_metrics.py

@@ -92,7 +92,11 @@ def test_metric_ignore_consistency(self, metric_class, kwargs):
         torch.testing.assert_close(res1, res2, msg=f"Failed for {metric_class.__name__}")
 
     @parameterized.expand(
-        [(metric_class, kwargs, ignore_index) for metric_class, kwargs in TEST_METRICS + SCIPY_METRICS for ignore_index in (0, 1)]
+        [
+            (metric_class, kwargs, ignore_index)
+            for metric_class, kwargs in TEST_METRICS + SCIPY_METRICS
+            for ignore_index in (0, 1)
+        ]
     )
     def test_metric_ignore_class_index(self, metric_class, kwargs, ignore_index):
         metric = metric_class(ignore_index=ignore_index, **kwargs)