Clean up some lints

Summary: Trying out a new tool and cleaning up some linter issues it flagged.

Reviewed By: esantorella

Differential Revision: D65274533

botorch/acquisition/acquisition.py

@@ -53,6 +53,7 @@ def set_X_pending(self, X_pending: Tensor | None = None) -> None:
                     "Pending points require a gradient but the acquisition function"
                     " will not provide a gradient to these points.",
                     BotorchWarning,
+                    stacklevel=2,
                 )
             self.X_pending = X_pending.detach().clone()
         else:

botorch/acquisition/cached_cholesky.py

@@ -87,6 +87,7 @@ def __init__(
             warnings.warn(
                 _get_cache_root_not_supported_message(type(model)),
                 RuntimeWarning,
+                stacklevel=3,
             )
             cache_root = False
         self._cache_root = cache_root
@@ -150,6 +151,7 @@ def _get_f_X_samples(self, posterior: GPyTorchPosterior, q_in: int) -> Tensor:
                     "ill-conditioned covariance matrix. "
                     "Falling back to standard sampling.",
                     BotorchWarning,
+                    stacklevel=3,
                 )
 
         # TODO: improve efficiency for multi-task models

botorch/acquisition/cost_aware.py

@@ -209,6 +209,7 @@ def forward(
                 warnings.warn(
                     "Encountered negative cost values in InverseCostWeightedUtility",
                     CostAwareWarning,
+                    stacklevel=2,
                 )
         # clamp (away from zero) and sum cost across elements of the q-batch -
         # this will be of shape `num_fantasies x batch_shape` or `batch_shape`

botorch/acquisition/decoupled.py

@@ -110,6 +110,7 @@ def set_X_pending(
                     "Pending points require a gradient but the acquisition function"
                     " will not provide a gradient to these points.",
                     BotorchWarning,
+                    stacklevel=2,
                 )
             self.X_pending = X_pending.detach().clone()
             if X_pending_evaluation_mask is not None:

botorch/acquisition/multi_objective/utils.py

@@ -61,7 +61,11 @@ def get_default_partitioning_alpha(num_objectives: int) -> float:
     if num_objectives <= 4:
         return 0.0
     elif num_objectives > 6:
-        warnings.warn("EHVI works best for less than 7 objectives.", BotorchWarning)
+        warnings.warn(
+            "EHVI works best for less than 7 objectives.",
+            BotorchWarning,
+            stacklevel=3,
+        )
     return 10 ** (-8 + num_objectives)
 
 

botorch/acquisition/multi_step_lookahead.py

@@ -552,6 +552,7 @@ def _construct_inner_samplers(
                 warnings.warn(
                     "inner_mc_samples is ignored for analytic acquisition functions",
                     BotorchWarning,
+                    stacklevel=3,
                 )
             inner_samplers.append(None)
         else:

botorch/models/model.py

@@ -236,6 +236,7 @@ def _set_transformed_inputs(self) -> None:
                     "attribute. Make sure that the `input_transform` is applied to "
                     "both the train inputs and test inputs.",
                     RuntimeWarning,
+                    stacklevel=3,
                 )
 
     def _revert_to_original_inputs(self) -> None:

botorch/utils/multi_objective/hypervolume.py

@@ -789,6 +789,7 @@ def set_X_pending(self, X_pending: Tensor | None = None) -> None:
                     "Pending points require a gradient but the acquisition function"
                     " will not provide a gradient to these points.",
                     BotorchWarning,
+                    stacklevel=2,
                 )
             X_pending = X_pending.detach().clone()
             if self.cache_pending:

botorch/utils/transforms.py

@@ -190,6 +190,7 @@ def _verify_output_shape(acqf: Any, X: Tensor, output: Tensor) -> bool:
             f"of X, but got output with shape {output.shape} for X with shape "
             f"{X.shape}. Make sure that this is the intended behavior!",
             RuntimeWarning,
+            stacklevel=3,
         )
         return True
 

test/exceptions/test_warnings.py

@@ -43,7 +43,7 @@ def test_botorch_warnings(self):
             UserInputWarning,
         ):
             with warnings.catch_warnings(record=True) as ws, settings.debug(True):
-                warnings.warn("message", WarningClass)
+                warnings.warn("message", WarningClass, stacklevel=1)
                 self.assertEqual(len(ws), 1)
                 self.assertTrue(issubclass(ws[-1].category, WarningClass))
                 self.assertTrue("message" in str(ws[-1].message))

test/test_settings.py

@@ -31,12 +31,12 @@ def test_debug(self):
         with settings.debug(False):
             with warnings.catch_warnings(record=True) as ws:
                 if settings.debug.on():
-                    warnings.warn("test", BotorchWarning)
+                    warnings.warn("test", BotorchWarning, stacklevel=1)
             self.assertEqual(len(ws), 0)
         # Check that warnings are not suppressed outside of context manager.
         with warnings.catch_warnings(record=True) as ws:
             if settings.debug.on():
-                warnings.warn("test", BotorchWarning)
+                warnings.warn("test", BotorchWarning, stacklevel=1)
         self.assertEqual(len(ws), 1)
 
         # Turn off debug.
@@ -45,12 +45,12 @@ def test_debug(self):
         with settings.debug(True):
             with warnings.catch_warnings(record=True) as ws:
                 if settings.debug.on():
-                    warnings.warn("test", BotorchWarning)
+                    warnings.warn("test", BotorchWarning, stacklevel=1)
             self.assertEqual(len(ws), 1)
         # Check that warnings are suppressed outside of context manager.
         with warnings.catch_warnings(record=True) as ws:
             if settings.debug.on():
-                warnings.warn("test", BotorchWarning)
+                warnings.warn("test", BotorchWarning, stacklevel=1)
         self.assertEqual(len(ws), 0)