formatting, build

Author: seashoo

botorch/sampling/pathwise/features/maps.py

@@ -128,8 +128,8 @@ def forward(self, x: Tensor, **kwargs: Any) -> Tensor:
 
         blocks: list[Tensor] = []
 
-        shape = self.raw_output_shape              # target output shape
-        ndim = len(shape)                          # #feature dimensions incl. batch
+        shape = self.raw_output_shape  # target output shape
+        ndim = len(shape)  # #feature dimensions incl. batch
 
         for feature_map in self:
             # 1. Evaluate (dense) features for the current sub-map.
@@ -152,7 +152,7 @@ def forward(self, x: Tensor, **kwargs: Any) -> Tensor:
                 # only up to such a scaling).
                 num_copies = prod(tile_shape)
                 if num_copies > 1:
-                    block = block * (num_copies ** -0.5)
+                    block = block * (num_copies**-0.5)
 
                 # ``multi_index`` inserts ``None`` (i.e. `None` in slice syntax)
                 # so that broadcasting expands the tensor along the new axes
@@ -174,8 +174,8 @@ def forward(self, x: Tensor, **kwargs: Any) -> Tensor:
 
     @property
     def raw_output_shape(self) -> Size:
-    # If the container is empty (e.g. DirectSumFeatureMap([])), treat the
-    # output as 0-D until feature maps are added.
+        # If the container is empty (e.g. DirectSumFeatureMap([])), treat the
+        # output as 0-D until feature maps are added.
         if not self:
             return Size([])
 

botorch/sampling/pathwise/prior_samplers.py

@@ -6,7 +6,6 @@
 
 from __future__ import annotations
 
-from copy import deepcopy
 from typing import Any, Callable, List
 
 import torch

botorch/sampling/pathwise/update_strategies.py

@@ -7,7 +7,6 @@
 from __future__ import annotations
 
 from collections.abc import Callable
-from copy import deepcopy
 from types import NoneType
 from typing import Any
 
@@ -167,7 +166,9 @@ def _draw_kernel_feature_paths_MultiTaskGP(
     # Determine total input dimensionality and identify the task-feature index.
     num_inputs = points.shape[-1]
     task_index = (
-        num_inputs + model._task_feature if model._task_feature < 0 else model._task_feature
+        num_inputs + model._task_feature
+        if model._task_feature < 0
+        else model._task_feature
     )
 
     # MTGP should always provide a ProductKernel = data × task. Enforce that
@@ -177,9 +178,11 @@ def _draw_kernel_feature_paths_MultiTaskGP(
 
     if not isinstance(model.covar_module, ProductKernel):
         raise RuntimeError(
-            "MultiTaskGP `covar_module` is expected to be a ProductKernel (data × task) "
+            "MultiTaskGP `covar_module` is expected to be a ProductKernel "
+            "(data × task) "
             f"but found {type(model.covar_module).__name__}. If you build a custom "
-            "MTGP variant please wrap the two kernels with gpytorch.kernels.ProductKernel."
+            "MTGP variant please wrap the two kernels with "
+            "gpytorch.kernels.ProductKernel."
         )
 
     combined_kernel = model.covar_module

test/sampling/pathwise/test_prior_samplers.py

@@ -235,6 +235,7 @@ def test_multitask_gp_kernel_handling(self):
         model2.covar_module = simple_kernel  # Non-ProductKernel
 
         import pytest
+
         with pytest.raises(RuntimeError):
             draw_kernel_feature_paths(model2, sample_shape=Size([1]))
 

test/sampling/pathwise/test_update_strategies.py

@@ -312,6 +312,7 @@ def test_multitask_gp_kernel_handling(self):
         model2.covar_module = simple_kernel  # Non-ProductKernel
 
         import pytest
+
         with pytest.raises(RuntimeError):
             gaussian_update(model=model2, sample_values=sample_values)
 

website/notebooks_community.json

@@ -1,18 +1 @@
-[
-  {
-    "id": "clf_constrained_bo",
-    "title": "BO with Binary Black-box Constraint"
-  },
-  {
-    "id": "hentropy_search",
-    "title": "H-Entropy Search"
-  },
-  {
-    "id": "multi_source_bo",
-    "title": "Multi-Information Source BO with Augmented GP"
-  },
-  {
-    "id": "vbll_thompson_sampling",
-    "title": "Variational Bayesian Last Layers for BO"
-  }
-]
+[]