lint: flake8 line lengths and other package level issues

Author: CompRhys

botorch/acquisition/analytic.py

@@ -982,7 +982,7 @@ def _log_ei_helper(u: Tensor) -> Tensor:
     if not (u.dtype == torch.float32 or u.dtype == torch.float64):
         raise TypeError(
             f"LogExpectedImprovement only supports torch.float32 and torch.float64 "
-            f"dtypes, but received {u.dtype = }."
+            f"dtypes, but received {u.dtype=}."
         )
     # The function has two branching decisions. The first is u < bound, and in this
     # case, just taking the logarithm of the naive _ei_helper implementation works.

botorch/acquisition/logei.py

@@ -540,7 +540,7 @@ def check_tau(tau: FloatOrTensor, name: str) -> FloatOrTensor:
     """Checks the validity of the tau arguments of the functions below, and returns
     `tau` if it is valid."""
     if isinstance(tau, Tensor) and tau.numel() != 1:
-        raise ValueError(name + f" is not a scalar: {tau.numel() = }.")
+        raise ValueError(name + f" is not a scalar: {tau.numel()=}.")
     if not (tau > 0):
-        raise ValueError(name + f" is non-positive: {tau = }.")
+        raise ValueError(name + f" is non-positive: {tau=}.")
     return tau

botorch/acquisition/multi_objective/hypervolume_knowledge_gradient.py

@@ -557,7 +557,7 @@ def _split_hvkg_fantasy_points(
     """
     if n_f * num_pareto > X.size(-2):
         raise ValueError(
-            f"`n_f*num_pareto` ({n_f*num_pareto}) must be less than"
+            f"`n_f*num_pareto` ({n_f * num_pareto}) must be less than"
             f" the `q`-batch dimension of `X` ({X.size(-2)})."
         )
     split_sizes = [X.size(-2) - n_f * num_pareto, n_f * num_pareto]

botorch/optim/optimize.py

@@ -122,10 +122,10 @@ def __post_init__(self) -> None:
             ):
                 if len(batch_initial_conditions_shape) == 2:
                     warnings.warn(
-                        "If using `batch_initial_conditions` together with `raw_samples`, "
-                        "the `batch_initial_conditions` must have 3 dimensions. "
-                        "Defaulting to old behavior of ignoring `raw_samples` by setting "
-                        "it to None.",
+                        "If using `batch_initial_conditions` together with "
+                        "`raw_samples`, the `batch_initial_conditions` must "
+                        "have 3 dimensions. Defaulting to old behavior of ignoring "
+                        "`raw_samples` by setting it to None.",
                         RuntimeWarning,
                     )
                     # Use object.__setattr__ to bypass immutability and set a value
@@ -137,11 +137,11 @@ def __post_init__(self) -> None:
                     and batch_initial_conditions_shape[0] not in [1, self.num_restarts]
                 ):
                     warnings.warn(
-                        "If using `batch_initial_conditions` together with `raw_samples`, "
-                        "the first repeat dimension of `batch_initial_conditions` must "
-                        "match `num_restarts` or be 1 to allow repeat matching. "
-                        "Defaulting to old behavior of ignoring `raw_samples` by setting "
-                        "it to None.",
+                        "If using `batch_initial_conditions` together with "
+                        "`raw_samples`, the first repeat dimension of "
+                        "`batch_initial_conditions` must match `num_restarts` "
+                        "or be 1 to allow repeat matching. Defaulting to old "
+                        "behavior of ignoring `raw_samples` by setting it to None.",
                         RuntimeWarning,
                     )
                     # Use object.__setattr__ to bypass immutability and set a value
@@ -280,7 +280,7 @@ def _optimize_acqf_sequential_q(
             if base_X_pending is not None
             else candidates
         )
-        logger.info(f"Generated sequential candidate {i+1} of {opt_inputs.q}")
+        logger.info(f"Generated sequential candidate {i + 1} of {opt_inputs.q}")
     opt_inputs.acq_function.set_X_pending(base_X_pending)
     return candidates, torch.stack(acq_value_list)
 
@@ -332,7 +332,8 @@ def _optimize_acqf_batch(opt_inputs: OptimizeAcqfInputs) -> tuple[Tensor, Tensor
             required_raw_samples -= opt_inputs.batch_initial_conditions.shape[-2]
 
         if required_raw_samples > 0:
-            # pyre-ignore[28]: Unexpected keyword argument `acq_function` to anonymous call.
+            # pyre-ignore[28]: Unexpected keyword argument `acq_function`
+            # to anonymous call.
             generated_initial_conditions = opt_inputs.get_ic_generator()(
                 acq_function=opt_inputs.acq_function,
                 bounds=opt_inputs.bounds,
@@ -402,7 +403,7 @@ def _optimize_batch_candidates() -> tuple[Tensor, Tensor, list[Warning]]:
             opt_warnings += ws
             batch_candidates_list.append(batch_candidates_curr)
             batch_acq_values_list.append(batch_acq_values_curr)
-            logger.info(f"Generated candidate batch {i+1} of {len(batched_ics)}.")
+            logger.info(f"Generated candidate batch {i + 1} of {len(batched_ics)}.")
 
         batch_candidates = torch.cat(batch_candidates_list)
         has_scalars = batch_acq_values_list[0].ndim == 0

botorch/utils/probability/utils.py

@@ -161,7 +161,7 @@ def log_ndtr(x: Tensor) -> Tensor:
     if not (x.dtype == torch.float32 or x.dtype == torch.float64):
         raise TypeError(
             f"log_Phi only supports torch.float32 and torch.float64 "
-            f"dtypes, but received {x.dtype = }."
+            f"dtypes, but received {x.dtype=}."
         )
     neg_inv_sqrt_2, log_2 = get_constants_like((_neg_inv_sqrt_2, _log_2), x)
     return log_erfc(neg_inv_sqrt_2 * x) - log_2
@@ -181,7 +181,7 @@ def log_erfc(x: Tensor) -> Tensor:
     if not (x.dtype == torch.float32 or x.dtype == torch.float64):
         raise TypeError(
             f"log_erfc only supports torch.float32 and torch.float64 "
-            f"dtypes, but received {x.dtype = }."
+            f"dtypes, but received {x.dtype=}."
         )
     is_pos = x > 0
     x_pos = x.masked_fill(~is_pos, 0)

test/acquisition/multi_objective/test_hypervolume_knowledge_gradient.py

@@ -436,7 +436,7 @@ def test_split_hvkg_fantasy_points(self):
         n_f = 100
         num_pareto = 3
         msg = (
-            rf".*\({n_f*num_pareto}\) must be less than"
+            rf".*\({n_f * num_pareto}\) must be less than"
             rf" the `q`-batch dimension of `X` \({X.size(-2)}\)\."
         )
         with self.assertRaisesRegex(ValueError, msg):

test/optim/test_optimize.py

@@ -259,7 +259,7 @@ def test_optimize_acqf_joint(
                 cnt += 1
                 self.assertEqual(mock_gen_batch_initial_conditions.call_count, cnt)
 
-                # test generation with provided initial conditions larger than raw_samples
+                # test case where provided initial conditions larger than raw_samples
                 candidates, acq_vals = optimize_acqf(
                     acq_function=mock_acq_function,
                     bounds=bounds,