Ignore problems with method overrides: #1065

sbi/inference/posteriors/ensemble_posterior.py

@@ -143,7 +143,7 @@ def weights(self, weights: Optional[Union[List[float], Tensor]]) -> None:
         else:
             raise TypeError
 
-    def sample(
+    def sample(  # pyright: ignore[reportIncompatibleMethodOverride]
         self, sample_shape: Shape = torch.Size(), x: Optional[Tensor] = None, **kwargs
     ) -> Tensor:
         r"""Return samples from posterior ensemble.
@@ -271,7 +271,7 @@ def potential(
             theta.to(self._device), track_gradients=track_gradients
         )
 
-    def map(
+    def map(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         x: Optional[Tensor] = None,
         num_iter: int = 1_000,
@@ -399,7 +399,7 @@ def __init__(
         self.potential_fns = potential_fns
         super().__init__(prior, x_o, device)
 
-    def allow_iid_x(self) -> bool:
+    def allow_iid_x(self) -> bool:  # pyright: ignore[reportIncompatibleMethodOverride]
         # in case there is different kinds of posteriors, this will produce an error
         # in `set_x()`
         return any(

sbi/inference/posteriors/importance_posterior.py

@@ -153,7 +153,7 @@ def estimate_normalization_constant(
 
         return self._normalization_constant.to(self._device)  # type: ignore
 
-    def sample(
+    def sample(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         sample_shape: Shape = torch.Size(),
         x: Optional[Tensor] = None,

sbi/inference/posteriors/mcmc_posterior.py

@@ -187,7 +187,7 @@ def log_prob(
             theta.to(self._device), track_gradients=track_gradients
         )
 
-    def sample(
+    def sample(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         sample_shape: Shape = torch.Size(),
         x: Optional[Tensor] = None,

sbi/inference/posteriors/rejection_posterior.py

@@ -99,7 +99,7 @@ def log_prob(
             theta.to(self._device), track_gradients=track_gradients
         )
 
-    def sample(
+    def sample(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         sample_shape: Shape = torch.Size(),
         x: Optional[Tensor] = None,

sbi/inference/posteriors/vi_posterior.py

@@ -274,7 +274,7 @@ def set_vi_method(self, method: str) -> "VIPosterior":
         self._optimizer_builder = get_VI_method(method)
         return self
 
-    def sample(
+    def sample(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         sample_shape: Shape = torch.Size(),
         x: Optional[Tensor] = None,

sbi/inference/snle/mnle.py

@@ -65,7 +65,7 @@ def __init__(
         kwargs = del_entries(locals(), entries=("self", "__class__"))
         super().__init__(**kwargs)
 
-    def train(
+    def train(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         training_batch_size: int = 50,
         learning_rate: float = 5e-4,

sbi/inference/snle/snle_base.py

@@ -67,7 +67,7 @@ def __init__(
         else:
             self._build_neural_net = density_estimator
 
-    def append_simulations(
+    def append_simulations(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         theta: Tensor,
         x: Tensor,
@@ -112,7 +112,7 @@ def append_simulations(
             data_device=data_device,
         )
 
-    def train(
+    def train(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         training_batch_size: int = 50,
         learning_rate: float = 5e-4,

sbi/inference/snpe/snpe_a.py

@@ -94,7 +94,7 @@ def __init__(
         )
         super().__init__(**kwargs)
 
-    def train(
+    def train(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         final_round: bool = False,
         training_batch_size: int = 50,
@@ -271,7 +271,7 @@ def correct_for_proposal(
         )
         return wrapped_density_estimator
 
-    def build_posterior(
+    def build_posterior(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         density_estimator: Optional[TorchModule] = None,
         prior: Optional[Distribution] = None,
@@ -423,7 +423,7 @@ def __init__(
         # Take care of z-scoring, pre-compute and store prior terms.
         self._set_state_for_mog_proposal()
 
-    def log_prob(self, inputs: Tensor, condition: Tensor, **kwargs) -> Tensor:
+    def log_prob(self, inputs: Tensor, condition: Tensor, **kwargs) -> Tensor:  # pyright: ignore[reportIncompatibleMethodOverride]
         """Compute the log-probability of the approximate posterior.
 
         Args:

sbi/inference/snpe/snpe_b.py

@@ -39,7 +39,7 @@ def __init__(
         kwargs = del_entries(locals(), entries=("self", "__class__"))
         super().__init__(**kwargs)
 
-    def _log_prob_proposal_posterior(
+    def _log_prob_proposal_posterior(  # pyright: ignore[reportIncompatibleMethodOverride]
         self, theta: Tensor, x: Tensor, masks: Tensor
     ) -> Tensor:
         """

sbi/inference/snpe/snpe_base.py

@@ -84,7 +84,7 @@ def __init__(
         self._proposal_roundwise = []
         self.use_non_atomic_loss = False
 
-    def append_simulations(
+    def append_simulations(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         theta: Tensor,
         x: Tensor,
@@ -201,7 +201,7 @@ def append_simulations(
 
         return self
 
-    def train(
+    def train(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         training_batch_size: int = 50,
         learning_rate: float = 5e-4,

sbi/inference/snpe/snpe_c.py

@@ -83,7 +83,7 @@ def __init__(
         kwargs = del_entries(locals(), entries=("self", "__class__"))
         super().__init__(**kwargs)
 
-    def train(
+    def train(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         num_atoms: int = 10,
         training_batch_size: int = 50,
@@ -286,7 +286,7 @@ def _log_prob_proposal_posterior(
                     "The density estimator must be a MDNtext for non-atomic loss."
                 )
 
-            return self._log_prob_proposal_posterior_mog(theta, x, proposal)
+            return self._log_prob_proposal_posterior_mog(theta, x, proposal)  # pyright: ignore[reportIncompatibleMethodOverride]
         else:
             if not hasattr(self._neural_net, "log_prob"):
                 raise ValueError(

sbi/inference/snre/bnre.py

@@ -50,7 +50,7 @@ def __init__(
         kwargs = del_entries(locals(), entries=("self", "__class__"))
         super().__init__(**kwargs)
 
-    def train(
+    def train(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         regularization_strength: float = 100.0,
         training_batch_size: int = 50,
@@ -104,7 +104,7 @@ def train(
         }
         return super().train(**kwargs)
 
-    def _loss(
+    def _loss(  # pyright: ignore[reportIncompatibleMethodOverride]
         self, theta: Tensor, x: Tensor, num_atoms: int, regularization_strength: float
     ) -> Tensor:
         """Returns the binary cross-entropy loss for the trained classifier.

sbi/inference/snre/snre_a.py

@@ -47,7 +47,7 @@ def __init__(
         kwargs = del_entries(locals(), entries=("self", "__class__"))
         super().__init__(**kwargs)
 
-    def train(
+    def train(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         training_batch_size: int = 50,
         learning_rate: float = 5e-4,

sbi/inference/snre/snre_b.py

@@ -47,7 +47,7 @@ def __init__(
         kwargs = del_entries(locals(), entries=("self", "__class__"))
         super().__init__(**kwargs)
 
-    def train(
+    def train(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         num_atoms: int = 10,
         training_batch_size: int = 50,

sbi/inference/snre/snre_base.py

@@ -79,7 +79,7 @@ def __init__(
         else:
             self._build_neural_net = classifier
 
-    def append_simulations(
+    def append_simulations(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         theta: Tensor,
         x: Tensor,
@@ -123,7 +123,7 @@ def append_simulations(
             data_device=data_device,
         )
 
-    def train(
+    def train(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         num_atoms: int = 10,
         training_batch_size: int = 50,

sbi/inference/snre/snre_c.py

@@ -61,7 +61,7 @@ def __init__(
         kwargs = del_entries(locals(), entries=("self", "__class__"))
         super().__init__(**kwargs)
 
-    def train(
+    def train(  # pyright: ignore[reportIncompatibleMethodOverride]
         self,
         num_classes: int = 5,
         gamma: float = 1.0,
@@ -123,7 +123,7 @@ def train(
         kwargs["loss_kwargs"] = {"gamma": kwargs.pop("gamma")}
         return super().train(**kwargs)
 
-    def _loss(
+    def _loss(  # pyright: ignore[reportIncompatibleMethodOverride]
         self, theta: Tensor, x: Tensor, num_atoms: int, gamma: float
     ) -> torch.Tensor:
         r"""Return cross-entropy loss (via ''multi-class sigmoid'' activation) for