Move the CallablePotentialWrapper to make posteriors pickleable

sbi/inference/posteriors/base_posterior.py

@@ -8,11 +8,14 @@
 import torch.distributions.transforms as torch_tf
 from torch import Tensor
 
+from sbi.inference.potentials.base_potential import (
+    BasePotential,
+    CallablePotentialWrapper,
+)
 from sbi.types import Array, Shape, TorchTransform
 from sbi.utils import gradient_ascent
 from sbi.utils.torchutils import ensure_theta_batched, process_device
 from sbi.utils.user_input_checks import process_x
-from sbi.inference.potentials.base_potential import BasePotential
 
 
 class NeuralPosterior(ABC):
@@ -41,16 +44,10 @@ def __init__(
         """
         if not isinstance(potential_fn, BasePotential):
             callable_potential = potential_fn
-
-            class CallablePotentialWrapper(BasePotential):
-                """If `potential_fn` is a callable it gets wrapped as this."""
-                allow_iid_x = True
-
-                def __call__(self, theta, **kwargs):
-                    return callable_potential(theta=theta, x_o=self.x_o, **kwargs)
-
-            potential_fn = CallablePotentialWrapper(None, None)
-
+            potential_device = "cpu" if device is None else device
+            potential_fn = CallablePotentialWrapper(
+                callable_potential, prior=None, x_o=None, device=potential_device
+            )
 
         # Ensure device string.
         self._device = process_device(potential_fn.device if device is None else device)

sbi/inference/posteriors/mcmc_posterior.py

@@ -31,7 +31,6 @@
 from sbi.types import Shape, TorchTransform
 from sbi.utils import pyro_potential_wrapper, tensor2numpy, transformed_potential
 from sbi.utils.torchutils import ensure_theta_batched
-from sbi.inference.potentials.base_potential import BasePotential
 
 
 class MCMCPosterior(NeuralPosterior):

sbi/inference/potentials/base_potential.py

@@ -9,7 +9,10 @@
 
 class BasePotential(metaclass=ABCMeta):
     def __init__(
-        self, prior: Optional[Distribution], x_o: Optional[Tensor] = None, device: str = "cpu"
+        self,
+        prior: Optional[Distribution],
+        x_o: Optional[Tensor] = None,
+        device: str = "cpu",
     ):
         """Initialize potential function.
 
@@ -61,3 +64,22 @@ def return_x_o(self) -> Optional[Tensor]:
         `self._x_o` is `None`.
         """
         return self._x_o
+
+
+class CallablePotentialWrapper(BasePotential):
+    """If `potential_fn` is a callable it gets wrapped as this."""
+
+    allow_iid_x = True
+
+    def __init__(
+        self,
+        callable_potential,
+        prior: Optional[Distribution],
+        x_o: Optional[Tensor] = None,
+        device: str = "cpu",
+    ):
+        super().__init__(prior, x_o, device)
+        self.callable_potential = callable_potential
+
+    def __call__(self, theta, **kwargs):
+        return self.callable_potential(theta=theta, x_o=self.x_o, **kwargs)

tests/potential_test.py

@@ -21,6 +21,7 @@
     [ImportanceSamplingPosterior, MCMCPosterior, RejectionPosterior, VIPosterior],
 )
 def test_callable_potential(sampling_method):
+    """Test whether callable potentials can be used to sample from a Gaussian."""
     dim = 2
     mean = 2.5
     cov = 2.0