add errors for MAP and iid data, adapt tests

sbi/inference/posteriors/score_posterior.py

@@ -331,6 +331,9 @@ def map(
         Returns:
             The MAP estimate.
         """
+        raise NotImplementedError(
+            "MAP estimation is currently not working accurately for ScorePosterior."
+        )
         return super().map(
             x=x,
             num_iter=num_iter,

sbi/inference/potentials/score_based_potential.py

@@ -160,6 +160,9 @@ def gradient(
                     input=theta, condition=self.x_o, time=time
                 )
             else:
+                raise NotImplementedError(
+                    "Score accumulation for IID data is not yet fully implemented."
+                )
                 if self.prior is None:
                     raise ValueError(
                         "Prior must be provided when interpreting the data as IID."

tests/linearGaussian_npse_test.py

@@ -1,3 +1,5 @@
+from typing import List
+
 import pytest
 import torch
 from torch import eye, ones, zeros
@@ -30,7 +32,7 @@
     ],
 )
 def test_c2st_npse_on_linearGaussian(
-    sde_type, num_dim: int, prior_str: str, sample_with: list[str]
+    sde_type, num_dim: int, prior_str: str, sample_with: List[str]
 ):
     """Test whether NPSE infers well a simple example with available ground truth."""
 
@@ -78,7 +80,7 @@ def test_c2st_npse_on_linearGaussian(
         check_c2st(
             samples,
             target_samples,
-            alg=f"npse-{sde_type or "vp"}-{prior_str}-{num_dim}D-{method}",
+            alg=f"npse-{sde_type or 'vp'}-{prior_str}-{num_dim}D-{method}",
         )
 
     # Checks for log_prob()
@@ -157,7 +159,12 @@ def simulator(theta):
     check_c2st(samples, target_samples, alg="npse_different_dims_and_resume_training")
 
 
-@pytest.mark.xfail(reason="iid_bridge not working.")
+@pytest.mark.xfail(
+    reason="iid_bridge not working.",
+    raises=NotImplementedError,
+    strict=True,
+    match="Score accumulation*",
+)
 @pytest.mark.parametrize("num_trials", [2, 10])
 def test_npse_iid_inference(num_trials):
     """Test whether NPSE infers well a simple example with available ground truth."""