[Fix] slicing BinnedSpikeTrain with scipy>=1.15.0 (#653)

elephant/conversion.py

@@ -692,7 +692,17 @@ def __getitem__(self, item):
             the returned binned sparse matrix will affect the original data.
         """
         # taken from csr_matrix.__getitem__
-        row, col = self.sparse_matrix._validate_indices(item)
+        valid_indices = self.sparse_matrix._validate_indices(item)
+        # TODO: The following is a hot fix to compensate an API change in SciPy 1.15 (#653)
+        # Currently, we cannot set scipy>=1.15 since this would not allow for Python 3.9.
+        # Once we phase out support for Python 3.9, remove the if statement and the else 
+        # branch and require scipy>=1.15.
+        if isinstance(valid_indices[0], tuple):
+            # New version of SciPy (1.15 and later)
+            row, col = valid_indices[0]
+        else:
+            # Previous version of SciPy
+            row, col = valid_indices
         spmat = self.sparse_matrix[item]
         if np.isscalar(spmat):
             # data with one element