Merge branch 'main' into remove_reff_resolver

CHANGELOG.md

@@ -8,6 +8,7 @@
 * datetime objects now can be validated as conversion options [#1139](https://github.com/catalystneuro/neuroconv/pull/1126)
 
 ## Features
+* Propagate the `unit_electrode_indices` argument from the spikeinterface tools to `BaseSortingExtractorInterface`. This allows users to map units to the electrode table when adding sorting data [PR #1124](https://github.com/catalystneuro/neuroconv/pull/1124)
 * Imaging interfaces have a new conversion option `always_write_timestamps` that can be used to force writing timestamps even if neuroconv's heuristics indicates regular sampling rate [PR #1125](https://github.com/catalystneuro/neuroconv/pull/1125)
 * Added .csv support to DeepLabCutInterface [PR #1140](https://github.com/catalystneuro/neuroconv/pull/1140)
 

src/neuroconv/datainterfaces/ecephys/basesortingextractorinterface.py

@@ -288,6 +288,7 @@ def add_to_nwbfile(
         write_as: Literal["units", "processing"] = "units",
         units_name: str = "units",
         units_description: str = "Autogenerated by neuroconv.",
+        unit_electrode_indices: Optional[list[list[int]]] = None,
     ):
         """
         Primary function for converting the data in a SortingExtractor to NWB format.
@@ -312,9 +313,15 @@ def add_to_nwbfile(
         units_name : str, default: 'units'
             The name of the units table. If write_as=='units', then units_name must also be 'units'.
         units_description : str, default: 'Autogenerated by neuroconv.'
+        unit_electrode_indices : list of lists of int, optional
+            A list of lists of integers indicating the indices of the electrodes that each unit is associated with.
+            The length of the list must match the number of units in the sorting extractor.
         """
         from ...tools.spikeinterface import add_sorting_to_nwbfile
 
+        if metadata is None:
+            metadata = self.get_metadata()
+
         metadata_copy = deepcopy(metadata)
         if write_ecephys_metadata:
             self.add_channel_metadata_to_nwb(nwbfile=nwbfile, metadata=metadata_copy)
@@ -346,4 +353,5 @@ def add_to_nwbfile(
             write_as=write_as,
             units_name=units_name,
             units_description=units_description,
+            unit_electrode_indices=unit_electrode_indices,
         )

src/neuroconv/tools/spikeinterface/spikeinterface.py

@@ -1368,7 +1368,7 @@ def add_units_table_to_nwbfile(
     write_in_processing_module: bool = False,
     waveform_means: Optional[np.ndarray] = None,
     waveform_sds: Optional[np.ndarray] = None,
-    unit_electrode_indices=None,
+    unit_electrode_indices: Optional[list[list[int]]] = None,
     null_values_for_properties: Optional[dict] = None,
 ):
     """
@@ -1405,15 +1405,23 @@ def add_units_table_to_nwbfile(
         Waveform standard deviation for each unit. Shape: (num_units, num_samples, num_channels).
     unit_electrode_indices : list of lists of int, optional
         For each unit, a list of electrode indices corresponding to waveform data.
-    null_values_for_properties: dict, optional
-        A dictionary mapping properties to null values to use when the property is not present
+    unit_electrode_indices : list of lists of int, optional
+        A list of lists of integers indicating the indices of the electrodes that each unit is associated with.
+        The length of the list must match the number of units in the sorting extractor.
     """
     unit_table_description = unit_table_description or "Autogenerated by neuroconv."
 
     assert isinstance(
         nwbfile, pynwb.NWBFile
     ), f"'nwbfile' should be of type pynwb.NWBFile but is of type {type(nwbfile)}"
 
+    if unit_electrode_indices is not None:
+        electrodes_table = nwbfile.electrodes
+        if electrodes_table is None:
+            raise ValueError(
+                "Electrodes table is required to map units to electrodes. Add an electrode table to the NWBFile first."
+            )
+
     null_values_for_properties = dict() if null_values_for_properties is None else null_values_for_properties
 
     if not write_in_processing_module and units_table_name != "units":
@@ -1668,7 +1676,7 @@ def add_sorting_to_nwbfile(
     units_description: str = "Autogenerated by neuroconv.",
     waveform_means: Optional[np.ndarray] = None,
     waveform_sds: Optional[np.ndarray] = None,
-    unit_electrode_indices=None,
+    unit_electrode_indices: Optional[list[list[int]]] = None,
 ):
     """Add sorting data (units and their properties) to an NWBFile.
 
@@ -1703,7 +1711,8 @@ def add_sorting_to_nwbfile(
     waveform_sds : np.ndarray, optional
         Waveform standard deviation for each unit. Shape: (num_units, num_samples, num_channels).
     unit_electrode_indices : list of lists of int, optional
-        For each unit, a list of electrode indices corresponding to waveform data.
+        A list of lists of integers indicating the indices of the electrodes that each unit is associated with.
+        The length of the list must match the number of units in the sorting extractor.
     """
 
     if skip_features is not None:

tests/test_ecephys/test_ecephys_interfaces.py

@@ -54,6 +54,41 @@ def test_propagate_conversion_options(self, setup_interface):
         assert nwbfile.units is None
         assert "processed_units" in ecephys.data_interfaces
 
+    def test_electrode_indices(self, setup_interface):
+
+        recording_interface = MockRecordingInterface(num_channels=4, durations=[0.100])
+        recording_extractor = recording_interface.recording_extractor
+        recording_extractor = recording_extractor.rename_channels(new_channel_ids=["a", "b", "c", "d"])
+        recording_extractor.set_property(key="property", values=["A", "B", "C", "D"])
+        recording_interface.recording_extractor = recording_extractor
+
+        nwbfile = recording_interface.create_nwbfile()
+
+        unit_electrode_indices = [[3], [0, 1], [1], [2]]
+        expected_properties_matching = [["D"], ["A", "B"], ["B"], ["C"]]
+        self.interface.add_to_nwbfile(nwbfile=nwbfile, unit_electrode_indices=unit_electrode_indices)
+
+        unit_table = nwbfile.units
+
+        for unit_row, electrode_indices, property in zip(
+            unit_table.to_dataframe().itertuples(index=False),
+            unit_electrode_indices,
+            expected_properties_matching,
+        ):
+            electrode_table_region = unit_row.electrodes
+            electrode_table_region_indices = electrode_table_region.index.to_list()
+            assert electrode_table_region_indices == electrode_indices
+
+            electrode_table_region_properties = electrode_table_region["property"].to_list()
+            assert electrode_table_region_properties == property
+
+    def test_electrode_indices_assertion_error_when_missing_table(self, setup_interface):
+        with pytest.raises(
+            ValueError,
+            match="Electrodes table is required to map units to electrodes. Add an electrode table to the NWBFile first.",
+        ):
+            self.interface.create_nwbfile(unit_electrode_indices=[[0], [1], [2], [3]])
+
 
 class TestRecordingInterface(RecordingExtractorInterfaceTestMixin):
     data_interface_cls = MockRecordingInterface