Group integrity (LorenFrankLab#1026)

* restrict populate calls to sort interval

* prevent adding new pieces to existing groups in create functions

* update changelog

* fix lint

* fix lint version

CHANGELOG.md

@@ -33,6 +33,7 @@ PositionGroup.alter()
     `processes > 1` #1001
 - Speed up fetch_nwb calls through merge tables #1017
 - Allow `ModuleNotFoundError` or `ImportError` for optional dependencies #1023
+- Ensure integrity of group tables #1026
 
 ### Pipelines
 

src/spyglass/decoding/v1/core.py

@@ -115,6 +115,11 @@ def create_group(
             "nwb_file_name": nwb_file_name,
             "position_group_name": group_name,
         }
+        if self & group_key:
+            raise ValueError(
+                f"Group {nwb_file_name}: {position_group_name} already exists",
+                "please delete the group before creating a new one",
+            )
         self.insert1(
             {
                 **group_key,

src/spyglass/spikesorting/analysis/v1/group.py

@@ -69,6 +69,12 @@ def create_group(
             "nwb_file_name": nwb_file_name,
             "unit_filter_params_name": unit_filter_params_name,
         }
+        if self & group_key:
+            raise ValueError(
+                f"Group {nwb_file_name}: {group_name} already exists",
+                "please delete the group before creating a new one",
+            )
+
         parts_insert = [{**key, **group_key} for key in keys]
 
         self.insert1(

src/spyglass/spikesorting/v0/spikesorting_populator.py

@@ -194,16 +194,16 @@ def spikesorting_pipeline_populator(
         )
         SpikeSortingRecordingSelection.insert1(ssr_key, skip_duplicates=True)
 
-    SpikeSortingRecording.populate(interval_dict)
+    SpikeSortingRecording.populate(sort_dict)
 
     # Artifact detection
     logger.info("Running artifact detection")
     artifact_keys = [
         {**k, "artifact_params_name": artifact_parameters}
-        for k in (SpikeSortingRecordingSelection() & interval_dict).fetch("KEY")
+        for k in (SpikeSortingRecordingSelection() & sort_dict).fetch("KEY")
     ]
     ArtifactDetectionSelection().insert(artifact_keys, skip_duplicates=True)
-    ArtifactDetection.populate(interval_dict)
+    ArtifactDetection.populate(sort_dict)
 
     # Spike sorting
     logger.info("Running spike sorting")

src/spyglass/spikesorting/v1/recording.py

@@ -73,7 +73,7 @@ def set_group_by_shank(
             Optional. If True, no sort groups are defined for unitrodes.
         """
         # delete any current groups
-        # (SortGroup & {"nwb_file_name": nwb_file_name}).delete()
+        (SortGroup & {"nwb_file_name": nwb_file_name}).delete()
         # get the electrodes from this NWB file
         electrodes = (
             Electrode()