Add ignore_timestamps_errors to OpenEphys interfaces
#579
Conversation
|
when |
| if ignore_timestamps_errors and self.recording_extractor.has_time_vector(): | ||
| for segment_index in range(self.recording_extractor.get_num_segments()): | ||
| self.recording_extractor._recording_segments[segment_index].time_vector = None | ||
| warn( | ||
| "The recording has a time vector but the timestamps might be discontinuous. " | ||
| "The time vector was reset to None to avoid writing these timestamps with the electrical series." | ||
| ) | ||
|
|
There was a problem hiding this comment.
@CodyCBakerPhD how about this?
There was a problem hiding this comment.
Intuitively seems perfectly fine - only thing to check is if the logic on the SI side relies on value of the .time_vector is None rather than on the logic hasattr(recording_segment, "time_vector")
There was a problem hiding this comment.
No matter what we have time to polish this; this shouldn't be merged to main until both neo and SI get new releases with the upstream fixes
Tye lab repo will have to pin to either this branch (to stay up-to-date with other things updated from main) or a commit from this branch
This reverts commit b474499.
Codecov Report
@@ Coverage Diff @@
## main #579 +/- ##
==========================================
- Coverage 90.78% 90.74% -0.05%
==========================================
Files 103 103
Lines 5374 5378 +4
==========================================
+ Hits 4879 4880 +1
- Misses 495 498 +3
Flags with carried forward coverage won't be shown. Click here to find out more.
|
|
@weiglszonja Can you remind me; do we still need or even want this exposure, or is the plan still to try to fix |
|
@CodyCBakerPhD I think we shouldn't expose it for now, can be closed. The plan is to fix the underlying files |
To fix catalystneuro/tye-lab-to-nwb#29
ignore_timestamps_errorstoOpenEphysLegacyRecordingExtractorAddignore_timestamps_errorstoOpenEphysLegacyRecordingExtractorSpikeInterface/spikeinterface#2043