Conversation
When a recording has no CPR activity, CprDepth and CprFrequency columns are all zeros. After filtering out zeros, timestamps1 is empty, causing IndexError on timestamps1[-1] and subsequent KeyError on CompDisp/CompRate accesses that assumed CPR data was always present. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Channels that are present in the recording but carry only zero values (e.g. Ecg2/3/4 leads connected but not recording, CprWaveVal without CPR activity) are now stored as empty DataFrames instead of being skipped or loaded with large arrays of meaningless zeros. This preserves channel presence for downstream detection while avoiding data bloat.
|
This reviewer fix adds a synthetic regression test for ZOLL CSV imports with |
|
Also adds a regression test. If you really need the empty channels, we can change this fix in a follow-up PR; for now I'd like to get this merged. Other PRs in this chain to follow. |
When a recording has no CPR activity, CprDepth and CprFrequency columns are all zeros. After filtering out zeros, timestamps1 is empty, causing IndexError on timestamps1[-1] and subsequent KeyError on CompDisp/CompRate accesses that assumed CPR data was always present.
#276
Additionally omits import of vast zero arrays.
Example File for ZOLL R Series containing ECG an PPG:
20260331124955_00147042.zip