You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Differing channel names across images are only a warning (channels pooled by index) #37
Data-flow audit 2026-10-29 (H1/H4), MAJOR. At v0.3.1.
What happens:
metadata.py:450-475 reports channel_names_differ as a warning only (contract §8).
Histograms are matched by channel index.
MorphoFeature (histogram_fit.py:240-246) turns its own channel names into indices and never compares them with the manifest's per-file channel_names.
Failure: image A has OME channel 0 "DAPI", image B has channel 0 "Actin". They are pooled as one channel, and MorphoFeature fits "DAPI" scaling on Actin pixels.
Fix: make differing channel names within one experiment an error. Consumers should check their channel names against the manifest's channel_names when these are present. I'll track the MorphoFeature side with its other findings.
Activity
Keep copies' unnamed channels unnamed; name whitespace-only differences
Trim leading/trailing whitespace of channel names at read time
Refuse OME-Zarr channel labels that are not text
RGB header override: rgb_header_override (Spec amendment RH1, P36)
RGB header override: owner answers of 14 November 2026 (review blocke…