Closed jeffersoncasimir closed 6 months ago
@jeffersoncasimir who would best the best person to review this PR?
@cmadjar I think all the developers who were given a test-25 clone VM (https://redmine.cbrain.mcgill.ca/issues/20684) should be able to test this with relatively little set-up time.
The list, excluding myself: @laemtl, @nicolasbrossard, @CamilleBeau, @regisoc
@laemtl @nicolasbrossard @CamilleBeau @regisoc could one of you review this PR? Thank you :-)
This PR resolves the data shift issue mentioned in aces/Loris#8892 by adding the number of valid samples in the last chunk to the index.json.
It has a Loris PR counterpart: aces/Loris#8999 to use this value and address the fact that some calculations were being made by multiplying the number of chunks by the number of values per chunk, assuming they were all full.