Fix Version/s: None
From Alex Broughton on slack:
I ran into an error where the expIdMask is not the same length as the bfk.rawXcorrs here: https://github.com/lsst/cp_pipe/blob/main/python/lsst/cp/pipe/makeBrighterFatterKernel.py#L239
With the PTC dataset I’m using, bfk.rawMeans and bfk.rawVariances are all the same length as the mask.
The mask appears to be print(len(inputPtc.expIdMask['C00'])) = 338, but len(inputPtc.covariances['C00']) = 326
bfk.rawXcorrs is set equal to inputPtc.covariances, which has shape (326, 8, 8 )
It seems like the expIdMask has already been applied to generate the rawXcorrs because the expIdMask has a total of 12 ‘False’ entries.
FYI I am using w_2022_32
There was a similar issue previously where the means and mask disagreed,
DM-31809, and I wonder if this has the same cause.
Looks good, thanks for the fix. I only wonder if we need to change the "raw" prefix in the name, as the PTC code has "raw" when a mask has not been applied.