Disclaimer this bug occured along side of several bugs so it might not be reproducible in the absence of other bugs
It seems that the subtract stellar polarization primitive crashes when there are nan's in the region it uses to calculate the stellar/instrumental polarization. I think this is because it the 'meanclip' function from the external directory, which currently is not configured to handle nans.
It seems that the subtract stellar polarization primitive crashes when there are nan's in the region it uses to calculate the stellar/instrumental polarization. I think this is because it the 'meanclip' function from the external directory, which currently is not configured to handle nans.