Closed dstansby closed 2 months ago
Hello @dstansby, thanks for trying out the code! We have been unsure what to do about the uncertainty type. The uncertainties given in the EIS observations are derived from the CCD read noise of EIS, which are technically neither the standard deviation nor the variance of the data (the only two types specified in astropy, last time I checked). In this case, would you recommend we create our own subclass of NDUncertanity
or just use the stddev (since it at least has the correct units)?
When I load some data, astropy prints a message that says
I'm guessing this is because when uncertainties are loaded the
uncertainty_type
(absolute or relative from memory) isn't specified. Would be good to specify this to remove this message!