Closed teozec closed 1 year ago
Hi @teozec , oops! I removed check_overflow
this morning because I missed this PR from yours… For me it's fine to get rid of check_overflow
completely, keeping it active would just confuse users.
Ok perfect, thanks! Should I also, while we are touching this function, remove the
if not detector:
detector = ["Q1", "Q2", "U1", "U2"]
check and simply use a default parameter detector = ["Q1", "Q2", "U1", "U2"]
? I think it would break compatibility only if someone, instead of not providing the parameter, called the function with detector=[]
, None
, False
etc.
Ok perfect, thanks! Should I also, while we are touching this function, remove the
if not detector: detector = ["Q1", "Q2", "U1", "U2"]
check and simply use a default parameter
detector = ["Q1", "Q2", "U1", "U2"]
? I think it would break compatibility only if someone, instead of not providing the parameter, called the function withdetector=[]
,None
,False
etc.
I checked the code and it does not seem that detector=None
is used anywhere, so ok!
Done! Let me know if I should change anything else. Thanks.
Hi. This PR aims at fixing the problem in the
DataFile.load_sci()
function, which failed at converting unsigned data to signed. This seems to be due to the h5py library, therefore I moved the conversion after the column selection, which returns a numpy array. This allows us to use the.astype()
method from numpy instead of the one from h5py. At the same time, I removed the now useless check for overflow (a quick search in the striptease library shows that it is not used anywhere in our programs).Let me know if you prefer to leave a dummy
check_overflow
parameter to preserve compatibility with any possible program that is not in striptease and I am not aware of.Thanks!