NIRSpec IFU flatfield step doesn't set outside-slice values properly
stscijgbot-jp opened this issue · comments
Issue JP-3587 was created on JIRA by David Law:
At the moment the NIRSpec IFU flatfield code doesn't set values outside the slice footprint correctly. I.e., it should use SCI=ERR=NaN and DQ=DO_NOT_USE, but at present it doesn't when constructing the flatfield. (In brief, this is because an empty array is instantiated for all of these, and then populated slice by slice using each of the different flatfield components. As a result values can differ from those within slices by a factor of 1e11.
At present this isn't taken care of until the end of the pathloss step, but it should be handled properly within the flatfield step itself.
Addressed by #8385
Note that this is distinct from the related issue that the DQ bits aren't being set properly within the IFU footprint either; see https://jira.stsci.edu/browse/JP-3507
Comment by Howard Bushouse on JIRA:
Fixed by #8385