Handling of STIS tag files

[jcv109]

Dear Justin/Sean,
I send to you this message regarding the processing of STIS tag files with your package (which runs fine with COS files). The package seems to properly identify stis filetypes (as _tag file). Hence, the routine io.py calls "extract_stis" (from stis.py). This routine calls in turn "extract_index" (from cos.py). However, this cos-routine attempts to filter the hdu data by coordinates (XCORR). Unfortunately, the STIS tag files do not contain the XCORR keyword. Hence, the program raises an exception about keyword not found. I wonder if a fix or hint or workaround could be provided. Could XCORR be replaced by AXIS1 when processing STIS files? Thanks in advance again for your advice, that is really appreciated.

[justincely]

hey @jcv109 interestingly, it does actually go for AXIS1 but i think the if statement then may be the wrong default. Sorry, this has been quite some time since I touched this you're inputting a corrtag or a rawtag?

[jcv109]

I am trying to process a "_tag" STIS file (time tag event list) as input file, and the current version seems to apply the same processing ("extract_stis") both to "_tag" files and "_corrtag" files. Should a "_tag"-only processing be in place, reading AXIS1 instead of XCORR? Or do you mean that it was expected to manually use another routine, "stis_corrtag"?, for converting the STIS "_tag" files into "_corrtag" files before attempting to extract the lightcurve?
(by the way, the "extract_stis" routine seems to be designed for warning the user about the computed flux and background results, whatever they came from "_tag" or "_corrtag" files?)
Thanks again for your time and support!

[sean-lockwood]

@justincely -
I believe not all STIS tag files contain a DETAXIS1 column -- depending on if it's spectroscopic or imaging.

[jcv109]

Dear Justin/Sean,
Just to complete the reporting after coming back to the STIS processing after a while. One can get the lightcurves corresponding to the STIS _tag.files when one firstly calls to "stis_corrtag", and creates the COS-like corrtag files from the _tag files, hence creates the XCORR keyword using AXIS1.
For completeness, a few modifications seem to be needed for running stis.py in the latest (?) astropy: the "fits.new_table" seems to be deprecated and must be replaced by "fits.BinTableHDU.from_columns" and the "clobber" option should be replaced by "overwrite". Once these changes are in place, lightcurves are nicely produced.
A (hopefully) final question.
These lightcurves are created without calibrated units, because FLUXTAB keyword is searched but not found in STIS files. Should this keyword be present in STIS files? It seems crds does not create it, and it is not present in original _tag files (?).
As usual, thank you very much in advance for your support, which is much appreciated.