Fix the commentary portion of the FITS

[mdlpstsci]

The ACS team has requested that certain specified keywords have the commentary portion of the FITS card set back to the original text in an email from S. Hoffman (21 November 2019). While some of the keyword comments are set by astropy, we have control over the remainder of the keyword comments in question. The full context of their issue including Warren's response is as follows.

Thank you for this list of keywords for which you identify that the comments need to be corrected. Several of those keywords, however, have comments generated by astropy.wcs using comments defined by the astropy community; specifically,

CRPIX1, CRPIX2 CRVAL1, CRVAL2 CTYPE1, CTYPE2 WCSAXES (at least, it should...) You can see this for yourself by using the '.to_header()' method on any HSTWCS object. Please also realize that all solutions provided by the astrometry database use the 'new' comments, and it is not clear whether it would be practical to change the headerlet code to only apply the values, something that would allow the pipeline to restore the comments without replacing all the solutions already in the database.

The remainder of your keywords are defined by the STWCS updatewcs task and the headerlet module (for all headerlet-related operations) where we directly control what comments are used for these keywords. It is assumed that you would prefer to have those keywords marked with '(bad)' updated to restore the 'old' comments for those keywords, something we can see about implementing.

We will see about getting these changes in for the single-visit processing, but it is pretty clear that we can not do much better than that at this time.

Thanks again for reporting this issue,

Warren

On 11/21/19 1:34 PM, Samantha Hoffmann wrote: Hi all,

Below is a list of the header keyword comment changes that Tyler found in the ACS regression test. I’ve marked each one as bad, where we prefer the original text, or good, where the changes are fine with us. Hope that helps!

Cheers, Samantha

New header keyword descriptions:

1. NPOLFILE (bad) new = "origin of non-polynmial distortion model" old = "Non-polynomial Offsets Reference File"
2. IDCTAB (bad) new = "origin of Polynomial Distortion" old = "image distortion correction table"
3. D2IMFILE (bad) new = "origin of detector to image correction" old = "Column Correction Reference File"
4. CD1_1 (bad) new = EMPTY old = "partial of first axis coordinate w.r.t. x"
5. CD1_2 (bad) new = EMPTY old = "partial of first axis coordinate w.r.t. y"
6. CD2_1 (bad) new = EMPTY old = "partial of second axis coordinate w.r.t. x"
7. CD2_2 (bad) new = EMPTY old = "partial of second axis coordinate w.r.t. y"
8. CRPIX1 (bad) new = "Pixel coordinate of reference point" old = "x-coordinate of reference pixel"
9. CRPIX2 (bad) new = "Pixel coordinate of reference point" old = "y-coordinate of reference pixel"
10. CRVAL1 (bad) new = "[deg] Coordinate value at reference point" old = "first axis value at reference pixel"
11. CRVAL2 (bad) new = "[deg] Coordinate value at reference point" old = "second axis value at reference pixel"
12. CTYPE1 (bad) new = "TAN (gnomonic) projection + SIP distortions" old = "the coordinate type for the first axis"
13. CTYPE2 (bad) new = "TAN (gnomonic) projection + SIP distortions" old = "the coordinate type for the first axis"
14. D2IM1.EXTVER (good) new = "Version number of WCSDVARR extension" old = "Version number of WCSDVARR extension containing d2im loo"
15. D2IM2.EXTVER (good) new = "Version number of WCSDVARR extension" old = "Version number of WCSDVARR extension containing d2im loo"
16. DP1.EXTVER (good) new = "Version number of WCSDVARR extension" old = "Version number of WCSDVARR extension containing d2im loo"
17. DP2.EXTVER (good) new = "Version number of WCSDVARR extension" old = "Version number of WCSDVARR extension containing d2im loo"
18. WCSAXES (bad) new = EMPTY old = "number of World Coordinate System axes"