astropy.wcs issue with 2MASS image headers

[David Rosario]

Hi everyone,


  I’m trying to use the WCS module within Astropy to do some simple object location work on Extended Source image cubes from the 2MASS survey, downloaded from IRSA. I keep running into an error that I can’t seem to understand or overcome, which seems to relate to the interaction with the external wcslib C library. FYI, I’m using Astropy 1.1.2 on Python 3.4.4 through iPython 3.2.0.

I’ve attached an example image (2.2 MB) for your reference, which I use for the code snippets below. The data in the primary extension of the fits file is 3-dimensional (a cube), which is a stack of images in various bands. This is the way that these Extended Source cubes are distributed by IRSA.

  I can use Astropy io.fits utilities to open the fits file and read in the header.

---------
from astropy.io import fits
hdulist = fits.open('2MASXJ22091625-4709599.fits’)
header = hdulist[0].header  # astropy header instance
---------

 Then I can use the astropy.wcs.WCS constructor to extract the WCS from this header

---------
from astropy import was
astr = wcs.WCS(header)

At this point, the code fails with the following traceback:

WARNING: FITSFixedWarning: 'celfix' made the change 'Unrecognized projection code (-SI in CTYPE2)'. [astropy.wcs.wcs]
---------------------------------------------------------------------------
InconsistentAxisTypesError                Traceback (most recent call last)
<ipython-input-22-cc051d3537d1> in <module>()
----> 1 astr = wcs.WCS(header)

/Users/davidrosario/anaconda/lib/python3.4/site-packages/astropy/wcs/wcs.py in __init__(self, header, fobj, key, minerr, relax, naxis, keysel, colsel, fix, translate_units, _do_set)
   498
   499         if _do_set:
--> 500             self.wcs.set()
   501
   502         for fd in close_fds:

InconsistentAxisTypesError: ERROR 4 in wcs_types() at line 2432 of file cextern/wcslib/C/wcs.c:
Unrecognized projection code (-SI in CTYPE2).


  I can’t seem to find anything particularly strange in the header - it’s actually quite simple, with no distortion corrections needed. The value of the keyword CTYPE3 is possibly non-standard. However, even if I delete that keyword from the FITS header, I get the same problem. Given that the 2MASS images are widely used in the community, I was puzzled by this issue. Could you please let me know if I’ve done something incorrectly or suggest a fix?

  Thanks very much for any help you could send me.

   -- David Rosario

[Adam Ginsburg]

Hi David,

    This one's straightforward, though perhaps surprising - did you get the data direct from IRSA?   The CTYPE keywords are required to have 8 characters, and yours looks like this:

CTYPE1  = 'RA---SIN'           / Orthographic Projection

CTYPE2  = 'DEC---SIN'          / Orthographic Projection

it should be DEC--SIN.  If you got this off of their archive, you should report it as an issue.

--
You received this message because you are subscribed to the Google Groups "astropy-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to astropy-dev...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Adam Ginsburg
Fellow, European Southern Observatory
http://www.adamgginsburg.com/

[David Rosario]

Thanks Adam! Repairing the header  seems to have done the trick.

Yes, these images were downloaded directly from IRSA, so it's a bit surprising about the malformed header. I'll get in touch with the folks there.

   Cheers and best wishes,

      -- David