[AstroPy] Problem running pywcs example
Thu Aug 19 08:23:48 CDT 2010
On 08/18/2010 03:56 PM, Peter Erwin wrote:
> Hi Mike,
> Thanks for the quick reply!
> It looks like what happened is that *was* an earlier version of pywcs hiding
> in /Users/erwin/lib/python [as opposed to
> /Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6 ,
> which is where pywcs 1.9 got installed]. Possibly it got installed as part of
> an earlier installation of stsci_python ...
> Anyway, I nuked the earlier installation, and now pywcs seems to work.
> Am I missing something, or does the documentation for pywcs mention anywhere what
> "sky coordinates" actually means (i.e., allowed formats, types, order, etc.?).
The order is described in the wcs_sky2pix docstring:
Either two or three arguments may be provided.
* 2 arguments: An /N/ x /naxis/ array of /x/- and
/y/-coordinates, and an /origin/.
* 3 arguments: 2 one-dimensional arrays of /x/ and /y/
coordinates, and an /origin/.
As for types, it may be any numeric type (and the underlying
calculations are performed in doubles).
> (From my experiments so far, I'm guessing that the output of WCS.wcs_pix2sky()
> for a single pair of pixel coordinates is [[Dec, RA]], with both in decimal degrees...?)
The fact that it's decimal degrees is somewhat buried (in the
description of the lower-level wcslib functions.) I will move this
information up to the higher-level functions -- and that should make it
into the next version of the docs.
> On Aug 18, 2010, at 9:04 PM, Michael Droettboom wrote:
>> This problem occurs due to a version incompatibility between pyfits and
>> Recent versions of pyfits (2.3 and later) were refactored to no longer
>> support numarray. pywcs required a change to be compatible with this
>> (this change first appeared in pywcs 1.8).
>> Given the versions you list, you should be ok -- however it seems that
>> you are not running the version of pywcs you think you are. Perhaps you
>> need to blitz the installation directory (in your case,
>> /Users/erwin/lib/python/pywcs) and reinstall.
>> On 08/18/2010 02:40 PM, Peter Erwin wrote:
>>> I decided to check out pywcs (in part because using Pyraf to get pixel values from
>>> RA,Dec pairs seemed to be overkill, and wcwcstran was failing to work for SDSS
>>> images with declinations close to zero).
>>> The installation ("python setup.py install") seemd to go fine, but I ran into the following when I tried
>>> running the example from http://stsdas.stsci.edu/astrolib/pywcs/examples.html
>>> <<begin quote>>
>>> In : hdulist = pyfits.open("pgc1905g_94-220.fits") # this is an SDSS g-band image
>>> In : wcs = pywcs.WCS(hdulist.header)
>>> AttributeError Traceback (most recent call last)
>>> /Users/erwin/Documents/Working/ESO_observing/WFI/targets/pgc1905_808/<ipython console> in<module>()
>>> /Users/erwin/lib/python/pywcs/pywcs.py in __init__(self, header, fobj, key, minerr, relax, naxis)
>>> 176 self.naxis = wcsprm.naxis
>>> --> 178 det2im = self._read_det2im_kw(header, fobj)
>>> 179 cpdis = self._read_distortion_kw(
>>> 180 header, fobj, key=key,dist='CPDIS', err=minerr)
>>> /Users/erwin/lib/python/pywcs/pywcs.py in _read_det2im_kw(self, header, fobj)
>>> 278 cdelt = [1.,1.]
>>> --> 280 if not isinstance(fobj, pyfits.NP_pyfits.HDUList):
>>> 281 return (None, None)
>>> AttributeError: 'module' object has no attribute 'NP_pyfits'
>>> <<end quote>>
>>> My system is: MacBook Pro (2009) running Mac OS 10.6.4, Python 2.6.5 (standard framework install
>>> from python.org) via ipython 0.10
>>> numpy 1.4.1
>>> pyfits 2.3.1 [same problem occurred for pyfits 2.2.2; I
>>> pywcs v1.9-4.4.4
>>> (All packages are installed inside the framework directory, *except* for pyfits, which installed in
>>> ~/lib/python )
>>> Any suggestions?
>>> -- Peter
>>> Peter Erwin Max-Planck-Insitute for Extraterrestrial
>>> firstname.lastname@example.org Physics, Giessenbachstrasse
>>> tel. +49 (0)89 30000 3695 85748 Garching, Germany
>>> fax +49 (0)89 30000 3495 http://www.mpe.mpg.de/~erwin
>>> AstroPy mailing list
>> Michael Droettboom
>> Science Software Branch
>> Space Telescope Science Institute
>> Baltimore, Maryland, USA
>> AstroPy mailing list
> Peter Erwin Max-Planck-Insitute for Extraterrestrial
> email@example.com Physics, Giessenbachstrasse
> tel. +49 (0)89 30000 3695 85748 Garching, Germany
> fax +49 (0)89 30000 3495 http://www.mpe.mpg.de/~erwin
> AstroPy mailing list
Science Software Branch
Space Telescope Science Institute
Baltimore, Maryland, USA
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the AstroPy