[AstroPy] Bug in wcs_world2pix?

[Michael Droettboom]

I've brought this to Mark Calabretta's attention -- he may have an idea.

Mike

On 09/19/2014 08:14 AM, David Berry wrote:
> Am I missing something? This is a TAN projection centred on CRVAL1=16,
> CRVAL2=23. Since it is a TAN projection, the projection plane only
> includes half the sky - i.e. the hemisphere centred on (16,23). But
> the test point (211,-26) is not contained within this hemisphere and
> so is not included in the TAN projection plane. In other words, no
> pixel has the coordinates RA=211 Dec=-26, and so the results of  using
> wcs.wcs_world2pix should be NaN, or "invalid", or something.
>
> Unless I'm missing something...
>
> David
>
>
> On 19 September 2014 12:58, Bob Garwood <bgarwood at nrao.edu> wrote:
>> I suspect what's going on here has to do with the specific CD matrix, which
>> includes rotation. I think because of that rotation the world coordinate
>> actually maps back in to the grid at the antipode. I haven't checked this by
>> hand to see if that's the case.
>>
>> Bob
>>
>> On September 19, 2014 7:47:30 AM EDT, Howard Bushouse <bushouse at stsci.edu>
>> wrote:
>>> Huh. But that’s not how astronomers measure RA and Dec, even when
>>> translated to degrees. If I’ve got an object with an RA of 17:45:00.0
>>> (hh:mm:ss.w) the RA in degrees is 266.25. We *never* use a negative RA.
>>> Instead of a range of -180 to 180, we use 0 to 360. Dec is the only one that
>>> ever goes negative (-90 to +90, as you indicated).
>>>
>>> -hb
>>>
>>> On Sep 18, 2014, at 8:26 PM, Michael Droettboom <mdroe at stsci.edu> wrote:
>>>
>>>>   The range of ra and dec is (-180, 180) and (-90, 90) respectively.
>>>>   (211, -26) is equivalent to (31, 26).
>>>>
>>>>   Mike
>>>>
>>>>   On 09/18/2014 05:04 PM, Maik Riechert wrote:
>>>>>   Hi,
>>>>>
>>>>>   I'm having some great trouble currently and am a bit confused to!
>>>>>   o:
>>>>>
>>>>>
>>>>>   wcs = WCS(header)
>>>>>   x,y = wcs.wcs_world2pix(211, -26, 0)
>>>>>   print x,y # 836.316942718 658.26364248
>>>>>   ra,dec = wcs.wcs_pix2world(x, y, 0)
>>>>>   print ra, dec # 31.0 26.0
>>>>>
>>>>>   It seems like world2pix gives me bogus results. I was expecting values
>>>>>   way outside my image area (image size is 4256x2832). Am I doing
>>>>>   something obvious wrong?
>>>>>
>>>>>   The WCS headers can be seen at http://pastebin.com/JrdiL949
>>>>>
>>>>>   Thanks for the help,
>>>>>   Maik
>>>>> ________________________________
>>>>>
>>>>>   AstroPy mailing list
>>>>>   AstroPy at scipy.org
>>>>>   http://mail.scipy.org/mailman/listinfo/astropy
>>>>
>>>>
>>>>   --
>>>>   Michael Droettboom
>>>>   Science Software Branch
>>>>   Space Telescope Science Institute
>>>>
>>>>   http://www.droettboom.com
>>>>
>>>> ________________________________
>>>>
>>>>   AstroPy mailing list
>>>>   AstroPy at scipy.or!
>>>>   g
>>>>
>>>> http://mail.scipy.org/mailman/listinfo/astropy
>>>
>>> ________________________________
>>>
>>> AstroPy mailing list
>>> AstroPy at scipy.org
>>> http://mail.scipy.org/mailman/listinfo/astropy
>>
>> _______________________________________________
>> AstroPy mailing list
>> AstroPy at scipy.org
>> http://mail.scipy.org/mailman/listinfo/astropy
>>
> _______________________________________________
> AstroPy mailing list
> AstroPy at scipy.org
> http://mail.scipy.org/mailman/listinfo/astropy


-- 
Michael Droettboom
Science Software Branch
Space Telescope Science Institute

http://www.droettboom.com