[AstroPy] Problem running pywcs example

[Phil Hodge]

Peter,

> Thanks for the "sub" suggestion... however, that seems to have the side effect of reversing
> the x,y output order from new_wcs.wcs_sky2pix when the FITS header has CTYPE1="DEC--TAN"
>
> i.e., the output from new_wsc.wcs_sky2pix(ra, dec) is [ array([y]), array([x]) ] when 
> CTYPE1="DEC--TAN", instead of [ array([x]), array([y]) ]


That sounds like a bug.  Mike and I talked about this (the order of the 
results depending on CTYPEi) yesterday.  I can understand that when you 
have other coordinate types (e.g. wavelength) together with celestial 
coordinates, the least confusing result is that the order of the output 
world coordinates be the same as CTYPEi.  If you extract just the 
celestial coordinates and specify the order with the sub() method, 
however, you really shouldn't have to care about the order of CTYPEi in 
the header; the celestial coordinates in the world coordinate vector 
should agree with what you specified in sub(), and the pixel coordinates 
should always be in the order [x, y].  (In my humble opinion.)

Phil