# astropy.wcs History¶

astropy.wcs began life as pywcs. Earlier version numbers refer to that package.

## pywcs Version 1.11¶

• Updated to wcslib version 4.8, which gives much more detailed error messages.
• Added functions get_pc() and get_cdelt(). These provide a way to always get the canonical representation of the linear transformation matrix, whether the header specified it in PC, CD or CROTA form.
• Long-running process will now release the Python GIL to better support Python multithreading.
• The dimensions of the cd and pc matrices were always returned as 2x2. They now are sized according to naxis.
• Supports Python 3.x
• Builds on Microsoft Windows without severely patching wcslib.
• Lots of new unit tests
• pywcs will now run without pyfits, though the SIP and distortion lookup table functionality is unavailable.
• Setting cunit will now verify that the values are valid unit strings.

## pywcs Version 1.10¶

• Adds a UnitConversion class, which gives access to wcslib’s unit conversion functionality. Given two convertible unit strings, pywcs can convert arrays of values from one to the other.
• Now uses wcslib 4.7
• Changes to some wcs values would not always calculate secondary values.

## pywcs Version 1.9¶

• Support binary image arrays and pixel list format WCS by presenting a way to call wcslib’s wcsbth()

• Updated underlying wcslib to version 4.5, which fixes the following:

• Fixed the interpretation of VELREF when translating AIPS-convention spectral types. Such translation is now handled by a new special- purpose function, spcaips(). The wcsprm struct has been augmented with an entry for velref which is filled by wcspih() and wcsbth(). Previously, selection by VELREF of the radio or optical velocity convention for type VELO was not properly handled.

## pywcs Version 1.3a1¶

Earlier versions of pywcs had two versions of every conversion method:

X(...)      -- treats the origin of pixel coordinates at (0, 0)
X_fits(...) -- treats the origin of pixel coordinates at (1, 1)


From version 1.3 onwards, there is only one method for each conversion, with an ‘origin’ argument:

• 0: places the origin at (0, 0), which is the C/Numpy convention.
• 1: places the origin at (1, 1), which is the Fortran/FITS convention.