Differences between funtools and iraf regions
Describes the differences between Funtools/ds9 regions and the old \s-1IRAF/PROS\s0 regions.
We have tried to make Funtools regions compatible with their predecessor, \s-1IRAF/PROS\s0 regions. For simple regions and simple boolean algebra between regions, there should be no difference between the two implementations. The following is a list of differences and incompatibilities between the two:
If a pixel is covered by two different regions expressions, Funtools assigns the mask value of the first region that contains that pixel. That is, successive regions do not overwrite previous regions in the mask, as was the case with the original \s-1PROS\s0 regions. This means that one must define overlapping regions in the reverse order in which they were defined in \s-1PROS\s0. If region N is fully contained within region M, then N should be defined before M, or else it will be \*(L"covered up\*(R" by the latter. This change is necessitated by the use of optimized filter compilation, i.e., Funtools only tests individual regions until a proper match is made.
The \s-1PANDA\s0 region has replaced the old \s-1PROS\s0 syntax in which a \s-1PIE\s0 accelerator was combined with an \s-1ANNULUS\s0 accelerator using \s-1AND\s0. That is,
ANNULUS(20,20,0,15,n=4) & PIE(20,20,0,360,n=3)
has been replaced by: PANDA(20,20,0,360,3,0,15,4) The \s-1PROS\s0 syntax was inconsistent with the meaning of the \s-1AND\s0 operator.
The meaning of pure numbers (i.e., without format specifiers) in regions has been clarified, as has the syntax for specifying coordinate systems. See the general discussion on Spatial Region Filtering for more information.
See funtools(7) for a list of Funtools help pages