[ITK] Definition of Composite Transform

[Matt McCormick]

Hi Floris,

Thanks for helping to improve our documentation.  Yes, it does appear
that the numeric order in the comments is inconsistent.  Improvements
are welcome :-).

Thanks,
Matt

> I think there is some inconsistency between comments in the code of the
> composite transform and its (doxygen) documentation.
> The documentation states that a transformation is composed in the following
> way:
> T_0 o T_1 = T_0(T_1(x)), with T_1 the transform added latest.
> https://github.com/InsightSoftwareConsortium/ITK/blob/master/Modules/Core/Transform/include/itkCompositeTransform.h
>
> However, some (very insightful) comments in the code state that:
> T is composed of T1(p1,x), T2(p2,x) and T3(p3, x) as: T3(p3, T2(p2, T1(p1,
> x0)))
> https://github.com/InsightSoftwareConsortium/ITK/blob/master/Modules/Core/Transform/include/itkCompositeTransform.hxx#L620-L659
>
> The confusing part is that the ordering is reversed (and starting from 1
> instead of 0). This may also explain the next comment:
> // to do: why parameters are listed from N-1 to 1???
> https://github.com/InsightSoftwareConsortium/ITK/blob/master/Modules/Core/Transform/include/itkCompositeTransform.hxx#L674
>
> The first definition is consistent with the code as far as I can see. Can
> anyone confirm this finding? If so, I'm happy to provide a patch via gerrit
> to solve the comments.
>
> Best,
> Floris
>
>
>
>
> _______________________________________________
> Community mailing list
> Community at itk.org
> http://public.kitware.com/mailman/listinfo/community
>