As a follow-up to Automatic naming of tensor components, I’d like to suggest to remove the automatic naming of components when a vector of length 6 is found.
The reasons for that are
The assumption that a vector of length 6 represents a symmetric 3x3 tensor is probably often, but certainly not always, right.
Different conventions for ordering the shear components are used, a common notation is Voigt notation [11,22,33,23,13,12], which is not what ParaView uses [11,22,33,12,23,13].
So, in agreement with
Mathieu Westphal (Kitware) mwestphal
I agree, and I mean to say that maybe the automatic naming of symmetric tensor is already confusing people. I added the naming some years ago thinking it was considered standard.
I believe that the automatic naming has more disadvantages than advantages. Explicit names provided by the application that generates the data is the way to go.
Note that automatic naming for vectors (‘X’, ‘Y’, ‘Z’) is still there. This is IMHO slightly inconsistent but reasonable because the chance that a vector of length 3 is represents spatial coordinates is pretty high. Still, if there is consensus that it’s better to avoid assumptions that might be wrong in rare cases, this would be a good moment to remove it.