evaluation of moments improved #20
Conversation
There was a problem hiding this comment.
The check was useless in its original position:
(a) the checked arrays were not allocated at this point, therefore size() would give any kind of output back
(b) it never threw an error nonetheless, as the .not. was not properly placed, so that the total output would actually only be true, if verts (what mostly the case due to (a)) did not align with n_tetra AND simultaneously, neighbour/neighbour_faces DID match (by some unlikely chance, as they were also not allocated before).
The position/check idea is properly an artifact of an earlies version of the code.
There was a problem hiding this comment.
The changes are okay, and the result of the reversibility test (GORILLA_APPLETS) still work with the new moment calculation. Suggestion for the future: relocate the moments + interface declaration to an external file/module and call the interface from the calc_optional_quantities routine.
No description provided.