> b) The actual number of direct offsprings of a face can be obtained via > face->n_children(). > > c) If you are on an active cell and and consider one of the faces of that > cell, then face->number_of_children() gives the total number of active > offsprings of that face, which (for a refined face) is always two in 2d, but > could be two, three or four in 3d. This is the number of subfaces to > integrate over in DG methods, for example. >
You mean there is a semantic difference between n_children() and number_of_children()? I do not think this is really a good idea. But I agree that we have a problem with naming here. I suggest to rename the second function n_active_children(), to be consistent with the remainder of the library and to avoid confusion. After all, this is, what it does according to the documentation. If I run through all indices from zero to this number, is this what I plug into neighbor_child_on_subface as second argument? Best, Guido _______________________________________________
