[Fwd: debugging ogs - nod_val_vector.size()]

unfortunately I originally posted in the "wrong" mailing list so please
excuse my double posting.
Since i am already writing this I might attach a few extra notes on the
way: I tested the benchmarks
and they seem to work now. I am still working on the last two though.

Have a good weekend!
Wolf Tilmann

untitled-[2] (3.44 KB)


---------------------------- Original Message ----------------------------
Subject: debugging ogs - nod_val_vector.size()


I am currently working on the coupling between the latest OGS version
and Eclipse (Up to this day the coupling only works with the 5.3.3
version of OGS).
While debugging I encountered something which caused problems in several
functions/loops and therefore I was asked to share it with you all since
it might affect you as well.
In the old version of OGS (5.3.3) the array nod_val_vector, or to be
exact its .size(), was used to run loops over all nodes of the mesh
(e.g. copying the pressure data of all nodes from Eclipse to OGS).
However, compared to the old 5.3.3 version the nod_val_vector array is
flipped in the current trunk version of OGS
(the size of each entry of nod_val_vector now corresponds to the total
number of nodes in the mesh and not the number of entries in
nod_val_vector as such).
So if you have to run a loop over all nodes of the mesh apossible
solution to the problem is to save the number of nodes in the mesh in a
long int like this:

      long nnodes = fem_msh_vector[0]->GetNodesNumber(false);

and use it instead of nod_val_vector.size(). Or of course something

Happy debugging and kindest regards,
Wolf Tilmann

Wolf Tilmann Pfeiffer
Institute of Geoscience
Christian-Albrechts-University of Kiel
Ludewig-Meyn-Strasse 10, 24118 Kiel, Germany
phone.: 0049 (0)431 880-4805