counter intuitive behaviour of place_receivers?

[Thomas-Ulrich]

I would expect place_receivers to discard receivers outside the mesh, but this is not the case.

di73yeq4@login03:/hppfs/scratch/0A/di73yeq4/issue_receivers> /hppfs/work/pr45fi/di73yeq4/Meshing/place_receivers/build/place_receivers -d 1 -r tmp/seissol_station.txt -m tmp/mesh.puml.h5 -o tmp/seissol_station_placed.txt
Read 18 receivers.
Wed Jan 08 09:51:38, Info:  Found 460084 cells 
Wed Jan 08 09:51:38, Info:  Found 77088 vertices 
Processing partition 0 to 1

gives 18 receivers in tmp/seissol_station_placed.txt.

files to be find in:
https://github.com/user-attachments/files/18036326/issue_receivers.tar.gz

[davschneller]

Could be due to #8 ; i.e. it could probably be avoided by not specifying the z coordinate. (i.e.: for a point to be ignored in the output, we currently (w/o #54 ) need it to have no z coordinate, and not be found by a free surface; the former is not fulfilled for out-of-domain points. That can be changed to only not being found by a free surface, as #54 does—it'll automatically exclude points which are outside the domain)

Does maybe #54 fix the semantics correctly—or should points with a given z coordinate which could not be mapped still be kept in the result file?