Hello, i fiddled more with the data, but to no avail. It seems its something in lochs extrapolation code using misaligned data or something... Is this reproducible at your setup?
Am 31.03.2020 um 18:26 schrieb Benedikt Hallinger <beni@hallinger.org>:
Hello, in a dataset i get a strange behavior regarding the loch model at passage ends:
At station 1.10 a weird spike is produced in the loch model, that should not be there. If one comments out some branch that is not located near 1.10, the problem goes away.
What causes these spikes?
Attached a test set with comented section, thus producing correct ending of the tunnel. Also, an image showing the problem (left: wrong spike, right: correct result with the branch commented)
Thanks alot, Beni <test.tar.gz> <testesult.png>