Daniel has been using the easi python binding on his laptop.
He installed "manually" easi (1.6.2), asagi (latest, i.e. 132da427f297ae0a31ff941874cc464a4feeb985), and ImpalaJIT (well I guess he does not need that but it is installed).
He uses numpy==2.2.4 and python==3.13.3.
When using this script,
https://github.com/SeisSol/SeisSol/blob/master/preprocessing/science/read_ini_fault_parameter.py
e.g.with
python ~/SeisSol/SeisSol/preprocessing/science/read_ini_fault_parameter.py angle1_output-fault.xdmf initial_stress.yaml --ref " -0.1,0,-1" --par "s_zz"
He gets very strange results (e.g., constant s_zz even when s_zz is depth-dependent).
initial_stress.yaml does not depend on ASAGI.
I tried running the same command on our local server (cachemiss, where easi is installed with spack), and it works well.
(also with the exact same numpy version he uses (but with python 3.11.14).
Parameter files:
sig_zz.yaml.txt
initial_stress.yaml.txt
Here is the CMakeCache.txt of the easi install (if that helps).
CMakeCache.txt
Daniel has been using the easi python binding on his laptop.
He installed "manually" easi (1.6.2), asagi (latest, i.e. 132da427f297ae0a31ff941874cc464a4feeb985), and ImpalaJIT (well I guess he does not need that but it is installed).
He uses numpy==2.2.4 and python==3.13.3.
When using this script,
https://github.com/SeisSol/SeisSol/blob/master/preprocessing/science/read_ini_fault_parameter.py
e.g.with
python ~/SeisSol/SeisSol/preprocessing/science/read_ini_fault_parameter.py angle1_output-fault.xdmf initial_stress.yaml --ref " -0.1,0,-1" --par "s_zz"He gets very strange results (e.g., constant s_zz even when s_zz is depth-dependent).
initial_stress.yaml does not depend on ASAGI.
I tried running the same command on our local server (cachemiss, where easi is installed with spack), and it works well.
(also with the exact same numpy version he uses (but with python 3.11.14).
Parameter files:
sig_zz.yaml.txt
initial_stress.yaml.txt
Here is the CMakeCache.txt of the easi install (if that helps).
CMakeCache.txt