[dschwoerer]

This allows for a runtime switch of the underlying method for further testing much easier.

[github-actions]

clang-tidy made some suggestions

[ZedThree]

I'm wondering if there's actually enough options here now to use the generic Factory somehow? We could just make the operator a class?

[dschwoerer]

I'm wondering if there's actually enough options here now to use the generic Factory somehow? We could just make the operator a class?

Maybe? That seems though a different PR ...

[dschwoerer]

@bendudson it seems the support operator does not converge:

Running command: ./fci_mms MYG=1 mesh:ddy:first=C2 mesh:paralleltransform:xzinterpolation:type=lagrange4pt mesh:file=mesh.8.fci.nc ... done in 0.472s
grad_par errors: l-2 0.007596 l-inf 0.011683
grad2_par2 errors: l-2 0.104717 l-inf 0.144821
div_par errors: l-2 0.007596 l-inf 0.011683
div_par_K_grad_par errors: l-2 0.073924 l-inf 0.127443
div_par_K_grad_par_mod_Original errors: l-2 0.070800 l-inf 0.122122
div_par_K_grad_par_mod_PetscOperator errors: l-2 0.049332 l-inf 0.156445
laplace_par errors: l-2 0.104717 l-inf 0.144821
FV_div_par_mod errors: l-2 0.021949 l-inf 0.044925
FV_div_par_fvv errors: l-2 0.067395 l-inf 0.161695
Running command: ./fci_mms MYG=1 mesh:ddy:first=C2 mesh:paralleltransform:xzinterpolation:type=lagrange4pt mesh:file=mesh.16.fci.nc ... done in 0.521s
grad_par errors: l-2 0.001570 l-inf 0.002509
grad2_par2 errors: l-2 0.030554 l-inf 0.044746
div_par errors: l-2 0.001570 l-inf 0.002509
div_par_K_grad_par errors: l-2 0.021575 l-inf 0.041754
div_par_K_grad_par_mod_Original errors: l-2 0.021353 l-inf 0.041218
div_par_K_grad_par_mod_PetscOperator errors: l-2 0.044648 l-inf 0.213109
laplace_par errors: l-2 0.030554 l-inf 0.044746
FV_div_par_mod errors: l-2 0.004186 l-inf 0.009372
FV_div_par_fvv errors: l-2 0.023426 l-inf 0.061165
Running command: ./fci_mms MYG=1 mesh:ddy:first=C2 mesh:paralleltransform:xzinterpolation:type=lagrange4pt mesh:file=mesh.32.fci.nc ... done in 0.745s
grad_par errors: l-2 0.000370 l-inf 0.000595
grad2_par2 errors: l-2 0.007939 l-inf 0.011790
div_par errors: l-2 0.000370 l-inf 0.000595
div_par_K_grad_par errors: l-2 0.005606 l-inf 0.011136
div_par_K_grad_par_mod_Original errors: l-2 0.005590 l-inf 0.011091
div_par_K_grad_par_mod_PetscOperator errors: l-2 0.043008 l-inf 0.285402
laplace_par errors: l-2 0.007939 l-inf 0.011790
FV_div_par_mod errors: l-2 0.000969 l-inf 0.002131
FV_div_par_fvv errors: l-2 0.009590 l-inf 0.024281
Running command: ./fci_mms MYG=1 mesh:ddy:first=C2 mesh:paralleltransform:xzinterpolation:type=lagrange4pt mesh:file=mesh.64.fci.nc ... done in 1.61s
grad_par errors: l-2 0.000091 l-inf 0.000147
grad2_par2 errors: l-2 0.002004 l-inf 0.002986
div_par errors: l-2 0.000091 l-inf 0.000147
div_par_K_grad_par errors: l-2 0.001415 l-inf 0.002827
div_par_K_grad_par_mod_Original errors: l-2 0.001414 l-inf 0.002822
div_par_K_grad_par_mod_PetscOperator errors: l-2 0.052484 l-inf 0.521644
laplace_par errors: l-2 0.002004 l-inf 0.002986
FV_div_par_mod errors: l-2 0.000238 l-inf 0.000518
FV_div_par_fvv errors: l-2 0.004446 l-inf 0.010604