search

OrchardLANL / DPFEHM.jl

DPFEHM: A Differentiable Subsurface Flow Simulator
Other
30 stars 13 forks source link

ensure examples all run #13

Closed WilkAndy closed 2 years ago

WilkAndy commented 2 years ago

It would be nice to have a script/action/whatever to run all the examples. Preferably, this could be run on every PullRequest, or at least, manually run to check for regressions. Currently, i'm going through all the examples, and many of them don't run without warnings/errors

fracture_network_solver_scaling/... often complains about "Package FEHM does not have DelimitedFiles in its dependencies..." gw_multigrid_inverse/inputdeck.jl:30 file not found piml/theislike_piml_setup.jl:73 params not defined

I haven't checked others

Refs https://github.com/openjournals/joss-reviews/issues/4560

WilkAndy commented 2 years ago

pressure_management/ex.jl ERROR: LoadError: UndefVarError: flatten not defined at line 82

richards/inputdeck.jl ERROR: LoadError: BoundsError: attempt to access 640800×640800 SparseArrays.SparseMatrixCSC{Float64, Int64} with 3200798 stored entries at index [1:640800, 640801]

richards_steadystate_3d/inputdeck.jl Similar BoundsError

scott_example/ex.jl might be running - i killed it after about 20mins of no output.

theislike/ex.jl ERROR: LoadError: DimensionMismatch("arrays could not be broadcast to a common size; got a dimension with lengths 2401 and 2601") at line 75

I believe all the other examples are working.

omalled commented 2 years ago

OK, I went through those examples and either got it running or deleted the example (some of there were basically deprecated by newer examples). So, I'll close the issue.

There are still some warnings, which I'm OK with. I can try to fix the warnings if you insist, but, e.g., fixing the FEHM warning requires a fix in a different package. I really appreciate you going through these examples! Those Richards examples that weren't working highlighted a testing gap, which I also fixed here.