Simone Silvestri comments

Results 667 comments of


                                            Simone Silvestri

Need interpretable error if `LagrangianParticles` are initialized with invalid positions

Future work on the particles module is also to write some tests for immersed boundary particle tracking

Adds grid metrics to NetCDF output

Hmmm, no we do not have that but we have the `inactive_cell`, `inactive_node` and `peripheral_node` functions, which kind of do that. (would be half if `peripheral_node` and zero if `inactive_node`)

`ShallowWaterModel.bathymetry`?

The bathymetric term is already in #2522

GPU + no-slip BC

We could remove tupled halo filling until we figure out what the Julia compiler does not like in passing a tuple of boundary conditions

GPU + no-slip BC

Ok I can remove it at the moment, then we have more time to investigate the issue

GPU + no-slip BC

I think it only depends on the combination of Boundary conditions you give to the GPU. Anyways, it's a weird error, I don't seem to be able to reproduce it...

GPU + no-slip BC

Also I am using CUDA 3.12 if it makes a difference

Construct grid parameters with `BigFloat` to avoid round-off error

It seems like with this method the grid's length is less accurate... I ll take a look

Construct grid parameters with `BigFloat` to avoid round-off error

By removing the conversion to `Float32` worked for me...

Construct grid parameters with `BigFloat` to avoid round-off error

Thanks @simonbyrne, that works beautifully!