Keaton Burns

Ok, we've already written functions for writing BCC files from Python arrays, so we can definitely go that route for now, but of course it would be great to have...

+1 for the icicle plots, especially if they can be modified (from that demo) to maintain the same number of rows as the focus moves down the tree, which I...

Thanks for the suggestions but I think we are unlikely to merge this, as we want the tests to be available in a standard pip install.

I'm not sure it's really all that standard for numerical packages, e.g. numpy and scipy ship with built-in tests. Also here the test requirements are actually the lightest -- they...

Pytest is required to run the test suite which tests the local c-extensions builds, including linking to FFTW and MPI, which happen during the pip installation (the c-extensions are not...

Yes I think in general there would be a lot of interest in adaptive schemes, but the complication is the simplest types of explicit schemes won't work -- we need...

The standard RKDP integrator is fully explicit. Dedalus requires IMEX schemes that support linearly implicit integration for algebraic constraints (like boundary conditions) and stiff terms (like diffusivities).

That's great! Definitely feel free to open up a pull request for more detailed code discussion. There should be a test in "test_ivp.py" that loops over the implemented timestepping schemes,...

Feel free to open up a PR if you want to discuss code details. Happy to review it and chat, even if its just draft form.

We are seeing that every for extremely sparse (1e-7) matrices, repeated solves of the cupy SuperLU factorization are spending lots of time in `spsm_analysis`. According to the [nvidia help forums](https://forums.developer.nvidia.com/t/cusparsespsv-solve-function-extremely-slow/312190/2),...