Peter Brady

Ah. Now I see the failure. This stems from custom comparison and hashing routines which need to be carefully wrapped in PyRecursion calls. I think the best thing to do...

Thanks @ngoldbaum. This sounds like a great idea but I'm not sure when I'll get around to it. Any interest in pursuing this yourself?

Absolutely!

That is a lot more cuda than I thought!

Thanks for the links. I can certainly see why the original choice was made. In my own field of computational physics some still prefer the CUDA route but libraries like...

PR #34 fixes this issue.

Unless I'm misunderstanding something, this is a straightforward change to `std::invoke_result_t`. Is this more about dropping c++14?

I guess it's not a complete drop-in replacement. I've got a fix in #1124.

Seems like a CI failure: ```console Error: Could not find file '/home/runner/work/_actions/_temp_68ccd911-c253-4309-82e1-094e899b9ceb/_staging/gregsdennis-dependencies-action-f4fa194/node_modules/.bin/esparse'. ```

I've updated the docs and changelog. Computing `c` is more of an algorithm than a formula so I referenced the paper instead of repeating the algorithm.