Closed AhmedSalih3d closed 2 years ago
Idea is cool, but does not fit main scope of this project.
Perhaps if some day we would be using different kernels through a simulation or for different equations, so we would have like 10-20 expressions, this could make sense - for now let us not divert attention.
Hello
I will not write too much here, but the basic idea is that it would be nice to give the user the option to specify a kernel and then using ForwardDiff.jl, provide numerically equivivalent derivative, gradient, divergence and curl operators. For more about the idea see here:
https://discourse.julialang.org/t/can-i-expect-forwarddiff-to-give-the-same-performance-in-this-case/67894/4
@LudwigBoess I hope this is fine to have here, I just don't want to lose this idea. I might fork this repos in the future, trying to implement this idea for my favourite kernel you have seen a few times (the one in the link) :)
Kind regards