Daniel VandenHeuvel

Thanks for the report @henry2004y. I'll have a bit more to say about this but is the line `triangle_vals = itp(_x, _y; method=Triangle())` intentional in your timings? You seem to...

Some thoughts - You aren't exactly comparing the same things here. With `derivatives=true` you are also timing derivative generation which Matplotlib wouldn't be doing (I've never used Matplotlib, so correct...

I did find one issue #35, and another that can't be fixed without a new major release, but they don't make much difference. I'm honestly not sure what the main...

I probably can't look at that for thinking of changes depending on Matplotlib's license, I'd have to double check that. I was thinking though: I've had to do something similar...

Well, I got around to looking at caching - see https://github.com/DanielVandH/NaturalNeighbours.jl/tree/triangle. It doesn't really make things much better unfortunately. ```julia julia> @benchmark nn_benchmark() BenchmarkTools.Trial: 443 samples with 1 evaluation. Range...

All that said, if you are happy with the comparison between the performance with `parallel=true` compared to Matplotlib as I've demonstrated above, feel free to close this. Otherwise we can...

The caching work should be registered soon (~15 minutes) as v1.3.4. By default, it will now cache the necessary coordinates for `Triangle()` (so you don't actually need `allow_cache=true` as I...

@henry2004y With DelaunayTriangulation 1.1.0 you can more easily change how predicates are computed by using the `predicates` keyword, which might also help your performance. e.g. ```julia julia> function nn_benchmark(kernel) itp...

Thanks for the issue! Technically yes this is "user error" but, it's probably closer to a bug since I actually do want to avoid duplicate points now as of https://github.com/JuliaGeometry/DelaunayTriangulation.jl/pull/218....

@dgleich: I realise now that I think I misunderstood your issue. My thought was that you were having an issue with your boundary nodes being something like `[p, p1, p2,...