WARNING: Currently under development and might be broken at any moment
I tiny Kd-Tree implementation that is pretty fast. It only stores a pointer to your items and returns indices into your list instead of cloning and allocating a bunch.
The main function that needs optimizing is nearest_within since it does some
unpredictable allocations as it grows a vec. Still faster than many other implementations though!
I would like to make that into an iterator instead if possible..
Key things that differ keyde from others and make it fast: - No recursion, only iterative implementations - No cloning of your data, everything is refered to by indices into your data