search

KristofferC / NearestNeighbors.jl

High performance nearest neighbor data structures (KDTree and BallTree) and algorithms for Julia.
Other
425 stars 66 forks source link

Nearest Neighbors in a very large dataset. #140

Closed compleathorseplayer closed 2 years ago

compleathorseplayer commented 2 years ago

Pardon me if this is a stupid question (or covered under another heading which I did not see), but is there a version here (or a way of using) that works for relatively large (~250k points of 20 dims) dataset? KDTree freezes my machine at this scale. I tried increasing the leafsize, but it didn't help and anyway, I don't have a feel for how to sensibly adjust.

I would be grateful for any suggestions.

Thanks in advance! PS I am really mainly interested in Regression/Classification PPS Apologies - I have tried so many methods - each with different calling sequence - I had transposed the matrix :-b Works OK the right way round

KristofferC commented 2 years ago

My guess is you mixed up dimensions and number of points?