Benchmarks

These benchmarks showcase the performances of the KeOps routines as the number of samples/points varies (typical use cases should be from 100 to 1,000,000).

Comparison with other related projects

KeOps is fast! You may find here a benchmark that compare the performances of pyKeOps with other related projects.

pyKeOps benchmarks

Note

If you run a KeOps script for the first time, the internal engine may take a few minutes to compile all the relevant formulas. This work is done once as KeOps stores the resulting shared object files (*.so) in a cache directory.

K-Nearest Neighbors search

K-Nearest Neighbors search

Datasets for the benchmarks

Datasets for the benchmarks

Mixed-precision and accuracy settings

Mixed-precision and accuracy settings

Radial kernels convolutions

Radial kernels convolutions

Gradient of Radial kernels convolutions

Gradient of Radial kernels convolutions

Benchmarking Gaussian convolutions in high dimensions

Benchmarking Gaussian convolutions in high dimensions

Solving positive definite linear systems

Solving positive definite linear systems

Scaling up Gaussian convolutions on 3D point clouds

Scaling up Gaussian convolutions on 3D point clouds