Spatial algorithms and data structures (cupyx.scipy.spatial)#

Hint

SciPy API Reference: Spatial (scipy.spatial)

Note

The spatial module uses the cuVS library as a backend. You need to install cuVS package <https://anaconda.org/rapidsai/cuvs> from rapidsai Conda channel to use features listed on this page.

Note

Currently, the spatial module is not supported on AMD ROCm platforms.

Nearest-neighbor queries#

KDTree(data[, leafsize, compact_nodes, ...])

KDTree(data, leafsize=16, compact_nodes=True, copy_data=False,

Delaunay triangulation#

Delaunay(points[, furthest_site, incremental])

Delaunay tessellation in 2 dimensions.

Functions#

distance_matrix(x, y[, p])

Compute the distance matrix.