GriSPy: Fixed-radius nearest neighbors grid search in Python
Abstract
GriSPy (Grid Search in Python) uses a regular grid search algorithm for quick fixed-radius nearest-neighbor lookup. It indexes a set of k-dimensional points in a regular grid providing a fast approach for nearest neighbors queries. Optional periodic boundary conditions can be provided for each axis individually. GriSPy implements three types of queries: bubble, shell and the nth-nearest, and offers three different metrics of interest in astronomy: the Euclidean and two distance functions in spherical coordinates of varying precision, haversine and Vincenty. It also provides a custom distance function. GriSPy is particularly useful for large datasets where a brute-force search is not practical.
- Publication:
-
Astrophysics Source Code Library
- Pub Date:
- December 2019
- Bibcode:
- 2019ascl.soft12013C
- Keywords:
-
- Software