Title: Efficiently handling spherical data in Postgres
Authors: O.Bartunov, S.Koposov, S.Karpov
Abstract: Spherical data appear in a lot of problems in modern science and technology, ranging from geoinformational systems to astronomical catalogues. Efficient handling of such data in Postgres is necessary, and Postgres do have support for it, though not in core, through its ability to define fully indexable custom data types and functional indices.
In the talk we review existing implementations of spherical data support available for Postgres in Q3C, pgSphere, PostGIS etc and compare its performance on the range of use-cases - from bulk loading to range queries to live update. We demonstrate how different indices used (B-tree in Q3C, GiST+R-tree in pgSphere and PostGIS) help to optimize data handling for a specific problem in hands.
Back to top