Opened 3 years ago
Last modified 2 years ago
#5032 closed defect
ST_EstimatedExtent gives bogus results on combined gist indexes — at Initial Version
Reported by: | Björn Harrtell | Owned by: | pramsey |
---|---|---|---|
Priority: | medium | Milestone: | PostGIS 3.0.7 |
Component: | postgis | Version: | master |
Keywords: | Cc: |
Description
Reproduction:
drop table test; with vals(k, geom) as (values (1, 'POINT(1 1)'::geometry), (2, 'POINT(1 2)'::geometry)) select * into table test from vals; analyze test; select st_estimatedextent('test', 'geom'); -- BOX(1 0.995000004768372,1 2.005000114440918) create index test_geom_idx on test using gist (geom); select st_estimatedextent('test', 'geom'); -- BOX(1 1,1 2) drop index test_geom_idx; create index test_k_geom_idx on test using gist (k, geom); select st_estimatedextent('test', 'geom'); -- BOX(1.401298464324817E-45 1,2.802596928649634E-45 1)
This is on:
POSTGIS="3.2.0dev 3.2.0beta2-11-g27f44ecf6" [EXTENSION] PGSQL="120" GEOS="3.9.0-CAPI-1.16.2" PROJ="7.2.1" LIBXML="2.9.12" LIBJSON="0.15" (core procs from "3.2.0dev 3.2.0beta2-3-geda70258f" need upgrade)
But I think it's an old bug.
Note:
See TracTickets
for help on using tickets.