As far as I understand it (not much), gist index over spatial data is in fact gist index over range(x), range(y).
This is why Gist works in n-dimension. It always works on range (conceptually). In fact rectangle are the intersection of a range on x and a range on y (literally) same, a 3D box is the intersection of range on x,y,z You could go further by adding time, etc. Cheers, Rémi-C 2015-04-01 9:00 GMT+02:00 Magnus Hagander <mag...@hagander.net>: > On Sat, Mar 28, 2015 at 7:52 AM, Rebecca Zahra <rebeccaza...@gmail.com> > wrote: > >> Good morning, >> >> I am Rebecca Zahra and I am currently in my final year of Masters studies >> at the University of Malta. My thesis is about the usage of indexes for >> multi-dimensional data. >> >> I was going through the posts regarding GIST indexes and I came across >> the following >> http://dba.stackexchange.com/questions/39589/optimizing-queries-on-a-range-of-timestamps-two-columns >> >> I was wondering if maybe you can help me with a question. I know that an >> R-Tree index implementation is used on top of GIST to index spatial data. >> Can you please tell me what type of index is used on top of GIST to index >> *range >> types*? >> >> > PostgreSQL has had indexable range types for quite some time now: > http://www.postgresql.org/docs/current/static/rangetypes.html#RANGETYPES-GIST > > Indexable with gist or spgist. I don't think the docs cover the actual > implementation internals though - you'll probably have to go to the source > if you need that. > > -- > Magnus Hagander > Me: http://www.hagander.net/ > Work: http://www.redpill-linpro.com/ >
