Return-Path: pg_adm@postgres.berkeley.edu
Received: by postgres.Berkeley.EDU (5.61/1.29)
	id AA02346; Fri, 26 Mar 93 02:55:15 -0800
Date: Fri, 26 Mar 93 02:55:15 -0800
Message-Id: <9303261055.AA02346@postgres.Berkeley.EDU>
From: Christian Millour <chris@etca.fr>
Subject: request/index problem
To: postgres@postgres.berkeley.edu
Sender: pg_adm@postgres.berkeley.edu

Hi knowledgeable netters,

I have this little problem:

given 3000+ rectangles, similar size, irregular distribution, possibly
overlapping, relatively tiny with respect to the universe (let's say the
universe edge size is about 600 times that of one rectangle), and 5000+ 2D
point locations, also irregularly distributed, I'd like to compute which
rectangle(s) (if any) each point falls in, and which and how many points
are covered by each rectangle.

1) What is the Right Thing to do ?
2) What is the Right Thing to do with Postgres ?

I figure this will boil down to an index/access method issue, however I
lack experience in this domain. Would R-Trees help ? How ? 

Thanks in advance for any advice and pointers.

- Christian Millour         ETCA/CTME/GIS/CEPIA 
  chris@etca.fr             16bis Ave Prieur de la Cote d'Or
  [33] (1) 42 31 96 45      94114 Arcueil Cedex -- France
All opinions expressed are my own. Standard disclaimers apply.