Re: [GENERAL] speeding up CUBE queries

2007-10-08 Thread Rajarshi Guha
On Oct 8, 2007, at 9:15 PM, Oleg Bartunov wrote: I'm not sure, but probably mtree will be better ? Thanks - do you have any pointers to this? Google only turns up an mtree utility that seems to be related to file system operations --

Re: [GENERAL] speeding up CUBE queries

2007-10-08 Thread Oleg Bartunov
I'm not sure, but probably mtree will be better ? On Mon, 8 Oct 2007, Rajarshi Guha wrote: Hi, I'm running Postgres 8.2.4 and have installed the CUBE extension. Using this I had a 10M row table populated with 12-dimensional zero-volume cubes (i.e., 12D points). My queries are of the form se

[GENERAL] speeding up CUBE queries

2007-10-08 Thread Rajarshi Guha
Hi, I'm running Postgres 8.2.4 and have installed the CUBE extension. Using this I had a 10M row table populated with 12-dimensional zero- volume cubes (i.e., 12D points). My queries are of the form select * from ctab where '(x1,x2,x3,...,x12), (y1,y2,y3,...,y12)'::cube @> cubeField; So e