I have just been going through a process where I am working with using a joined table to intersect with another data set. I found that if you want to do further joins and further operations like intersections on selections (subsets or one table), then you are much better off to save the subset/ join as a temporary table and work with that than the original. It was at least 10 times faster even with all joining fields indexed.
r
On Thu, 26 Feb 2004 09:58:32 -0500, Jacques Paris <[EMAIL PROTECTED]> wrote:
I have a MI table with the geo base and some tabular data, and a DBF table
registered in MI; I wille be dealing with tables containing between 10 000
and 100 000 "rows". I plan to make several operations involving a joint.
Both tables contain a column with unique identifiers.
I would like to learn about your experience in that area. I am looking for
the setup that will guarantee the fastest execution. I have identified two
areas where choices are possible: the type of the identifier and the use of
indexes.
Should the identifier column be string or integer? Is one "faster" than the
other?
Should indexing be done on the identifier columns? Are gains in speed worth
the time for spent in indexing and the enlarged IND files?
Any input will be appreciated.
Jacques Paris e-mail [EMAIL PROTECTED] MapBasic-MapInfo support http://www.paris-pc-gis.com
--------------------------------------------------------------------- List hosting provided by Directions Magazine | www.directionsmag.com | To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] Message number: 10610
--
Robert Crossley Agtrix P/L 9 Short St New Brighton 2483 Far Southern Queensland AUSTRALIA
153.549004 E 28.517344 S
P: 02 6680 1309 F: New Connection M: 0419 718 642 E: [EMAIL PROTECTED] W: www.agtrix.com W: www.wotzhere.com
--------------------------------------------------------------------- List hosting provided by Directions Magazine | www.directionsmag.com | To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] Message number: 10619
