> -----Original Message----- > From: [email protected] [mailto:[email protected]] On Behalf > Of Mike Preece > Sent: Tuesday, January 05, 2010 2:16 PM > To: jBASE > Subject: [SPAM] Re: Performance issue on files with INDEX > > > > On Jan 5, 4:29 pm, pat <[email protected]> wrote: > > Au Contraire > > > > Using an Index with a limited number of differing values will reduce > > the SELECT time when Selecting a specific value Say for example in > the > > extreme case of ( only ) four different 'CURRENCIES' within the items > > in your file, eg : > > > > 1 million items with a 'CURRENCY' of 'GBP' > > 1 million items with a 'CURRENCY' of 'USD' > > 1 million items with a 'CURRENCY' of 'YEN' > > and > > 1 thousand items with a 'CURRENCY' of 'FF' > > > > SELECTing the million item ids for items with a 'CURRENCY' of 'USD' > > will be instantaneous via the index, compared with processing the > > '3,001,000' items looking for a 'CURRENCY' of 'USD' > > > > I await with interest the actual figures from the OP. I wonder whether > your suggested figures for an "extreme case" reflects real life. > Perhaps the figures for one currency will almost always represent way > more than 30% of data. >
Yes, but if you append the item ID to the indexed data and modify the query to account for it, you won't have any problems with building it anyway. Though it is really better to program to the indexes in jBC (OPENINDEX etc), where you then have complete control over how you traverse. Using SELECT is really just a 'cop-out' from doing it programmatically (I think you already said that too). Jim
-- Please read the posting guidelines at: http://groups.google.com/group/jBASE/web/Posting%20Guidelines IMPORTANT: Type T24: at the start of the subject line for questions specific to Globus/T24 To post, send email to [email protected] To unsubscribe, send email to [email protected] For more options, visit this group at http://groups.google.com/group/jBASE?hl=en
