Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.misc -> Re: Definition of "Selectivity"

Re: Definition of "Selectivity"

From: Fred Casale <bigfred_at_geocities.com>
Date: Fri, 4 Jun 1999 14:43:44 -0400
Message-ID: <7j96op$26n$1@autumn.news.rcn.net>

Fred Casale <bigfred_at_geocities.com> wrote in message news:7j8v3a$mq3$1_at_autumn.news.rcn.net...
>
> I'm a bit confused over Oracle's definition
> of index selectivity....
>
> The documentation says that you divide the
> number of rows by the cardinality of the index
> and the results should be expressed as a
> percentage.
>
> They define selectivity as "the percentage of
> rows in a table having the same value for the
> indexed column"....
>
> By using the formula above on a column with
> 10 rows and cardinality of 2:
>
> 10 / 2 = 5 <--- but this isn't a percentage!
>
> To me, the formula based on the definition should
> be:
>
> rows / cardinality / rows
>
> -or-
>
> the inverse of the cardinality....
>
> In the example, you'd get 20% which is the correct
> answer assuming uniform distribution....

Sorry, I meant 50%!

>
> Where am I going wrong?
>
> Thanks.
>
>
Received on Fri Jun 04 1999 - 13:43:44 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US