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

Home -> Community -> Mailing Lists -> Oracle-L -> Re: "data is skewed" - what does it mean ?

Re: "data is skewed" - what does it mean ?

From: orababy <orababy_at_mail15.com>
Date: Fri, 25 Jul 2003 09:17:57 +0400 (MSD)
Message-Id: <26007.339511@fatcity.com>


Gogala,

so , if there is an index on a column like SEX , then it means data ( in that index ) is skewed. am i right ?

for an index , if cardinality is low OR clustering_factor is high --> it means data is skewed.

sorry if i sound stupid :-|

Is "SKEW" used with respect to index or table or both ?

>>..data is the index is not distributed evenly

i cannot understand this.
index data is not disturbed evenly with respect to **** ?

can u explain ( if possible ) with an example plz ?

 Mladen Gogala <mgogala_at_adelphia.net>:

> It's actually a substitute for another word which is not to be
repeated on
> this list because of the presence of the ladies, much
like "heck", "darn"
> or
> "frigging". It means that the data in the index (you found that
word in the
> chapter about indexes, statistics and the Curses Based
Optimizer, am I
> right)
> is not distributed evenly, with small variations in the number
of rows for
> any
> given key, but that there is a small subset of keys which
corresponds to
> significant part of the table. Speaking mathematically, there
exists at
> least
> one value in the indexed columns so that the number of rows
corresponding
> to
> this value is significantly larger then the average number of
rows per
> value.
> In other words, it means that the application relying on such
Received on Fri Jul 25 2003 - 00:17:57 CDT

Original text of this message

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