[
https://issues.apache.org/jira/browse/LUCENE-4607?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13528017#comment-13528017
]
Robert Muir commented on LUCENE-4607:
-------------------------------------
The other idea (just for discussion) would be "number of i/os".
So for example, phrasequery's formula would likely use totalTermFreq rather
than docFreq.
This would reflect the fact that its much more expensive than a conjunction.
> Add estimateDocCount to DocIdSetIterator
> ----------------------------------------
>
> Key: LUCENE-4607
> URL: https://issues.apache.org/jira/browse/LUCENE-4607
> Project: Lucene - Core
> Issue Type: Bug
> Components: core/search
> Affects Versions: 4.0
> Reporter: Simon Willnauer
> Fix For: 4.1, 5.0
>
> Attachments: LUCENE-4607.patch
>
>
> this is essentially a spinnoff from LUCENE-4236
> We currently have no way to make any decsision on how costly a DISI is
> neither when we apply filters nor when we build conjunctions in BQ. Yet we
> have most of the information already and can easily expose them via a cost
> API such that BS and FilteredQuery can apply optimizations on per segment
> basis.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
