[
https://issues.apache.org/jira/browse/LUCENE-4782?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13583060#comment-13583060
]
Tommaso Teofili commented on LUCENE-4782:
-----------------------------------------
thanks Robert, that may make sense but I'm not really sure as in 4.x we still
support 3x codecs in general.
Also I noticed that other methods used in the classification module that depend
on the underlying codec may return -1 (unsupported).
I'll have a deeper look and see if it really make sense to 'fallback' such
calls or either it'd be safer and reasonable to follow your suggestion.
> Let the NaiveBayes classifier have a fallback docCount method if codec
> doesn't support Terms#docCount()
> -------------------------------------------------------------------------------------------------------
>
> Key: LUCENE-4782
> URL: https://issues.apache.org/jira/browse/LUCENE-4782
> Project: Lucene - Core
> Issue Type: Bug
> Components: modules/classification
> Reporter: Tommaso Teofili
> Assignee: Tommaso Teofili
> Fix For: 4.2, 5.0
>
>
> In _SimpleNaiveBayesClassifier_ _docsWithClassSize_ variable is initialized
> to _MultiFields.getTerms(this.atomicReader,
> this.classFieldName).getDocCount()_ which may be -1 if the codec doesn't
> support doc counts, therefore there should be an alternative way to
> initialize such a variable with the documents count.
--
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
