lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Wolf Siberski <siber...@l3s.de>
Subject Re: SortTest failing
Date Tue, 26 Apr 2005 08:15:29 GMT
Chuck Williams wrote:
> Otis Gospodnetic wrote:
>> And this makes me think that this broke during my last commit of Wolf's
>> patch for MultiSearcher and docFreq stuff.  However I did run 'ant
>> test' before commit and did see BUILD SUCCESSFUL, so I'm not 100% sure.
>>
>>
>> Anyone else seeing this error?
>>  
>>
> Otis, I think that is a bug in the test case that Wolf's patch has 
> exposed. It creates an index consisting of two copies of "full". This 
> should be equivalent to an index where every document is duplicated 
> (occurs twice), not to an index where each document occurs only once. 
> With the new correct idf normalizaiton, this increases all the docfreq's 
> for the terms, which changes the score. Scores on the duplicated index 
> should not be the same as scores on the unduplicated index.

Exactly. The test relied on the 'wrong' scoring of the previous MultiSearcher
implementation, and thus fails now. For some reason, the test didn't fail
when I submitted the patch. A correction is included as part of the addition
I submitted recently (http://issues.apache.org/bugzilla/show_bug.cgi?id=31841).

--Wolf


---------------------------------------------------------------------
To unsubscribe, e-mail: java-dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-dev-help@lucene.apache.org


Mime
View raw message