[
https://issues.apache.org/jira/browse/SOLR-9549?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15513070#comment-15513070
]
Joel Bernstein commented on SOLR-9549:
--------------------------------------
I suspect the problem is in this block of code:
{code}
public void collect(int doc) throws IOException{
int valuesDocID = leafOutcomeValue.docID();
if (valuesDocID < doc) {
valuesDocID = leafOutcomeValue.advance(valuesDocID);
}
int outcome;
if (valuesDocID == doc) {
outcome = (int) leafOutcomeValue.longValue();
} else {
outcome = 0;
}
outcome = trainingParams.positiveLabel == outcome? 1 : 0;
if (outcome == 1) {
positiveDocsSet.set(context.docBase + doc);
}
docsSet.set(context.docBase+doc);
}
{code
> StreamExpressionTest failures
> -----------------------------
>
> Key: SOLR-9549
> URL: https://issues.apache.org/jira/browse/SOLR-9549
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Alan Woodward
>
> Reproduces for me on master:
> ant test -Dtestcase=StreamExpressionTest
> -Dtests.method=testBasicTextLogitStream -Dtests.seed=DB749AA9C9E30657
> -Dtests.multiplier=3 -Dtests.slow=true -Dtests.locale=os
> -Dtests.timezone=Asia/Bahrain -Dtests.asserts=true -Dtests.file.encoding=UTF-8
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
