[ https://issues.apache.org/jira/browse/MATH-1089?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gilles updated MATH-1089: ------------------------- Attachment: MATH-1089.patch I propose the attached patch; overall it is probably faster, though the absolute speed difference will be marginal in most cases. > Precision.round() returns different results when provided negative zero as > double or float > ------------------------------------------------------------------------------------------ > > Key: MATH-1089 > URL: https://issues.apache.org/jira/browse/MATH-1089 > Project: Commons Math > Issue Type: Bug > Reporter: Thomas Neidhart > Fix For: 3.3 > > Attachments: MATH-1089.patch > > > Precision.round(-0.0d, x) = 0.0 > Precision.round(-0.0f, x) = -0.0 > After discussion on the mailinglist, the result should always be -0.0. -- This message was sent by Atlassian JIRA (v6.1.5#6160)