[ https://issues.apache.org/jira/browse/MATH-1276?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14804592#comment-14804592 ]
Gilles commented on MATH-1276: ------------------------------ I'd suggest that the clean-up of the current code is done in a separate commit: * remove needless declaration of a local variable "p" with the same value as the "probabilityOfSuccess" field * remove needless local variable "ret" ("return" statements can be used directly in each way of the alternatives). * define new instance variable "log1mProbabilityOfSuccess" * define new instance variable "logProbabilityOfSuccess" (not in the proposed patch but would be a similar improvement as the above) The change in method "probability" should be in a separate commit too. Personally, I think that "reading noise" is lowered by not using unnecessary number "suffixes": "1" rather than "1.0" or "1.". > Override GeometricDistribution.inverseCumulativeProbability() to speed up > sampling > ---------------------------------------------------------------------------------- > > Key: MATH-1276 > URL: https://issues.apache.org/jira/browse/MATH-1276 > Project: Commons Math > Issue Type: Improvement > Reporter: Otmar Ertl > Attachments: MATH-1276.patch > > -- This message was sent by Atlassian JIRA (v6.3.4#6332)