commons-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Luc Maisonobe (JIRA)" <>
Subject [jira] Commented: (MATH-393) Method "getResult()" in "MultiStartUnivariateRealOptimizer"
Date Sun, 25 Jul 2010 08:39:50 GMT


Luc Maisonobe commented on MATH-393:

You're right.
Don't forget to fix also the Javadoc in the interface which is also misleading when you fiw
this. I have checked other implementations of this interface (BrentMinimizer) and what it
returns is really the optimum.

> Method "getResult()" in "MultiStartUnivariateRealOptimizer"
> -----------------------------------------------------------
>                 Key: MATH-393
>                 URL:
>             Project: Commons Math
>          Issue Type: Bug
>            Reporter: Gilles
>            Assignee: Gilles
>            Priority: Trivial
> In "MultiStartUnivariateRealOptimizer" (package "optimization"), the method "getResult"
returns the result of the last run of the "underlying" optimizer; this last result might not
be the best one, in which case it will not correspond to the value returned by the "optimize"
method. This is confusing and does not seem very useful. I think that "getResult" should be
defined as
> {code} 
> public double getResult() {
>     return optima[0];
> }
> {code}
> and similarly
> {code}
> public double getFunctionValue() {
>     return optimaValues[0];
> }
> {code}

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message