You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@commons.apache.org by "Konstantin Berlin (JIRA)" <ji...@apache.org> on 2012/11/17 15:50:16 UTC
[jira] [Created] (MATH-902)
org.apache.commons.math3.exception.TooManyEvaluationsException should
return best found result
Konstantin Berlin created MATH-902:
--------------------------------------
Summary: org.apache.commons.math3.exception.TooManyEvaluationsException should return best found result
Key: MATH-902
URL: https://issues.apache.org/jira/browse/MATH-902
Project: Commons Math
Issue Type: Bug
Affects Versions: 3.1
Environment: any
Reporter: Konstantin Berlin
The org.apache.commons.math3.exception.TooManyEvaluationsException should contain within it the best found solution so far. There are numerous examples when the optimization might not have converged to the stopping condition but the minimum point is better than the starting point that was provided. The user should have the ability to at least decide if it is good enough, or use it as a starting point into a different optimization run.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (MATH-902)
org.apache.commons.math3.exception.TooManyEvaluationsException should
return best found result
Posted by "Gilles (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MATH-902?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gilles updated MATH-902:
------------------------
Fix Version/s: 3.1
Assignee: Gilles
> org.apache.commons.math3.exception.TooManyEvaluationsException should return best found result
> ----------------------------------------------------------------------------------------------
>
> Key: MATH-902
> URL: https://issues.apache.org/jira/browse/MATH-902
> Project: Commons Math
> Issue Type: Wish
> Environment: any
> Reporter: Konstantin Berlin
> Assignee: Gilles
> Priority: Minor
> Fix For: 3.1
>
>
> The org.apache.commons.math3.exception.TooManyEvaluationsException should contain within it the best found solution so far. There are numerous examples when the optimization might not have converged to the stopping condition but the minimum point is better than the starting point that was provided. The user should have the ability to at least decide if it is good enough, or use it as a starting point into a different optimization run.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MATH-902)
org.apache.commons.math3.exception.TooManyEvaluationsException should
return best found result
Posted by "Gilles (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MATH-902?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13501163#comment-13501163 ]
Gilles commented on MATH-902:
-----------------------------
Unless there are objections, I'll implement the suggestion from [this message on the "dev" ML|http://markmail.org/message/f4mq4mpai3ydybm7] into the "Simple...Checker" classes (package "o.a.c.m.optimization").
> org.apache.commons.math3.exception.TooManyEvaluationsException should return best found result
> ----------------------------------------------------------------------------------------------
>
> Key: MATH-902
> URL: https://issues.apache.org/jira/browse/MATH-902
> Project: Commons Math
> Issue Type: Wish
> Environment: any
> Reporter: Konstantin Berlin
> Priority: Minor
>
> The org.apache.commons.math3.exception.TooManyEvaluationsException should contain within it the best found solution so far. There are numerous examples when the optimization might not have converged to the stopping condition but the minimum point is better than the starting point that was provided. The user should have the ability to at least decide if it is good enough, or use it as a starting point into a different optimization run.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Resolved] (MATH-902)
org.apache.commons.math3.exception.TooManyEvaluationsException should
return best found result
Posted by "Gilles (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MATH-902?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gilles resolved MATH-902.
-------------------------
Resolution: Fixed
> org.apache.commons.math3.exception.TooManyEvaluationsException should return best found result
> ----------------------------------------------------------------------------------------------
>
> Key: MATH-902
> URL: https://issues.apache.org/jira/browse/MATH-902
> Project: Commons Math
> Issue Type: Wish
> Environment: any
> Reporter: Konstantin Berlin
> Assignee: Gilles
> Priority: Minor
> Fix For: 3.1
>
>
> The org.apache.commons.math3.exception.TooManyEvaluationsException should contain within it the best found solution so far. There are numerous examples when the optimization might not have converged to the stopping condition but the minimum point is better than the starting point that was provided. The user should have the ability to at least decide if it is good enough, or use it as a starting point into a different optimization run.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (MATH-902)
org.apache.commons.math3.exception.TooManyEvaluationsException should
return best found result
Posted by "Gilles (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MATH-902?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gilles updated MATH-902:
------------------------
Priority: Minor (was: Major)
Affects Version/s: (was: 3.1)
Issue Type: Wish (was: Bug)
Changing the issue type: this is definitely not a bug, perhaps a shortcoming of the chosen API.
It would help focusing the discussion (that happens on the "dev" ML) if you could provide an actual use-case.
> org.apache.commons.math3.exception.TooManyEvaluationsException should return best found result
> ----------------------------------------------------------------------------------------------
>
> Key: MATH-902
> URL: https://issues.apache.org/jira/browse/MATH-902
> Project: Commons Math
> Issue Type: Wish
> Environment: any
> Reporter: Konstantin Berlin
> Priority: Minor
>
> The org.apache.commons.math3.exception.TooManyEvaluationsException should contain within it the best found solution so far. There are numerous examples when the optimization might not have converged to the stopping condition but the minimum point is better than the starting point that was provided. The user should have the ability to at least decide if it is good enough, or use it as a starting point into a different optimization run.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MATH-902)
org.apache.commons.math3.exception.TooManyEvaluationsException should
return best found result
Posted by "Gilles (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MATH-902?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13501397#comment-13501397 ]
Gilles commented on MATH-902:
-----------------------------
New constructors committed in revision 1411807.
Please check that this solves your problem.
> org.apache.commons.math3.exception.TooManyEvaluationsException should return best found result
> ----------------------------------------------------------------------------------------------
>
> Key: MATH-902
> URL: https://issues.apache.org/jira/browse/MATH-902
> Project: Commons Math
> Issue Type: Wish
> Environment: any
> Reporter: Konstantin Berlin
> Assignee: Gilles
> Priority: Minor
> Fix For: 3.1
>
>
> The org.apache.commons.math3.exception.TooManyEvaluationsException should contain within it the best found solution so far. There are numerous examples when the optimization might not have converged to the stopping condition but the minimum point is better than the starting point that was provided. The user should have the ability to at least decide if it is good enough, or use it as a starting point into a different optimization run.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MATH-902)
org.apache.commons.math3.exception.TooManyEvaluationsException should
return best found result
Posted by "Phil Steitz (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MATH-902?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13499511#comment-13499511 ]
Phil Steitz commented on MATH-902:
----------------------------------
I understand the practical objective here, but I don't think the exception class is the right place to address it. Also, this is probably something to discuss on the mailing list. Lets discuss on the list.
> org.apache.commons.math3.exception.TooManyEvaluationsException should return best found result
> ----------------------------------------------------------------------------------------------
>
> Key: MATH-902
> URL: https://issues.apache.org/jira/browse/MATH-902
> Project: Commons Math
> Issue Type: Bug
> Affects Versions: 3.1
> Environment: any
> Reporter: Konstantin Berlin
>
> The org.apache.commons.math3.exception.TooManyEvaluationsException should contain within it the best found solution so far. There are numerous examples when the optimization might not have converged to the stopping condition but the minimum point is better than the starting point that was provided. The user should have the ability to at least decide if it is good enough, or use it as a starting point into a different optimization run.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MATH-902)
org.apache.commons.math3.exception.TooManyEvaluationsException should
return best found result
Posted by "Konstantin Berlin (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MATH-902?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13502111#comment-13502111 ]
Konstantin Berlin commented on MATH-902:
----------------------------------------
It's not clear to me now what takes priority, the maxEvals in the optimize function or the number I give in the Simple...Checker. What if the numbers are equal?
> org.apache.commons.math3.exception.TooManyEvaluationsException should return best found result
> ----------------------------------------------------------------------------------------------
>
> Key: MATH-902
> URL: https://issues.apache.org/jira/browse/MATH-902
> Project: Commons Math
> Issue Type: Wish
> Environment: any
> Reporter: Konstantin Berlin
> Assignee: Gilles
> Priority: Minor
> Fix For: 3.1
>
>
> The org.apache.commons.math3.exception.TooManyEvaluationsException should contain within it the best found solution so far. There are numerous examples when the optimization might not have converged to the stopping condition but the minimum point is better than the starting point that was provided. The user should have the ability to at least decide if it is good enough, or use it as a starting point into a different optimization run.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MATH-902)
org.apache.commons.math3.exception.TooManyEvaluationsException should
return best found result
Posted by "Gilles (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MATH-902?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13502427#comment-13502427 ]
Gilles commented on MATH-902:
-----------------------------
First, please note that in the checker, the {{converged}} method is passed the number of _iterations_ (main loop of the algorithm). In the {{optimize}} method, it is the number of _evaluations_ (calls to the objective function).
In an algorithm where there is one evaluation for each iteration, the checker will return {{true}} when it reaches "maxEval", leading the algorithm to return the current best point (no exception raised).
> org.apache.commons.math3.exception.TooManyEvaluationsException should return best found result
> ----------------------------------------------------------------------------------------------
>
> Key: MATH-902
> URL: https://issues.apache.org/jira/browse/MATH-902
> Project: Commons Math
> Issue Type: Wish
> Environment: any
> Reporter: Konstantin Berlin
> Assignee: Gilles
> Priority: Minor
> Fix For: 3.1
>
>
> The org.apache.commons.math3.exception.TooManyEvaluationsException should contain within it the best found solution so far. There are numerous examples when the optimization might not have converged to the stopping condition but the minimum point is better than the starting point that was provided. The user should have the ability to at least decide if it is good enough, or use it as a starting point into a different optimization run.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MATH-902)
org.apache.commons.math3.exception.TooManyEvaluationsException should
return best found result
Posted by "Gilles (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/MATH-902?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13503384#comment-13503384 ]
Gilles commented on MATH-902:
-----------------------------
Same changes applied in package "o.a.c.m.optimization.univariate" (revision 1413171).
> org.apache.commons.math3.exception.TooManyEvaluationsException should return best found result
> ----------------------------------------------------------------------------------------------
>
> Key: MATH-902
> URL: https://issues.apache.org/jira/browse/MATH-902
> Project: Commons Math
> Issue Type: Wish
> Environment: any
> Reporter: Konstantin Berlin
> Assignee: Gilles
> Priority: Minor
> Fix For: 3.1
>
>
> The org.apache.commons.math3.exception.TooManyEvaluationsException should contain within it the best found solution so far. There are numerous examples when the optimization might not have converged to the stopping condition but the minimum point is better than the starting point that was provided. The user should have the ability to at least decide if it is good enough, or use it as a starting point into a different optimization run.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira