Ticket #1426 (closed defect: done)

Opened 2 years ago

Last modified 2 years ago

Persistence of SVM results fails

Reported by: swinkler Owned by: gkronber
Priority: high Milestone: HeuristicLab 3.3.4
Component: Algorithms.DataAnalysis Version: 3.3.4
Keywords: Cc: vdorfer

Description (last modified by swinkler) (diff)

After running a SVM training algorithm and storing the algorithm with results, the loading of the saved algorithm fails. The thrown exception is shown below. To reproduce this problem do the following:

  • Create new SVM algorithm
  • Add a regression problem, for example the demo problem from the GP-SR demo example
  • Start the algorithm
  • After the end of the training process store the algorithm
  • Load the stored file -> exception

Attachments

SVMpersistenceException.jpg Download (81.8 KB) - added by swinkler 2 years ago.

Change History

Changed 2 years ago by swinkler

comment:1 Changed 2 years ago by swinkler

  • Description modified (diff)

comment:2 Changed 2 years ago by gkronber

Fixed with r5690 in data analysis refactoring branch.

comment:3 Changed 2 years ago by gkronber

r5692 merged r5690 from data analysis refactoring branch (see #1418) into trunk to fix persistence problems of SVMs.

comment:4 Changed 2 years ago by gkronber

  • Status changed from new to accepted

comment:5 Changed 2 years ago by gkronber

  • Owner changed from gkronber to swinkler
  • Status changed from accepted to reviewing

comment:6 Changed 2 years ago by swinkler

  • Owner changed from swinkler to gkronber
  • Status changed from reviewing to readytorelease

Saving and loading of SVM algorithms and results works perfectly. Thanks for fixing this.

comment:7 Changed 2 years ago by swagner

  • Status changed from readytorelease to closed
  • Version changed from 3.3.3 to 3.3.4
  • Resolution set to done
Note: See TracTickets for help on using tickets.