Orange Forum • View topic - SVM Evalulation Fails

SVM Evalulation Fails

Report bugs (or imagined bugs).
(Archived/read-only, please use our ticketing system for reporting bugs and their discussion.)
Forum rules
Archived/read-only, please use our ticketing system for reporting bugs and their discussion.

SVM Evalulation Fails

Postby Guest » Thu Aug 27, 2009 23:07

I attempt to train with the Heart Disease data set and predict the same data set by connecting the data file to the SVM module, the svm to the evaluate module as well as the original data set to the evaluate module.

The following error occurs:

Code: Select all
Unhandled exception of type AttributeError occured at 15:04:25:
Traceback:
  File: OWBaseWidget.py, line 573 in processSignals
  Code: handler(value, (widgetFrom, nameFrom, id))
    File: OWPredictions.py, line 304 in setPredictor
    Code: ov = getoutvar(self.predictors.values())
      File: OWPredictions.py, line 282 in getoutvar
      Code: ov = predictors[0].classVar
        File: orngSVM.py, line 134 in __getattr__
        Code: raise AttributeError(name)
          AttributeError: classVar

Postby Ales » Fri Aug 28, 2009 13:03

Fixed. Update from SVN, it should work now.

THANKS!

Postby dragoljub » Wed Sep 02, 2009 1:31

Thanks for the quick fix!

It would be nice to highlight misclassified samples in red when viewing the prediction results!

-Gagi


Return to Bugs



cron