On Monday, March 15, 2004, at 10:52 AM, Jack Tanner wrote:
"Ken Williams" <[EMAIL PROTECTED]> wrote in message news:[EMAIL PROTECTED]I haven't yet tried it - a patch would be great. What does the failure
look like?
Actually, it looks like all the errors boil down to SMO being at a different
path in the recent Weka release (3.4.1). SMO is now
weka.classifiers.functions.supportVector class SMO. Could you update this?
Oh, just a testing error then. I'll probably just remove the 'weka_classifier' parameter from the tests, since the default NaiveBayes classifier should be found fine. (Right?)
-Ken
