#3 Numeric array constructed DataSet + train = Bug

[Anonymous]

The following code doesnt work :

>m = RandomArray.random( ( 10, 20 ) )
>l = [ "-1" for x in range( 5 ) ]
>l.extend( [ "1" for x in range( 5 ) ] )

>s = PyML.svm.SVM()

>d0 = PyML.datafunc.DataSet( X = m, L=l )
>s.train( d0 )

>d1 = PyML.datafunc.SparseDataSet( X = m, L=l )
>s.train( d1 )

>d2 = PyML.datafunc.SparseCDataSet( X = m, L=l )
>s.train( d2 )

d1 et d2 work, but d0 doesnt. THe error msg is :

Traceback (most recent call last):
File "testPyML.py", line 95, in ?
s.train( d0 )
File
"/home/cast/progs/lib/python2.3/site-packages/PyML/svm.py",
line 95, in train
Classifier.train(self, data, **args)
File
"/home/cast/progs/lib/python2.3/site-packages/PyML/baseClassifiers.py",
line 99, in train
self.featureKeyDict =
copy.deepcopy(data.featureKeyDict)
AttributeError: 'DataSet' object has no attribute
'featureKeyDict'