def __init__(self, inp, target, nb_classes=0, class_labels=None):
"""Initialize an empty dataset.
`inp` is used to specify the dimensionality of the input. While the
number of targets is given by implicitly by the training samples, it can
also be set explicity by `nb_classes`. To give the classes names, supply
an iterable of strings as `class_labels`."""
# FIXME: hard to keep nClasses synchronized if appendLinked() etc. is used.
SequentialDataSet.__init__(self, inp, target)
# we want integer class numbers as targets
self.convertField('target', int)
if len(self) > 0:
# calculate class histogram, if we already have data
self.calculateStatistics()
self.nClasses = nb_classes
self.class_labels = range(self.nClasses) if class_labels is None else class_labels
# copy classes (targets may be changed into other representation)
self.setField('class', self.getField('target'))
def __init__(self, indim, targetdim):
SequentialDataSet.__init__(self, indim, targetdim)
self.addField('importance', targetdim)
self.link.append('importance')
def removeSequence(self, index):
"""Remove sequence (including class field) from the dataset."""
self.assignClasses()
self.linkFields(['input', 'target', 'class'])
SequentialDataSet.removeSequence(self, index)
self.unlinkFields(['class'])
