def add_multiclass_classifier(self, classifier):
"""
Add information about an additional weak learner to the queue of classifiers
to be blended.
"""
formatted = classifier
if(formatted.__class__ == numpy.ndarray):
formatted = formatted.tolist()
# TODO: check the size of the input
LPBoostMulticlassClassifier_wrap.add_multiclass_classifier(self, formatted)
def __init__(self, number_of_classes, nu, **kwargs):
self.number_of_classes = number_of_classes
self.nu = nu
self.weight_sharing = kwargs.get("weight_sharing", True)
self.labels = kwargs.get("labels", range(0, self.number_of_classes))
if(self.labels.__class__ == numpy.ndarray):
self.labels = self.labels.tolist()
self.interior_point = kwargs.get("interior_point", False)
self.solver = kwargs.get("solver", "clp")
LPBoostMulticlassClassifier_wrap.__init__(self, self.number_of_classes, self.nu, self.weight_sharing)
self.initialize_boosting(self.labels, self.interior_point, self.solver)
