def extend(self, other, patterns = None) :
if self.__class__ != other.__class__ :
raise ValueError, 'datasets should be the same class'
if patterns is None : patterns = range(len(other))
# first check if ids have compatible hash values :
for id in other.featureID :
if (hash(id) in self.featureKeyDict and
id != self.featureID[self.featureKeyDict[hash(id)]]) :
raise ValueError, 'bad hash'
# recompute featureID related stuff
self.featureKey = misc.union(self.featureKey, other.featureKey)
self.featureKey.sort()
self.featureKeyDict.clear()
for i in range(len(self.featureKey)) :
self.featureKeyDict[self.featureKey[i]] = i
featureIDs = misc.union(self.featureID, other.featureID)
self.featureID = [None for i in range(len(self.featureKey))]
for id in featureIDs :
self.featureID[self.featureKeyDict[hash(id)]] = id
self.extendX(other, patterns)
self.labels.extend(other.labels, patterns)
def extend(self, other, patterns=None):
if self.__class__ != other.__class__:
raise ValueError, 'datasets should be the same class'
if patterns is None: patterns = range(len(other))
# first check if ids have compatible hash values :
for id in other.featureID:
if (hash(id) in self.featureKeyDict
and id != self.featureID[self.featureKeyDict[hash(id)]]):
raise ValueError, 'bad hash'
# recompute featureID related stuff
self.featureKey = misc.union(self.featureKey, other.featureKey)
self.featureKey.sort()
self.featureKeyDict.clear()
for i in range(len(self.featureKey)):
self.featureKeyDict[self.featureKey[i]] = i
featureIDs = misc.union(self.featureID, other.featureID)
self.featureID = [None for i in range(len(self.featureKey))]
for id in featureIDs:
self.featureID[self.featureKeyDict[hash(id)]] = id
self.extendX(other, patterns)
self.labels.extend(other.labels, patterns)
def selectFeatures(self, data, **args) :
labels = data.labels
features = []
for k in range(data.labels.numClasses) :
data2 = labels.oneAgainstRest(data, k)
features2 = self.featureSelector.selectFeatures(data2)
features = misc.union(features, features2)
data.attachLabels(labels)
return features
def selectFeatures(self, data, **args):
labels = data.labels
features = []
for k in range(data.labels.numClasses):
data2 = labels.oneAgainstRest(data, k)
features2 = self.featureSelector.selectFeatures(data2)
features = misc.union(features, features2)
data.attachLabels(labels)
return features