def pruneSF(data, widget, minExmplsInLeaf, progress_steps):
print "\t", "Pruning + Saturation Filter:"
#file.flush()
classifier = orngTree.TreeLearner(data, sameMajorityPruning=1, mForPruning=0, storeExamples=1)
print "\t\t", "Classifier complexity:\t", orngTree.countNodes(classifier), "nodes."
#file.flush()
## [noisyA, dataset] = excludePruned(data, classifier, minExmplsInLeaf)
[noisePruned, dataset] = excludePruned(data, classifier, minExmplsInLeaf)
print "\t\t", len(noisePruned), "example(s) were excluded by pruning."
#file.flush()
classifier2 = orngTree.TreeLearner(dataset, sameMajorityPruning=1, mForPruning=0, storeExamples=1)
print "\t\t", "Pruned Classifier complexity:", orngTree.countNodes(classifier2), "nodes. "
#file.flush()
# Saturation filtering
## [noisy_data, filtered_data] = saturation(dataset, "tree")
n = len(data)
#widget.progress = int(len(noisePruned)*1.0/len(data)*100)
widget.progress = int(sum([n-i for i in range(len(noisePruned))])*1.0/progress_steps*100)
widget.save()
print "progress:", widget.progress
#[noiseSF, filtered_data] = saturation(dataset, widget)#, "tree")
noiseSF = saturation(dataset, widget)#, "tree")
#print "\t\t", "Size of filtered dataset:", len(filtered_data)
print "\t\t", "Noisy examples (", len(noiseSF["inds"])+len(noisePruned),"(",len(noisePruned),"pruned,",\
len(noiseSF["inds"]), "SF ))\n"#: (class, id)"
#file.flush()
#noisy_data.sort(meta_id)
#noiseSF.sort()
# Merge both obtained sets of noisy examples
#noisyA.extend(noisy_data)
noisePruned.extend(noiseSF["inds"])
#return noisyA
return {"inds" : sorted(noisePruned), "name" : "PruneSF"}
def prune_sf(data, minExmplsInLeaf, progress_steps, widget=None):
"""Prune Saturation Filter
:param data:
:param minExmplsInLeaf:
:param progress_steps:
:param widget:
:return:
"""
print "\t", "Pruning + Saturation Filter:"
#file.flush()
classifier = orngTree.TreeLearner(data,
sameMajorityPruning=1,
mForPruning=0,
storeExamples=1)
print "\t\t", "Classifier complexity:\t", orngTree.countNodes(
classifier), "nodes."
#file.flush()
## [noisyA, dataset] = exclude_pruned(data, classifier, minExmplsInLeaf)
[noisePruned, dataset] = exclude_pruned(data, classifier, minExmplsInLeaf)
print "\t\t", len(noisePruned), "example(s) were excluded by pruning."
#file.flush()
classifier2 = orngTree.TreeLearner(dataset,
sameMajorityPruning=1,
mForPruning=0,
storeExamples=1)
print "\t\t", "Pruned Classifier complexity:", orngTree.countNodes(
classifier2), "nodes. "
#file.flush()
# Saturation filtering
## [noisy_data, filtered_data] = saturation(dataset, "tree")
n = len(data)
#widget.progress = int(len(noisePruned)*1.0/len(data)*100)
if not (widget == None):
widget.progress = int(
sum([n - i for i in range(len(noisePruned))]) * 1.0 /
progress_steps * 100)
widget.save()
print "progress:", widget.progress
#[noiseSF, filtered_data] = saturation(dataset, widget)#, "tree")
noiseSF = saturation(dataset, widget) #, "tree")
#print "\t\t", "Size of filtered dataset:", len(filtered_data)
print "\t\t", "Noisy examples (", len(noiseSF["inds"])+len(noisePruned),"(",len(noisePruned),"pruned,",\
len(noiseSF["inds"]), "SF ))\n"#: (class, id)"
#file.flush()
#noisy_data.sort(meta_id)
#noiseSF.sort()
# Merge both obtained sets of noisy examples
#noisyA.extend(noisy_data)
noisePruned.extend(noiseSF["inds"])
#return noisyA
return {"inds": sorted(noisePruned), "name": "PruneSF"}
def testMinSplit(self):
decisionTree = DecisionTree()
decisionTree.setMinSplit(20)
decisionTree.learnModel(self.X, self.y)
size = orngTree.countNodes(decisionTree.getClassifier())
#orngTree.printTree(decisionTree.getClassifier())
decisionTree.setMinSplit(0)
decisionTree.learnModel(self.X, self.y)
size2 = orngTree.countNodes(decisionTree.getClassifier())
#orngTree.printTree(decisionTree.getClassifier())
self.assertTrue(size < size2)
def testMinSplit(self):
decisionTree = DecisionTree()
decisionTree.setMinSplit(20)
decisionTree.learnModel(self.X, self.y)
size = orngTree.countNodes(decisionTree.getClassifier())
#orngTree.printTree(decisionTree.getClassifier())
decisionTree.setMinSplit(0)
decisionTree.learnModel(self.X, self.y)
size2 = orngTree.countNodes(decisionTree.getClassifier())
#orngTree.printTree(decisionTree.getClassifier())
self.assertTrue(size < size2)
def setClassificationTree(self, tree):
self.closeContext()
if tree and (not tree.classVar
or tree.classVar.varType != orange.VarTypes.Discrete):
self.error(
"This viewer only shows trees with discrete classes.\nThere is another viewer for regression trees"
)
self.tree = None
else:
self.error()
self.tree = tree
self.setTreeView()
self.sliderChanged()
self.targetCombo.clear()
if tree:
self.treeNodes, self.treeLeaves = orngTree.countNodes(
tree), orngTree.countLeaves(tree)
self.infoa.setText('Number of nodes: %i' % self.treeNodes)
self.infob.setText('Number of leaves: %i' % self.treeLeaves)
self.targetCombo.addItems(
[name for name in tree.tree.examples.domain.classVar.values])
self.targetClass = 0
self.openContext("", tree.domain)
else:
self.treeNodes = self.treeLeaves = 0
self.infoa.setText('No tree on input.')
self.infob.setText('')
self.openContext("", None)
def ctree(self, tree=None):
self.clear()
if not tree:
self.centerRootButton.setDisabled(1)
self.centerNodeButton.setDisabled(0)
self.infoa.setText('No tree.')
self.infob.setText('')
self.tree=None
self.rootNode = None
else:
self.tree=tree.tree
self.infoa.setText('Number of nodes: ' + str(orngTree.countNodes(tree)))
self.infob.setText('Number of leaves: ' + str(orngTree.countLeaves(tree)))
if hasattr(self.scene, "colorPalette"):
self.scene.colorPalette.setNumberOfColors(len(self.tree.distribution))
# self.scene.setDataModel(GraphicsTree(self.tree))
self.rootNode=self.walkcreate(self.tree, None)
# self.scene.addItem(self.rootNode)
self.scene.fixPos(self.rootNode,self.HSpacing,self.VSpacing)
self.activateLoadedSettings()
self.sceneView.centerOn(self.rootNode.x(), self.rootNode.y())
self.updateNodeToolTips()
self.centerRootButton.setDisabled(0)
self.centerNodeButton.setDisabled(1)
self.scene.update()
def ctree(self, tree=None):
self.clear()
if not tree:
self.centerRootButton.setDisabled(1)
self.centerNodeButton.setDisabled(0)
self.infoa.setText('No tree.')
self.infob.setText('')
self.tree = None
self.rootNode = None
else:
self.tree = tree.tree
self.infoa.setText('Number of nodes: ' +
str(orngTree.countNodes(tree)))
self.infob.setText('Number of leaves: ' +
str(orngTree.countLeaves(tree)))
if hasattr(self.scene, "colorPalette"):
self.scene.colorPalette.setNumberOfColors(
len(self.tree.distribution))
# self.scene.setDataModel(GraphicsTree(self.tree))
self.rootNode = self.walkcreate(self.tree, None)
# self.scene.addItem(self.rootNode)
self.scene.fixPos(self.rootNode, self.HSpacing, self.VSpacing)
self.activateLoadedSettings()
self.sceneView.centerOn(self.rootNode.x(), self.rootNode.y())
self.updateNodeToolTips()
self.centerRootButton.setDisabled(0)
self.centerNodeButton.setDisabled(1)
self.scene.update()
def setClassificationTree(self, tree):
self.closeContext()
if tree and (not tree.classVar or tree.classVar.varType != orange.VarTypes.Discrete):
self.error("This viewer only shows trees with discrete classes.\nThere is another viewer for regression trees")
self.tree = None
else:
self.error()
self.tree = tree
self.setTreeView()
self.sliderChanged()
self.targetCombo.clear()
if tree:
self.treeNodes, self.treeLeaves = orngTree.countNodes(tree), orngTree.countLeaves(tree)
self.infoa.setText('Number of nodes: %i' % self.treeNodes)
self.infob.setText('Number of leaves: %i' % self.treeLeaves)
self.targetCombo.addItems([name for name in tree.tree.examples.domain.classVar.values])
self.targetClass = 0
self.openContext("", tree.domain)
else:
self.treeNodes = self.treeLeaves = 0
self.infoa.setText('No tree on input.')
self.infob.setText('')
self.openContext("", None)
def updateTree(self):
self.setTreeView()
self.learner = FixedTreeLearner(self.tree, self.captionTitle)
self.infoa.setText("Number of nodes: %i" % orngTree.countNodes(self.tree))
self.infob.setText("Number of leaves: %i" % orngTree.countLeaves(self.tree))
# self.send("Data", self.tree)
self.send("Classifier", self.tree)
self.send("Tree Learner", self.learner)
def testMinSplit(self):
randomForest = RandomForest()
randomForest.setMinSplit(20)
randomForest.learnModel(self.X, self.y)
size = numpy.zeros(100)
i = 0
for c in randomForest.getClassifier().classifiers:
size[i] = orngTree.countNodes(c)
i += 1
size = numpy.mean(size)
#orngTree.printTree(randomForest.getClassifier())
randomForest.setMinSplit(0)
randomForest.learnModel(self.X, self.y)
size2 = numpy.zeros(100)
i = 0
for c in randomForest.getClassifier().classifiers:
size2[i] = orngTree.countNodes(c)
i += 1
size2 = numpy.mean(size2)
self.assertTrue(size < size2)
def testMinSplit(self):
randomForest = RandomForest()
randomForest.setMinSplit(20)
randomForest.learnModel(self.X, self.y)
size = numpy.zeros(100)
i = 0
for c in randomForest.getClassifier().classifiers:
size[i] = orngTree.countNodes(c)
i += 1
size = numpy.mean(size)
#orngTree.printTree(randomForest.getClassifier())
randomForest.setMinSplit(0)
randomForest.learnModel(self.X, self.y)
size2 = numpy.zeros(100)
i = 0
for c in randomForest.getClassifier().classifiers:
size2[i] = orngTree.countNodes(c)
i += 1
size2 = numpy.mean(size2)
self.assertTrue(size < size2)
def cmplx(set):
classifier = orngTree.TreeLearner(set, sameMajorityPruning=1, mForPruning=0)
return orngTree.countNodes(classifier)
def sendReport(self):
self.reportData(self.data)
self.treeNodes, self.treeLeaves = orngTree.countNodes(
self.tree), orngTree.countLeaves(self.tree)
super(OWITree, self).sendReport()
def sendReport(self):
self.reportData(self.data)
self.treeNodes, self.treeLeaves = orngTree.countNodes(self.tree), orngTree.countLeaves(self.tree)
super(OWITree, self).sendReport()
# Description: Defines a tree learner (trunks of depth less than 5) and uses them in forest tree, prints out the number of nodes in each tree
# Category: classification, ensembles
# Classes: RandomForestLearner
# Uses: bupa.tab
# Referenced: orngEnsemble.htm
import orange, orngTree, orngEnsemble
data = orange.ExampleTable('bupa.tab')
tree = orngTree.TreeLearner(storeNodeClassifier = 0, storeContingencies=0, \
storeDistributions=1, minExamples=5, ).instance()
gini = orange.MeasureAttribute_gini()
tree.split.discreteSplitConstructor.measure = \
tree.split.continuousSplitConstructor.measure = gini
tree.maxDepth = 5
tree.split = orngEnsemble.SplitConstructor_AttributeSubset(tree.split, 3)
forestLearner = orngEnsemble.RandomForestLearner(learner=tree, trees=50)
forest = forestLearner(data)
for c in forest.classifiers:
print orngTree.countNodes(c),
print
def prune(self):
for action in self.actionList:
self.feaList[action], self.DynamicLearner[action].treeList = self.DynamicLearner[action].prune(self.feaList[action])
print "action learner ", action, " nodes: ", orngTree.countNodes(self.DynamicLearner[action].treeList[0])
print "action learner leaves: ", self.DynamicLearner[action].treeList[0].count_leaves()
def cmplx(set):
classifier = orngTree.TreeLearner(set,
sameMajorityPruning=1,
mForPruning=0)
return orngTree.countNodes(classifier)
