def testSetMaxDepth(self):
maxDepth = 20
decisionTree = DecisionTree()
decisionTree.setMaxDepth(maxDepth)
decisionTree.learnModel(self.X, self.y)
#self.assertTrue(DecisionTree.depth(decisionTree.getClassifier().tree) <= maxDepth+1)
maxDepth = 5
decisionTree = DecisionTree()
decisionTree.setMaxDepth(maxDepth)
decisionTree.learnModel(self.X, self.y)
def testSetWeight(self):
#Try weight = 0 and weight = 1
decisionTree = DecisionTree()
decisionTree.setWeight(0.0)
decisionTree.learnModel(self.X, self.y)
predY = decisionTree.predict(self.X)
self.assertTrue((predY == numpy.zeros(predY.shape[0])).all())
decisionTree.setWeight(1.0)
decisionTree.learnModel(self.X, self.y)
predY = decisionTree.predict(self.X)
self.assertTrue((predY == numpy.ones(predY.shape[0])).all())
def testMinSplit(self):
decisionTree = DecisionTree()
decisionTree.setMinSplit(20)
decisionTree.learnModel(self.X, self.y)
size = decisionTree.getTree().node_count
#orngTree.printTree(decisionTree.getClassifier())
decisionTree.setMinSplit(1)
decisionTree.learnModel(self.X, self.y)
size2 = decisionTree.getTree().node_count
#orngTree.printTree(decisionTree.getClassifier())
self.assertTrue(size < size2)
def testPredict(self):
decisionTree = DecisionTree()
decisionTree.learnModel(self.X, self.y)
predY = decisionTree.predict(self.X)
inds = numpy.random.permutation(self.X.shape[0])
predY2 = decisionTree.predict(self.X[inds, :])
self.assertTrue((predY[inds] == predY2).all())
#Let's test on -1, +1 labels
y2 = (self.y*2)-1
decisionTree.learnModel(self.X, y2)
predY2 = decisionTree.predict(self.X)
self.assertTrue((predY2 == predY*2-1).all())
def testLearnModel(self):
decisionTree = DecisionTree()
decisionTree.learnModel(self.X, self.y)
tree = decisionTree.getClassifier()
