def testSaveLearnedKnowledge(self):
testValues = [-0.0310651210442,
-0.618841896127,
0.649230972775,
0.649230972775,
-0.618841896127]
alpha = DataVector(len(testValues))
for i in xrange(len(testValues)):
alpha[i] = testValues[i]
learnedKnowledge = LearnedKnowledge()
learnedKnowledge.update(alpha)
controller = CheckpointController("saveknowledge", pathlocal)
controller.setLearnedKnowledge(learnedKnowledge)
controller.saveLearnedKnowledge(0)
sampleLines = list()
f = gzip.open(pathlocal + "/saveknowledge.0.arff.gz", "r")
try:
for line in f.readlines():
if len(line)>1 and "@" not in line:
sampleLines.append(float(line))
finally:
f.close()
self.assertEqual(testValues, [float(i) for i in sampleLines])
def loadLearnedKnowledge(self, iteration):
knowledgeFile = self.composeName(iteration) + ".arff.gz"
knowledgeMemento = LearnedKnowledgeFormatter().deserializeFromFile(
knowledgeFile)
knowledge = LearnedKnowledge()
knowledge.setMemento(knowledgeMemento)
return knowledge
def setUp(self):
level = 2
dim = 2
l = 0.00001
self.classifier = Classifier()
dataContainer = ARFFAdapter(
pathlocal + "/datasets/classifier.train.arff").loadData()
self.classifier.setDataContainer(dataContainer)
foldingPolicy = FoldingPolicy(dataContainer)
self.classifier.setFoldingPolicy(foldingPolicy)
grid = Grid.createLinearGrid(dim)
storage = grid.createGridGenerator()
storage.regular(level)
self.classifier.setGrid(grid)
self.classifier.setLearnedKnowledge(LearnedKnowledge())
spec = TrainingSpecification()
spec.setL(l)
spec.setCOperator(createOperationLaplace(grid))
self.classifier.setSpecification(spec)
stopPolicy = TrainingStopPolicy()
stopPolicy.setAdaptiveIterationLimit(0)
self.classifier.setStopPolicy(stopPolicy)
solver = CGSolver()
#solver.attachEventController(InfoToScreen())
solver.setImax(500)
solver.setReuse(True)
self.classifier.setSolver(solver)
def testLearnDataWithFolding(self, ):
correct = [
0.6612903226, 0.1428571429, 0.5741935484, 0.9142857143,
0.6193548387, 0.5142857143, 0.5870967742, 0.7714285714,
0.6032258065, 0.5714285714, 0.6387096774, 0.4000000000,
0.5935483871, 0.7428571429, 0.6193548387, 0.5142857143,
0.5903225806, 0.7714285714, 0.6063492063, 0.5666666667
]
level = 2
dim = 6
l = 0.00001
self.classifier = Classifier()
dataContainer = ARFFAdapter(
pathlocal +
"/../../../datasets/liver/liver-disorders_normalized.arff"
).loadData()
self.classifier.setDataContainer(dataContainer)
foldingPolicy = SequentialFoldingPolicy(dataContainer, 10)
self.classifier.setFoldingPolicy(foldingPolicy)
grid = Grid.createLinearGrid(dim)
storage = grid.getGenerator()
storage.regular(level)
self.classifier.setGrid(grid)
self.classifier.setLearnedKnowledge(LearnedKnowledge())
spec = TrainingSpecification()
spec.setL(l)
spec.setCOperator(createOperationLaplace(grid))
spec.setCOperatorType('laplace')
self.classifier.setSpecification(spec)
stopPolicy = TrainingStopPolicy()
stopPolicy.setAdaptiveIterationLimit(0)
self.classifier.setStopPolicy(stopPolicy)
self.classifier.setSolver(CGSolver())
self.classifier.learnDataWithFolding()
for i in range(10):
self.assertAlmostEqual(correct[2 * i],
self.classifier.trainAccuracy[i])
self.assertAlmostEqual(correct[2 * i + 1],
self.classifier.testAccuracy[i])
def loadLearnedKnowledge(self, iteration):
knowledgeFile = self.composeName(iteration) + ".arff.gz"
knowledgeMemento = LearnedKnowledgeFormatter().deserializeFromFile(knowledgeFile)
knowledge = LearnedKnowledge()
knowledge.setMemento(knowledgeMemento)
return knowledge