def testSetTau(self): tau = 0.0 tol = 10 - 6 kernel = LinearKernel() predictor = PrimalDualCCARegression(kernel, tau, tau) self.assertEquals(predictor.getTau1(), tau) predictor.setTau1(0.1) self.assertEquals(predictor.getTau1(), 0.1)
def testPredict2(self): # Test predicting on low-rank matrices numExamples = 10 numFeatures = 5 X = numpy.random.rand(numExamples, numFeatures) Y = numpy.random.rand(numExamples, numFeatures) tau = 0.0 tol = 10 - 6 kernel = LinearKernel() predictor = PrimalDualCCARegression(kernel, tau, tau) A = predictor.learnModel(X, Y) predY = predictor.predict(X)
def testPredict2(self): #Test predicting on low-rank matrices numExamples = 10 numFeatures = 5 X = numpy.random.rand(numExamples, numFeatures) Y = numpy.random.rand(numExamples, numFeatures) tau = 0.0 tol = 10 - 6 kernel = LinearKernel() predictor = PrimalDualCCARegression(kernel, tau, tau) A = predictor.learnModel(X, Y) predY = predictor.predict(X)
def testPredict(self): numExamples = 10 numFeatures = 10 X = numpy.random.rand(numExamples, numFeatures) Y = X tau = 0.0 tol = 10 - 6 kernel = LinearKernel() predictor = PrimalDualCCARegression(kernel, tau, tau) A = predictor.learnModel(X, Y) testX = X[0:5, :] predY = predictor.predict(testX) self.assertTrue(numpy.linalg.norm(predY - Y[0:5, :]) < tol)
def testLearnModel(self): numExamples = 10 numFeatures = 10 X = numpy.random.rand(numExamples, numFeatures) Y = X tau = 0.0 tol = 10 - 6 kernel = LinearKernel() predictor = PrimalDualCCARegression(kernel, tau, tau) A = predictor.learnModel(X, Y) self.assertTrue(numpy.linalg.norm(numpy.dot(numpy.dot(X, X.T), A) - Y) < tol) self.assertTrue(numpy.linalg.norm(predictor.predict(X) - Y) < tol)
def testLearnModel(self): numExamples = 10 numFeatures = 10 X = numpy.random.rand(numExamples, numFeatures) Y = X tau = 0.0 tol = 10 - 6 kernel = LinearKernel() predictor = PrimalDualCCARegression(kernel, tau, tau) A = predictor.learnModel(X, Y) self.assertTrue( numpy.linalg.norm(numpy.dot(numpy.dot(X, X.T), A) - Y) < tol) self.assertTrue(numpy.linalg.norm(predictor.predict(X) - Y) < tol)