def test_kernel_builing(self):
X = (SP.random.rand(5, 10) > .5)*1.0
kernel = utils.estimateKernel(X, scale=False)
small_kernel = utils.estimateKernel(X[:, 0:5], scale=False)
small_kernel_test = utils.update_Kernel(kernel, X[:, 5:], scale=False)
self.assertAlmostEqual((small_kernel -
small_kernel_test).sum(), 0)
def test_normalization_kernel(self):
#SP.random.seed(42)
n = 50
m = 100
X = (SP.random.rand(n, m) > .5)*1.
X_test = (SP.random.rand(10, m) > .5)*1.
K = utils.estimateKernel(X)
y = SP.random.rand(n, 1)
SP.random.seed(1)
mf = MF(kernel=K)
mf.fit(X, y)
results_1 = mf.predict(X_test)
X -= X.mean(axis=0)
X /= X.std(axis=0)
X_test -= X_test.mean(axis=0)
X_test /= X_test.std(axis=0)
SP.random.seed(1)
mf = MF(kernel=K)
mf.fit(X, y)
results_2 = mf.predict(X_test)
self.assertEqual(results_1.sum(), results_2.sum())
def test_normalization_kernel(self):
#SP.random.seed(42)
n = 50
m = 100
X = (SP.random.rand(n, m) > .5) * 1.
X_test = (SP.random.rand(10, m) > .5) * 1.
K = utils.estimateKernel(X)
y = SP.random.rand(n, 1)
SP.random.seed(1)
mf = MF(kernel=K)
mf.fit(X, y)
results_1 = mf.predict(X_test)
X -= X.mean(axis=0)
X /= X.std(axis=0)
X_test -= X_test.mean(axis=0)
X_test /= X_test.std(axis=0)
SP.random.seed(1)
mf = MF(kernel=K)
mf.fit(X, y)
results_2 = mf.predict(X_test)
self.assertEqual(results_1.sum(), results_2.sum())
def test_kernel_builing(self):
X = (SP.random.rand(5, 10) > .5) * 1.0
kernel = utils.estimateKernel(X, scale=False)
small_kernel = utils.estimateKernel(X[:, 0:5], scale=False)
small_kernel_test = utils.update_Kernel(kernel, X[:, 5:], scale=False)
self.assertAlmostEqual((small_kernel - small_kernel_test).sum(), 0)
