def test_train(self):
stm = STM(8, 4, 4, 10)
parameters = stm._parameters()
stm.train(
randint(2, size=[stm.dim_in, 2000]),
randint(2, size=[stm.dim_out, 2000]),
parameters={
'verbosity': 0,
'max_iter': 0,
})
# parameters should not have changed
self.assertLess(max(abs(stm._parameters() - parameters)), 1e-20)
def callback(i, stm):
callback.counter += 1
return
callback.counter = 0
max_iter = 10
stm.train(
randint(2, size=[stm.dim_in, 10000]),
randint(2, size=[stm.dim_out, 10000]),
parameters={
'verbosity': 0,
'max_iter': max_iter,
'threshold': 0.,
'batch_size': 1999,
'callback': callback,
'cb_iter': 2,
})
self.assertEqual(callback.counter, max_iter / 2)
# test zero-dimensional nonlinear inputs
stm = STM(0, 5, 5)
glm = GLM(stm.dim_in_linear, LogisticFunction, Bernoulli)
glm.weights = randn(*glm.weights.shape)
input = randn(stm.dim_in_linear, 10000)
output = glm.sample(input)
stm.train(input, output, parameters={'max_iter': 20})
# STM should be able to learn GLM behavior
self.assertAlmostEqual(glm.evaluate(input, output), stm.evaluate(input, output), 1)
# test zero-dimensional inputs
stm = STM(0, 0, 10)
input = empty([0, 10000])
output = rand(1, 10000) < 0.35
stm.train(input, output)
self.assertLess(abs(mean(stm.sample(input)) - mean(output)), 0.1)
def test_sample(self): q = 0.92 N = 10000 stm = STM(0, 0, 1, 1) stm.biases = [log(q / (1. - q))] x = mean(stm.sample(empty([0, N]))) - q p = 2. - 2. * norm.cdf(abs(x), scale=sqrt(q * (1. - q) / N)) # should fail in about 1/1000 tests, but not more self.assertGreater(p, 0.0001)
def test_basics(self): dim_in_nonlinear = 10 dim_in_linear = 8 num_components = 7 num_features = 50 num_samples = 100 # create model stm = STM(dim_in_nonlinear, dim_in_linear, num_components, num_features) # generate output input_nonlinear = randint(2, size=[dim_in_nonlinear, num_samples]) input_linear = randint(2, size=[dim_in_linear, num_samples]) input = vstack([input_nonlinear, input_linear]) output = stm.sample(input) loglik = stm.loglikelihood(input, output) # check hyperparameters self.assertEqual(stm.dim_in, dim_in_linear + dim_in_nonlinear) self.assertEqual(stm.dim_in_linear, dim_in_linear) self.assertEqual(stm.dim_in_nonlinear, dim_in_nonlinear) self.assertEqual(stm.num_components, num_components) self.assertEqual(stm.num_features, num_features) # check parameters self.assertEqual(stm.biases.shape[0], num_components) self.assertEqual(stm.biases.shape[1], 1) self.assertEqual(stm.weights.shape[0], num_components) self.assertEqual(stm.weights.shape[1], num_features) self.assertEqual(stm.features.shape[0], dim_in_nonlinear) self.assertEqual(stm.features.shape[1], num_features) self.assertEqual(stm.predictors.shape[0], num_components) self.assertEqual(stm.predictors.shape[1], dim_in_nonlinear) self.assertEqual(stm.linear_predictor.shape[0], dim_in_linear) self.assertEqual(stm.linear_predictor.shape[1], 1) # check dimensionality of output self.assertEqual(output.shape[0], 1) self.assertEqual(output.shape[1], num_samples) self.assertEqual(loglik.shape[0], 1) self.assertEqual(loglik.shape[1], num_samples)
