def test_run(self):
# Tests complete run.
# Create log pdf
log_pdf = pints.toy.MultimodalGaussianLogPDF(
modes=[[0, 2], [0, 7], [5, 0], [4, 4]])
# Create non-adaptive mcmc
x0 = np.array([1, 1])
mcmc = pints.SliceRankShrinkingMCMC(x0)
mcmc.set_sigma_c(3)
# Run multiple iterations of the sampler
chain = []
while len(chain) < 100:
x = mcmc.ask()
fx, grad = log_pdf.evaluateS1(x)
reply = mcmc.tell((fx, grad))
if reply is not None:
y, fy, ac = reply
chain.append(y)
self.assertTrue(np.all(x == y))
self.assertEqual(fx, fy[0])
self.assertTrue(np.all(grad == fy[1]))
self.assertEqual(np.shape(chain), (100, 2))
def test_first_run(self):
# Tests the very first run of the sampler.
# Create log pdf
log_pdf = pints.toy.GaussianLogPDF([2, 4], [[1, 0], [0, 3]])
# Create mcmc
x0 = np.array([2., 4.])
mcmc = pints.SliceRankShrinkingMCMC(x0)
# Ask() should fail if _ready_for_tell flag is True
x = mcmc.ask()
with self.assertRaises(RuntimeError):
mcmc.ask()
# Check whether first iteration of ask() returns x0
self.assertTrue(np.all(x == x0))
# Calculate log pdf for x0
fx, grad = log_pdf.evaluateS1(x0)
# Test first iteration of tell(). The first point in the chain
# should be x0
sample = mcmc.tell((fx, grad))
self.assertTrue(np.all(sample == x0))
# We update the _current_log_pdf value used to generate the new slice
self.assertEqual(mcmc.current_log_pdf(), fx)
# Check that the new slice has been constructed appropriately
self.assertTrue(mcmc.current_slice_height() < mcmc.current_log_pdf())
# Tell() should fail when fx is infinite
mcmc = pints.SliceRankShrinkingMCMC(x0)
mcmc.ask()
with self.assertRaises(ValueError):
fx = np.inf
mcmc.tell((fx, grad))
# Tell() should fail when _ready_for_tell is False
with self.assertRaises(RuntimeError):
mcmc.tell((fx, grad))
# Test sensitivities
self.assertTrue(mcmc.needs_sensitivities())
def test_basic(self):
# Test basic methods of the class.
# Create mcmc
x0 = np.array([1, 1])
mcmc = pints.SliceRankShrinkingMCMC(x0)
# Test name
self.assertEqual(
mcmc.name(),
'Slice Sampling - Covariance-Adaptive: Rank Shrinking.')
# Test set_sigma_c(), sigma_c()
mcmc.set_sigma_c(6)
self.assertEqual(mcmc.sigma_c(), 6)
with self.assertRaises(ValueError):
mcmc.set_sigma_c(-1)
# Test number of hyperparameters
self.assertEqual(mcmc.n_hyper_parameters(), 1)
# Test setting hyperparameters
mcmc.set_hyper_parameters([33])
self.assertEqual(mcmc.sigma_c(), 33)