def test_default_generator():
step_gen = nd.MaxStepGenerator(num_steps=10)
h = np.array([h for h in step_gen(0)])
desired = 2.0 * 2.0**(-np.arange(10) + 0)
assert_array_almost_equal((h - desired) / desired, 0)
def test_max_step_generator_with_base_step01():
desired = 0.1
step_gen = nd.MaxStepGenerator(base_step=desired, num_steps=1, offset=0)
methods = ['forward', 'backward', 'central', 'complex']
lengths = [2, 2, 1, 1]
for n, method in zip(lengths, methods):
h = [h for h in step_gen(0, method=method)]
assert_array_almost_equal((h[0] - desired) / desired, 0)
assert_equal(n, len(h))
def test_run_hamiltonian(self):
# Important to restrict the step in order to avoid the
# discontinutiy at x=[0,0] of the hamiltonian
for method in ['central', 'complex']:
step = nd.MaxStepGenerator(base_step=1e-4)
hessian = nd.Hessian(None, step=step, method=method)
h, _error_estimate, true_h = run_hamiltonian(hessian,
verbose=False)
assert (np.abs((h - true_h) / true_h) < 1e-4).all()
def test_default_base_step():
step_gen = nd.MaxStepGenerator(num_steps=1, offset=0)
h = [h for h in step_gen(0)]
desired = 2.0
assert_array_almost_equal((h[0] - desired) / desired, 0)