def test_3():
transmat = np.array([[0.2, 0.3, 0.5], [0.4, 0.4, 0.2], [0.8, 0.2, 0.0]])
means = np.array([[0.0], [2.0], [4.0]])
kappas = np.array([[8.0], [8.0], [6.0]])
X = [create_timeseries(means, kappas, transmat) for i in range(20)]
# For each value of various options, create a 3 state HMM and see if it is correct.
for reversible_type in ('mle', 'transpose'):
model = VonMisesHMM(n_states=3, reversible_type=reversible_type, thresh=1e-4, n_iter=30)
model.fit(X)
validate_timeseries(means, kappas, transmat, model, 0.1, 0.5, 0.1)
assert abs(model.fit_logprob_[-1]-model.score(X)) < 0.5
def test_2_state():
transmat = np.array([[0.7, 0.3], [0.4, 0.6]])
means = np.array([[0.0], [2.0]])
kappas = np.array([[4.0], [8.0]])
X = [create_timeseries(means, kappas, transmat) for i in range(10)]
# For each value of various options,
# create a 2 state HMM and see if it is correct.
for reversible_type in ('mle', 'transpose'):
model = VonMisesHMM(n_states=2, reversible_type=reversible_type,
thresh=1e-4, n_iter=30)
model.fit(X)
validate_timeseries(means, kappas, transmat, model, 0.1, 0.5, 0.05)
assert abs(model.fit_logprob_[-1] - model.score(X)) < 0.5
