def test_compute_count_matrix_list(self):
"""Test computation of count matrix for a list of trajectories"""
check_mat = np.array([[0, 14, 0], [0, 0, 12], [12, 0, 0]])
chain = np.array([0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1])
count_mat = est.compute_count_matrix_list([chain, chain])
self.assertTrue(np.allclose(count_mat, check_mat, atol=1e-5))
def test_compute_count_matrix_list(self):
"""Test computation of count matrix for a list of trajectories"""
check_mat = np.array([[0, 14, 0], [0, 0, 12], [12, 0, 0]])
chain = np.array(
[0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1])
count_mat = est.compute_count_matrix_list([chain, chain])
self.assertTrue(np.allclose(count_mat, check_mat, atol=1e-5))
import numpy as np
import estimation.estimation as est
import naive_sampling as nsampl
trans_mat = np.array([[0.5, 0.25, 0.25], [0.5, 0.0, 0.5], [0.25, 0.25, 0.5]],
dtype=np.float64)
# trans_mat = np.array([[0.25, 0.75], [0.75, 0.25]], dtype=np.float64)
print trans_mat
chain = nsampl.evolve_chain(0, trans_mat, 20000)
# chain = np.array([0, 0, 1, 0, 1, 0, 1, 0, 1, 0, 1, 1, 1, 0, 1, 0, 1, 0, 0, 0])
print chain
chain_list = [chain, chain]
# comoute count matrix
c_mat = est.compute_count_matrix(chain, i_tau=1)
c_mat2 = est.compute_count_matrix_list(chain_list)
# estimate transition matrix
t_mat = est.estimate_transition_matrix(c_mat, 100000, 1e-3)
reversible_check = est.check_reversibility(t_mat)
print c_mat
print t_mat
print np.sum(t_mat, axis=1)
print np.linalg.norm(t_mat - trans_mat, ord='fro') / np.linalg.norm(trans_mat,
ord='fro')
print reversible_check