def test_tpm(self):
"""Checking transition probability matrix."""
tfm = np.array([[0, 3, 1], [1, 0, 4], [3, 2, 0]], dtype=np.float)
result = tfm / tfm.sum(axis=1)[:, None]
a = mc.MarkovChain().from_data(self.seq)
self.assertTrue(np.allclose(a.observed_p_matrix, result))
def test_mcsim(self):
'''Checking simulation process'''
a = mc.MarkovChain().from_data(self.seq)
si, sn = a.simulate(15, start=0, ret='both', seed=11)
si2 = a.simulate(15, start=0, ret='indices', seed=11)
sn2 = a.simulate(15, start=0, ret='states', seed=11)
self.assertTrue(np.allclose(si, si2) and np.all(sn == sn2))
def test_mcsim(self):
"""Checking simulation process."""
a = mc.MarkovChain().from_data(self.seq)
seed = np.random.randint(0, 15, 15)
si, sn = a.simulate(15, start=0, ret='both', seed=seed)
si2 = a.simulate(15,
tf=a.observed_matrix,
start=0,
ret='indices',
seed=seed)
sn2 = a.simulate(15,
start=0,
states=['A', 'B', 'C'],
ret='states',
seed=seed)
si3, sn3 = a.simulate(15, start='A', ret='both', seed=seed)
si4, sn4 = a.simulate(15, start='B', ret='both', seed=seed)
si5, sn5 = a.simulate(15, ret='both', seed=seed)
self.assertTrue(np.allclose(si, si2) and np.all(sn == sn2))
self.assertTrue(np.allclose(si, si3) and np.all(sn == sn3))
self.assertTrue(np.allclose(si4, si5) and np.all(sn4 == sn5))
def test_mcgraph(self):
"""Checking graphing process for MC."""
a = mc.MarkovChain().from_data(self.seq)
graph = a.graph_make()
self.assertTrue(isinstance(graph, Digraph))
def test_nom(self):
"""Checking raising a matrix to power N."""
a = mc.MarkovChain().from_data(self.seq)
a2 = a.n_order_matrix()
self.assertTrue(
np.allclose(a2, a.observed_p_matrix @ a.observed_p_matrix))
def test_prf(self):
"""Checking prob_to_freq method."""
a = mc.MarkovChain().from_data(self.seq)
self.assertTrue(np.allclose(a.prob_to_freq_matrix(),
a.observed_matrix))
def test_tfm(self):
"""Checking transition frequency matrix."""
result = np.array([[0, 3, 1], [1, 0, 4], [3, 2, 0]])
a = mc.MarkovChain().from_data(self.seq)
self.assertTrue(np.allclose(a.observed_matrix, result))
