def test_rho_simple():
K = 2
sequences = [
'tga',
]
m = create_model( K, sequences, psi, theta, phi, upsilon )
v = m.var_dist
v.eta = numpy.array( # a pssm that looks like G[G]A
[
[ 0.25, 0.25, 0.25, 0.25 ],
[ 0.01, 0.01, 0.97, 0.01 ],
[ 0.01, 0.01, 0.97, 0.01 ],
[ 0.97, 0.01, 0.01, 0.01 ]
]
)
print 'Updating'
for i in xrange( 100 ):
print m.update(
update_rho = True,
update_eta = False,
update_tau = False
),
print "\n".join( str( r ) for r in v.r_mode )
hmm.write_model_svg(
m,
name = 'rho_simple',
dir = 'test',
show_rev_comp = True,
show_dists = False,
edge_lengths = 3.0
)
return m, v
def test_simple():
K = 2
sequences = [
'tg',
'tg',
'tag',
'tag',
]
print theta
m = create_model(K, sequences, psi, theta, phi, upsilon)
#hmm.write_model_svg(
# m,
# name = '%d' % K,
# dir = 'test',
# show_rev_comp = True,
# show_dists = False,
# edge_lengths = 3.0
#)
#raise ""
v = m.var_dist
v.eta = numpy.array( # a pssm that looks like T[A]G
[[0.25, 0.25, 0.25, 0.25], [0.01, 0.01, 0.01, 0.97],
[0.97, 0.01, 0.01, 0.01], [0.01, 0.01, 0.97, 0.01]])
print 'Updating'
for i in xrange(100):
m.update(update_rho=True, update_eta=False, update_tau=True)
print "\n".join(str(r) for r in m.var_dist.r_mode)
hmm.write_model_svg(m,
name='simple',
dir='test',
show_rev_comp=True,
show_dists=True,
edge_lengths=3.0)
return m, v
def test_tau_simple():
K = 1
sequences = [
'tg',
]
m = create_model( K, sequences, psi, theta, phi, upsilon )
v = m.var_dist
v.rho = numpy.array(
[
[
[ .98, .01, .01 ],
[ .01, .98, .01 ]
]
]
)
print 'Updating'
for i in xrange( 100 ):
m.update(
update_rho = False,
update_eta = True,
update_tau = True
)
hmm.write_model_svg(
m,
name = 'tau_simple',
dir = 'test',
show_rev_comp = True,
show_dists = False,
edge_lengths = 3.0
)
return m, v
def test_tau_gapped():
K = 2
sequences = [
'tag',
'tg',
]
m = create_model(K, sequences, psi, theta, phi, upsilon)
v = m.var_dist
v.rho = [
numpy.array([
[.01, .94, .01, .01, .01, .01, .01],
[.01, .01, .94, .01, .01, .01, .01],
[.01, .01, .01, .94, .01, .01, .01],
]),
numpy.array([
[.01, .94, .01, .01, .01, .01, .01],
[.01, .01, .01, .94, .01, .01, .01],
])
]
print 'Updating'
for i in xrange(100):
m.update(update_rho=False, update_eta=True, update_tau=True)
hmm.write_model_svg(m,
name='tau_gapped',
dir='test',
show_rev_comp=True,
show_dists=False,
edge_lengths=3.0)
return m, v
def test_eta_gapped():
"Test update eta with gapped 2 base pssm"
K = 2
sequences = [
'tag', # 'a' is the gap
#'tg',
]
m = create_model(K, sequences, psi, theta, phi, upsilon)
v = m.var_dist
v.rho = [
numpy.array([
[.01, .94, .01, .01, .01, .01, .01],
[.01, .01, .94, .01, .01, .01, .01],
[.01, .01, .01, .94, .01, .01, .01],
]),
# numpy.array(
# [
# [ .01, .94, .01, .01, .01, .01, .01 ],
# [ .01, .01, .01, .94, .01, .01, .01 ],
# ]
# )
]
for i in xrange(100):
m.update(update_rho=False, update_eta=True, update_tau=False)
print 'Updating'
hmm.write_model_svg(m,
name='eta_gapped',
dir='test',
show_rev_comp=True,
show_dists=True,
edge_lengths=3.0)
return m, v
def test_write_models():
for k in xrange( 10 ):
model = hmm.Model(
hmm.ObservedData(
k + 1,
hmm.ObservedSequences( sequences )
)
)
hmm.write_model_svg(
model,
'states_%d' % (k + 1),
show_rev_comp = False
)
def test_transitions():
"Test that transition probs are set correctly. Can check by inspecting svg."
K = 3
sequences = [
'tgacg',
]
m = create_model(K, sequences, psi, theta, phi, upsilon)
v = m.var_dist
v.tau = numpy.array([[.90, .10], [.50, .50], [.80, .20]])
hmm.write_model_svg(m,
name='transitions',
dir='test',
show_rev_comp=True,
show_dists=False,
edge_lengths=3.0)
return m, v
def test_simple():
K = 2
sequences = [
'tg',
'tg',
'tag',
'tag',
]
print theta
m = create_model( K, sequences, psi, theta, phi, upsilon )
#hmm.write_model_svg(
# m,
# name = '%d' % K,
# dir = 'test',
# show_rev_comp = True,
# show_dists = False,
# edge_lengths = 3.0
#)
#raise ""
v = m.var_dist
v.eta = numpy.array( # a pssm that looks like T[A]G
[
[ 0.25, 0.25, 0.25, 0.25 ],
[ 0.01, 0.01, 0.01, 0.97 ],
[ 0.97, 0.01, 0.01, 0.01 ],
[ 0.01, 0.01, 0.97, 0.01 ]
]
)
print 'Updating'
for i in xrange( 100 ):
m.update(
update_rho = True,
update_eta = False,
update_tau = True
)
print "\n".join(str( r ) for r in m.var_dist.r_mode)
hmm.write_model_svg(
m,
name = 'simple',
dir = 'test',
show_rev_comp = True,
show_dists = True,
edge_lengths = 3.0
)
return m, v
def test_eta_simple():
"Test update eta with simple 1 base pssm"
K = 1
sequences = [
'tg',
]
m = create_model(K, sequences, psi, theta, phi, upsilon)
v = m.var_dist
v.rho = numpy.array([[[.98, .01, .01], [.01, .98, .01]]])
print 'Updating'
for i in xrange(100):
m.update(update_rho=False, update_eta=True, update_tau=False)
hmm.write_model_svg(m,
name='eta_simple',
dir='test',
show_rev_comp=True,
show_dists=True,
edge_lengths=3.0)
return m, v
def test_rho_simple():
K = 2
sequences = [
'tga',
]
m = create_model(K, sequences, psi, theta, phi, upsilon)
v = m.var_dist
v.eta = numpy.array( # a pssm that looks like G[G]A
[[0.25, 0.25, 0.25, 0.25], [0.01, 0.01, 0.97, 0.01],
[0.01, 0.01, 0.97, 0.01], [0.97, 0.01, 0.01, 0.01]])
print 'Updating'
for i in xrange(100):
print m.update(update_rho=True, update_eta=False, update_tau=False),
print "\n".join(str(r) for r in v.r_mode)
hmm.write_model_svg(m,
name='rho_simple',
dir='test',
show_rev_comp=True,
show_dists=False,
edge_lengths=3.0)
return m, v
def test_eta_gapped():
"Test update eta with gapped 2 base pssm"
K = 2
sequences = [
'tag', # 'a' is the gap
#'tg',
]
m = create_model( K, sequences, psi, theta, phi, upsilon )
v = m.var_dist
v.rho = [
numpy.array(
[
[ .01, .94, .01, .01, .01, .01, .01 ],
[ .01, .01, .94, .01, .01, .01, .01 ],
[ .01, .01, .01, .94, .01, .01, .01 ],
]
),
# numpy.array(
# [
# [ .01, .94, .01, .01, .01, .01, .01 ],
# [ .01, .01, .01, .94, .01, .01, .01 ],
# ]
# )
]
for i in xrange( 100 ):
m.update(
update_rho = False,
update_eta = True,
update_tau = False
)
print 'Updating'
hmm.write_model_svg(
m,
name = 'eta_gapped',
dir = 'test',
show_rev_comp = True,
show_dists = True,
edge_lengths = 3.0
)
return m, v
def test_transitions():
"Test that transition probs are set correctly. Can check by inspecting svg."
K = 3
sequences = [
'tgacg',
]
m = create_model( K, sequences, psi, theta, phi, upsilon )
v = m.var_dist
v.tau = numpy.array(
[
[ .90, .10 ],
[ .50, .50 ],
[ .80, .20 ]
]
)
hmm.write_model_svg(
m,
name = 'transitions',
dir = 'test',
show_rev_comp = True,
show_dists = False,
edge_lengths = 3.0
)
return m, v
def test_write_models():
for k in xrange(10):
model = hmm.Model(
hmm.ObservedData(k + 1, hmm.ObservedSequences(sequences)))
hmm.write_model_svg(model, 'states_%d' % (k + 1), show_rev_comp=False)