From 60c2a251e38bb90e1a5e886efc18324c7da9d5d6 Mon Sep 17 00:00:00 2001
From: Ron Weiss <ronweiss@gmail.com>
Date: Sun, 7 Nov 2010 16:32:39 -0500
Subject: [PATCH] fix pep8, pyflakes errprs
---
scikits/learn/gmm.py | 8 ++++----
scikits/learn/hmm.py | 22 +++++++++++-----------
scikits/learn/tests/test_gmm.py | 8 ++++----
scikits/learn/tests/test_hmm.py | 33 ++++++++++++++++++---------------
4 files changed, 37 insertions(+), 34 deletions(-)
diff --git a/scikits/learn/gmm.py b/scikits/learn/gmm.py
index 1ce12b6e4f..ae1046a319 100644
--- a/scikits/learn/gmm.py
+++ b/scikits/learn/gmm.py
@@ -474,7 +474,7 @@ class GMM(BaseEstimator):
cv, self._cvtype, self._n_states)
elif not hasattr(self, 'covars'):
self.covars = _distribute_covar_matrix_to_match_cvtype(
- np.eye(self.n_features), cvtype, n_states)
+ np.eye(self.n_features), self.cvtype, self.n_states)
# EM algorithm
logprob = []
@@ -562,9 +562,9 @@ def _lmvnpdffull(obs, means, covars):
log_prob = np.empty((nobs,nmix))
for c, (mu, cv) in enumerate(itertools.izip(means, covars)):
cv_chol = linalg.cholesky(cv, lower=True)
- cv_det = np.prod(np.diagonal(cv_chol))**2
- cv_sol = solve_triangular(cv_chol, (obs - mu).T, lower=True)
- log_prob[:, c] = -.5 * (np.sum(cv_sol**2, axis=0) + \
+ cv_det = np.prod(np.diagonal(cv_chol)) ** 2
+ cv_sol = solve_triangular(cv_chol, (obs - mu).T, lower=True)
+ log_prob[:, c] = -.5 * (np.sum(cv_sol ** 2, axis=0) + \
ndim * np.log(2 * np.pi) + np.log(cv_det))
return log_prob
diff --git a/scikits/learn/hmm.py b/scikits/learn/hmm.py
index f748d4108d..41a432ae32 100644
--- a/scikits/learn/hmm.py
+++ b/scikits/learn/hmm.py
@@ -274,7 +274,7 @@ class _BaseHMM(BaseEstimator):
currstate = (startprob_cdf > rand).argmax()
obs = [self._generate_sample_from_state(currstate)]
- for x in xrange(n-1):
+ for x in xrange(n - 1):
rand = np.random.rand()
currstate = (transmat_cdf[currstate] > rand).argmax()
obs.append(self._generate_sample_from_state(currstate))
@@ -403,8 +403,8 @@ class _BaseHMM(BaseEstimator):
lattice[0] = self._log_startprob + framelogprob[0]
for n in xrange(1, nobs):
- idx = self._prune_states(lattice[n-1], maxrank, beamlogprob)
- pr = self._log_transmat[idx].T + lattice[n-1,idx]
+ idx = self._prune_states(lattice[n - 1], maxrank, beamlogprob)
+ pr = self._log_transmat[idx].T + lattice[n - 1,idx]
lattice[n] = np.max(pr, axis=1) + framelogprob[n]
traceback[n] = np.argmax(pr, axis=1)
lattice[lattice <= ZEROLOGPROB] = -np.Inf
@@ -427,9 +427,9 @@ class _BaseHMM(BaseEstimator):
fwdlattice[0] = self._log_startprob + framelogprob[0]
for n in xrange(1, nobs):
- idx = self._prune_states(fwdlattice[n-1], maxrank, beamlogprob)
+ idx = self._prune_states(fwdlattice[n - 1], maxrank, beamlogprob)
fwdlattice[n] = (logsum(self._log_transmat[idx].T
- + fwdlattice[n-1,idx], axis=1)
+ + fwdlattice[n - 1,idx], axis=1)
+ framelogprob[n])
fwdlattice[fwdlattice <= ZEROLOGPROB] = -np.Inf
@@ -449,7 +449,7 @@ class _BaseHMM(BaseEstimator):
-50)
#beamlogprob)
#-np.Inf)
- bwdlattice[n-1] = logsum(self._log_transmat[:,idx]
+ bwdlattice[n - 1] = logsum(self._log_transmat[:,idx]
+ bwdlattice[n,idx] + framelogprob[n,idx],
axis=1)
bwdlattice[bwdlattice <= ZEROLOGPROB] = -np.Inf
@@ -514,7 +514,7 @@ class _BaseHMM(BaseEstimator):
stats['start'] += posteriors[0]
if 't' in params:
for t in xrange(len(framelogprob)):
- zeta = (fwdlattice[t-1][:,np.newaxis] + self._log_transmat
+ zeta = (fwdlattice[t - 1][:,np.newaxis] + self._log_transmat
+ framelogprob[t] + bwdlattice[t])
stats['trans'] += np.exp(zeta - logsum(zeta))
@@ -718,7 +718,7 @@ class GaussianHMM(_BaseHMM):
if 'c' in params:
if self._cvtype in ('spherical', 'diag'):
- stats['obs**2'] += np.dot(posteriors.T, obs**2)
+ stats['obs**2'] += np.dot(posteriors.T, obs ** 2)
elif self._cvtype in ('tied', 'full'):
for t, o in enumerate(obs):
obsobsT = np.outer(o, o)
@@ -754,10 +754,10 @@ class GaussianHMM(_BaseHMM):
meandiff = self._means - means_prior
if self._cvtype in ('spherical', 'diag'):
- cv_num = (means_weight * (meandiff)**2
+ cv_num = (means_weight * (meandiff) ** 2
+ stats['obs**2']
- 2 * self._means * stats['obs']
- + self._means**2 * denom)
+ + self._means ** 2 * denom)
cv_den = max(covars_weight - 1, 0) + denom
if self._cvtype == 'spherical':
self._covars = (covars_prior / cv_den.mean(axis=1)
@@ -971,7 +971,7 @@ class GMMHMM(_BaseHMM):
# XXX: Hotfit for n_mix that is incompatible with the scikit's
# BaseEstimator API
self.n_mix = n_mix
- self.cvtype= cvtype
+ self.cvtype = cvtype
if gmms is None:
gmms = []
for x in xrange(self.n_states):
diff --git a/scikits/learn/tests/test_gmm.py b/scikits/learn/tests/test_gmm.py
index 10f14f94ee..fd2089ca84 100644
--- a/scikits/learn/tests/test_gmm.py
+++ b/scikits/learn/tests/test_gmm.py
@@ -19,12 +19,14 @@ def _generate_random_spd_matrix(ndim):
randspd = np.dot(np.dot(U, 1.0 + np.diag(np.random.rand(ndim))), V)
return randspd
+
def test_logsum_1D():
A = np.random.rand(2) + 1.0
for axis in range(1):
Asum = gmm.logsum(A, axis)
assert_array_almost_equal(np.exp(Asum), np.sum(np.exp(A), axis))
+
def test_logsum_3D():
"""
Test also on a 3D matrix
@@ -85,7 +87,7 @@ def _naive_lmvnpdf_diag(obs, mu, cv):
ref = np.empty((len(obs), len(mu)))
stds = np.sqrt(cv)
for i, (m, std) in enumerate(itertools.izip(mu, stds)):
- ref[:, i] = np.log(stats.norm.pdf(obs, m, std)).sum(axis=1)
+ ref[:, i] = np.log(stats.norm.pdf(obs, m, std)).sum(axis=1)
return ref
@@ -118,7 +120,6 @@ def test_lmvnpdf_spherical():
assert_array_almost_equal(lpr, reference)
-
def test_lmvnpdf_full():
n_features, n_states, n_obs = 2, 3, 10
@@ -133,7 +134,6 @@ def test_lmvnpdf_full():
assert_array_almost_equal(lpr, reference)
-
def test_GMM_attributes():
n_states, n_features = 10, 4
cvtype = 'diag'
@@ -168,6 +168,7 @@ def test_GMM_attributes():
assert_raises(ValueError, gmm.GMM, n_states=20, cvtype='badcvtype')
+
class GMMTester():
n_states = 10
n_features = 4
@@ -181,7 +182,6 @@ class GMMTester():
'full': np.array([_generate_random_spd_matrix(n_features) + 5 * I
for x in xrange(n_states)])}
-
def test_eval(self):
g = gmm.GMM(self.n_states, self.cvtype)
# Make sure the means are far apart so posteriors.argmax()
diff --git a/scikits/learn/tests/test_hmm.py b/scikits/learn/tests/test_hmm.py
index 5a80dacdc6..1b63253a48 100644
--- a/scikits/learn/tests/test_hmm.py
+++ b/scikits/learn/tests/test_hmm.py
@@ -7,7 +7,7 @@ from .test_gmm import _generate_random_spd_matrix
from .. import hmm
-SKIP_FAILING = True # skip failing tests
+SKIP_FAILING = True # skip failing tests
class SeedRandomNumberGeneratorTestCase(TestCase):
@@ -221,11 +221,13 @@ class GaussianHMMParams(object):
transmat = np.random.rand(n_states, n_states)
transmat /= np.tile(transmat.sum(axis=1)[:,np.newaxis], (1, n_states))
means = np.random.randint(-20, 20, (n_states, n_features))
- covars = {'spherical': (1.0 + 2 * np.random.rand(n_states))**2,
- 'tied': _generate_random_spd_matrix(n_features) + np.eye(n_features),
- 'diag': (1.0 + 2 * np.random.rand(n_states, n_features))**2,
- 'full': np.array([_generate_random_spd_matrix(n_features)
- + np.eye(n_features) for x in xrange(n_states)])}
+ covars = {'spherical': (1.0 + 2 * np.random.rand(n_states)) ** 2,
+ 'tied': (_generate_random_spd_matrix(n_features)
+ + np.eye(n_features)),
+ 'diag': (1.0 + 2 * np.random.rand(n_states, n_features)) ** 2,
+ 'full': np.array(
+ [_generate_random_spd_matrix(n_features) + np.eye(n_features)
+ for x in xrange(n_states)])}
expanded_covars = {'spherical': [np.eye(n_features) * cov
for cov in covars['spherical']],
'diag': [np.diag(cov) for cov in covars['diag']],
@@ -328,11 +330,10 @@ class GaussianHMMTester(GaussianHMMParams):
% (self.cvtype, params, trainll, np.diff(trainll)))
self.assertTrue(np.all(np.diff(trainll) > -0.5))
-
def test_fit_with_priors(self, params='stmc', n_iter=10,
verbose=False):
- startprob_prior = 10*self.startprob + 2.0
- transmat_prior = 10*self.transmat + 2.0
+ startprob_prior = 10 * self.startprob + 2.0
+ transmat_prior = 10 * self.transmat + 2.0
means_prior = self.means
means_weight = 2.0
covars_weight = 2.0
@@ -343,8 +344,8 @@ class GaussianHMMTester(GaussianHMMParams):
h = hmm.GaussianHMM(self.n_states, self.cvtype)
h.startprob = self.startprob
h.startprob_prior = startprob_prior
- h.transmat=hmm.normalize(self.transmat
- + np.diag(np.random.rand(self.n_states)), 1)
+ h.transmat = hmm.normalize(self.transmat
+ + np.diag(np.random.rand(self.n_states)), 1)
h.transmat_prior = transmat_prior
h.means = 20 * self.means
h.means_prior = means_prior
@@ -516,10 +517,12 @@ class GMMHMMParams(object):
g = gmm.GMM(n_mix, cvtype=cvtype)
g.means = np.random.randint(-20, 20, (n_mix, n_features))
mincv = 0.1
- g.covars = {'spherical': (mincv + mincv * np.random.rand(n_mix))**2,
+ g.covars = {'spherical': (mincv
+ + mincv * np.random.rand(n_mix)) ** 2,
'tied': _generate_random_spd_matrix(n_features)
+ mincv * np.eye(n_features),
- 'diag': (mincv + mincv * np.random.rand(n_mix, n_features))**2,
+ 'diag': (mincv
+ + mincv * np.random.rand(n_mix, n_features)) ** 2,
'full': np.array([_generate_random_spd_matrix(n_features)
+ mincv * np.eye(n_features)
for x in xrange(n_mix)])}[cvtype]
@@ -534,8 +537,8 @@ class TestGMMHMM(GMMHMMParams, SeedRandomNumberGeneratorTestCase):
np.random.seed(self.seed)
self.gmms = []
for state in xrange(self.n_states):
- self.gmms.append(self.create_random_gmm(self.n_mix, self.n_features,
- self.cvtype))
+ self.gmms.append(self.create_random_gmm(
+ self.n_mix, self.n_features, self.cvtype))
def test_attributes(self):
h = hmm.GMMHMM(self.n_states, cvtype=self.cvtype)
--
GitLab