def test_skip(self):
for skip in [0, 3, 13]:
r1 = FeatureReader(self.trajfile, self.topfile)
out_with_skip = r1.get_output(skip=skip)[0]
r2 = FeatureReader(self.trajfile, self.topfile)
out = r2.get_output()[0]
np.testing.assert_almost_equal(out_with_skip, out[skip::],
err_msg="The first %s rows were skipped, but that did not "
"match the rows with skip=0 and sliced by [%s::]" % (skip, skip))
def test_skip_input_list(self):
for skip in [0, 3, 13]:
r1 = FeatureReader([self.trajfile, self.trajfile2], self.topfile)
out_with_skip = r1.get_output(skip=skip)
r2 = FeatureReader([self.trajfile, self.trajfile2], self.topfile)
out = r2.get_output()
np.testing.assert_almost_equal(out_with_skip[0], out[0][skip::],
err_msg="The first %s rows of the first file were skipped, but that did not "
"match the rows with skip=0 and sliced by [%s::]" % (skip, skip))
np.testing.assert_almost_equal(out_with_skip[1], out[1][skip::],
err_msg="The first %s rows of the second file were skipped, but that did not"
" match the rows with skip=0 and sliced by [%s::]" % (skip, skip))
def test_covariances_and_eigenvalues(self):
reader = FeatureReader(self.trajnames, self.temppdb)
trans = TICA(lag=1,
output_dimension=self.dim,
force_eigenvalues_le_one=True)
trans.data_producer = reader
for tau in [1, 10, 100, 1000, 2000]:
log.info('number of trajectories reported by tica %d' %
trans.number_of_trajectories())
trans.lag = tau
trans.parametrize()
data = trans.get_output()
# print '@@cov', trans.cov
# print '@@cov_tau', trans.cov_tau
log.info('max. eigenvalue: %f' % np.max(trans.eigenvalues))
self.assertTrue(np.all(trans.eigenvalues <= 1.0))
# check ICs
check = tica(data=data,
lag=tau,
dim=self.dim,
force_eigenvalues_le_one=True)
check.parametrize()
self.assertTrue(np.allclose(np.eye(self.dim), check.cov))
ic_cov_tau = np.zeros((self.dim, self.dim))
ic_cov_tau[np.diag_indices(self.dim)] = trans.eigenvalues
self.assertTrue(np.allclose(ic_cov_tau, check.cov_tau))
def testTimeLaggedIterator(self):
lag = 10
reader = FeatureReader(self.trajfile, self.topfile)
frames = 0
data = []
lagged = []
for _, X, Y in reader.iterator(lag=lag):
frames += X.shape[0]
data.append(X)
lagged.append(Y)
assert len(data) == len(lagged)
# .reshape(self.xyz.shape)
merged_lagged = np.concatenate(lagged, axis=0)
# reproduce outcome
xyz_s = self.xyz.shape
fake_lagged = self.xyz.reshape((xyz_s[0], -1))[lag:]
self.assertTrue(np.allclose(merged_lagged, fake_lagged))
# restore shape of input
data = np.vstack(data).reshape(self.xyz[lag:].shape)
self.assertEqual(frames, reader.trajectory_lengths()[0] - lag)
self.assertTrue(np.allclose(data, self.xyz[:len(self.xyz) - lag]))
def test_covariances_and_eigenvalues(self):
reader = FeatureReader(self.trajnames, self.temppdb)
for tau in [1, 10, 100, 1000, 2000]:
trans = tica(lag=tau, dim=self.dim, kinetic_map=False)
trans.data_producer = reader
log.info('number of trajectories reported by tica %d' %
trans.number_of_trajectories())
trans.parametrize()
data = trans.get_output()
# print '@@cov', trans.cov
# print '@@cov_tau', trans.cov_tau
log.info('max. eigenvalue: %f' % np.max(trans.eigenvalues))
self.assertTrue(np.all(trans.eigenvalues <= 1.0))
# check ICs
check = tica(data=data,
lag=tau,
dim=self.dim,
force_eigenvalues_le_one=True)
check.parametrize()
np.testing.assert_allclose(np.eye(self.dim), check.cov, atol=1e-8)
np.testing.assert_allclose(check.mean, 0.0, atol=1e-8)
ic_cov_tau = np.zeros((self.dim, self.dim))
ic_cov_tau[np.diag_indices(self.dim)] = trans.eigenvalues
np.testing.assert_allclose(ic_cov_tau, check.cov_tau, atol=1e-8)
def test_partial_fit(self):
reader = FeatureReader(self.trajnames, self.temppdb, chunksize=10000)
output = reader.get_output()
params = {'dim': self.dim, 'lag': 1001}
ref = api.tica(reader, **params)
partial = api.tica(**params)
for traj in output:
partial.partial_fit(traj)
np.testing.assert_allclose(partial.eigenvalues, ref.eigenvalues, atol=1e-3)
# only compare first two eigenvectors, because we only have two metastable processes
np.testing.assert_allclose(np.abs(partial.eigenvectors[:2]),
np.abs(ref.eigenvectors[:2]), rtol=1e-3, atol=1e-3)
def testIteratorAccess2(self):
reader = FeatureReader([self.trajfile, self.trajfile2], self.topfile)
reader.chunksize = 100
data = {itraj: [] for itraj in range(reader.number_of_trajectories())}
for i, X in reader:
data[i].append(X)
# restore shape of input
data[0] = np.vstack(data[0]).reshape(-1, 9)
data[1] = np.vstack(data[1]).reshape(-1, 9)
np.testing.assert_equal(data[0], self.xyz.reshape(-1, 9))
np.testing.assert_equal(data[1], self.xyz2.reshape(-1, 9))
def testIteratorAccess2(self):
reader = FeatureReader([self.trajfile, self.trajfile], self.topfile)
reader.chunksize = 100
frames = 0
data = []
for i, X in reader:
frames += X.shape[0]
data.append(X)
self.assertEqual(frames, reader.trajectory_lengths()[0] * 2)
# restore shape of input
data = np.array(data[0:reader.trajectory_lengths()[0] /
reader.chunksize]).reshape(self.xyz.shape)
self.assertTrue(np.allclose(data, self.xyz))
def test_covariances_and_eigenvalues(self):
reader = FeatureReader(self.trajnames, self.temppdb, chunksize=10000)
for lag in [1, 11, 101, 1001, 2001]: # avoid cos(w*tau)==0
trans = api.tica(data=reader, dim=self.dim, lag=lag)
log.info('number of trajectories reported by tica %d' % trans.number_of_trajectories())
log.info('tau = %d corresponds to a number of %f cycles' % (lag, self.w*lag/(2.0*np.pi)))
# analytical solution for C_ij(lag) is 0.5*A[i]*A[j]*cos(phi[i]-phi[j])*cos(w*lag)
ana_cov = 0.5*self.A[:, np.newaxis]*self.A*np.cos(self.phi[:, np.newaxis]-self.phi)
ana_cov_tau = ana_cov*np.cos(self.w*lag)
self.assertTrue(np.allclose(ana_cov, trans.cov, atol=1.E-3))
self.assertTrue(np.allclose(ana_cov_tau, trans.cov_tau, atol=1.E-3))
log.info('max. eigenvalue: %f' % np.max(trans.eigenvalues))
self.assertTrue(np.all(trans.eigenvalues <= 1.0))
def test_covariances_and_eigenvalues(self):
reader = FeatureReader(self.trajnames, self.temppdb)
for tau in [1, 10, 100, 1000, 2000]:
trans = tica(lag=tau, dim=self.dim, kinetic_map=False)
trans.estimate(reader)
data = trans.get_output()
log.info('max. eigenvalue: %f' % np.max(trans.eigenvalues))
self.assertTrue(np.all(trans.eigenvalues <= 1.0))
# check ICs
check = tica(data=data, lag=tau, dim=self.dim)
np.testing.assert_allclose(np.eye(self.dim), check.cov, atol=1e-8)
np.testing.assert_allclose(check.mean, 0.0, atol=1e-8)
ic_cov_tau = np.zeros((self.dim, self.dim))
ic_cov_tau[np.diag_indices(self.dim)] = trans.eigenvalues
np.testing.assert_allclose(ic_cov_tau, check.cov_tau, atol=1e-8)
def test_featurereader_xtc(self):
# cause cache failures
with settings(use_trajectory_lengths_cache=False):
reader = FeatureReader(xtcfiles, pdbfile)
results = {}
for f in xtcfiles:
traj_info = self.db[f, reader]
results[f] = traj_info.ndim, traj_info.length, traj_info.offsets
expected = {}
for f in xtcfiles:
with mdtraj.open(f) as fh:
length = len(fh)
ndim = fh.read(1)[0].shape[1]
offsets = fh.offsets if hasattr(fh, 'offsets') else []
expected[f] = ndim, length, offsets
np.testing.assert_equal(results, expected)