class Test0TaggedSet(unittest.TestCase):
def test_init(self):
ls = tl.TaggedSet()
ls = tl.TaggedSet(zip([1, 2, 3], [["a", "b"], "a", ["b", "c"]]))
self.assertListEqual(ls._data, [1, 2, 3])
self.assertListEqual(ls._tags, [{"a", "b"}, {"a"}, {"b", "c"}])
return cnt
def setUp(self):
ds = tl.TaggedSet()
ds.add(tl.Trajectory.fromArray([0, 0.75, 0.5, 0.3, 5.4, 5.5, 5.3, -2.0, 5.4]))
ds.add(tl.Trajectory.fromArray([1.2, 1.4, np.nan, np.nan, 10.0, 10.2]))
def setUp(self):
def traj():
return tl.Trajectory.fromArray(
np.cumsum(np.random.normal(size=(10, 3)), axis=0))
self.data = tl.TaggedSet((traj() for _ in range(10)), hasTags=False)
self.fit = msdfit.lib.SplineFit(self.data, ss_order=1, n=2)
class TestUtilClean(myTestCase):
def setUp(self):
# Split with threshold = 2 => trajectories of lengths [2, 3, 3]
self.traj = tl.Trajectory.fromArray([1., 2, 4.1, 4.5, 3, -0.5, -1, -0.7])
# Split with threshold = 3 => 3 trajectories with lengths [4, 5, 6]
self.ds = tl.TaggedSet()
self.assertListEqual(self.ls._data, [2, 3, 4])
def test_process(self):
# need a TaggedSet of mutable objects
mutls = tl.TaggedSet(zip(['hello', 'World', '!'], [["a", "b"], "a", ["b", "c"]]))
mutls.makeSelection(tags="a")
newls = mutls.process(lambda word : word+'_moo')
self.assertListEqual(newls._data, ['hello_moo', 'World_moo'])
def test_map_unique(self):
def funTrue(x):
return True
def fun2(x):
return 2*x
self.assertTrue(self.ls.map_unique(funTrue))
with self.assertRaises(RuntimeError):
self.assertEqual(self.ls[1], 2)
def test_mergein(self):
newls = tl.TaggedSet(zip([4, 5, 6], [["d"], [], "e"]))
self.ls.mergein(newls, additionalTags='new')
self.assertListEqual(self.ls._data, [1, 2, 3, 4, 5, 6])
self.assertSetEqual(self.ls.tagset(), {'a', 'b', 'c', 'd', 'e', 'new'})
self.ls |= newls
self.assertEqual(len(self.ls), 9)
self.ds.add(tl.Trajectory.fromArray([1.2, 1.4, np.nan, np.nan, 10.0, 10.2]))
def test_csv(self):
filename = "testdata/test_write.csv"
tl.io.write.csv(self.ds, filename)
with open(filename, 'r') as f:
self.assertTrue(f.read() == 'id\tframe\tx\n0\t0\t0.0\n0\t1\t0.75\n0\t2\t0.5\n0\t3\t0.3\n0\t4\t5.4\n0\t5\t5.5\n0\t6\t5.3\n0\t7\t-2.0\n0\t8\t5.4\n1\t0\t1.2\n1\t1\t1.4\n1\t4\t10.0\n1\t5\t10.2\n')
ds = tl.TaggedSet()
ds.add(tl.Trajectory.fromArray(np.arange(20).reshape((2, 5, 2))))
tl.io.write.csv(ds, filename)
def setUp(self):
model = tl.models.rouse.Model(10)
tracked = [2, 7]
def traj():
conf = model.conf_ss()
traj = []
for _ in range(10):
conf = model.evolve(conf)
traj.append(conf[tracked[0] - tracked[1]])
return tl.Trajectory.fromArray(traj)
self.data = tl.TaggedSet((traj() for _ in range(10)), hasTags=False)
def test_fitting_rouse(self):
modelfam = neda.ParametricFamily((0.1, ), [(1e-10, None)])
modelfam.get = lambda k: neda.models.RouseModel(20, 1, k)
true_params = (1, )
data = tl.TaggedSet()
mod_gen = modelfam.get(*true_params)
for state in 3 * [0, 1]:
lt = neda.Loopingtrace.fromStates(100 * [state])
data.add(
mod_gen.trajectory_from_loopingtrace(lt,
localization_error=0.1))
fitres = neda.models.fit(data, modelfam, maxfun=200)
self.assertTrue(fitres.success)
for true_param, est_param in zip(true_params, fitres.x):
self.assertAlmostEqual(true_param, est_param, delta=0.5)
def test_fitting_factorized(self):
modelfam = neda.ParametricFamily((0.1, 10), [(1e-10, None),
(1e-10, None)])
modelfam.get = lambda s0, s1: neda.models.FactorizedModel([
scipy.stats.maxwell(scale=s0),
scipy.stats.maxwell(scale=s1),
])
true_params = (1, 4)
data = tl.TaggedSet()
mod_gen = modelfam.get(*true_params)
for state in 3 * [0, 1]:
lt = neda.Loopingtrace.fromStates(100 * [state])
data.add(mod_gen.trajectory_from_loopingtrace(lt))
fitres = neda.models.fit(data, modelfam, maxfun=200)
self.assertTrue(fitres.success)
for true_param, est_param in zip(true_params, fitres.x):
self.assertAlmostEqual(true_param, est_param, delta=0.5)
class TestIOWrite(myTestCase):
def setUp(self):
self.ds = tl.TaggedSet()
class Test1TaggedSet(unittest.TestCase):
ls = tl.TaggedSet([1, 2, 3], hasTags=False)
def test_add(self):
ls = tl.TaggedSet()
ls.add(1)
def setUp(self):
def traj():
return tl.Trajectory.fromArray(
np.cumsum(np.random.normal(size=(10, 3)), axis=0))
self.data = tl.TaggedSet((traj() for _ in range(10)), hasTags=False)