class TestStructureContainer(unittest.TestCase):
"""Container for tests of the class functionality."""
def __init__(self, *args, **kwargs):
super(TestStructureContainer, self).__init__(*args, **kwargs)
prim = bulk('Ag', a=4.09)
chemical_symbols = ['Ag', 'Au']
self.cs = ClusterSpace(structure=prim,
cutoffs=[4.0, 4.0, 4.0],
chemical_symbols=chemical_symbols)
self.structure_list = []
self.user_tags = []
for k in range(4):
structure = prim.repeat(2)
symbols = [chemical_symbols[0]] * len(structure)
symbols[:k] = [chemical_symbols[1]] * k
structure.set_chemical_symbols(symbols)
self.structure_list.append(structure)
self.user_tags.append('Structure {}'.format(k))
self.properties_list = []
self.add_properties_list = []
for k, structure in enumerate(self.structure_list):
structure.set_calculator(EMT())
properties = {'energy': structure.get_potential_energy(),
'volume': structure.get_volume(),
'Au atoms': structure.get_chemical_symbols().count('Au')}
self.properties_list.append(properties)
add_properties = {'total_energy': structure.get_total_energy()}
self.add_properties_list.append(add_properties)
def shortDescription(self):
"""Silences unittest from printing the docstrings in test cases."""
return None
def setUp(self):
"""Instantiates class before each test."""
self.sc = StructureContainer(self.cs)
for structure, tag, props in zip(self.structure_list,
self.user_tags,
self.properties_list):
self.sc.add_structure(structure, tag, props)
def test_init(self):
"""Tests that initialization of tested class works."""
# check empty initialization
self.assertIsInstance(StructureContainer(self.cs), StructureContainer)
# check whether method raises Exceptions
with self.assertRaises(TypeError) as cm:
StructureContainer('my_sc.sc')
self.assertIn('cluster_space must be a ClusterSpace', str(cm.exception))
def test_len(self):
"""Tests length functionality."""
len_structure_container = self.sc.__len__()
self.assertEqual(len_structure_container, len(self.structure_list))
def test_getitem(self):
"""Tests getitem functionality."""
structure = self.sc.__getitem__(1)
self.assertIsNotNone(structure)
def test_get_structure_indices(self):
"""Tests get_structure_indices functionality."""
list_index = [x for x in range(len(self.structure_list))]
self.assertEqual(self.sc.get_structure_indices(), list_index)
def test_add_structure(self):
"""Tests add_structure functionality."""
# add structure with tag and property
structure = self.structure_list[0]
properties = self.properties_list[0]
tag = 'Structure 4'
self.sc.add_structure(structure, tag, properties)
self.assertEqual(len(self.sc), len(self.structure_list) + 1)
# add atom and read property from calculator
self.sc.add_structure(structure)
self.assertEqual(len(self.sc), len(self.structure_list) + 2)
self.assertEqual(self.sc[5].properties['energy'],
self.properties_list[0]['energy'] / len(structure))
# add atom and don't read property from calculator
structure_cpy = structure.copy()
structure_cpy.set_calculator(EMT())
self.sc.add_structure(structure_cpy)
self.assertEqual(len(self.sc), len(self.structure_list) + 3)
self.assertNotIn('energy', self.sc[6].properties)
# check that duplicate structure is not added.
with self.assertRaises(ValueError) as cm:
self.sc.add_structure(structure, allow_duplicate=False)
msg = 'Input structure and Structure 0 have identical ' \
'cluster vectors at index 0'
self.assertEqual(msg, str(cm.exception))
self.assertEqual(len(self.sc), len(self.structure_list) + 3)
symbols = ['Au' for i in range(len(structure))]
structure.set_chemical_symbols(symbols)
self.sc.add_structure(structure, 'Structure 5', allow_duplicate=False)
self.assertEqual(len(self.sc), len(self.structure_list) + 4)
def test_get_condition_number(self):
"""Tests get_condition_number functionality."""
target = np.linalg.cond(self.sc.get_fit_data()[0])
retval = self.sc.get_condition_number()
self.assertEqual(target, retval)
def test_get_fit_data(self):
"""Tests get_fit_data functionality."""
import numpy as np
cluster_vectors, properties = self.sc.get_fit_data()
# testing outputs have ndarray type
self.assertIsInstance(cluster_vectors, np.ndarray)
self.assertIsInstance(properties, np.ndarray)
# testing values of cluster_vectors and properties
for structure, cv in zip(self.structure_list, cluster_vectors):
retval = list(cv)
target = list(self.cs.get_cluster_vector(structure))
self.assertAlmostEqual(retval, target, places=9)
for target, retval in zip(self.properties_list, properties):
self.assertEqual(retval, target['energy'])
# passing a list of indexes
cluster_vectors, properties = self.sc.get_fit_data([0])
retval = list(cluster_vectors[0])
structure = self.structure_list[0]
target = list(self.cs.get_cluster_vector(structure))
self.assertAlmostEqual(retval, target, places=9)
retval2 = properties[0]
target2 = self.properties_list[0]
self.assertEqual(retval2, target2['energy'])
def test_repr(self):
"""Tests repr functionality."""
retval = self.sc.__repr__()
target = """
================================ Structure Container =================================
Total number of structures: 4
--------------------------------------------------------------------------------------
index | user_tag | n_atoms | chemical formula | Au atoms | energy | volume
--------------------------------------------------------------------------------------
0 | Structure 0 | 8 | Ag8 | 0 | 0.0127 | 136.8359
1 | Structure 1 | 8 | Ag7Au | 1 | -0.0073 | 136.8359
2 | Structure 2 | 8 | Ag6Au2 | 2 | -0.0255 | 136.8359
3 | Structure 3 | 8 | Ag5Au3 | 3 | -0.0382 | 136.8359
======================================================================================
""" # noqa
self.assertEqual(strip_surrounding_spaces(target),
strip_surrounding_spaces(retval))
# test representation of an empty structure container
sc = StructureContainer(self.cs)
self.assertEqual(sc.__repr__(), 'Empty StructureContainer')
def test_get_string_representation(self):
"""Tests _get_string_representation functionality."""
retval = self.sc._get_string_representation(print_threshold=2)
target = """
================================ Structure Container =================================
Total number of structures: 4
--------------------------------------------------------------------------------------
index | user_tag | n_atoms | chemical formula | Au atoms | energy | volume
--------------------------------------------------------------------------------------
0 | Structure 0 | 8 | Ag8 | 0 | 0.0127 | 136.8359
...
3 | Structure 3 | 8 | Ag5Au3 | 3 | -0.0382 | 136.8359
======================================================================================
""" # noqa
self.assertEqual(strip_surrounding_spaces(target),
strip_surrounding_spaces(retval))
def test_print_overview(self):
"""Tests print_overview functionality."""
with StringIO() as capturedOutput:
sys.stdout = capturedOutput # redirect stdout
self.sc.print_overview()
sys.stdout = sys.__stdout__ # reset redirect
self.assertTrue('Structure Container' in capturedOutput.getvalue())
def test_cluster_space(self):
"""Tests cluster space functionality."""
cs_onlyread = self.sc.cluster_space
self.assertEqual(str(cs_onlyread), str(self.cs))
def test_available_properties(self):
"""Tests available_properties property."""
available_properties = sorted(self.properties_list[0])
self.sc.add_structure(self.structure_list[0], properties=self.properties_list[0])
self.assertSequenceEqual(available_properties, self.sc.available_properties)
def test_read_write(self):
"""Tests the read and write functionality."""
temp_file = tempfile.NamedTemporaryFile()
# check before with a non-tar file
with self.assertRaises(TypeError) as context:
self.sc.read(temp_file)
self.assertTrue('{} is not a tar file'.format(str(temp_file.name))
in str(context.exception))
# save and read an empty structure container
sc = StructureContainer(self.cs)
sc.write(temp_file.name)
sc_read = StructureContainer.read(temp_file.name)
self.assertEqual(sc_read.__str__(), 'Empty StructureContainer')
# save to file
self.sc.write(temp_file.name)
# read from file object
sc_read = self.sc.read(temp_file.name)
# check data
self.assertEqual(len(self.sc), len(sc_read))
self.assertEqual(self.sc.__str__(), sc_read.__str__())
for fs, fs_read in zip(self.sc._structure_list, sc_read._structure_list):
self.assertEqual(list(fs.cluster_vector),
list(fs_read.cluster_vector))
self.assertEqual(fs.structure, fs_read.structure)
self.assertEqual(fs.user_tag, fs_read.user_tag)
self.assertEqual(fs.properties, fs_read.properties)
temp_file.close()
# step 1: Basic setup
db = connect('reference_data.db')
primitive_structure = db.get(id=1).toatoms() # primitive structure
# step 2: Set up the basic structure and a cluster space
cs = ClusterSpace(structure=primitive_structure,
cutoffs=[13.5, 6.0, 5.5],
chemical_symbols=['Ag', 'Pd'])
print(cs)
# step 3: Parse the input structures and set up a structure container
sc = StructureContainer(cluster_space=cs)
for row in db.select('natoms<=8'):
sc.add_structure(structure=row.toatoms(),
user_tag=row.tag,
properties={'mixing_energy': row.mixing_energy})
print(sc)
# step 4: Train parameters
opt = CrossValidationEstimator(fit_data=sc.get_fit_data(key='mixing_energy'),
fit_method='lasso')
opt.validate()
opt.train()
print(opt)
# step 5: Construct cluster expansion and write it to file
ce = ClusterExpansion(cluster_space=cs,
parameters=opt.parameters,
metadata=opt.summary)
print(ce)
ClusterExpansion)
from mchammer.ensembles import SemiGrandCanonicalEnsemble as SGCEnsemble
from mchammer.calculators import ClusterExpansionCalculator as CECalculator
from mchammer.observers import ClusterExpansionObserver as CEObserver
# step 1: Read cluster expansion for mixing energies from file
ce_mix_energies = ClusterExpansion.read('mixing_energy.ce')
chemical_symbols = ce_mix_energies.cluster_space.chemical_symbols[0]
cs = ce_mix_energies.cluster_space
# step 2: Read structures from database into a structure container
db = connect('reference_data.db')
sc = StructureContainer(cluster_space=cs)
for row in db.select('natoms<=8'):
sc.add_structure(structure=row.toatoms(),
user_tag=row.tag,
properties={'lattice_parameter': row.lattice_parameter})
# step 3: Construct cluster expansion for lattice parameter
fit_data = sc.get_fit_data(key='lattice_parameter')
opt = CrossValidationEstimator(fit_data=fit_data, fit_method='lasso')
opt.validate()
opt.train()
ce_latt_param = ClusterExpansion(cluster_space=cs, parameters=opt.parameters)
# step 4: Set up the calculator and a canonical ensemble
structure = cs.primitive_structure.repeat(3)
structure.set_chemical_symbols([chemical_symbols[0]] * len(structure))
calculator = CECalculator(structure=structure,
cluster_expansion=ce_mix_energies)
ensemble = SGCEnsemble(calculator=calculator,