def testConstructionWithCustomRates(self):
""" Test construction with custom rates. """
# A first process.
coords = [[1.0, 2.0, 3.4], [12.0, 13.0, -1.0], [1.1, 1.2, 1.3]]
types0 = ["A", "*", "B"]
types1 = ["B", "*", "A"]
rate_0_1 = 3.5
process_0 = KMCProcess(coords,
types0,
types1,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
# A second process.
coords = [[1.0, 2.0, 3.4], [1.1, 1.2, 1.3]]
types0 = ["A", "C"]
types1 = ["C", "A"]
rate_0_1 = 1.5
process_1 = KMCProcess(coords,
types0,
types1,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
processes = [process_0, process_1]
# Test that it fails with the wrong rate calculator.
interactions = KMCInteractions(processes=processes,
implicit_wildcards=True)
# Test that it fails if the rate calculator is of wrong type.
self.assertRaises(
Error, lambda: interactions.setRateCalculator(
rate_calculator="CustomRateCalculator"))
# Test that it fails if the rate calculator is instantiated.
self.assertRaises(
Error, lambda: interactions.setRateCalculator(
rate_calculator=CustomRateCalculator("DummyConfig")))
# Construct the interactions object with a custom rate calculator class.
kmc_interactions = KMCInteractions(processes=processes,
implicit_wildcards=True)
kmc_interactions.setRateCalculator(
rate_calculator=CustomRateCalculator)
# Test the stored name.
self.assertEqual(
kmc_interactions._KMCInteractions__rate_calculator_str,
"CustomRateCalculator")
# Test the stored rate calculator class.
self.assertTrue(
kmc_interactions._KMCInteractions__rate_calculator_class ==
CustomRateCalculator)
def testConstruction(self):
""" Test that the KMCInteractions class can be constructed. """
# A first process.
coords = [[1.0, 2.0, 3.4], [12.0, 13.0, -1.0], [1.1, 1.2, 1.3]]
types0 = ["A", "*", "B"]
types1 = ["B", "*", "A"]
rate_0_1 = 3.5
process_0 = KMCProcess(coords,
types0,
types1,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
# A second process.
coords = [[1.0, 2.0, 3.4], [1.1, 1.2, 1.3]]
types0 = ["A", "C"]
types1 = ["C", "A"]
rate_0_1 = 1.5
process_1 = KMCProcess(coords,
types0,
types1,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
processes = [process_0, process_1]
# Construct the interactions object.
kmc_interactions = KMCInteractions(processes=processes)
# Check the default implicit wildcard.
self.assertTrue(kmc_interactions.implicitWildcards())
# Construct again with non default value.
kmc_interactions = KMCInteractions(processes=processes,
implicit_wildcards=False)
# Check the wildcard again.
self.assertFalse(kmc_interactions.implicitWildcards())
# Check the processes stored on the object.
stored_processes = kmc_interactions._KMCInteractions__processes
# Checks that the address is the same.
self.assertTrue(stored_processes == processes)
def testConstructionBuckets(self):
""" Test that the KMCInteractions class can be constructed. """
# A first process.
coords = [[1.0, 2.0, 3.4], [12.0, 13.0, -1.0], [1.1, 1.2, 1.3]]
minimum_match = ["A", "*", "B"]
update = [[(-1, "A"), (1, "B")], "*", [(1, "A"), (-1, "B")]]
rate_0_1 = 3.5
process_0 = KMCBucketProcess(coordinates=coords,
minimum_match=minimum_match,
update=update,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
# A second process.
coords = [[1.0, 2.0, 3.4], [12.0, 13.0, -1.0], [1.1, 1.2, 1.3]]
minimum_match = ["A", "*", "C"]
update = [[(-1, "A"), (1, "C")], "*", [(1, "A"), (-1, "C")]]
rate_0_1 = 1.5
process_1 = KMCBucketProcess(coordinates=coords,
minimum_match=minimum_match,
update=update,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
processes = [process_0, process_1]
# Construct the interactions object.
kmc_interactions = KMCInteractions(processes=processes)
# Check the default implicit wildcard.
self.assertTrue(kmc_interactions.implicitWildcards())
# Construct again with non default value.
kmc_interactions = KMCInteractions(processes=processes,
implicit_wildcards=False)
# Check the wildcard again.
self.assertFalse(kmc_interactions.implicitWildcards())
# Check the processes stored on the object.
stored_processes = kmc_interactions._KMCInteractions__processes
# Checks that the address is the same.
self.assertTrue(stored_processes == processes)
def testConstructionFailsWrongWildcard(self):
""" Test that we fail to construct with other than bool input for the implicit wildcard flag """
# A first process.
coords = [[1.0, 2.0, 3.4], [12.0, 13.0, -1.0], [1.1, 1.2, 1.3]]
types0 = ["A", "*", "B"]
types1 = ["B", "*", "A"]
rate_0_1 = 3.5
process_0 = KMCProcess(coords,
types0,
types1,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
# A second process.
coords = [[1.0, 2.0, 3.4], [1.1, 1.2, 1.3]]
types0 = ["A", "C"]
types1 = ["C", "A"]
rate_0_1 = 1.5
process_1 = KMCProcess(coords,
types0,
types1,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
processes = [process_0, process_1]
# Construct and fail.
self.assertRaises(
Error,
lambda: KMCInteractions(processes=processes, implicit_wildcards=0))
self.assertRaises(
Error, lambda: KMCInteractions(processes=processes,
implicit_wildcards="True"))
self.assertRaises(
Error, lambda: KMCInteractions(processes=processes,
implicit_wildcards=[False]))
def testConstructionFailNoList(self):
""" Test that the construction fails if the interactions list is not a list. """
# Setup.
coords = [[1.0, 2.0, 3.4], [12.0, 13.0, -1.0], [1.1, 1.2, 1.3]]
types0 = ["A", "*", "B"]
types1 = ["B", "*", "A"]
rate_0_1 = 3.5
process_0 = KMCProcess(coords,
types0,
types1,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
# Construct and fail.
self.assertRaises(Error, lambda: KMCInteractions(processes=process_0))
def testBackendNoFailWrongBasisMatch(self):
""" Test for no failure when constructing backend with wrong n_basis """
# A first process.
coords = [[1.0, 2.0, 3.4], [1.1, 1.2, 1.3]]
types0 = ["A", "B"]
types1 = ["B", "A"]
rate_0_1 = 3.5
process_0 = KMCProcess(coords,
types0,
types1,
basis_sites=[0, 4],
rate_constant=rate_0_1)
# A second process.
types0 = ["A", "C"]
types1 = ["C", "A"]
rate_0_1 = 1.5
process_1 = KMCProcess(coords,
types0,
types1,
basis_sites=[0, 1],
rate_constant=rate_0_1)
processes = [process_0, process_1]
# Construct the interactions object.
kmc_interactions = KMCInteractions(processes=processes)
# Setup a dict with the possible types.
possible_types = {
"A": 13,
"D": 2,
"B": 3,
"J": 4,
"C": 5,
}
# Use a dummy argument for the configuration.
config = "DummyConfig"
# Get the backend - The [0,4] sites list for process_0 is simply ignored.
cpp_interactions = kmc_interactions._backend(possible_types, 2, config)
self.assertEqual(len(cpp_interactions.processes()[0].basisSites()), 1)
self.assertEqual(cpp_interactions.processes()[0].basisSites()[0], 0)
self.assertEqual(len(cpp_interactions.processes()[1].basisSites()), 2)
self.assertEqual(cpp_interactions.processes()[1].basisSites()[0], 0)
self.assertEqual(cpp_interactions.processes()[1].basisSites()[1], 1)
def testBackendBuckets(self):
"""
Test that the backend behaves as expected with bucket processes.
"""
# A first process.
coords = [[1.0, 2.0, 3.4], [12.0, 13.0, -1.0], [1.1, 1.2, 1.3]]
minimum_match = ["A", "*", "B"]
update = [[(-1, "A"), (1, "B")], [(0, "*")], [(1, "A"), (-1, "B")]]
rate_0_1 = 3.5
process_0 = KMCBucketProcess(coordinates=coords,
minimum_match=minimum_match,
update=update,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
# A second process.
coords = [[1.0, 2.0, 3.4], [12.0, 13.0, -1.0], [1.1, 1.2, 1.3]]
minimum_match = ["A", "*", "C"]
update = [[(-1, "A"), (1, "C")], [(0, "*")], [(1, "A"), (-1, "C")]]
rate_0_1 = 1.5
process_1 = KMCBucketProcess(coordinates=coords,
minimum_match=minimum_match,
update=update,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
processes = [process_0, process_1]
# Construct the interactions object.
kmc_interactions = KMCInteractions(processes=processes)
# Setup a dict with the possible types.
possible_types = {
"A": 0,
"B": 1,
"J": 2,
"C": 3,
}
# Use a dummy argument for the configuration.
config = "DummyConfig"
# Check that setting up the backend fails if we have types in the processes that do
# not corresponds to a type in the list of possible types.
self.assertRaises(
Error,
lambda: kmc_interactions._backend(possible_types, 2, config))
possible_types = {
"A": 0,
"B": 1,
"D": 2,
"J": 3,
"C": 4,
}
self.assertRaises(
Error,
lambda: kmc_interactions._backend(possible_types, 2, config))
possible_types = {
"A": 0,
"B": 1,
"F": 2,
"J": 3,
"C": 4,
}
self.assertRaises(
Error,
lambda: kmc_interactions._backend(possible_types, 2, config))
# Both "D" and "F" must be present, and the widcard "*".
possible_types["D"] = 5
possible_types["*"] = 6
# Get the backend.
cpp_interactions = kmc_interactions._backend(possible_types, 2, config)
# Check the type.
self.assertTrue(isinstance(cpp_interactions, Backend.Interactions))
# Get the processes out.
cpp_processes = cpp_interactions.processes()
# Check the length of the processes.
self.assertEqual(cpp_processes.size(), 2)
# Get the elements out of the second process.
update_types = cpp_processes[1].processMatchList()[0].update_types
# This is an "A"
match_types = cpp_processes[1].processMatchList()[0].match_types
self.assertEqual(match_types[0], 1)
self.assertEqual(match_types[1], 0)
self.assertEqual(match_types[2], 0)
self.assertEqual(match_types[3], 0)
self.assertEqual(match_types[4], 0)
self.assertEqual(match_types[5], 0)
self.assertEqual(match_types[6], 0)
# This is an "C"
match_types = cpp_processes[1].processMatchList()[1].match_types
self.assertEqual(match_types[0], 0)
self.assertEqual(match_types[1], 0)
self.assertEqual(match_types[2], 0)
self.assertEqual(match_types[3], 0)
self.assertEqual(match_types[4], 1)
self.assertEqual(match_types[5], 0)
self.assertEqual(match_types[6], 0)
# This is a "*"
match_types = cpp_processes[1].processMatchList()[2].match_types
self.assertEqual(match_types[0], 0)
self.assertEqual(match_types[1], 0)
self.assertEqual(match_types[2], 0)
self.assertEqual(match_types[3], 0)
self.assertEqual(match_types[4], 0)
self.assertEqual(match_types[5], 0)
self.assertEqual(match_types[6], 1)
def testBackend(self):
"""
Test that the generated backend object is what we expect.
"""
# A first process.
coords = [[1.0, 2.0, 3.4], [12.0, 13.0, -1.0], [1.1, 1.2, 1.3]]
types0 = ["A", "D", "B"]
types1 = ["B", "F", "A"]
rate_0_1 = 3.5
process_0 = KMCProcess(coords,
types0,
types1,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
# A second process.
coords = [[1.0, 2.0, 3.4], [1.1, 1.2, 1.3]]
types0 = ["A", "C"]
types1 = ["C", "A"]
rate_0_1 = 1.5
process_1 = KMCProcess(coords,
types0,
types1,
basis_sites=[0, 1, 3],
rate_constant=rate_0_1)
processes = [process_0, process_1]
# Construct the interactions object.
kmc_interactions = KMCInteractions(processes=processes)
# Setup a dict with the possible types.
possible_types = {
"A": 13,
"B": 3,
"J": 4,
"C": 5,
}
# Use a dummy argument for the configuration.
config = "DummyConfig"
# Check that setting up the backend fails if we have types in the processes that do
# not corresponds to a type in the list of possible types.
self.assertRaises(
Error,
lambda: kmc_interactions._backend(possible_types, 2, config))
possible_types = {
"A": 13,
"B": 3,
"D": 2,
"J": 4,
"C": 5,
}
self.assertRaises(
Error,
lambda: kmc_interactions._backend(possible_types, 2, config))
possible_types = {
"A": 13,
"B": 3,
"F": 2,
"J": 4,
"C": 5,
}
self.assertRaises(
Error,
lambda: kmc_interactions._backend(possible_types, 2, config))
# Both "D" and "F" must be present.
possible_types["D"] = 123
# Get the backend.
cpp_interactions = kmc_interactions._backend(possible_types, 2, config)
# Check the type.
self.assertTrue(isinstance(cpp_interactions, Backend.Interactions))
# Get the processes out.
cpp_processes = cpp_interactions.processes()
# Check the length of the processes.
self.assertEqual(cpp_processes.size(), 2)
# Get the elements out of the second process.
match_types = cpp_processes[1].processMatchList()[0].match_types
update_types = cpp_processes[1].processMatchList()[0].update_types
# Match type should be "A" -> 15 and update type "C" -> 5
self.assertEqual(match_types[13], 1)
self.assertEqual(update_types[5], 1)
# Get the elements out of the second process.
match_types = cpp_processes[1].processMatchList()[1].match_types
update_types = cpp_processes[1].processMatchList()[1].update_types
# Match type should be "C" -> 5 and update type "A" -> 13
self.assertEqual(match_types[5], 1)
self.assertEqual(update_types[13], 1)
def testBackendWithCustomRates(self):
""" Test that we can construct the backend object with custom rates. """
# A first process.
coords = [[1.0, 2.0, 3.4], [1.1, 1.2, 1.3]]
types0 = ["A", "B"]
types1 = ["B", "A"]
rate_0_1 = 3.5
process_0 = KMCProcess(coords,
types0,
types1,
basis_sites=[0],
rate_constant=rate_0_1)
# A second process.
types0 = ["A", "C"]
types1 = ["C", "A"]
rate_0_1 = 1.5
process_1 = KMCProcess(coords,
types0,
types1,
basis_sites=[0],
rate_constant=rate_0_1)
processes = [process_0, process_1]
# Set the custom rates class to use.
custom_rates_class = CustomRateCalculator
# Construct the interactions object.
kmc_interactions = KMCInteractions(processes=processes,
implicit_wildcards=False)
kmc_interactions.setRateCalculator(rate_calculator=custom_rates_class)
# Set the rate function on the custom rates calculator for testing.
ref_rnd = numpy.random.uniform(0.0, 1.0)
def testRateFunction(coords, types_before, types_after, rate_const,
process_number, global_coordinate):
return ref_rnd
kmc_interactions._KMCInteractions__rate_calculator.rate = testRateFunction
# Setup a dict with the possible types.
possible_types = {
"A": 13,
"D": 2,
"B": 3,
"J": 4,
"C": 5,
}
# Construct the backend.
cpp_interactions = kmc_interactions._backend(possible_types, 2)
# Get the rate calculator reference out of the C++ object and
# check that a call from C++ calls the Python extension.
cpp_coords = Backend.StdVectorDouble()
cpp_types1 = Backend.StdVectorString()
cpp_types2 = Backend.StdVectorString()
rate_constant = 543.2211
process_number = 33
coordinate = (1.2, 3.4, 5.6)
self.assertAlmostEqual(
cpp_interactions.rateCalculator().backendRateCallback(
cpp_coords,
cpp_coords.size() / 3, cpp_types1, cpp_types2, rate_constant,
process_number, coordinate[0], coordinate[1], coordinate[2]),
ref_rnd, 12)
self.assertAlmostEqual(
kmc_interactions._KMCInteractions__rate_calculator.
backendRateCallback(cpp_coords,
cpp_coords.size() / 3, cpp_types1, cpp_types2,
rate_constant, process_number, coordinate[0],
coordinate[1], coordinate[2]), ref_rnd, 12)
# Construct a C++ RateCalculator object directly and check that this object
# returns the rate given to it.
cpp_rate_calculator = Backend.RateCalculator()
self.assertAlmostEqual(
cpp_rate_calculator.backendRateCallback(
cpp_coords,
cpp_coords.size() / 3, cpp_types1, cpp_types2, rate_constant,
process_number, coordinate[0], coordinate[1], coordinate[2]),
rate_constant, 12)
def testRun2(self):
""" Test the run of an A-B flip model. """
# Cell.
cell_vectors = [[ 1.000000e+00, 0.000000e+00, 0.000000e+00],
[ 0.000000e+00, 1.000000e+00, 0.000000e+00],
[ 0.000000e+00, 0.000000e+00, 1.000000e+00]]
basis_points = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
unit_cell = KMCUnitCell(
cell_vectors=cell_vectors,
basis_points=basis_points)
# Lattice.
lattice = KMCLattice(
unit_cell=unit_cell,
repetitions=(10,10,1),
periodic=(True, True, False))
# Configuration.
types = ['B']*100
possible_types = ['A','B']
configuration = KMCConfiguration(
lattice=lattice,
types=types,
possible_types=possible_types)
# Interactions.
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
process_0 = KMCProcess(coordinates,
['A'],
['B'],
basis_sites=[0],
rate_constant=4.0)
process_1 = KMCProcess(coordinates,
['B'],
['A'],
basis_sites=[0],
rate_constant=1.0)
processes = [process_0, process_1]
interactions = KMCInteractions(processes)
# Setup the model.
ab_flip_model = KMCLatticeModel(configuration, interactions)
# Run the model with a trajectory file.
name = os.path.abspath(os.path.dirname(__file__))
name = os.path.join(name, "..", "TestUtilities", "Scratch")
trajectory_filename = os.path.join(name, "ab_flip_traj.py")
self.__files_to_remove.append(trajectory_filename)
# The control parameters.
control_parameters = KMCControlParameters(number_of_steps=1000,
dump_interval=500,
seed=2013)
# Run the model for 1000 steps.
ab_flip_model.run(control_parameters,
trajectory_filename=trajectory_filename)
# Read the first last frames from the trajectory file and check that
# the fraction of A is close to 20% in the last, and 0 in the first.
if MPICommons.isMaster():
global_dict = {}
local_dict = {}
execfile(trajectory_filename, global_dict, local_dict)
# Count the first frame.
elem = local_dict["types"][0]
nA = len([ ee for ee in elem if ee == "A" ])
nB = len([ ee for ee in elem if ee == "B" ])
self.assertEqual(nA, 0)
self.assertEqual(nB, 100)
# Count the last frame.
elem = local_dict["types"][-1]
nA = len([ ee for ee in elem if ee == "A" ])
nB = len([ ee for ee in elem if ee == "B" ])
# Note that the average should be 20.0% over a long run.
# It is pure luck that it is exact at this particular
# step with the presently used random number seed.
fraction = nA * 100.0 / (nA + nB)
target = 20.0
self.assertAlmostEqual(fraction, target, 3)
def testRunImplicitWildcards(self):
""" Test that ta valid model can run for a few steps. """
# Setup a unitcell.
unit_cell = KMCUnitCell(cell_vectors=numpy.array([[1.0,0.0,0.0],
[0.0,1.0,0.0],
[0.0,0.0,1.0]]),
basis_points=[[0.0,0.0,0.0]])
# And a lattice.
lattice = KMCLattice(unit_cell=unit_cell,
repetitions=(10,10,1),
periodic=(True,True,False))
# Set the stating configuration types.
types = ['B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B',
'B', 'B', 'B', 'B', 'B', 'A', 'B', 'B', 'B', 'B',
'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B',
'B', 'A', 'B', 'B', 'B', 'B', 'B', 'A', 'B', 'B',
'B', 'B', 'B', 'B', 'B', 'B', 'A', 'B', 'B', 'B',
'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B',
'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B',
'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B',
'B', 'B', 'B', 'B', 'B', 'B', 'B', 'A', 'B', 'B',
'A', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'A']
# Setup the configuration.
config = KMCConfiguration(lattice=lattice,
types=types,
possible_types=["A","B"])
# Generate the interactions with a distance so large that we get a
# layer of implicite wildcards in the C++ matchlists.
sites = [0]
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00],
[ 1.000000e+00, 0.000000e+00, 0.000000e+00],
[ -1.000000e+00, 0.000000e+00, 0.000000e+00],
[ 0.000000e+00, -1.000000e+00, 0.000000e+00],
[ 0.000000e+00, 1.000000e+00, 0.000000e+00],
[ 2.000000e+00, 2.000000e+00, 0.000000e+00]]
types0 = ['A', 'B', 'B', 'B', 'B', 'A']
types1 = ['B', 'B', 'A', 'B', 'B', 'A']
process_0 = KMCProcess(coordinates, types0, types1, None, sites, 1.0)
types0 = ['A', 'B', 'B', 'B', 'B', 'B']
types1 = ['B', 'B', 'A', 'B', 'B', 'B']
process_1 = KMCProcess(coordinates, types0, types1, None, sites, 1.0)
types0 = ['A', 'B', 'B', 'B', 'B', 'B']
types1 = ['B', 'B', 'B', 'A', 'B', 'B']
process_2 = KMCProcess(coordinates, types0, types1, None, sites, 1.0)
types0 = ['A', 'B', 'B', 'B', 'B', 'B']
types1 = ['B', 'B', 'B', 'B', 'A', 'B']
process_3 = KMCProcess(coordinates, types0, types1, None, sites, 1.0)
# Processes.
processes = [process_0,
process_1,
process_2,
process_3]
# No implicit wildcards.
interactions = KMCInteractions(processes=processes,
implicit_wildcards=False)
# Create the model.
model = KMCLatticeModel(config, interactions)
# Get the match types out.
match_types = [ l.match_type for l in model._backend().interactions().processes()[0].minimalMatchList() ]
# This does not have wildcards added.
ref_match_types = [1, 2, 2, 2, 2, 1]
self.assertEqual( match_types, ref_match_types )
# Create with implicit wildcards - this is default behavior.
interactions = KMCInteractions(processes=processes)
# Create the model.
model = KMCLatticeModel(config, interactions)
# Check the process matchlists again.
match_types = [ l.match_type for l in model._backend().interactions().processes()[0].minimalMatchList() ]
ref_match_types = [1, 2, 2, 2, 2,
0, 0, 0, 0, 0,
0, 0, 0, 0, 0,
0, 0, 0, 0, 0,
0, 0, 0, 0, 1]
# This one has the wildcards (zeroes) added.
self.assertEqual( match_types, ref_match_types )
# Setup the run paramters.
control_parameters = KMCControlParameters(number_of_steps=10,
dump_interval=1)
model.run(control_parameters)
def testRunFailAnalysis(self):
""" Test that the analyis plugins get called correctly. """
# Cell.
cell_vectors = [[ 1.000000e+00, 0.000000e+00, 0.000000e+00],
[ 0.000000e+00, 1.000000e+00, 0.000000e+00],
[ 0.000000e+00, 0.000000e+00, 1.000000e+00]]
basis_points = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
unit_cell = KMCUnitCell(
cell_vectors=cell_vectors,
basis_points=basis_points)
# Lattice.
lattice = KMCLattice(
unit_cell=unit_cell,
repetitions=(10,10,1),
periodic=(True, True, False))
# Configuration.
types = ['B']*100
possible_types = ['A','B']
configuration = KMCConfiguration(
lattice=lattice,
types=types,
possible_types=possible_types)
# Interactions.
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
process_0 = KMCProcess(coordinates,
['A'],
['B'],
basis_sites=[0],
rate_constant=4.0)
process_1 = KMCProcess(coordinates,
['B'],
['A'],
basis_sites=[0],
rate_constant=1.0)
processes = [process_0, process_1]
interactions = KMCInteractions(processes)
# Setup the model.
ab_flip_model = KMCLatticeModel(configuration, interactions)
# The control parameters.
control_parameters = KMCControlParameters(number_of_steps=1000,
dump_interval=500)
# Setup a valid minimal analysis object.
class AnalysisProxy1(KMCAnalysisPlugin):
def __init__(self):
pass
# Fail because of not instantitated analysis.
self.assertRaises( Error,
lambda : ab_flip_model.run(control_parameters,
analysis=[AnalysisProxy1]) )
# Fail because of empty analyis list.
self.assertRaises( Error,
lambda : ab_flip_model.run(control_parameters,
analysis=[]) )
# Fail because of not list.
self.assertRaises( Error,
lambda : ab_flip_model.run(control_parameters,
analysis=AnalysisProxy1()) )
# Fail because of wrong type.
self.assertRaises( Error,
lambda : ab_flip_model.run(control_parameters,
analysis=[AnalysisProxy1(), "AP3"]) )
def testCustomRatesRun(self):
""" Test the run of an A-B flip model with custom rates. """
# Cell.
cell_vectors = [[ 1.000000e+00, 0.000000e+00, 0.000000e+00],
[ 0.000000e+00, 1.000000e+00, 0.000000e+00],
[ 0.000000e+00, 0.000000e+00, 1.000000e+00]]
basis_points = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
unit_cell = KMCUnitCell(
cell_vectors=cell_vectors,
basis_points=basis_points)
# Lattice.
lattice = KMCLattice(
unit_cell=unit_cell,
repetitions=(10,10,1),
periodic=(True, True, False))
# Configuration.
types = ['B']*100
possible_types = ['A','B']
configuration = KMCConfiguration(
lattice=lattice,
types=types,
possible_types=possible_types)
# Interactions.
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
process_0 = KMCProcess(coordinates, ['A'], ['B'], None, [0], 4.0)
process_1 = KMCProcess(coordinates, ['B'], ['A'], None, [0], 1.0)
processes = [process_0, process_1]
interactions = KMCInteractions(processes,
implicit_wildcards=True)
# Custom rates.
rate_calculator = CustomRateCalculator
interactions.setRateCalculator(rate_calculator)
# Setup the model.
ab_flip_model = KMCLatticeModel(configuration, interactions)
# Run the model with a trajectory file.
name = os.path.abspath(os.path.dirname(__file__))
name = os.path.join(name, "..", "TestUtilities", "Scratch")
trajectory_filename = os.path.join(name, "ab_flip_traj_custom.py")
self.__files_to_remove.append(trajectory_filename)
# The control parameters.
control_parameters = KMCControlParameters(number_of_steps=1000,
dump_interval=500,
seed=2013)
# Run the model for 1000 steps.
ab_flip_model.run(control_parameters,
trajectory_filename=trajectory_filename)
# Read the first last frames from the trajectory file and check that
# the fraction of A is close to 75% in the last, and 0 in the first.
if MPICommons.isMaster():
global_dict = {}
local_dict = {}
execfile(trajectory_filename, global_dict, local_dict)
# Count the first frame.
elem = local_dict["types"][0]
nA = len([ ee for ee in elem if ee == "A" ])
nB = len([ ee for ee in elem if ee == "B" ])
self.assertEqual(nA, 0)
self.assertEqual(nB, 100)
# Count the last frame.
elem = local_dict["types"][-1]
nA = len([ ee for ee in elem if ee == "A" ])
nB = len([ ee for ee in elem if ee == "B" ])
# Note that the average over a long simulation should be
# 75% A using the modified rate function. In this particular
# step the A population is 74%.
value = 1.0 * nA / (nA + nB)
self.assertAlmostEqual(0.74, value, 2)
def testCalculationNonOrthogonal(self):
""" Test a calculation with the on-the-fly MSD analysis. """
# Setup a system, a periodic 10 atoms long 1D chain.
unit_cell = KMCUnitCell(cell_vectors=numpy.array([[2.0, 1.0, 0.0],
[0.0, 1.0, 0.0],
[0.0, 2.0, 1.0]]),
basis_points=[[0.0, 0.0, 0.0]])
# And a lattice.
lattice = KMCLattice(unit_cell=unit_cell,
repetitions=(10, 10, 10),
periodic=(True, True, True))
# Setup an initial configuration with one B in a sea of A:s.
types = ["A"] * 10 * 10 * 10
types[5] = "B"
config = KMCConfiguration(lattice=lattice,
types=types,
possible_types=["A", "B"])
# Setup a diffusion process to the left.
coordinates_p0 = [[0.0, 0.0, 0.0], [-1.0, 0.0, 0.0]]
p0 = KMCProcess(coordinates=coordinates_p0,
elements_before=["B", "A"],
elements_after=["A", "B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p1 = [[0.0, 0.0, 0.0], [1.0, 0.0, 0.0]]
p1 = KMCProcess(coordinates=coordinates_p1,
elements_before=["B", "A"],
elements_after=["A", "B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p2 = [[0.0, 0.0, 0.0], [0.0, -1.0, 0.0]]
p2 = KMCProcess(coordinates=coordinates_p2,
elements_before=["B", "A"],
elements_after=["A", "B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p3 = [[0.0, 0.0, 0.0], [0.0, 1.0, 0.0]]
p3 = KMCProcess(coordinates=coordinates_p3,
elements_before=["B", "A"],
elements_after=["A", "B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p4 = [[0.0, 0.0, 0.0], [0.0, 0.0, -1.0]]
p4 = KMCProcess(coordinates=coordinates_p4,
elements_before=["B", "A"],
elements_after=["A", "B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p5 = [[0.0, 0.0, 0.0], [0.0, 0.0, 1.0]]
p5 = KMCProcess(coordinates=coordinates_p5,
elements_before=["B", "A"],
elements_after=["A", "B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
interactions = KMCInteractions(processes=[p0, p1, p2, p3, p4, p5],
implicit_wildcards=True)
model = KMCLatticeModel(configuration=config,
interactions=interactions)
# Setup the analysis.
msd = OnTheFlyMSD(history_steps=400,
n_bins=10,
t_max=25.0,
track_type="B")
# Setup the control parameters.
control_parameters = KMCControlParameters(number_of_steps=4000,
dump_interval=100,
analysis_interval=1,
seed=2013)
# Run the model.
model.run(control_parameters=control_parameters, analysis=[msd])
# Get the results out.
results = msd.results()
time_steps = msd.timeSteps()
std_dev = msd.stdDev()
bin_counters = msd.binCounters()
# Compare against references.
ref_results = numpy.array(
[[
10.36856672, 28.37170047, 44.38285764, 58.94374243,
72.11121737, 85.02318805, 98.58352498, 115.27324368,
126.49573433, 138.47100039
],
[
16.05138132, 44.90792913, 68.59720617, 93.35093194,
123.68147594, 148.63693439, 172.54794759, 203.0082052,
232.5195128, 267.73366007
],
[
2.59976568, 7.35085681, 11.8568078, 16.56678774, 21.77700882,
26.34223625, 30.40786814, 35.09199292, 40.24434182,
46.48414664
],
[
26.41994804, 73.2796296, 112.98006381, 152.29467437,
195.79269332, 233.66012244, 271.13147258, 318.28144888,
359.01524713, 406.20466046
],
[
12.9683324, 35.72255728, 56.23966543, 75.51053018,
93.88822619, 111.3654243, 128.99139313, 150.3652366,
166.74007615, 184.95514703
],
[
18.651147, 52.25878594, 80.45401397, 109.91771968,
145.45848476, 174.97917064, 202.95581574, 238.10019812,
272.76385461, 314.21780672
],
[
29.01971372, 80.63048641, 124.8368716, 168.86146212,
217.56970213, 260.0023587, 301.53934072, 353.3734418,
399.25958894, 452.68880711
]])
diff = numpy.linalg.norm(ref_results - results)
self.assertAlmostEqual(diff, 0.0, 6)
ref_time_steps = numpy.array(
[1.25, 3.75, 6.25, 8.75, 11.25, 13.75, 16.25, 18.75, 21.25, 23.75])
diff = numpy.linalg.norm(ref_time_steps - time_steps)
self.assertAlmostEqual(diff, 0.0, 8)
ref_std_dev = numpy.array([
[
0.55062327, 2.49729593, 4.98805342, 7.82646477, 10.85497401,
14.13906215, 17.83387379, 22.42069464, 26.20482295, 30.35650154
],
[
0.85240944, 3.95282576, 7.70942988, 12.39500157, 18.61789684,
24.71780818, 31.21412347, 39.48518175, 48.16868093, 58.69429153
],
[
0.13806069, 0.6470273, 1.33255031, 2.19971409, 3.27811502,
4.38062279, 5.50081855, 6.82540746, 8.33700723, 10.19055301
],
[
0.99209394, 4.56092478, 8.97847654, 14.29873598, 20.84046684,
27.47595651, 34.68217147, 43.77406499, 52.59000894, 62.96841965
],
[
0.4869731, 2.22337228, 4.46934176, 7.08957906, 9.99360309,
13.09539481, 16.50011919, 20.68011712, 24.42476236, 28.67109723
],
[
0.70036814, 3.25258729, 6.3936455, 10.32002241, 15.48281847,
20.57569786, 25.96138447, 32.74653167, 39.95555528, 48.70894069
],
[
0.88975069, 4.09754088, 8.10024368, 12.94487456, 18.90879051,
24.96317829, 31.49377349, 39.68202523, 47.75293586, 57.29701788
]
])
diff = numpy.linalg.norm(ref_std_dev - std_dev)
self.assertAlmostEqual(diff, 0.0, 6)
ref_bin_counters = (58893, 58531, 57985, 58641, 57509, 57659, 57396,
57037, 56732, 56518)
self.assertEqual(bin_counters, ref_bin_counters)
def testCalculationNonOrthogonal(self):
""" Test a calculation with the on-the-fly MSD analysis. """
# Setup a system, a periodic 10 atoms long 1D chain.
unit_cell = KMCUnitCell(cell_vectors=numpy.array([[2.0,1.0,0.0],
[0.0,1.0,0.0],
[0.0,2.0,1.0]]),
basis_points=[[0.0,0.0,0.0]])
# And a lattice.
lattice = KMCLattice(unit_cell=unit_cell,
repetitions=(10,10,10),
periodic=(True,True,True))
# Setup an initial configuration with one B in a sea of A:s.
types = ["A"]*10*10*10
types[5] = "B"
config = KMCConfiguration(lattice=lattice,
types=types,
possible_types=["A","B"])
# Setup a diffusion process to the left.
coordinates_p0 = [[0.0, 0.0, 0.0],[-1.0, 0.0, 0.0]]
p0 = KMCProcess(coordinates=coordinates_p0,
elements_before=["B","A"],
elements_after=["A","B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p1 = [[0.0, 0.0, 0.0],[1.0, 0.0, 0.0]]
p1 = KMCProcess(coordinates=coordinates_p1,
elements_before=["B","A"],
elements_after=["A","B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p2 = [[0.0, 0.0, 0.0],[0.0,-1.0, 0.0]]
p2 = KMCProcess(coordinates=coordinates_p2,
elements_before=["B","A"],
elements_after=["A","B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p3 = [[0.0, 0.0, 0.0],[0.0, 1.0, 0.0]]
p3 = KMCProcess(coordinates=coordinates_p3,
elements_before=["B","A"],
elements_after=["A","B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p4 = [[0.0, 0.0, 0.0],[0.0, 0.0,-1.0]]
p4 = KMCProcess(coordinates=coordinates_p4,
elements_before=["B","A"],
elements_after=["A","B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p5 = [[0.0, 0.0, 0.0],[0.0, 0.0, 1.0]]
p5 = KMCProcess(coordinates=coordinates_p5,
elements_before=["B","A"],
elements_after=["A","B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
interactions = KMCInteractions(processes=[p0, p1, p2, p3, p4, p5],
implicit_wildcards=True)
model = KMCLatticeModel(configuration=config,
interactions=interactions)
# Setup the analysis.
msd = OnTheFlyMSD(history_steps=400,
n_bins=10,
t_max=25.0,
track_type="B")
# Setup the control parameters.
control_parameters = KMCControlParameters(number_of_steps=4000,
dump_interval=100,
analysis_interval=1,
seed=2013)
# Run the model.
model.run(control_parameters=control_parameters,
analysis=[msd])
# Get the results out.
results = msd.results()
time_steps = msd.timeSteps()
std_dev = msd.stdDev()
bin_counters = msd.binCounters()
# Compare against references.
ref_results = numpy.array([[ 11.10849324 , 32.89332622 , 56.86314552 , 84.47799379 , 110.30394149,
135.26081658, 162.25612518, 189.40268295, 219.32538873, 252.00244119],
[ 17.33153679 , 50.69982999 , 82.74958435 , 111.33382274 , 136.44824258,
164.84499628, 193.94367236, 234.25340515, 272.64453361, 302.94924061],
[ 2.98786918 , 8.65022668 , 14.01808716 , 18.62678885 , 22.65708384,
26.03107861, 30.03937616, 35.1483 , 40.68319007, 47.21074474],
[ 28.44003004 , 83.59315621 , 139.61272987 , 195.81181652 , 246.75218407,
300.10581285, 356.19979754, 423.65608811, 491.96992234, 554.95168181],
[ 14.09636242 , 41.5435529 , 70.88123268 , 103.10478264 , 132.96102533,
161.29189519, 192.29550134, 224.55098295, 260.00857879, 299.21318593],
[ 20.31940597 , 59.35005667 , 96.76767151 , 129.96061158 , 159.10532643,
190.87607489, 223.98304852, 269.40170515, 313.32772368, 350.15998535],
[ 31.42789922 , 92.24338289 , 153.63081703 , 214.43860537 , 269.40926791,
326.13689146, 386.2391737 , 458.8043881 , 532.65311241, 602.16242655]])
diff = numpy.linalg.norm(ref_results - results)
self.assertAlmostEqual(diff, 0.0, 6)
ref_time_steps = numpy.array([ 1.25, 3.75, 6.25, 8.75, 11.25, 13.75, 16.25, 18.75, 21.25, 23.75])
diff = numpy.linalg.norm(ref_time_steps - time_steps)
self.assertAlmostEqual(diff, 0.0, 8)
ref_std_dev = numpy.array([[ 0.59348826 , 2.92235801 , 6.47651272 , 11.36067446 , 16.79740936,
22.78018427, 29.72954492, 37.27728411, 45.97782684, 55.88661276],
[ 0.92596389 , 4.50435001 , 9.42488725 , 14.97226982 , 20.7787406,
27.76265504, 35.53552825, 46.10457783, 57.15527613, 67.18509081],
[ 0.15963149 , 0.76851636 , 1.59661093 , 2.50494685 , 3.45028752,
4.38406911 , 5.5039955 , 6.91771175 , 8.52853689 , 10.46993274],
[ 1.07441492 , 5.2514756 , 11.24398775 , 18.62020347 , 26.57035045,
35.73918442, 46.14937581, 58.95988 , 72.92611647, 87.02483617],
[ 0.53253608 , 2.60984229 , 5.70856048 , 9.80447485 , 14.31728377,
19.20802777, 24.91387536, 31.25058126, 38.54181941, 46.9211633 ],
[ 0.76763185 , 3.72847956 , 7.7933761 , 12.35825842 , 17.13251009,
22.73116664, 29.01932554, 37.49242051, 46.4454696 , 54.91039375],
[ 0.96941939 , 4.731515 , 10.1024813 , 16.6495642 , 23.68662473,
31.71206536, 40.85854085, 52.13448317, 64.46787783, 77.10029967]])
diff = numpy.linalg.norm(ref_std_dev - std_dev)
self.assertAlmostEqual(diff, 0.0, 6)
ref_bin_counters = (59930, 59996, 59545, 59256, 59064, 59076, 58284, 58294, 57817, 57349)
self.assertEqual(bin_counters, ref_bin_counters)
def testRunRngTypeDevice(self):
""" Test to use the PRNG DEVICE. """
# Cell.
cell_vectors = [[ 1.000000e+00, 0.000000e+00, 0.000000e+00],
[ 0.000000e+00, 1.000000e+00, 0.000000e+00],
[ 0.000000e+00, 0.000000e+00, 1.000000e+00]]
basis_points = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
unit_cell = KMCUnitCell(
cell_vectors=cell_vectors,
basis_points=basis_points)
# Lattice.
lattice = KMCLattice(
unit_cell=unit_cell,
repetitions=(4,4,1),
periodic=(True, True, False))
# Configuration.
types = ['B']*16
possible_types = ['A','B']
configuration = KMCConfiguration(
lattice=lattice,
types=types,
possible_types=possible_types)
# Interactions.
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
process_0 = KMCProcess(coordinates,
['A'],
['B'],
basis_sites=[0],
rate_constant=4.0)
process_1 = KMCProcess(coordinates,
['B'],
['A'],
basis_sites=[0],
rate_constant=1.0)
processes = [process_0, process_1]
interactions = KMCInteractions(processes)
# Setup the model.
ab_flip_model_DEVICE = KMCLatticeModel(configuration, interactions)
support_device = False
if (not support_device):
# If DEVICE is not supported on your system this is the test you should run.
self.assertRaises( Error,
lambda: ab_flip_model_DEVICE.run(KMCControlParameters(number_of_steps=10000,
dump_interval=5000,
seed=2013,
rng_type="DEVICE")))
else:
# If DEVICE is supported the aboove test will fail, and you should run this tests instead.
# Run the model for 10000 steps with DEVICE.
ab_flip_model_DEVICE.run(KMCControlParameters(number_of_steps=10000,
dump_interval=5000,
seed=2013,
rng_type="DEVICE"))
# Get the simulation time out.
t_DEVICE = ab_flip_model_DEVICE._KMCLatticeModel__cpp_timer.simulationTime()
self.assertTrue(t_DEVICE < 410.0 and t_DEVICE > 370.0)
def testScript(self):
""" Test that the KMCInteractions can generate its own script. """
# A first process.
coords = [[1.0, 2.0, 3.4], [1.1, 1.2, 1.3]]
types0 = ["A", "B"]
types1 = ["B", "A"]
rate_0_1 = 3.5
process_0 = KMCProcess(coords,
types0,
types1,
basis_sites=[0],
rate_constant=rate_0_1)
# A second process.
types0 = ["A", "C"]
types1 = ["C", "A"]
rate_0_1 = 1.5
process_1 = KMCProcess(coords,
types0,
types1,
basis_sites=[0],
rate_constant=rate_0_1)
processes = [process_0, process_1]
# Construct the interactions object.
kmc_interactions = KMCInteractions(processes=processes)
script = kmc_interactions._script()
ref_script = """
# -----------------------------------------------------------------------------
# Interactions
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00],
[ 1.000000e-01, -8.000000e-01, -2.100000e+00]]
elements_before = ['A','B']
elements_after = ['B','A']
move_vectors = [( 0,[ 1.000000e-01, -8.000000e-01, -2.100000e+00]),
( 1,[ -1.000000e-01, 8.000000e-01, 2.100000e+00])]
basis_sites = [0]
rate_constant = 3.500000e+00
process_0 = KMCProcess(
coordinates=coordinates,
elements_before=elements_before,
elements_after=elements_after,
move_vectors=move_vectors,
basis_sites=basis_sites,
rate_constant=rate_constant)
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00],
[ 1.000000e-01, -8.000000e-01, -2.100000e+00]]
elements_before = ['A','C']
elements_after = ['C','A']
move_vectors = [( 0,[ 1.000000e-01, -8.000000e-01, -2.100000e+00]),
( 1,[ -1.000000e-01, 8.000000e-01, 2.100000e+00])]
basis_sites = [0]
rate_constant = 1.500000e+00
process_1 = KMCProcess(
coordinates=coordinates,
elements_before=elements_before,
elements_after=elements_after,
move_vectors=move_vectors,
basis_sites=basis_sites,
rate_constant=rate_constant)
processes = [process_0,
process_1]
interactions = KMCInteractions(
processes=processes,
implicit_wildcards=True)
"""
self.assertEqual(script, ref_script)
# Get a script with another name and another number of processes.
processes = [process_0]
kmc_interactions = KMCInteractions(processes=processes,
implicit_wildcards=False)
script = kmc_interactions._script(variable_name="my_kmc_interactions")
ref_script = """
# -----------------------------------------------------------------------------
# Interactions
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00],
[ 1.000000e-01, -8.000000e-01, -2.100000e+00]]
elements_before = ['A','B']
elements_after = ['B','A']
move_vectors = [( 0,[ 1.000000e-01, -8.000000e-01, -2.100000e+00]),
( 1,[ -1.000000e-01, 8.000000e-01, 2.100000e+00])]
basis_sites = [0]
rate_constant = 3.500000e+00
process_0 = KMCProcess(
coordinates=coordinates,
elements_before=elements_before,
elements_after=elements_after,
move_vectors=move_vectors,
basis_sites=basis_sites,
rate_constant=rate_constant)
processes = [process_0]
my_kmc_interactions = KMCInteractions(
processes=processes,
implicit_wildcards=False)
"""
self.assertEqual(script, ref_script)
def testRunRngType(self):
""" Test that it is possible to run with each of the supported PRNG:s. """
# Cell.
cell_vectors = [[ 1.000000e+00, 0.000000e+00, 0.000000e+00],
[ 0.000000e+00, 1.000000e+00, 0.000000e+00],
[ 0.000000e+00, 0.000000e+00, 1.000000e+00]]
basis_points = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
unit_cell = KMCUnitCell(
cell_vectors=cell_vectors,
basis_points=basis_points)
# Lattice.
lattice = KMCLattice(
unit_cell=unit_cell,
repetitions=(4,4,1),
periodic=(True, True, False))
# Configuration.
types = ['B']*16
possible_types = ['A','B']
configuration = KMCConfiguration(
lattice=lattice,
types=types,
possible_types=possible_types)
# Interactions.
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
process_0 = KMCProcess(coordinates,
['A'],
['B'],
basis_sites=[0],
rate_constant=4.0)
process_1 = KMCProcess(coordinates,
['B'],
['A'],
basis_sites=[0],
rate_constant=1.0)
processes = [process_0, process_1]
interactions = KMCInteractions(processes)
# Setup the models.
ab_flip_model_MT = KMCLatticeModel(configuration, interactions)
ab_flip_model_RANLUX24 = KMCLatticeModel(configuration, interactions)
ab_flip_model_RANLUX48 = KMCLatticeModel(configuration, interactions)
ab_flip_model_MINSTD = KMCLatticeModel(configuration, interactions)
ab_flip_model_DEVICE = KMCLatticeModel(configuration, interactions)
# Run the model for 10000 steps with MT.
ab_flip_model_MT.run(KMCControlParameters(number_of_steps=10000,
dump_interval=5000,
seed=2013,
rng_type="MT"))
# Get the simulation time out.
t_MT = ab_flip_model_MT._KMCLatticeModel__cpp_timer.simulationTime()
# Run the model for 10000 steps with RANLUX24.
ab_flip_model_RANLUX24.run(KMCControlParameters(number_of_steps=10000,
dump_interval=5000,
seed=2013,
rng_type="RANLUX24"))
# Get the simulation time out.
t_RANLUX24 = ab_flip_model_RANLUX24._KMCLatticeModel__cpp_timer.simulationTime()
# Run the model for 10000 steps with RANLUX48.
ab_flip_model_RANLUX48.run(KMCControlParameters(number_of_steps=10000,
dump_interval=5000,
seed=2013,
rng_type="RANLUX48"))
# Get the simulation time out.
t_RANLUX48 = ab_flip_model_RANLUX48._KMCLatticeModel__cpp_timer.simulationTime()
# Run the model for 10000 steps with MINSTD.
ab_flip_model_MINSTD.run(KMCControlParameters(number_of_steps=10000,
dump_interval=5000,
seed=2013,
rng_type="MINSTD"))
# Get the simulation time out.
t_MINSTD = ab_flip_model_MINSTD._KMCLatticeModel__cpp_timer.simulationTime()
# These values should be simillar but not equal. Check against hardcoded values.
self.assertAlmostEqual(t_MT, 394.569398158, 5)
self.assertAlmostEqual(t_RANLUX24, 389.712162523, 5)
self.assertAlmostEqual(t_RANLUX48, 390.544423280, 5)
self.assertAlmostEqual(t_MINSTD, 384.712302086, 5)
def testRunTrajectory(self):
""" Test the run of an A-B flip model with trajectory output. """
# Cell.
cell_vectors = [[ 1.000000e+00, 0.000000e+00, 0.000000e+00],
[ 0.000000e+00, 1.000000e+00, 0.000000e+00],
[ 0.000000e+00, 0.000000e+00, 1.000000e+00]]
basis_points = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
unit_cell = KMCUnitCell(
cell_vectors=cell_vectors,
basis_points=basis_points)
# Lattice.
lattice = KMCLattice(
unit_cell=unit_cell,
repetitions=(4,4,1),
periodic=(True, True, False))
# Configuration.
types = ['B']*16
possible_types = ['A','B']
configuration = KMCConfiguration(
lattice=lattice,
types=types,
possible_types=possible_types)
# Interactions.
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
process_0 = KMCProcess(coordinates,
['A'],
['B'],
basis_sites=[0],
rate_constant=4.0)
process_1 = KMCProcess(coordinates,
['B'],
['A'],
basis_sites=[0],
rate_constant=1.0)
processes = [process_0, process_1]
interactions = KMCInteractions(processes)
# Setup the model.
ab_flip_model = KMCLatticeModel(configuration, interactions)
# Construct the trajectory fileames.
name = os.path.abspath(os.path.dirname(__file__))
name = os.path.join(name, "..", "TestUtilities", "Scratch")
lattice_trajectory_filename = os.path.join(name, "ab_flip_traj_lattice.py")
xyz_trajectory_filename = os.path.join(name, "ab_flip_traj_xyz.xyz")
self.__files_to_remove.append(lattice_trajectory_filename)
self.__files_to_remove.append(xyz_trajectory_filename)
# The control parameters.
control_parameters = KMCControlParameters(number_of_steps=1000,
dump_interval=500,
seed=2013)
# Run the model for 1000 steps with a lattice trajectory.
ab_flip_model.run(control_parameters,
trajectory_filename=lattice_trajectory_filename,
trajectory_type='lattice')
# Check the file content.
ref_lattice = """
sites=[[ 0.000000, 0.000000, 0.000000],
[ 0.000000, 1.000000, 0.000000],
[ 0.000000, 2.000000, 0.000000],
[ 0.000000, 3.000000, 0.000000],
[ 1.000000, 0.000000, 0.000000],
[ 1.000000, 1.000000, 0.000000],
[ 1.000000, 2.000000, 0.000000],
[ 1.000000, 3.000000, 0.000000],
[ 2.000000, 0.000000, 0.000000],
[ 2.000000, 1.000000, 0.000000],
[ 2.000000, 2.000000, 0.000000],
[ 2.000000, 3.000000, 0.000000],
[ 3.000000, 0.000000, 0.000000],
[ 3.000000, 1.000000, 0.000000],
[ 3.000000, 2.000000, 0.000000],
[ 3.000000, 3.000000, 0.000000]]
times=[]
steps=[]
types=[]
times.append( 0.0000000000e+00)
steps.append(0)
types.append(["B","B","B","B","B","B","B","B","B","B","B","B","B","B","B","B"])
times.append( 1.9093684175e+01)
steps.append(500)
types.append(["B","A","B","B","A","B","B","B","A","B","B","B","B","B","B","A"])
times.append( 4.0622006972e+01)
steps.append(1000)
types.append(["B","A","A","B","B","B","B","A","B","B","B","A","B","B","B","B"])
"""
with open(lattice_trajectory_filename, "r") as t:
lattice_data = t.read()
# Check with "in" to avoid comparing dates.
self.assertTrue(ref_lattice in lattice_data)
# Run the model for 1000 steps with an xyz trajectory.
ab_flip_model.run(control_parameters,
trajectory_filename=xyz_trajectory_filename,
trajectory_type='xyz')
ref_xyz = """KMCLib XYZ FORMAT VERSION 2013.10.15
CELL VECTORS
a: 1.0000000000e+00 0.0000000000e+00 0.0000000000e+00
b: 0.0000000000e+00 1.0000000000e+00 0.0000000000e+00
c: 0.0000000000e+00 0.0000000000e+00 1.0000000000e+00
REPETITIONS 4 1 1
PERIODICITY True True False
STEP 0
16
TIME 4.0622006972e+01
B 0.0000000000e+00 0.0000000000e+00 0.0000000000e+00 0
A 0.0000000000e+00 1.0000000000e+00 0.0000000000e+00 1
A 0.0000000000e+00 2.0000000000e+00 0.0000000000e+00 2
B 0.0000000000e+00 3.0000000000e+00 0.0000000000e+00 3
B 1.0000000000e+00 0.0000000000e+00 0.0000000000e+00 4
B 1.0000000000e+00 1.0000000000e+00 0.0000000000e+00 5
B 1.0000000000e+00 2.0000000000e+00 0.0000000000e+00 6
A 1.0000000000e+00 3.0000000000e+00 0.0000000000e+00 7
B 2.0000000000e+00 0.0000000000e+00 0.0000000000e+00 8
B 2.0000000000e+00 1.0000000000e+00 0.0000000000e+00 9
B 2.0000000000e+00 2.0000000000e+00 0.0000000000e+00 10
A 2.0000000000e+00 3.0000000000e+00 0.0000000000e+00 11
B 3.0000000000e+00 0.0000000000e+00 0.0000000000e+00 12
B 3.0000000000e+00 1.0000000000e+00 0.0000000000e+00 13
B 3.0000000000e+00 2.0000000000e+00 0.0000000000e+00 14
B 3.0000000000e+00 3.0000000000e+00 0.0000000000e+00 15
STEP 500
16
TIME 5.9660935122e+01
B 0.0000000000e+00 0.0000000000e+00 0.0000000000e+00 0
A 0.0000000000e+00 1.0000000000e+00 0.0000000000e+00 1
B 0.0000000000e+00 2.0000000000e+00 0.0000000000e+00 2
B 0.0000000000e+00 3.0000000000e+00 0.0000000000e+00 3
B 1.0000000000e+00 0.0000000000e+00 0.0000000000e+00 4
B 1.0000000000e+00 1.0000000000e+00 0.0000000000e+00 5
B 1.0000000000e+00 2.0000000000e+00 0.0000000000e+00 6
A 1.0000000000e+00 3.0000000000e+00 0.0000000000e+00 7
B 2.0000000000e+00 0.0000000000e+00 0.0000000000e+00 8
B 2.0000000000e+00 1.0000000000e+00 0.0000000000e+00 9
B 2.0000000000e+00 2.0000000000e+00 0.0000000000e+00 10
B 2.0000000000e+00 3.0000000000e+00 0.0000000000e+00 11
A 3.0000000000e+00 0.0000000000e+00 0.0000000000e+00 12
A 3.0000000000e+00 1.0000000000e+00 0.0000000000e+00 13
B 3.0000000000e+00 2.0000000000e+00 0.0000000000e+00 14
B 3.0000000000e+00 3.0000000000e+00 0.0000000000e+00 15
STEP 1000
16
TIME 8.1189257919e+01
A 0.0000000000e+00 0.0000000000e+00 0.0000000000e+00 0
B 0.0000000000e+00 1.0000000000e+00 0.0000000000e+00 1
B 0.0000000000e+00 2.0000000000e+00 0.0000000000e+00 2
B 0.0000000000e+00 3.0000000000e+00 0.0000000000e+00 3
B 1.0000000000e+00 0.0000000000e+00 0.0000000000e+00 4
A 1.0000000000e+00 1.0000000000e+00 0.0000000000e+00 5
B 1.0000000000e+00 2.0000000000e+00 0.0000000000e+00 6
B 1.0000000000e+00 3.0000000000e+00 0.0000000000e+00 7
B 2.0000000000e+00 0.0000000000e+00 0.0000000000e+00 8
B 2.0000000000e+00 1.0000000000e+00 0.0000000000e+00 9
A 2.0000000000e+00 2.0000000000e+00 0.0000000000e+00 10
A 2.0000000000e+00 3.0000000000e+00 0.0000000000e+00 11
B 3.0000000000e+00 0.0000000000e+00 0.0000000000e+00 12
B 3.0000000000e+00 1.0000000000e+00 0.0000000000e+00 13
B 3.0000000000e+00 2.0000000000e+00 0.0000000000e+00 14
B 3.0000000000e+00 3.0000000000e+00 0.0000000000e+00 15
"""
with open(xyz_trajectory_filename, "r") as t:
xyz_data = t.read()
self.assertEqual(ref_xyz, xyz_data)
# Running with wrong trajectory_type input fails.
self.assertRaises( Error,
lambda : ab_flip_model.run(control_parameters,
trajectory_filename=xyz_trajectory_filename,
trajectory_type='abc') )
self.assertRaises( Error,
lambda : ab_flip_model.run(control_parameters,
trajectory_filename=xyz_trajectory_filename,
trajectory_type=123) )
def testConstruction(self):
""" Test the construction of the lattice model """
# Setup a unitcell.
unit_cell = KMCUnitCell(cell_vectors=numpy.array([[2.8,0.0,0.0],
[0.0,3.2,0.0],
[0.0,0.5,3.0]]),
basis_points=[[0.0,0.0,0.0],
[0.5,0.5,0.5],
[0.25,0.25,0.75]])
# Setup the lattice.
lattice = KMCLattice(unit_cell=unit_cell,
repetitions=(4,4,1),
periodic=(True,True,False))
types = ['A','A','A','A','B','B',
'A','A','A','B','B','B',
'B','B','A','A','B','A',
'B','B','B','A','B','A',
'B','A','A','A','B','B',
'B','B','B','B','B','B',
'A','A','A','A','B','B',
'B','B','A','B','B','A']
# Setup the configuration.
config = KMCConfiguration(lattice=lattice,
types=types,
possible_types=['A','C','B'])
# A first process.
coords = [[1.0,2.0,3.4],[1.1,1.2,1.3]]
types0 = ["A","B"]
types1 = ["B","A"]
sites = [0,1,2]
rate_0_1 = 3.5
process_0 = KMCProcess(coords,
types0,
types1,
basis_sites=sites,
rate_constant=rate_0_1)
# A second process.
coords = [[1.0,2.0,3.4],[1.1,1.2,1.3]]
types0 = ["A","C"]
types1 = ["C","A"]
sites = [0,1,2]
rate_0_1 = 1.5
process_1 = KMCProcess(coords,
types0,
types1,
basis_sites=sites,
rate_constant=rate_0_1)
# Construct the interactions object.
processes = [process_0, process_1]
interactions = KMCInteractions(processes=processes)
# Construct the model.
model = KMCLatticeModel(config, interactions)
# Check that it has the attribute _backend which is None
self.assertTrue(hasattr(model,"_KMCLatticeModel__backend"))
self.assertTrue(model._KMCLatticeModel__backend is None)
# Check that it has the correct interactions stored.
self.assertTrue(model._KMCLatticeModel__interactions == interactions)
# Check that it has the correct configuration stored.
self.assertTrue(model._KMCLatticeModel__configuration == config)
def testBackendWithCustomRates(self):
""" Test that we can construct the backend object with custom rates. """
# A first process.
coords = [[1.0, 2.0, 3.4], [1.1, 1.2, 1.3]]
types0 = ["A", "B"]
types1 = ["B", "A"]
rate_0_1 = 3.5
process_0 = KMCProcess(coords,
types0,
types1,
basis_sites=[0],
rate_constant=rate_0_1)
# A second process.
types0 = ["A", "C"]
types1 = ["C", "A"]
rate_0_1 = 1.5
process_1 = KMCProcess(coords,
types0,
types1,
basis_sites=[0],
rate_constant=rate_0_1)
processes = [process_0, process_1]
# Set the custom rates class to use.
custom_rates_class = CustomRateCalculator
# Set the rate function on the custom rates calculator for testing.
ref_rnd = numpy.random.uniform(0.0, 1.0)
def testRateFunction(obj, coords, types_before, types_after,
rate_const, process_number, global_coordinate):
return ref_rnd
# Store the original.
custom_rate_function = custom_rates_class.rate
try:
custom_rates_class.rate = testRateFunction
# Construct the interactions object.
kmc_interactions = KMCInteractions(processes=processes,
implicit_wildcards=False)
# Setup a dict with the possible types.
possible_types = {
"A": 13,
"D": 2,
"B": 3,
"J": 4,
"C": 5,
}
# Use a dummy argument for the configuration.
config = "DummyConfig"
# Test that it fails if the rate calculator is of wrong class.
kmc_interactions.setRateCalculator(rate_calculator=Error)
self.assertRaises(
Error,
lambda: kmc_interactions._backend(possible_types, 2, config))
# Test that it fails if the rate calculator is the base class.
kmc_interactions.setRateCalculator(
rate_calculator=KMCRateCalculatorPlugin)
self.assertRaises(
Error,
lambda: kmc_interactions._backend(possible_types, 2, config))
# But this should work.
kmc_interactions.setRateCalculator(
rate_calculator=custom_rates_class)
# Construct the backend.
cpp_interactions = kmc_interactions._backend(
possible_types, 2, config)
# Test that the configuration on the custom rate class is the one given.
self.assertTrue(
kmc_interactions._KMCInteractions__rate_calculator.
configuration == config) # <- Check by reference.
# Get the rate calculator reference out of the C++ object and
# check that a call from C++ calls the Python extension.
cpp_coords = Backend.StdVectorDouble()
cpp_types1 = Backend.StdVectorString()
cpp_types2 = Backend.StdVectorString()
rate_constant = 543.2211
process_number = 33
coordinate = (1.2, 3.4, 5.6)
self.assertAlmostEqual(
cpp_interactions.rateCalculator().backendRateCallback(
cpp_coords,
cpp_coords.size() / 3, cpp_types1, cpp_types2,
rate_constant, process_number, coordinate[0],
coordinate[1], coordinate[2]), ref_rnd, 12)
self.assertAlmostEqual(
kmc_interactions._KMCInteractions__rate_calculator.
backendRateCallback(cpp_coords,
cpp_coords.size() / 3, cpp_types1,
cpp_types2, rate_constant, process_number,
coordinate[0], coordinate[1],
coordinate[2]), ref_rnd, 12)
finally:
# Reset the class.
custom_rates_class.rate = custom_rate_function
# Construct a C++ RateCalculator object directly and check that this object
# returns the rate given to it.
cpp_rate_calculator = Backend.RateCalculator()
self.assertAlmostEqual(
cpp_rate_calculator.backendRateCallback(
cpp_coords,
cpp_coords.size() / 3, cpp_types1, cpp_types2, rate_constant,
process_number, coordinate[0], coordinate[1], coordinate[2]),
rate_constant, 12)
def testCalculation(self):
""" Test a calculation with the on-the-fly MSD analysis. """
# Setup a system, a periodic 10 atoms long 1D chain.
unit_cell = KMCUnitCell(cell_vectors=numpy.array([[1.0,0.0,0.0],
[0.0,1.0,0.0],
[0.0,0.0,1.0]]),
basis_points=[[0.0,0.0,0.0]])
# And a lattice.
lattice = KMCLattice(unit_cell=unit_cell,
repetitions=(10,10,10),
periodic=(True,True,True))
# Setup an initial configuration with one B in a sea of A:s.
types = ["A"]*10*10*10
types[5] = "B"
config = KMCConfiguration(lattice=lattice,
types=types,
possible_types=["A","B"])
# Setup a diffusion process to the left.
coordinates_p0 = [[0.0, 0.0, 0.0],[-1.0, 0.0, 0.0]]
p0 = KMCProcess(coordinates=coordinates_p0,
elements_before=["B","A"],
elements_after=["A","B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p1 = [[0.0, 0.0, 0.0],[1.0, 0.0, 0.0]]
p1 = KMCProcess(coordinates=coordinates_p1,
elements_before=["B","A"],
elements_after=["A","B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p2 = [[0.0, 0.0, 0.0],[0.0,-1.0, 0.0]]
p2 = KMCProcess(coordinates=coordinates_p2,
elements_before=["B","A"],
elements_after=["A","B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p3 = [[0.0, 0.0, 0.0],[0.0, 1.0, 0.0]]
p3 = KMCProcess(coordinates=coordinates_p3,
elements_before=["B","A"],
elements_after=["A","B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p4 = [[0.0, 0.0, 0.0],[0.0, 0.0,-1.0]]
p4 = KMCProcess(coordinates=coordinates_p4,
elements_before=["B","A"],
elements_after=["A","B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p5 = [[0.0, 0.0, 0.0],[0.0, 0.0, 1.0]]
p5 = KMCProcess(coordinates=coordinates_p5,
elements_before=["B","A"],
elements_after=["A","B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
interactions = KMCInteractions(processes=[p0, p1, p2, p3, p4, p5],
implicit_wildcards=True)
model = KMCLatticeModel(configuration=config,
interactions=interactions)
# Setup the analysis.
msd = OnTheFlyMSD(history_steps=400,
n_bins=10,
t_max=25.0,
track_type="B")
# Setup the control parameters.
control_parameters = KMCControlParameters(number_of_steps=4000,
dump_interval=100,
analysis_interval=1,
seed=2013)
# Run the model.
model.run(control_parameters=control_parameters,
analysis=[msd])
# Get the results out.
results = msd.results()
time_steps = msd.timeSteps()
std_dev = msd.stdDev()
bin_counters = msd.binCounters()
# Compare against references.
ref_results = numpy.array([[ 2.77712331 , 8.22333156 , 14.21578638 , 21.11949845 , 27.57598537,
33.81520414, 40.5640313 , 47.35067074, 54.83134718, 63.0006103 ],
[ 2.80810946 , 8.48503234 , 13.7997313 , 17.62812205 , 22.37202018,
29.12157221, 35.40750463, 41.44944591, 48.94392653, 56.67111894],
[ 2.98786918 , 8.65022668 , 14.01808716 , 18.62678885 , 22.65708384,
26.03107861, 30.03937616, 35.1483 , 40.68319007, 47.21074474],
[ 5.58523277 , 16.70836389 , 28.01551768 , 38.74762049 , 49.94800555,
62.93677636, 75.97153593, 88.80011665, 103.77527371, 119.67172924],
[ 5.76499249 , 16.87355824 , 28.23387354 , 39.7462873 , 50.23306921,
59.84628275, 70.60340745, 82.49897073, 95.51453725, 110.21135504],
[ 5.79597864 , 17.13525902 , 27.81781846 , 36.2549109 , 45.02910402,
55.15265082, 65.44688079, 76.59774591, 89.62711659, 103.88186368],
[ 8.57310195 , 25.35859057 , 42.03360484 , 57.37440934 , 72.60508939,
88.96785497, 106.01091209, 123.94841665, 144.45846377, 166.88247398]])
diff = numpy.linalg.norm(ref_results - results)
self.assertAlmostEqual(diff, 0.0, 6)
ref_time_steps = numpy.array([ 1.25, 3.75, 6.25, 8.75, 11.25, 13.75, 16.25, 18.75, 21.25, 23.75])
diff = numpy.linalg.norm(ref_time_steps - time_steps)
self.assertAlmostEqual(diff, 0.0, 8)
ref_std_dev = numpy.array( [[ 0.14837207, 0.7305895 , 1.61912818, 2.84016862, 4.19935234,
5.69504607, 7.43238623, 9.31932103, 11.49445671, 13.97165319],
[ 0.15002755, 0.75383991, 1.57174096, 2.37064526, 3.40687718,
4.90455993, 6.48757634, 8.15787162, 10.26025939, 12.5679611 ],
[ 0.15963149, 0.76851636, 1.59661093, 2.50494685, 3.45028752,
4.38406911, 5.5039955 , 6.91771175, 8.52853689, 10.46993274],
[ 0.21100039, 1.04965011, 2.25628521, 3.68460183, 5.37841647,
7.49505328, 9.84289993, 12.35824143, 15.38290727, 18.76634124],
[ 0.2177914 , 1.06002792, 2.27387093, 3.77956739, 5.40911222,
7.12701069, 9.14740325, 11.48131598, 14.15839455, 17.28281115],
[ 0.218962 , 1.07646844, 2.24036311, 3.44756425, 4.84874766,
6.56805258, 8.47932177, 10.66004723, 13.28568526, 16.29025096],
[ 0.26444438, 1.30073885, 2.76405291, 4.45469653, 6.38348309,
8.65082886, 11.21442742, 14.08440462, 17.48404426, 21.36747194]])
diff = numpy.linalg.norm(ref_std_dev - std_dev)
self.assertAlmostEqual(diff, 0.0, 6)
ref_bin_counters = (59930, 59996, 59545, 59256, 59064, 59076, 58284, 58294, 57817, 57349)
self.assertEqual(bin_counters, ref_bin_counters)
def testBackend(self):
""" Test that the backend object is correctly constructed. """
# Setup a unitcell.
unit_cell = KMCUnitCell(cell_vectors=numpy.array([[2.8,0.0,0.0],
[0.0,3.2,0.0],
[0.0,0.5,3.0]]),
basis_points=[[0.0,0.0,0.0],
[0.5,0.5,0.5],
[0.25,0.25,0.75]])
# Setup the lattice.
lattice = KMCLattice(unit_cell=unit_cell,
repetitions=(4,4,1),
periodic=(True,True,False))
types = ['A','A','A','A','B','B',
'A','A','A','B','B','B',
'B','B','A','A','B','A',
'B','B','B','A','B','A',
'B','A','A','A','B','B',
'B','B','B','B','B','B',
'A','A','A','A','B','B',
'B','B','A','B','B','A']
# Setup the configuration.
config = KMCConfiguration(lattice=lattice,
types=types,
possible_types=['A','C','B'])
# A first process.
coords = [[1.0,2.0,3.4],[1.1,1.2,1.3]]
types0 = ["A","B"]
types1 = ["B","A"]
sites = [0,1,2,3,4]
rate_0_1 = 3.5
process_0 = KMCProcess(coords,
types0,
types1,
basis_sites=sites,
rate_constant=rate_0_1)
# A second process.
types0 = ["A","C"]
types1 = ["C","A"]
rate_0_1 = 1.5
process_1 = KMCProcess(coords,
types0,
types1,
basis_sites=sites,
rate_constant=rate_0_1)
# Construct the interactions object.
processes = [process_0, process_1]
interactions = KMCInteractions(processes=processes)
# Construct the model.
model = KMCLatticeModel(config, interactions)
# Get the c++ backend out.
cpp_model = model._backend()
# Check that this backend object is stored on the class.
self.assertTrue(model._KMCLatticeModel__backend == cpp_model)
def testRunWithAnalysis(self):
""" Test that the analyis plugins get called correctly. """
# Cell.
cell_vectors = [[ 1.000000e+00, 0.000000e+00, 0.000000e+00],
[ 0.000000e+00, 1.000000e+00, 0.000000e+00],
[ 0.000000e+00, 0.000000e+00, 1.000000e+00]]
basis_points = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
unit_cell = KMCUnitCell(
cell_vectors=cell_vectors,
basis_points=basis_points)
# Lattice.
lattice = KMCLattice(
unit_cell=unit_cell,
repetitions=(10,10,1),
periodic=(True, True, False))
# Configuration.
types = ['B']*100
possible_types = ['A','B']
configuration = KMCConfiguration(
lattice=lattice,
types=types,
possible_types=possible_types)
# Interactions.
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00]]
process_0 = KMCProcess(coordinates,
['A'],
['B'],
basis_sites=[0],
rate_constant=4.0)
process_1 = KMCProcess(coordinates,
['B'],
['A'],
basis_sites=[0],
rate_constant=1.0)
processes = [process_0, process_1]
interactions = KMCInteractions(processes)
# Setup the model.
ab_flip_model = KMCLatticeModel(configuration, interactions)
# Run the model with a trajectory file.
name = os.path.abspath(os.path.dirname(__file__))
name = os.path.join(name, "..", "TestUtilities", "Scratch")
trajectory_filename = os.path.join(name, "ab_flip_traj.py")
self.__files_to_remove.append(trajectory_filename)
# The control parameters.
control_parameters = KMCControlParameters(number_of_steps=1000,
dump_interval=500,
analysis_interval=300)
# Setup a valid minimal analysis object.
class AnalysisProxy1(KMCAnalysisPlugin):
def __init__(self):
pass
# Setup a slightly larger analyis object.
class AnalysisProxy2(KMCAnalysisPlugin):
def __init__(self):
self.setup_called = False
self.finalize_called = False
self.register_step_counts = 0
def setup(self, step, time, configuration):
self.setup_called = True
def registerStep(self, step, time, configuration):
self.register_step_counts += 1
def finalize(self):
self.finalize_called = True
ap2 = AnalysisProxy2()
analysis = [ AnalysisProxy1(), ap2 ]
# Run the model for 1000 steps with the analysis objects.
# With dump interval 500 the analysis objects should be
# called on startup, the at step 300, 600 and step 900 and
# a finalization after that.
ab_flip_model.run(control_parameters,
trajectory_filename=trajectory_filename,
analysis=analysis)
self.assertTrue(ap2.setup_called)
self.assertTrue(ap2.finalize_called)
self.assertEqual(ap2.register_step_counts, 3)
def testScript(self):
""" Test that a script can be created. """
# Setup a unitcell.
unit_cell = KMCUnitCell(cell_vectors=numpy.array([[2.8,0.0,0.0],
[0.0,3.2,0.0],
[0.0,0.5,3.0]]),
basis_points=[[0.0,0.0,0.0],
[0.5,0.5,0.5],
[0.25,0.25,0.75]])
# Setup the lattice.
lattice = KMCLattice(unit_cell=unit_cell,
repetitions=(4,4,1),
periodic=(True,True,False))
types = ['A','A','A','A','B','B',
'A','A','A','B','B','B',
'B','B','A','A','B','A',
'B','B','B','A','B','A',
'B','A','A','A','B','B',
'B','B','B','B','B','B',
'A','A','A','A','B','B',
'B','B','A','B','B','A']
# Setup the configuration.
config = KMCConfiguration(lattice=lattice,
types=types,
possible_types=['A','C','B'])
# A first process.
coords = [[1.0,2.0,3.4],[1.1,1.2,1.3]]
types0 = ["A","B"]
types1 = ["B","A"]
rate_0_1 = 3.5
process_0 = KMCProcess(coords,
types0,
types1,
basis_sites=[0],
rate_constant=rate_0_1)
# A second process.
types0 = ["A","C"]
types1 = ["C","A"]
rate_0_1 = 1.5
process_1 = KMCProcess(coords,
types0,
types1,
basis_sites=[0],
rate_constant=rate_0_1)
# Construct the interactions object.
processes = [process_0, process_1]
interactions = KMCInteractions(processes)
# Construct the model.
model = KMCLatticeModel(config, interactions)
# Get the script.
script = model._script()
ref_script = """
# -----------------------------------------------------------------------------
# Unit cell
cell_vectors = [[ 2.800000e+00, 0.000000e+00, 0.000000e+00],
[ 0.000000e+00, 3.200000e+00, 0.000000e+00],
[ 0.000000e+00, 5.000000e-01, 3.000000e+00]]
basis_points = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00],
[ 5.000000e-01, 5.000000e-01, 5.000000e-01],
[ 2.500000e-01, 2.500000e-01, 7.500000e-01]]
unit_cell = KMCUnitCell(
cell_vectors=cell_vectors,
basis_points=basis_points)
# -----------------------------------------------------------------------------
# Lattice
lattice = KMCLattice(
unit_cell=unit_cell,
repetitions=(4,4,1),
periodic=(True, True, False))
# -----------------------------------------------------------------------------
# Configuration
types = ['A','A','A','A','B','B','A','A','A','B','B','B','B',
'B','A','A','B','A','B','B','B','A','B','A','B','A',
'A','A','B','B','B','B','B','B','B','B','A','A','A',
'A','B','B','B','B','A','B','B','A']
possible_types = ['A','C','B']
configuration = KMCConfiguration(
lattice=lattice,
types=types,
possible_types=possible_types)
# -----------------------------------------------------------------------------
# Interactions
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00],
[ 1.000000e-01, -8.000000e-01, -2.100000e+00]]
elements_before = ['A','B']
elements_after = ['B','A']
move_vectors = [( 0,[ 1.000000e-01, -8.000000e-01, -2.100000e+00]),
( 1,[ -1.000000e-01, 8.000000e-01, 2.100000e+00])]
basis_sites = [0]
rate_constant = 3.500000e+00
process_0 = KMCProcess(
coordinates=coordinates,
elements_before=elements_before,
elements_after=elements_after,
move_vectors=move_vectors,
basis_sites=basis_sites,
rate_constant=rate_constant)
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00],
[ 1.000000e-01, -8.000000e-01, -2.100000e+00]]
elements_before = ['A','C']
elements_after = ['C','A']
move_vectors = [( 0,[ 1.000000e-01, -8.000000e-01, -2.100000e+00]),
( 1,[ -1.000000e-01, 8.000000e-01, 2.100000e+00])]
basis_sites = [0]
rate_constant = 1.500000e+00
process_1 = KMCProcess(
coordinates=coordinates,
elements_before=elements_before,
elements_after=elements_after,
move_vectors=move_vectors,
basis_sites=basis_sites,
rate_constant=rate_constant)
processes = [process_0,
process_1]
interactions = KMCInteractions(
processes=processes,
implicit_wildcards=True)
# -----------------------------------------------------------------------------
# Lattice model
model = KMCLatticeModel(
configuration=configuration,
interactions=interactions)
"""
self.assertEqual(script, ref_script)
# Get the script again, with a different variable name.
script = model._script(variable_name="my_model")
ref_script = """
# -----------------------------------------------------------------------------
# Unit cell
cell_vectors = [[ 2.800000e+00, 0.000000e+00, 0.000000e+00],
[ 0.000000e+00, 3.200000e+00, 0.000000e+00],
[ 0.000000e+00, 5.000000e-01, 3.000000e+00]]
basis_points = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00],
[ 5.000000e-01, 5.000000e-01, 5.000000e-01],
[ 2.500000e-01, 2.500000e-01, 7.500000e-01]]
unit_cell = KMCUnitCell(
cell_vectors=cell_vectors,
basis_points=basis_points)
# -----------------------------------------------------------------------------
# Lattice
lattice = KMCLattice(
unit_cell=unit_cell,
repetitions=(4,4,1),
periodic=(True, True, False))
# -----------------------------------------------------------------------------
# Configuration
types = ['A','A','A','A','B','B','A','A','A','B','B','B','B',
'B','A','A','B','A','B','B','B','A','B','A','B','A',
'A','A','B','B','B','B','B','B','B','B','A','A','A',
'A','B','B','B','B','A','B','B','A']
possible_types = ['A','C','B']
configuration = KMCConfiguration(
lattice=lattice,
types=types,
possible_types=possible_types)
# -----------------------------------------------------------------------------
# Interactions
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00],
[ 1.000000e-01, -8.000000e-01, -2.100000e+00]]
elements_before = ['A','B']
elements_after = ['B','A']
move_vectors = [( 0,[ 1.000000e-01, -8.000000e-01, -2.100000e+00]),
( 1,[ -1.000000e-01, 8.000000e-01, 2.100000e+00])]
basis_sites = [0]
rate_constant = 3.500000e+00
process_0 = KMCProcess(
coordinates=coordinates,
elements_before=elements_before,
elements_after=elements_after,
move_vectors=move_vectors,
basis_sites=basis_sites,
rate_constant=rate_constant)
coordinates = [[ 0.000000e+00, 0.000000e+00, 0.000000e+00],
[ 1.000000e-01, -8.000000e-01, -2.100000e+00]]
elements_before = ['A','C']
elements_after = ['C','A']
move_vectors = [( 0,[ 1.000000e-01, -8.000000e-01, -2.100000e+00]),
( 1,[ -1.000000e-01, 8.000000e-01, 2.100000e+00])]
basis_sites = [0]
rate_constant = 1.500000e+00
process_1 = KMCProcess(
coordinates=coordinates,
elements_before=elements_before,
elements_after=elements_after,
move_vectors=move_vectors,
basis_sites=basis_sites,
rate_constant=rate_constant)
processes = [process_0,
process_1]
interactions = KMCInteractions(
processes=processes,
implicit_wildcards=True)
# -----------------------------------------------------------------------------
# Lattice model
my_model = KMCLatticeModel(
configuration=configuration,
interactions=interactions)
"""
# Check.
self.assertEqual(script, ref_script)
def testCalculation(self):
""" Test a calculation with the on-the-fly MSD analysis. """
# Setup a system, a periodic 10 atoms long 1D chain.
unit_cell = KMCUnitCell(cell_vectors=numpy.array([[1.0, 0.0, 0.0],
[0.0, 1.0, 0.0],
[0.0, 0.0, 1.0]]),
basis_points=[[0.0, 0.0, 0.0]])
# And a lattice.
lattice = KMCLattice(unit_cell=unit_cell,
repetitions=(10, 10, 10),
periodic=(True, True, True))
# Setup an initial configuration with one B in a sea of A:s.
types = ["A"] * 10 * 10 * 10
types[5] = "B"
config = KMCConfiguration(lattice=lattice,
types=types,
possible_types=["A", "B"])
# Setup a diffusion process to the left.
coordinates_p0 = [[0.0, 0.0, 0.0], [-1.0, 0.0, 0.0]]
p0 = KMCProcess(coordinates=coordinates_p0,
elements_before=["B", "A"],
elements_after=["A", "B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p1 = [[0.0, 0.0, 0.0], [1.0, 0.0, 0.0]]
p1 = KMCProcess(coordinates=coordinates_p1,
elements_before=["B", "A"],
elements_after=["A", "B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p2 = [[0.0, 0.0, 0.0], [0.0, -1.0, 0.0]]
p2 = KMCProcess(coordinates=coordinates_p2,
elements_before=["B", "A"],
elements_after=["A", "B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p3 = [[0.0, 0.0, 0.0], [0.0, 1.0, 0.0]]
p3 = KMCProcess(coordinates=coordinates_p3,
elements_before=["B", "A"],
elements_after=["A", "B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p4 = [[0.0, 0.0, 0.0], [0.0, 0.0, -1.0]]
p4 = KMCProcess(coordinates=coordinates_p4,
elements_before=["B", "A"],
elements_after=["A", "B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
coordinates_p5 = [[0.0, 0.0, 0.0], [0.0, 0.0, 1.0]]
p5 = KMCProcess(coordinates=coordinates_p5,
elements_before=["B", "A"],
elements_after=["A", "B"],
move_vectors=None,
basis_sites=[0],
rate_constant=1.0)
interactions = KMCInteractions(processes=[p0, p1, p2, p3, p4, p5],
implicit_wildcards=True)
model = KMCLatticeModel(configuration=config,
interactions=interactions)
# Setup the analysis.
msd = OnTheFlyMSD(history_steps=400,
n_bins=10,
t_max=25.0,
track_type="B")
# Setup the control parameters.
control_parameters = KMCControlParameters(number_of_steps=4000,
dump_interval=100,
analysis_interval=1,
seed=2013)
# Run the model.
model.run(control_parameters=control_parameters, analysis=[msd])
# Get the results out.
results = msd.results()
time_steps = msd.timeSteps()
std_dev = msd.stdDev()
bin_counters = msd.binCounters()
# Compare against references.
ref_results = numpy.array([
[
2.59214168, 7.09292512, 11.09571441, 14.73593561, 18.02780434,
21.25579701, 24.64588125, 28.81831092, 31.62393358, 34.6177501
],
[
2.65374493, 7.10317609, 10.90626886, 14.87948705, 18.25169973,
21.00603548, 22.97611332, 25.57822115, 28.87848128, 32.2240879
],
[
2.59976568, 7.35085681, 11.8568078, 16.56678774, 21.77700882,
26.34223625, 30.40786814, 35.09199292, 40.24434182, 46.48414664
],
[
5.24588661, 14.19610121, 22.00198327, 29.61542266, 36.27950408,
42.2618325, 47.62199456, 54.39653208, 60.50241486, 66.841838
],
[
5.19190736, 14.44378193, 22.9525222, 31.30272335, 39.80481316,
47.59803326, 55.05374939, 63.91030384, 71.8682754, 81.10189674
],
[
5.2535106, 14.45403291, 22.76307666, 31.44627479, 40.02870855,
47.34827174, 53.38398146, 60.67021407, 69.1228231, 78.70823454
],
[
7.84565228, 21.54695802, 33.85879107, 46.1822104, 58.05651289,
68.60406875, 78.02986271, 89.48852499, 100.74675668,
113.32598464
]
])
diff = numpy.linalg.norm(ref_results - results)
self.assertAlmostEqual(diff, 0.0, 6)
ref_time_steps = numpy.array(
[1.25, 3.75, 6.25, 8.75, 11.25, 13.75, 16.25, 18.75, 21.25, 23.75])
diff = numpy.linalg.norm(ref_time_steps - time_steps)
self.assertAlmostEqual(diff, 0.0, 8)
ref_std_dev = numpy.array(
[[
0.13765582, 0.62432398, 1.24701335, 1.95661619, 2.7137435,
3.53476554, 4.45846845, 5.60517366, 6.55120574, 7.58912539
],
[
0.14092726, 0.62522628, 1.22572215, 1.97567674, 2.7474467,
3.49323106, 4.1564055, 4.97497483, 5.98245856, 7.06437139
],
[
0.13806069, 0.6470273, 1.33255031, 2.19971409, 3.27811502,
4.38062279, 5.50081855, 6.82540746, 8.33700723, 10.19055301
],
[
0.19698798, 0.88356546, 1.74848804, 2.780551, 3.86164462,
4.96954405, 6.09163579, 7.48129474, 8.86263902, 10.36158694
],
[
0.19496101, 0.89898111, 1.82402696, 2.93896933, 4.23688379,
5.59702476, 7.04227937, 8.7897482, 10.52755635, 12.57213116
],
[
0.19727427, 0.89961913, 1.8089718, 2.95244717, 4.26071555,
5.56765545, 6.82868861, 8.34413034, 10.12539137, 12.20107405
],
[
0.24054939, 1.09498957, 2.19698279, 3.54031591, 5.04564022,
6.58676948, 8.14969886, 10.04910241, 12.04968783, 14.34371883
]])
diff = numpy.linalg.norm(ref_std_dev - std_dev)
self.assertAlmostEqual(diff, 0.0, 6)
ref_bin_counters = (58893, 58531, 57985, 58641, 57509, 57659, 57396,
57037, 56732, 56518)
self.assertEqual(bin_counters, ref_bin_counters)