def setUp(self):
self.natoms = 4
self.k = 1
self.Emax = 2 #this choice is fundamentally arbitrary, it's only used to generate the initial configuration
#self.temperatures = [0.2,0.27,0.362,0.487,0.65,0.88,1.18,1.6]
self.temperatures = [0.2, 0.362, 0.65, 1.18]
self.stepsize = 1
self.niter = 2e6
self.adjustf = 0.9
self.adjust_niter = 0.3 * self.niter
self.radius = 1e10
self.tolerance = 2e-1
self.bdim = 3
self.ndim = self.natoms * self.bdim
self.origin = np.zeros(self.ndim)
self.seeds = dict(takestep=42, metropolis=44)
self.potential = Harmonic(self.origin,
self.k,
bdim=self.bdim,
com=False)
self.mcrunner = Metropolis_MCrunner(self.potential,
self.origin,
1,
self.stepsize,
self.niter,
hEmax=100,
adjustf=self.adjustf,
adjustf_niter=self.adjust_niter,
radius=self.radius,
bdim=self.bdim,
single=False,
seeds=self.seeds)
def test_heat_capacity_2D(self):
self.bdim = 2
self.ndim = self.natoms * self.bdim
self.origin = np.zeros(self.ndim)
self.potential = Harmonic(self.origin,
self.k,
bdim=self.bdim,
com=False)
print "2D"
for T in self.temperatures:
mcrunner = Metropolis_MCrunner(self.potential,
self.origin,
T,
self.stepsize,
self.niter,
hEmax=100,
adjustf=self.adjustf,
adjustf_niter=self.adjust_niter,
radius=self.radius,
bdim=self.bdim)
#MCMC
mcrunner.run()
#collect the results
binenergy, hist, mean, variance = mcrunner.get_histogram()
cv = variance / (T * T)
cv_true = self.natoms * self.bdim / 2.0
print cv, cv_true
self.assertLess(abs(cv - cv_true), self.tolerance,
'failed for temperature {}, cv = {}'.format(T, cv))
def test_heat_capacity_2D_com(self):
self.bdim = 2
self.ndim = self.natoms * self.bdim
self.origin = np.zeros(self.ndim)
potential = Harmonic(self.origin, self.k, bdim=self.bdim, com=True)
#self.start_coords = vector_random_uniform_hypersphere(self.ndim) * np.sqrt(2*self.Emax) #coordinates sampled from Pow(ndim)
print "2D com"
for T in self.temperatures:
mcrunner = Metropolis_MCrunner(potential,
self.origin,
T,
self.stepsize,
self.niter,
hEmax=100,
adjustf=self.adjustf,
adjustf_niter=self.adjust_niter,
radius=self.radius,
bdim=self.bdim)
#MCMC
mcrunner.run()
#collect the results
binenergy, hist, mean, variance = mcrunner.get_histogram()
cv = variance / (T * T)
cv_true = (self.natoms - 1) * self.bdim / 2.0
print cv, cv_true
self.assertLess(abs(cv - cv_true), self.tolerance,
'failed for temperature {}, cv = {}'.format(T, cv))
def test_heat_capacity_3D_com(self):
self.bdim = 3
self.ndim = self.natoms*self.bdim
self.origin = np.zeros(self.ndim)
potential = Harmonic(self.origin, self.k, bdim=self.bdim, com=True)
mcrunner = Metropolis_MCrunner(potential, self.origin, 1, self.stepsize, self.niter, hEmax = 100,
adjustf = self.adjustf, adjustf_niter = self.adjust_niter, radius=self.radius,
bdim=self.bdim, seeds=self.seeds)
mcrunner.run()
print "3D com"
for T in self.temperatures:
mcrunner = Metropolis_MCrunner(potential, self.origin, T, self.stepsize, self.niter, hEmax = 100,
adjustf = self.adjustf, adjustf_niter = self.adjust_niter, radius=self.radius,
bdim=self.bdim, seeds=self.seeds)
#MCMC
mcrunner.run()
#collect the results
binenergy, hist, mean, variance = mcrunner.get_histogram()
cv = variance/(T*T)
cv_true = (self.natoms-1)*self.bdim/2.0
print cv, cv_true
self.assertLess(abs(cv-cv_true),self.tolerance,'failed for temperature {}, cv = {}'.format(T,cv))
def test_heat_capacity_2D_com(self):
self.bdim = 2
self.ndim = self.natoms*self.bdim
self.origin = np.zeros(self.ndim)
potential = Harmonic(self.origin,self.k, bdim=self.bdim,com=True)
#self.start_coords = vector_random_uniform_hypersphere(self.ndim) * np.sqrt(2*self.Emax) #coordinates sampled from Pow(ndim)
print "2D com"
for T in self.temperatures:
mcrunner = Metropolis_MCrunner(potential, self.origin, T, self.stepsize, self.niter, hEmax = 100,
adjustf = self.adjustf, adjustf_niter = self.adjust_niter, radius=self.radius,
bdim=self.bdim)
#MCMC
mcrunner.run()
#collect the results
binenergy, hist, mean, variance = mcrunner.get_histogram()
cv = variance/(T*T)
cv_true = (self.natoms-1)*self.bdim/2.0
print cv, cv_true
self.assertLess(abs(cv-cv_true),self.tolerance,'failed for temperature {}, cv = {}'.format(T,cv))
def setUp(self):
self.natoms = 4
self.k=1
self.Emax = 2 #this choice is fundamentally arbitrary, it's only used to generate the initial configuration
#self.temperatures = [0.2,0.27,0.362,0.487,0.65,0.88,1.18,1.6]
self.temperatures = [0.2, 0.362, 0.65, 1.18]
self.stepsize = 1
self.niter = 2e6
self.adjustf = 0.9
self.adjust_niter = 0.3 * self.niter
self.radius = 1e10
self.tolerance = 2e-1
self.bdim = 3
self.ndim = self.natoms * self.bdim
self.origin = np.zeros(self.ndim)
self.potential = Harmonic(self.origin, self.k, bdim=self.bdim, com=False)
self.mcrunner = Metropolis_MCrunner(self.potential, self.origin, 1, self.stepsize, self.niter, hEmax = 100,
adjustf = self.adjustf, adjustf_niter = self.adjust_niter, radius=self.radius,
bdim=self.bdim, single=True)
class TestMetropolisGlobal(unittest.TestCase):
def setUp(self):
self.natoms = 4
self.k=1
self.Emax = 2 #this choice is fundamentally arbitrary, it's only used to generate the initial configuration
#self.temperatures = [0.2,0.27,0.362,0.487,0.65,0.88,1.18,1.6]
self.temperatures = [0.2, 0.362, 0.65, 1.18]
self.stepsize = 1
self.niter = 2e6
self.adjustf = 0.9
self.adjust_niter = 0.3 * self.niter
self.radius = 1e10
self.tolerance = 2e-1
self.bdim = 3
self.ndim = self.natoms * self.bdim
self.origin = np.zeros(self.ndim)
self.seeds = dict(takestep=42, metropolis=44)
self.potential = Harmonic(self.origin, self.k, bdim=self.bdim, com=False)
self.mcrunner = Metropolis_MCrunner(self.potential, self.origin, 1, self.stepsize, self.niter, hEmax = 100,
adjustf = self.adjustf, adjustf_niter = self.adjust_niter, radius=self.radius,
bdim=self.bdim, single=False, seeds=self.seeds)
def test_set_control(self):
self.mcrunner.set_control(10)
self.mcrunner.run()
self.assertEqual(self.mcrunner.get_temperature(), self.mcrunner.temperature)
def test_dump_histogram(self):
self.mcrunner.run()
mean, variance = self.mcrunner.dump_histogram("test_histogram.dat")
data = np.genfromtxt("test_histogram.dat")
#print data
binenergy, hist, mean2, variance2 = self.mcrunner.get_histogram()
self.assertListEqual(np.ndarray.tolist(binenergy), np.ndarray.tolist(data[:,0]))
self.assertListEqual(np.ndarray.tolist(hist), np.ndarray.tolist(data[:,1]))
self.assertEqual(mean, mean2)
self.assertEqual(variance, variance2)
count = self.mcrunner.histogram.get_count()
self.assertEqual(count, 1400000)
def test_heat_capacity_3D_com(self):
self.bdim = 3
self.ndim = self.natoms*self.bdim
self.origin = np.zeros(self.ndim)
potential = Harmonic(self.origin, self.k, bdim=self.bdim, com=True)
mcrunner = Metropolis_MCrunner(potential, self.origin, 1, self.stepsize, self.niter, hEmax = 100,
adjustf = self.adjustf, adjustf_niter = self.adjust_niter, radius=self.radius,
bdim=self.bdim, seeds=self.seeds)
mcrunner.run()
print "3D com"
for T in self.temperatures:
mcrunner = Metropolis_MCrunner(potential, self.origin, T, self.stepsize, self.niter, hEmax = 100,
adjustf = self.adjustf, adjustf_niter = self.adjust_niter, radius=self.radius,
bdim=self.bdim, seeds=self.seeds)
#MCMC
mcrunner.run()
#collect the results
binenergy, hist, mean, variance = mcrunner.get_histogram()
cv = variance/(T*T)
cv_true = (self.natoms-1)*self.bdim/2.0
print cv, cv_true
self.assertLess(abs(cv-cv_true),self.tolerance,'failed for temperature {}, cv = {}'.format(T,cv))
def test_heat_capacity_3D(self):
self.bdim = 3
self.ndim = self.natoms*self.bdim
self.origin = np.zeros(self.ndim)
potential = Harmonic(self.origin, self.k, bdim=self.bdim, com=False)
print "3D"
for T in self.temperatures:
mcrunner = Metropolis_MCrunner(potential, self.origin, T, self.stepsize, self.niter, hEmax = 100,
adjustf = self.adjustf, adjustf_niter = self.adjust_niter, radius=self.radius,
bdim=self.bdim, seeds=self.seeds)
#MCMC
mcrunner.run()
#collect the results
binenergy, hist, mean, variance = mcrunner.get_histogram()
cv = variance/(T*T)
cv_true = self.natoms*self.bdim/2.0
print cv, cv_true
self.assertLess(abs(cv-cv_true),self.tolerance,'failed for temperature {}, cv = {}'.format(T,cv))
def test_heat_capacity_2D_com(self):
self.bdim = 2
self.ndim = self.natoms*self.bdim
self.origin = np.zeros(self.ndim)
potential = Harmonic(self.origin,self.k, bdim=self.bdim,com=True)
#self.start_coords = vector_random_uniform_hypersphere(self.ndim) * np.sqrt(2*self.Emax) #coordinates sampled from Pow(ndim)
print "2D com"
for T in self.temperatures:
mcrunner = Metropolis_MCrunner(potential, self.origin, T, self.stepsize, self.niter, hEmax = 100,
adjustf = self.adjustf, adjustf_niter = self.adjust_niter, radius=self.radius,
bdim=self.bdim, seeds=self.seeds)
#MCMC
mcrunner.run()
#collect the results
binenergy, hist, mean, variance = mcrunner.get_histogram()
cv = variance/(T*T)
cv_true = (self.natoms-1)*self.bdim/2.0
print cv, cv_true
self.assertLess(abs(cv-cv_true),self.tolerance,'failed for temperature {}, cv = {}'.format(T,cv))
def test_heat_capacity_2D(self):
self.bdim = 2
self.ndim = self.natoms*self.bdim
self.origin = np.zeros(self.ndim)
self.potential = Harmonic(self.origin, self.k, bdim=self.bdim, com=False)
print "2D"
for T in self.temperatures:
mcrunner = Metropolis_MCrunner(self.potential, self.origin, T, self.stepsize, self.niter, hEmax = 100,
adjustf = self.adjustf, adjustf_niter = self.adjust_niter, radius=self.radius,
bdim=self.bdim, seeds=self.seeds)
#MCMC
mcrunner.run()
#collect the results
binenergy, hist, mean, variance = mcrunner.get_histogram()
cv = variance/(T*T)
cv_true = self.natoms*self.bdim/2.0
print cv, cv_true
self.assertLess(abs(cv-cv_true),self.tolerance,'failed for temperature {}, cv = {}'.format(T,cv))
def test_composite_moves(self):
self.bdim = 2
self.ndim = self.natoms*self.bdim
self.origin = np.zeros(self.ndim)
self.potential = Harmonic(self.origin, self.k, bdim=self.bdim, com=False)
natoms = 4
bdim = 3
k = 1
origin = np.zeros(natoms * bdim)
potential = Harmonic(origin, k, bdim=bdim, com=False)
path = args.base_directory
#Parallel Tempering
temperature = 1.0
stepsize = 1
niter = 1e4
mcrunner = Metropolis_MCrunner(potential,
origin,
temperature,
stepsize,
niter,
hEmax=100,
adjustf=0.9,
adjustf_niter=3000,
radius=100000,
seeds=dict(metropolis=44, takestep=42))
ptrunner = MPI_PT_RLhandshake(mcrunner,
0.2,
1.6,
max_ptiter=501,
pfreq=10,
base_directory=path,
verbose=False,
suppress_histogram=False)
ptrunner.run()
class TestMetropolisGlobal(unittest.TestCase):
def setUp(self):
self.natoms = 4
self.k = 1
self.Emax = 2 #this choice is fundamentally arbitrary, it's only used to generate the initial configuration
#self.temperatures = [0.2,0.27,0.362,0.487,0.65,0.88,1.18,1.6]
self.temperatures = [0.2, 0.362, 0.65, 1.18]
self.stepsize = 1
self.niter = 2e6
self.adjustf = 0.9
self.adjust_niter = 0.3 * self.niter
self.radius = 1e10
self.tolerance = 2e-1
self.bdim = 3
self.ndim = self.natoms * self.bdim
self.origin = np.zeros(self.ndim)
self.seeds = dict(takestep=42, metropolis=44)
self.potential = Harmonic(self.origin,
self.k,
bdim=self.bdim,
com=False)
self.mcrunner = Metropolis_MCrunner(self.potential,
self.origin,
1,
self.stepsize,
self.niter,
hEmax=100,
adjustf=self.adjustf,
adjustf_niter=self.adjust_niter,
radius=self.radius,
bdim=self.bdim,
single=False,
seeds=self.seeds)
def test_set_control(self):
self.mcrunner.set_control(10)
self.mcrunner.run()
self.assertEqual(self.mcrunner.get_temperature(),
self.mcrunner.temperature)
def test_dump_histogram(self):
self.mcrunner.run()
mean, variance = self.mcrunner.dump_histogram("test_histogram.dat")
data = np.genfromtxt("test_histogram.dat")
#print data
binenergy, hist, mean2, variance2 = self.mcrunner.get_histogram()
self.assertListEqual(np.ndarray.tolist(binenergy),
np.ndarray.tolist(data[:, 0]))
self.assertListEqual(np.ndarray.tolist(hist),
np.ndarray.tolist(data[:, 1]))
self.assertEqual(mean, mean2)
self.assertEqual(variance, variance2)
count = self.mcrunner.histogram.get_count()
self.assertEqual(count, 1400000)
def test_heat_capacity_3D_com(self):
self.bdim = 3
self.ndim = self.natoms * self.bdim
self.origin = np.zeros(self.ndim)
potential = Harmonic(self.origin, self.k, bdim=self.bdim, com=True)
mcrunner = Metropolis_MCrunner(potential,
self.origin,
1,
self.stepsize,
self.niter,
hEmax=100,
adjustf=self.adjustf,
adjustf_niter=self.adjust_niter,
radius=self.radius,
bdim=self.bdim,
seeds=self.seeds)
mcrunner.run()
print "3D com"
for T in self.temperatures:
mcrunner = Metropolis_MCrunner(potential,
self.origin,
T,
self.stepsize,
self.niter,
hEmax=100,
adjustf=self.adjustf,
adjustf_niter=self.adjust_niter,
radius=self.radius,
bdim=self.bdim,
seeds=self.seeds)
#MCMC
mcrunner.run()
#collect the results
binenergy, hist, mean, variance = mcrunner.get_histogram()
cv = variance / (T * T)
cv_true = (self.natoms - 1) * self.bdim / 2.0
print cv, cv_true
self.assertLess(abs(cv - cv_true), self.tolerance,
'failed for temperature {}, cv = {}'.format(T, cv))
def test_heat_capacity_3D(self):
self.bdim = 3
self.ndim = self.natoms * self.bdim
self.origin = np.zeros(self.ndim)
potential = Harmonic(self.origin, self.k, bdim=self.bdim, com=False)
print "3D"
for T in self.temperatures:
mcrunner = Metropolis_MCrunner(potential,
self.origin,
T,
self.stepsize,
self.niter,
hEmax=100,
adjustf=self.adjustf,
adjustf_niter=self.adjust_niter,
radius=self.radius,
bdim=self.bdim,
seeds=self.seeds)
#MCMC
mcrunner.run()
#collect the results
binenergy, hist, mean, variance = mcrunner.get_histogram()
cv = variance / (T * T)
cv_true = self.natoms * self.bdim / 2.0
print cv, cv_true
self.assertLess(abs(cv - cv_true), self.tolerance,
'failed for temperature {}, cv = {}'.format(T, cv))
def test_heat_capacity_2D_com(self):
self.bdim = 2
self.ndim = self.natoms * self.bdim
self.origin = np.zeros(self.ndim)
potential = Harmonic(self.origin, self.k, bdim=self.bdim, com=True)
#self.start_coords = vector_random_uniform_hypersphere(self.ndim) * np.sqrt(2*self.Emax) #coordinates sampled from Pow(ndim)
print "2D com"
for T in self.temperatures:
mcrunner = Metropolis_MCrunner(potential,
self.origin,
T,
self.stepsize,
self.niter,
hEmax=100,
adjustf=self.adjustf,
adjustf_niter=self.adjust_niter,
radius=self.radius,
bdim=self.bdim,
seeds=self.seeds)
#MCMC
mcrunner.run()
#collect the results
binenergy, hist, mean, variance = mcrunner.get_histogram()
cv = variance / (T * T)
cv_true = (self.natoms - 1) * self.bdim / 2.0
print cv, cv_true
self.assertLess(abs(cv - cv_true), self.tolerance,
'failed for temperature {}, cv = {}'.format(T, cv))
def test_heat_capacity_2D(self):
self.bdim = 2
self.ndim = self.natoms * self.bdim
self.origin = np.zeros(self.ndim)
self.potential = Harmonic(self.origin,
self.k,
bdim=self.bdim,
com=False)
print "2D"
for T in self.temperatures:
mcrunner = Metropolis_MCrunner(self.potential,
self.origin,
T,
self.stepsize,
self.niter,
hEmax=100,
adjustf=self.adjustf,
adjustf_niter=self.adjust_niter,
radius=self.radius,
bdim=self.bdim,
seeds=self.seeds)
#MCMC
mcrunner.run()
#collect the results
binenergy, hist, mean, variance = mcrunner.get_histogram()
cv = variance / (T * T)
cv_true = self.natoms * self.bdim / 2.0
print cv, cv_true
self.assertLess(abs(cv - cv_true), self.tolerance,
'failed for temperature {}, cv = {}'.format(T, cv))
def test_composite_moves(self):
self.bdim = 2
self.ndim = self.natoms * self.bdim
self.origin = np.zeros(self.ndim)
self.potential = Harmonic(self.origin,
self.k,
bdim=self.bdim,
com=False)
