from MAST.structopt.tools.StemCalc import ConvStem
from MAST.structopt import inp_out
from mpi4py import MPI
import numpy as np
import math
import time
import scipy.interpolate
rank = MPI.COMM_WORLD.Get_rank()
if rank==0:
# experimental device information
aber=[[0,0],[0,0],[22.56,-20.1],[22.08,-7.5],[0.1198,0],[0.9018,-170.1],[0.04964,20.9],[28.43,-120.6],[11.84,153.8],[8.456,76.1],[0.622,0],[2.811,-125.5]]
autostemparameters={'Electron energy': 200,'Spherical aberration': 1.4,'Defocus':0,'Aperture semiangle': 24.5,'Source size': 0.882,'Slice size': 25.0,'Pixels':976,'Chromatic aberration Coefficient':1.4,'Delta E':0.73,'aber':aber,'Scale Factor':0.00570113}
# based on precalculated PSF("PSF.txt") and structure information("STEM_ref"), calculate the reference STEM image
A = ConvStem(parameters=autostemparameters,calc_exp=False)
atoms_ref=inp_out.read_xyz('STEM_ref',0)
nk = autostemparameters['Pixels']
A.psf = np.empty([nk,nk],dtype=float)
fileobj = open('PSF.txt', 'r')
lines = fileobj.readlines()
for x in range(0,nk):
A.psf[x] = lines[x].split()
fileobj.close()
STEM_ref = A.get_image(A.psf, atoms_ref, autostemparameters['Slice size'], autostemparameters['Pixels'])
autostemparameters['Exp_Image'] = STEM_ref
#setup parameters
autostemparameters['Grid_sim2exp'] = 1
autostemparameters = {
'Electron energy': 200,
'Spherical aberration': 1.4,
'Defocus': 0,
'Aperture semiangle': 24.5,
'Source size': 0.882,
'Slice size': 25.0,
'Pixels': 976,
'Chromatic aberration Coefficient': 1.4,
'Delta E': 0.73,
'aber': aber,
'Scale Factor': 0.00570113
}
# based on precalculated PSF("PSF.txt") and structure information("STEM_ref"), calculate the reference STEM image
A = ConvStem(parameters=autostemparameters, calc_exp=False)
atoms_ref = inp_out.read_xyz('STEM_ref', 0)
nk = autostemparameters['Pixels']
A.psf = np.empty([nk, nk], dtype=float)
fileobj = open('PSF.txt', 'r')
lines = fileobj.readlines()
for x in range(0, nk):
A.psf[x] = lines[x].split()
fileobj.close()
STEM_ref = A.get_image(A.psf, atoms_ref, autostemparameters['Slice size'],
autostemparameters['Pixels'])
autostemparameters['Exp_Image'] = STEM_ref
#setup parameters
"Electron energy": 200,
"Spherical aberration": 1.4,
"Defocus": 0,
"Aperture semiangle": 24.5,
"Source size": 0.882,
"Slice size": 25.0,
"Pixels": 976,
"Chromatic aberration Coefficient": 1.4,
"Delta E": 0.73,
"aber": aber,
"Scale Factor": 0.00570113,
}
# based on precalculated PSF("PSF.txt") and structure information(a xyz format file), calculate the STEM image
autostemparameters["Grid_sim2exp"] = 1
A = ConvStem(parameters=autostemparameters, calc_exp=False)
nk = autostemparameters["Pixels"]
A.psf = np.empty([nk, nk], dtype=float)
fileobj = open("PSF.txt", "r")
lines = fileobj.readlines()
for x in range(0, nk):
A.psf[x] = lines[x].split()
fileobj.close()
# plt.imshow(A.psf)
# plt.show()
Au = inp_out.read_xyz("Output-rank0/indiv00.xyz", -1)
# Au=inp_out.read_xyz('STEM_ref',0)
imAu = A.get_image(A.psf, Au, autostemparameters["Slice size"], autostemparameters["Pixels"])
IPlot = imAu.T
Au1 = gpb([('Au', 10, 0, 0), ('Cu', 10, 0, 0)], size)
aber = [[0, 0], [0, 0], [22.56, -20.1], [22.08, -7.5], [0.1198, 0],
[0.9018, -170.1], [0.04964, 20.9], [28.43, -120.6], [11.84, 153.8],
[8.456, 76.1], [0.622, 0], [2.811, -125.5]]
autostemparameters = {
'Electron energy': 200,
'Spherical aberration': 1.4,
'Defocus': 0,
'Aperture semiangle': 24.5,
'Source size': 0.882,
'Slice size': 20.0,
'Pixels': 800,
'aber': aber,
'Scale Factor': 0.00570113
}
A = ConvStem(parameters=autostemparameters, calc_exp=False)
#fig = plt.figure()
#ax = fig.add_subplot(111)
#psf = numpy.fft.fftshift(A.psf)
#ax.imshow(psf,cmap=cm.hot)
#plt.show()
imAu1 = A.get_image(A.psf, Au1, autostemparameters['Slice size'],
autostemparameters['Pixels'], 1)
#fig2 = plt.figure()
#ax2 = fig2.add_subplot(111)
#ax2.imshow(imAu1,cmap=cm.hot)
#plt.show()
autostemparameters['Exp_Image'] = imAu1
optparameters = {
'structure': 'Cluster',
'optimizer_type': 'BH',
#import matplotlib.pyplot as plt
#import matplotlib.cm as cm
from MAST.structopt.tools.StemCalc import ConvStem
from MAST.structopt.generate.gen_pop_box import gen_pop_box as gpb
from MAST.structopt import Optimizer
import numpy
size = 10.0
Au1 = gpb([('Au',10,0,0),('Cu',10,0,0)],size)
aber=[[0,0],[0,0],[22.56,-20.1],[22.08,-7.5],[0.1198,0],[0.9018,-170.1],[0.04964,20.9],[28.43,-120.6],[11.84,153.8],[8.456,76.1],[0.622,0],[2.811,-125.5]]
autostemparameters={'Electron energy': 200,'Spherical aberration': 1.4,'Defocus':0,'Aperture semiangle': 24.5,'Source size': 0.882,'Slice size': 20.0,'Pixels':800,'aber':aber,'Scale Factor':0.00570113}
A = ConvStem(parameters=autostemparameters,calc_exp=False)
#fig = plt.figure()
#ax = fig.add_subplot(111)
#psf = numpy.fft.fftshift(A.psf)
#ax.imshow(psf,cmap=cm.hot)
#plt.show()
imAu1 = A.get_image(A.psf, Au1, autostemparameters['Slice size'], autostemparameters['Pixels'], 1)
#fig2 = plt.figure()
#ax2 = fig2.add_subplot(111)
#ax2.imshow(imAu1,cmap=cm.hot)
#plt.show()
autostemparameters['Exp_Image'] = imAu1
optparameters = {'structure': 'Cluster', 'optimizer_type': 'BH', 'size':10,'atomlist': [('Au', 10, 0, 0), ('Cu', 10, 0, 0)],'FIT_SCHEME':'STEM_Cost','STEM_Parameters':autostemparameters}
B = Optimizer(optparameters)
B.algorithm_serial()
def read_parameter_input(input, logger):
"""Function to convert input string, file, or dictionary to a dictionary to contain
the parameters for use by the optimizer class.
input:
input : Can be dictionary, string, or filename
logfile : Name of logfile to write to. Default is None
output:
parameters : dictionary for defining parameters for optimizer with defaults
"""
parameters = None
if isinstance(input, dict):
# If supplied input is already a dictionary set parameters equal to that dictionary
parameters = input
else:
#Check to see if input is a filename
if os.path.exists(input):
try:
#Check if input file is already formatted as a python dictionary
parameters = eval('{%s}' % open(input).read())
except:
f = open(input, 'r')
lines = f.readlines()
f.close()
else:
#lines = input
#Split lines in input string
lines = input.rstrip().split('\n')
if parameters == None:
parameters = {}
for one in lines:
if one != '\n':
p, v = one.strip().split('=')
try:
#Convert numbers to floats
parameters[p.strip()] = float(v.strip())
except:
try:
#Convert list, tuple, and booleans
if p.strip() != 'calc_method':
parameters[p.strip()] = eval(v.strip())
else:
parameters[p.strip()] = v.strip()
except:
try:
#Leave remaining as strings
parameters[p.strip()] = v.strip()
except:
print 'Trouble with input line: ', one
if 'parallel' not in parameters:
parameters['parallel'] = False
else:
parameters['parallel'] = bool(parameters['parallel'])
if 'filename' not in parameters:
parameters['filename'] = 'Output'
else:
parameters['filename'] = str(parameters['filename'])
try:
rank = MPI.COMM_WORLD.Get_rank()
except:
rank = 0
if logger:
if 'loggername' not in parameters:
parameters['loggername'] = '{0}-rank{1}-{2}.log'.format(
parameters['filename'], rank, time.strftime("%Y_%m%d_%H%M%S"))
logger = initialize_logger(parameters['loggername'])
else:
parameters['loggername'] = str(parameters['loggername'])
logger = initialize_logger(parameters['loggername'])
#logging.getLogger(parameters['loggername'])
else:
parameters['loggername'] = None
logger = dummy_logger_no_write()
if ('atomlist' not in parameters):
#Stop program if atom list parameter not in input
logger.critical(
"Input file/string/dictionary must include an atomlist defined as 'atomlist':[('Xx',Concentration,Mass,Chemical Potential)]"
)
logger.critical("Current parameters include:\n" + repr(parameters))
raise RuntimeError(
"Input file/string/dictionary must include an atomlist defined as 'atomlist':[('Xx',Concentration,Mass,Chemical Potential)]"
)
else:
if isinstance(parameters['atomlist'], str):
parameters['atomlist'] = eval(parameters['atomlist'].strip())
if not isinstance(parameters['atomlist'], list):
logger.critical(
'Something is wrong with atomlist parameter: {0}'.format(
parameters['atomlist']))
raise RuntimeError(
"Input file/string/dictionary must include an atomlist defined as 'atomlist':[('Xx',Concentration,Mass,Chemical Potential)]: {0}"
.format(parameters['atomlist']))
if 'structure' not in parameters:
#Stop program if structure parameter not in input
logger.critical(
"Input file/dictionary must include a structure for the simulation as 'structure':'Cluster/Crystal/Defect'"
)
logger.debug("Current parameters include:\n" + repr(parameters))
raise RuntimeError(
"Input file/dictionary must include a structure for the simulation as 'structure':'Cluster/Crystal/Defect'"
)
for one in parameters['atomlist']:
if len(one) != 4:
#Stop program if atom list parameter not properly formatted
logger.critical(
'Format of atom list not correct. Must be [(symbol,concentration,mass,potential)]'
)
logger.debug('Issue in section : {0}'.format(one))
logger.debug('Current atomlist is formatted as : {0}'.format(
parameters['atomlist']))
raise RuntimeError(
'Format of atom list not correct. Must be [(symbol,concentration,mass,potential)]'
)
if 'natoms' not in parameters:
parameters['natoms'] = int(
sum([abs(c) for ind, c, m, u in parameters['atomlist']]))
if rank == 0:
logger.warning('Number of atoms in simulation not set')
logger.warning('Assuming natoms = {0}'.format(
parameters['natoms']))
else:
parameters['natoms'] = int(parameters['natoms'])
parameters['atomlist'] = check_atomlist_concentration(
parameters['atomlist'], parameters['natoms'], parameters['loggername'])
if 'optimizer_type' not in parameters:
parameters['optimizer_type'] = 'Random'
if rank == 0:
logger.info(
'optimizer_type not set. Default values set to Random.')
#Set parameter defaults based on optimizer structure
if parameters['optimizer_type'] == 'GA':
nindiv = 10
genealogy = True
nbests = 100
algtype = 'lambda+mu'
cxpb = 0.8
mutpb = 0.15
natselectscheme = 'tournament'
else:
nindiv = 1
genealogy = False
nbests = 100
algtype = parameters['optimizer_type']
cxpb = 0.0
mutpb = 1.0
if parameters['optimizer_type'] == 'SA':
natselectscheme = 'metropolis'
predator = 'adapting'
elif parameters['optimizer_type'] == 'BH':
natselectscheme = 'metropolis'
else:
natselectscheme = 'best'
if 'nindiv' not in parameters:
parameters['nindiv'] = nindiv
if rank == 0:
logger.info(
'Setting number of individuals in population (nindiv) = {0}'.
format(parameters['nindiv']))
else:
parameters['nindiv'] = int(parameters['nindiv'])
#Parameters for output
if 'genealogy' not in parameters:
parameters['genealogy'] = genealogy
if rank == 0:
logger.info('Setting genealogy = {0}'.format(
parameters['genealogy']))
else:
parameters['genealogy'] = bool(parameters['genealogy'])
if 'output_format' not in parameters:
parameters['output_format'] = 'fitness'
if rank == 0:
logger.info('Setting output format = {0}'.format(
parameters['output_format']))
else:
parameters['output_format'] = str(parameters['output_format']).lower()
if 'allenergyfile' not in parameters:
parameters['allenergyfile'] = False
if rank == 0:
logger.info('Setting allenergyfile = {0}'.format(
parameters['allenergyfile']))
else:
parameters['allenergyfile'] = bool(parameters['allenergyfile'])
if 'best_inds_list' not in parameters:
parameters['best_inds_list'] = True
if rank == 0:
logger.info('Setting best_inds_list = {0}'.format(
parameters['best_inds_list']))
else:
parameters['best_inds_list'] = bool(parameters['best_inds_list'])
if 'number_of_bests' not in parameters:
parameters['number_of_bests'] = nbests
if rank == 0:
logger.info('Setting number_of_bests = {0}'.format(
parameters['number_of_bests']))
else:
parameters['number_of_bests'] = int(parameters['number_of_bests'])
if 'indiv_defect_write' not in parameters:
parameters['indiv_defect_write'] = False
if rank == 0:
logger.info('Setting indiv_defect_write = {0}'.format(
parameters['indiv_defect_write']))
else:
parameters['indiv_defect_write'] = bool(
parameters['indiv_defect_write'])
if 'vacancy_output' not in parameters:
parameters['vacancy_output'] = False
if rank == 0:
logger.info('Setting vacancy_output = {0}'.format(
parameters['vacancy_output']))
else:
parameters['vacancy_output'] = bool(parameters['vacancy_output'])
if 'restart_optimizer' not in parameters:
parameters['restart_optimizer'] = False
else:
parameters['restart_optimizer'] = bool(parameters['restart_optimizer'])
#Parameters for post-processing
if 'lattice_concentration' not in parameters:
parameters['lattice_concentration'] = False
if rank == 0:
logger.info('Setting lattice_concentration = {0}'.format(
parameters['lattice_concentration']))
else:
parameters['lattice_concentration'] = bool(
parameters['lattice_concentration'])
if 'postprocessing' not in parameters:
parameters['postprocessing'] = False
if rank == 0:
logger.info('Setting postprocessing = {0}'.format(
parameters['postprocessing']))
else:
parameters['postprocessing'] = bool(parameters['postprocessing'])
if 'genealogytree' not in parameters:
parameters['genealogytree'] = False
if rank == 0:
logger.info('Setting genealogytree = {0}'.format(
parameters['genealogytree']))
else:
parameters['genealogytree'] = bool(parameters['genealogytree'])
#Parameters for general algorithm
if 'seed' not in parameters:
parameters['seed'] = random.randint(0, 10)
if rank == 0:
logger.info('Setting Random number seed (seed) to {0}'.format(
parameters['seed']))
else:
parameters['seed'] = int(parameters['seed'])
if 'forcing' not in parameters:
parameters['forcing'] = 'Concentration'
if rank == 0:
logger.info('Setting forcing = {0}'.format(parameters['forcing']))
logger.info('Assuming forcing concentration control')
else:
parameters['forcing'] = str(parameters['forcing'])
if 'debug' not in parameters:
parameters['debug'] = ['None']
if rank == 0:
logger.info('Setting debug = {0}'.format(parameters['debug']))
else:
parameters['debug'] = list(parameters['debug'])
if 'None' not in parameters['debug']:
print '***** DEBUGGING RUN *****'
if 'algorithm_type' not in parameters:
parameters['algorithm_type'] = algtype
if rank == 0:
logger.info('Setting algorithm type = {0}'.format(
parameters['algorithm_type']))
else:
parameters['algorithm_type'] = str(parameters['algorithm_type'])
if 'migration_intervals' not in parameters:
parameters['migration_intervals'] = 5
if rank == 0:
logger.info('Setting migration_intervals = '.format(
parameters['migration_intervals']))
else:
parameters['migration_intervals'] = int(
parameters['migration_intervals'])
if 'migration_percent' not in parameters:
parameters['migration_percent'] = 0.05
if rank == 0:
logger.info('Setting migration_percent = {0}'.format(
parameters['migration_percent']))
else:
parameters['migration_percent'] = float(
parameters['migration_percent'])
if 'fingerprinting' not in parameters:
parameters['fingerprinting'] = False
if rank == 0:
logger.info('Setting fingerprinting = {0}'.format(
parameters['fingerprinting']))
else:
parameters['fingerprinting'] = bool(parameters['fingerprinting'])
if 'fpbin' not in parameters:
parameters['fpbin'] = 0.25
if rank == 0:
logger.info('Setting fingerprint bin to {0}'.format(
parameters['fpbin']))
else:
parameters['fpbin'] = float(parameters['fpbin'])
if 'fpcutoff' not in parameters:
parameters['fpcutoff'] = 15.0
if rank == 0:
logger.info('Setting fingerprint cutoff distance to {0}'.format(
parameters['fpcutoff']))
else:
parameters['fpcutoff'] = float(parameters['fpcutoff'])
parameters['bulkfp'] = None
if 'fixed_region' not in parameters:
parameters['fixed_region'] = False
if rank == 0:
logger.info('Setting fixed_region = {0}'.format(
parameters['fixed_region']))
else:
parameters['fixed_region'] = bool(parameters['fixed_region'])
if 'rattle_atoms' not in parameters:
parameters['rattle_atoms'] = False
if rank == 0:
logger.info('Setting rattle_atoms = {0}'.format(
parameters['rattle_atoms']))
else:
parameters['rattle_atoms'] = bool(parameters['rattle_atoms'])
if 'constrain_position' not in parameters:
parameters['constrain_position'] = False
if rank == 0:
logger.info('Setting constrain_position = {0}'.format(
parameters['constrain_position']))
else:
parameters['constrain_position'] = bool(
parameters['constrain_position'])
if 'restart' not in parameters:
parameters['restart'] = False
if rank == 0:
logger.info('Setting restart = {0}'.format(parameters['restart']))
else:
parameters['restart'] = bool(parameters['restart'])
if 'restart_ints' not in parameters:
parameters['restart_ints'] = 0
if rank == 0:
if parameters['restart']:
logger.info('Setting restart_ints = {0}'.format(
parameters['restart_ints']))
else:
parameters['restart_ints'] = int(parameters['restart_ints'])
# Parameters to generate the population and individual
if 'r_ab' not in parameters:
parameters['r_ab'] = 2.5
if rank == 0:
logger.info('Setting r_ab = {0}'.format(parameters['r_ab']))
else:
parameters['r_ab'] = float(parameters['r_ab'])
if 'size' not in parameters:
parameters['size'] = parameters['natoms']**0.33333 * parameters['r_ab']
if rank == 0:
logger.info('Setting size to (natoms)^(1/3)*r_ab = {0}'.format(
parameters['size']))
else:
parameters['size'] = float(parameters['size'])
if 'generate_flag' not in parameters:
parameters['generate_flag'] = 'box'
if rank == 0:
logger.info('Setting default generation scheme = {0}'.format(
parameters['generate_flag']))
else:
parameters['generate_flag'] = str(parameters['generate_flag']).lower()
parameters['solidbulk'] = None
if 'sf' not in parameters:
parameters['sf'] = 1.75
if rank == 0:
if parameters['structure'] == 'Defect':
logger.info('Setting size factor for Defect (sf) = {0}'.format(
parameters['sf']))
else:
parameters['sf'] = float(parameters['sf'])
if 'supercell' not in parameters:
parameters['supercell'] = (1, 1, 1)
if rank == 0:
if parameters['structure'] == 'Defect':
logger.info(
'Setting supercell for Defect (supercell) = {0}'.format(
parameters['supercell']))
else:
parameters['supercell'] = tuple(parameters['supercell'])
if 'solidfile' not in parameters:
if parameters['structure'] == 'Defect':
logger.critical(
'Must provide a file for bulk solid if running a defect simulation.'
)
raise RuntimeError(
'Error: Bulk for Defect not specified. Enter name of file for bulk structure as solidfile parameter'
)
else:
parameters['solidfile'] = None
else:
parameters['solidfile'] = str(parameters['solidfile'])
if 'solidcell' not in parameters:
if rank == 0:
if parameters['structure'] == 'Defect':
logger.warning(
'Warning cell size for Bulk Solid not specified assuming distance between 1st and 3rd atom'
)
parameters['solidcell'] = None
else:
try:
parameters['solidcell'] = eval('numpy.' + parameters['solidcell'])
except:
try:
parameters['solidcell'] = list(parameters['solidcell'])
except:
if rank == 0:
if parameters['structure'] == 'Defect':
logger.warning(
'Warning cell size for Bulk Solid not recognized assuming distance between 1st and 3rd atom'
)
logger.debug('solidcell input as: {0}'.format(
parameters['solidcell']))
parameters['solidcell'] = None
if 'evalsolid' not in parameters:
parameters['evalsolid'] = False
if rank == 0:
if parameters['structure'] == 'Defect':
logger.info('Not evaluating Solid')
else:
parameters['evalsolid'] = bool(parameters['evalsolid'])
if 'finddefects' not in parameters:
parameters['finddefects'] = True
if rank == 0:
if parameters['structure'] == 'Defect':
logger.info('Setting finddefects = {0}'.format(
parameters['finddefects']))
else:
parameters['finddefects'] = bool(parameters['finddefects'])
if 'trackvacs' not in parameters:
parameters['trackvacs'] = False
if rank == 0:
if parameters['structure'] == 'Defect':
logger.info('Setting trackvacs = {0}'.format(
parameters['trackvacs']))
else:
parameters['trackvacs'] = bool(parameters['trackvacs'])
if 'trackswaps' not in parameters:
parameters['trackswaps'] = False
if rank == 0:
if parameters['structure'] == 'Defect':
logger.info('Setting trackswaps = {0}'.format(
parameters['trackswaps']))
else:
parameters['trackswaps'] = bool(parameters['trackswaps'])
if 'random_loc_start' not in parameters:
parameters['random_loc_start'] = False
if rank == 0:
if parameters['structure'] == 'Defect':
logger.info('Setting random_loc_start = {0}'.format(
parameters['random_loc_start']))
else:
parameters['random_loc_start'] = bool(parameters['random_loc_start'])
if 'random_vac_start' not in parameters:
parameters['random_vac_start'] = False
if rank == 0:
if parameters['structure'] == 'Defect':
logger.info('Setting random_vac_start = {0}'.format(
parameters['random_vac_start']))
else:
parameters['random_vac_start'] = bool(parameters['random_vac_start'])
if 'purebulkenpa' not in parameters:
if parameters['structure'] == 'Defect':
parameters['purebulkenpa'] = None
else:
parameters['purebulkenpa'] = 0
if 'natomsbulk' not in parameters:
if parameters['structure'] == 'Defect':
parameters['natomsbulk'] = None
else:
parameters['natomsbulk'] = 0
if 'surfacefile' not in parameters:
if parameters['structure'] == 'Surface':
logger.critical(
'Must provide a file for bulk solid if running a defect simulation.'
)
raise RuntimeError(
'Error: Bulk for Defect not specified. Enter name of file for bulk structure as solidfile parameter'
)
else:
parameters['surfacefile'] = None
if 'surfacecell' not in parameters:
parameters['surfacecell'] = None
if 'surftopthick' not in parameters:
parameters['surftopthick'] = 0
if 'cell_shape_options' not in parameters:
parameters['cell_shape_options'] = ['cubic', 'hexagonal', 'triclinic', \
'monoclinic', 'orthorhombic', 'tetragonal']
else:
parameters['cell_shape_options'] = list(
parameters['cell_shape_options'])
if parameters['cell_shape_options'] == 'all':
parameters['cell_shape_options']=['cubic', 'hexagonal', 'triclinic', \
'monoclinic', 'orthorhombic', 'tetragonal']
if rank == 0:
if parameters['structure'] == 'Crystal':
logger.info('Assuming following cell shape options: {0}'.format(
parameters['cell_shape_options']))
if 'alloy' not in parameters:
parameters['alloy'] = True
if rank == 0:
logger.info('Setting alloy = {0}'.format(parameters['alloy']))
else:
parameters['alloy'] = bool(parameters['alloy'])
#Parameters to Evaluate an Individual
if 'calc_method' not in parameters:
if 'vaspcalc' in parameters:
parameters['calc_method'] = 'VASP'
if rank == 0:
logger.info(
'Setting calculation method to VASP. calc_method = {0}'.
format(parameters['calc_method']))
elif 'pair_style' in parameters:
parameters['calc_method'] = 'LAMMPS'
if rank == 0:
logger.info(
'Setting calculation method to LAMMPS. calc_method = {0}'.
format(parameters['calc_method']))
elif 'lammps_min' in parameters:
parameters['calc_method'] = 'LAMMPS'
if rank == 0:
logger.info(
'Setting calculation method to LAMMPS. calc_method = {0}'.
format(parameters['calc_method']))
else:
parameters['calc_method'] = 'LennardJones'
if rank == 0:
logger.info(
'Calculation method not specified assuming ase LennardJones. calc_method = {0}'
.format(parameters['calc_method']))
else:
parameters['calc_method'] = str(parameters['calc_method'])
if 'vaspcalc' not in parameters:
parameters['vaspcalc'] = "Vasp()"
if rank == 0:
if parameters['calc_method'] == 'VASP':
logger.info('Setting vaspcalc = {0}'.format(
parameters['vaspcalc']))
if 'pair_style' not in parameters:
if rank == 0:
if parameters['calc_method'] == 'LAMMPS':
logger.info(
'No pair_style for LAMMPS specified. Assuming Lennard-Jones'
)
parameters['pair_style'] = None
if 'bopcutoff' not in parameters:
if parameters['pair_style'] == 'bop':
logger.critical('BOP potential requires a cutoff distance.')
logger.debug('Parameters = {0}'.format(parameters))
raise RuntimeError(
"ERROR:Cutoff distance must be supplied with bop potential! Specify with bopcutoff=N"
)
else:
parameters['bopcutoff'] = 0
else:
parameters['bopcutoff'] = float(parameters['bopcutoff'])
if 'buckcutoff' not in parameters:
parameters['buckcutoff'] = 1
if rank == 0:
if parameters['pair_style'] == 'buck':
logger.warning('Setting buckingham potential cutoff = ' +
repr(parameters['buckcutoff']))
else:
parameters['buckcutoff'] = int(parameters['buckcutoff'])
if 'buckparameters' not in parameters:
parameters['buckparameters'] = ['* * 100.00 1.5 200.0']
if rank == 0:
if parameters['pair_style'] == 'buck':
logger.warning('Setting Buckingham parameters = {0}'.format(
parameters['buckparameters']))
if 'ps_name' not in parameters:
parameters['ps_name'] = None
if rank == 0:
if parameters['pair_style'] == 'other':
logger.info('Setting ps_name = {0}'.format(
parameters['ps_name']))
if 'pair_coeff' not in parameters:
if parameters['pair_style'] == 'other':
logger.critical(
'Must provide pair_coeff for use with pair_style = other')
raise RuntimeError(
"ERROR:Coefficients for structure=other potential not specifed. Use pair_coeff"
)
else:
parameters['pair_coeff'] = None
if 'ps_other' not in parameters:
parameters['ps_other'] = None
if rank == 0:
if parameters['pair_style'] == 'other':
logger.info('Setting ps_other = {0}'.format(
parameters['ps_other']))
if 'pot_file' not in parameters:
parameters['pot_file'] = None
if rank == 0:
if parameters['calc_method'] == 'LAMMPS':
logger.warning('No potential file for LAMMPS specified.')
else:
parameters['pot_file'] = str(parameters['pot_file'])
if 'lammps_keep_files' not in parameters:
parameters['lammps_keep_files'] = False
if rank == 0:
if parameters['calc_method'] == 'LAMMPS':
logger.info('Setting lammps_keep_files = {0}'.format(
parameters['lammps_keep_files']))
else:
parameters['lammps_keep_files'] = bool(parameters['lammps_keep_files'])
if 'lammps_thermo_steps' not in parameters:
parameters['lammps_thermo_steps'] = 1
if rank == 0:
if parameters['calc_method'] == 'LAMMPS':
logger.info('Setting lammps_thermo_steps = {0}'.format(
parameters['lammps_thermo_steps']))
else:
parameters['lammps_thermo_steps'] = int(
parameters['lammps_thermo_steps'])
if ('lammps_min' in parameters) and (parameters['lammps_min'] != None):
parameters['ase_min'] = False
else:
parameters['lammps_min'] = None
if rank == 0:
if parameters['calc_method'] == 'LAMMPS':
logger.info('No Local minimization implemented in LAMMPS')
if 'lammps_min_style' not in parameters:
if parameters['lammps_min']:
parameters['lammps_min_style'] = 'cg'
if rank == 0:
if parameters['calc_method'] == 'LAMMPS':
logger.info(
'Setting lammps_min_style = {0}'.format(
'lammps_min_style') +
'. Using LAMMPS conjugate gradient local minimizer')
else:
parameters['lammps_min_style'] = None
if 'large_box_size' not in parameters:
parameters['large_box_size'] = 500.0
if rank == 0:
if parameters['structure'] == 'Cluster':
logger.info('Setting large_box_size to {0}'.format(
parameters['large_box_size']))
else:
parameters['large_box_size'] = float(parameters['large_box_size'])
if 'ase_min' not in parameters:
if parameters['calc_method'] == 'LennardJones':
parameters['ase_min'] = True
else:
parameters['ase_min'] = False
if rank == 0:
logger.info('Setting ase_min = {0}'.format(parameters['ase_min']))
else:
parameters['ase_min'] = bool(parameters['ase_min'])
if 'ase_min_fmax' not in parameters:
parameters['ase_min_fmax'] = 0.01
if rank == 0:
if parameters['ase_min']:
logger.info(
'Setting ASE maximum force value (ase_min_fmax) = {0}'.
format(parameters['ase_min_fmax']))
else:
parameters['Sovlerfmax'] = float(parameters['ase_min_fmax'])
if 'ase_min_maxsteps' not in parameters:
parameters['ase_min_maxsteps'] = 2500
if rank == 0:
if parameters['ase_min']:
logger.info(
'Setting maximum number of steps for BFGS solver (ase_min_maxsteps) = {0}'
.format(parameters['ase_min_maxsteps']))
else:
parameters['SovlerMxSteps'] = int(parameters['ase_min_maxsteps'])
#Parameters for Crossovers
if 'cxpb' not in parameters:
parameters['cxpb'] = cxpb
if rank == 0:
logger.info('Setting crossover probability (cxpb) = {0}'.format(
parameters['cxpb']))
else:
parameters['cxpb'] = float(parameters['cxpb'])
if 'cx_scheme' not in parameters:
parameters['cx_scheme'] = 'cxtp'
if rank == 0:
logger.info(
'Assuming two-point crossover. Setting cx_scheme = {0}'.
format(parameters['cx_scheme']))
else:
parameters['cx_scheme'] = str(parameters['cx_scheme']).lower()
if 'selection_scheme' not in parameters:
parameters['selection_scheme'] = 'tournament2'
if rank == 0:
logger.info('Setting selection_scheme = {0}'.format(
parameters['selection_scheme']))
else:
parameters['selection_scheme'] = str(
parameters['selection_scheme']).lower()
#Parameters for Mutations
if 'mutpb' not in parameters:
parameters['mutpb'] = mutpb
if rank == 0:
logger.info('Setting mutation probability (mutpb) = {0}'.format(
parameters['mutpb']))
else:
parameters['mutpb'] = float(parameters['mutpb'])
if 'mutation_options' not in parameters:
if parameters['structure'] == 'Cluster':
parameters['mutation_options']=['lattice_alteration','rotation',\
'permutation','scale_size']
elif parameters['structure'] == 'Crystal':
parameters['mutation_options']=['lattice_alteration','rotation',\
'permutation','scale_size', 'cell_shape', 'lammps_box_relax']
elif parameters['structure'] == 'Defect':
parameters['mutation_options'] = [
'lattice_alteration', 'rotation', 'permutation'
]
if rank == 0:
logger.info('Setting mutations options = {0}'.format(
parameters['mutation_options']))
else:
parameters['mutation_options'] = list(parameters['mutation_options'])
for i in range(len(parameters['mutation_options'])):
parameters['mutation_options'][i] = parameters['mutation_options'][
i].lower()
BHFlag = False
for one in parameters['mutation_options']:
if 'basin_hop' in one:
BHFlag = True
if 'bh_steps' not in parameters:
parameters['bh_steps'] = 100
if rank == 0:
if BHFlag == True:
logger.warning(
'Max steps not specified for Basin Hop mutation, setting bh_steps = {0}'
.format(parameters['bh_steps']))
else:
parameters['bh_steps'] = int(parameters['bh_steps'])
if 'bh_temp' not in parameters:
parameters['bh_temp'] = 1000 * 8.617385692256675e-05
if rank == 0:
if BHFlag == True:
logger.warning(
'Temperature not set for Basin Hop mutation, setting bh_temp in kT = {0}'
.format(parameters['bh_temp']))
else:
parameters['bh_temp'] = float(parameters['bh_temp'])
if 'mutant_add' not in parameters:
parameters['mutant_add'] = False
if rank == 0:
logger.info('Setting mutant_add = {0}'.format(
parameters['mutant_add']))
else:
parameters['mutant_add'] = bool(parameters['mutant_add'])
if 'quench_max_temp' not in parameters:
parameters['quench_max_temp'] = 1000
if rank == 0:
if 'quench' in parameters['mutation_options']:
logger.info('Setting quench_max_temp = {0}'.format(
parameters['quench_max_temp']))
else:
parameters['quench_max_temp'] = int(parameters['quench_max_temp'])
if 'quench_min_temp' not in parameters:
parameters['quench_min_temp'] = 2
if rank == 0:
if 'quench' in parameters['mutation_options']:
logger.info('Seting quench_min_temp = {0}'.format(
parameters['quench_min_temp']))
else:
parameters['quench_min_temp'] = int(parameters['quench_min_temp'])
if 'quench_step_size' not in parameters:
parameters['quench_step_size'] = 0.01
if rank == 0:
if 'quench' in parameters['mutation_options']:
logger.info('Setting quench_step_size = {0}'.format(
parameters['quench_step_size']))
else:
parameters['quench_step_size'] = float(parameters['quench_step_size'])
if 'quench_n_steps_1' not in parameters:
parameters['quench_n_steps_1'] = 10000
if rank == 0:
if 'quench' in parameters['mutation_options']:
logger.info('Setting quench_n_steps_1 = {0}'.format(
parameters['quench_n_steps_1']))
else:
parameters['quench_n_steps_1'] = int(parameters['quench_n_steps_1'])
if 'quench_n_steps_2' not in parameters:
parameters['quench_n_steps_2'] = parameters['quench_n_steps_1'] * 2
if rank == 0:
if 'quench' in parameters['mutation_options']:
logger.info('Setting quench_n_steps_2 = {0}'.format(
parameters['quench_n_steps_2']))
else:
parameters['quench_n_steps_2'] = int(parameters['quench_n_steps_2'])
if 'isolate_mutation' not in parameters:
parameters['isolate_mutation'] = False
if rank == 0:
logger.info('Setting isolate_mutation flag = {0}'.format(
parameters['isolate_mutation']))
else:
parameters['isolate_mutation'] = bool(parameters['isolate_mutation'])
#Parameters for Selection
if 'fitness_scheme' not in parameters:
parameters['fitness_scheme'] = 'totalenfit'
if rank == 0:
logger.info('Setting fitness_scheme = {0}'.format(
parameters['fitness_scheme']))
else:
parameters['fitness_scheme'] = str(
parameters['fitness_scheme']).lower()
if 'energy_cutoff_factor' not in parameters:
parameters['energy_cutoff_factor'] = 10.0
if rank == 0:
logger.info('Setting energy_cutoff_factor = {0}'.format(
parameters['energy_cutoff_factor']))
else:
parameters['energy_cutoff_factor'] = float(
parameters['energy_cutoff_factor'])
if 'stem_parameters' not in parameters:
if 'stem' in parameters['fitness_scheme']:
logger.critical(
'Must provide stem_parameters for STEM_Cost calculation')
raise RuntimeError(
"STEM parameters not specified. Cannot simulate image files")
else:
parameters['stem_parameters'] = {}
else:
stemparams = dict(parameters['stem_parameters'])
if 'stem_keep_files' not in parameters:
parameters['stem_keep_files'] = True
if rank == 0:
if 'stem' in parameters['fitness_scheme']:
logger.info('Setting stem_keep_files = {0}'.format(
parameters['stem_keep_files']))
else:
parameters['stem_keep_files'] = parameters['stem_keep_files']
if 'stem_coeff' not in parameters:
parameters['stem_coeff'] = None
if rank == 0:
if 'stem' in parameters['fitness_scheme']:
logger.info('Setting stem_coeff with first individual')
else:
try:
parameters['stem_coeff'] = float(parameters['stem_coeff'])
except:
if rank == 0:
if 'stem' in parameters['fitness_scheme']:
logger.warning(
'Trouble reading stem_coeff input. stem_coeff = {0}'.
format(parameters['stem_coeff']))
parameters['stem_coeff'] = None
if 'stem' in parameters['fitness_scheme']:
#Initialize function for experimental image
from MAST.structopt.tools.StemCalc import ConvStem
logger.info('Initializing ConvStem Calculator')
parameters['stemcalc'] = ConvStem(
parameters=stemparams,
tmp_dir='/' + os.getcwd() + '/ConvStemImages/',
keep_files=parameters['stem_keep_files'])
else:
parameters['stemcalc'] = None
if 'constrain_swaps' not in parameters:
if 'IntSwap' in parameters['mutation_options']:
parameters['swaplist'] = None
if rank == 0:
logger.info('Setting swaplist for IntSwap = None')
elif parameters['fitness_scheme'] == 'chempotswap':
parameters['swaplist'] = None
if rank == 0:
logger.info('Setting swaplist for chempotswap = None')
else:
parameters['swaplist'] = False
parameters['constrain_swaps'] = False
if rank == 0:
logger.info('Setting swaplist = False')
else:
parameters['swaplist'] = parameters['constrain_swaps']
if 'natural_selection_scheme' not in parameters:
parameters['natural_selection_scheme'] = natselectscheme
if rank == 0:
logger.info('Setting natural_selection_scheme = {0}'.format(
parameters['natural_selection_scheme']))
else:
parameters['natural_selection_scheme'] = str(
parameters['natural_selection_scheme']).lower()
if 'tournsize' not in parameters:
parameters['tournsize'] = 3
if 'tournament' in parameters[
'selection_scheme'] or 'tournament' in parameters[
'natural_selection_scheme']:
if rank == 0:
logger.info('Setting Tournament size (tournsize) = {0}'.format(
parameters['tournsize']))
else:
parameters['tournsize'] = int(parameters['tournsize'])
if 'fusslimit' not in parameters:
parameters['fusslimit'] = 10.0
if rank == 0:
logger.info('Setting FUSS limit (fusslimit) = {0}'.format(
parameters['fusslimit']))
else:
parameters['fusslimit'] = float(parameters['fusslimit'])
if 'metropolis_temp' not in parameters:
parameters['metropolis_temp'] = 30.0
if rank == 0:
logger.info('Setting metropolis_temp = {0}'.format(
parameters['metropolis_temp']))
else:
parameters['metropolis_temp'] = float(parameters['metropolis_temp'])
if 'mark' not in parameters:
parameters['mark'] = None
#Parameters for Convergence
if 'convergence_scheme' not in parameters:
parameters['convergence_scheme'] = 'max_gen'
if rank == 0:
logger.info(
'Setting convergence scheme (convergence_scheme) = {0}'.format(
parameters['convergence_scheme']))
else:
parameters['convergence_scheme'] = str(
parameters['convergence_scheme']).lower()
if 'maxgen' not in parameters:
parameters['maxgen'] = 5
if rank == 0:
logger.info(
'Setting Max Number of generations (maxgen) = {0}'.format(
parameters['maxgen']))
else:
parameters['maxgen'] = int(parameters['maxgen'])
if 'reqrep' not in parameters:
parameters['reqrep'] = 10
if rank == 0:
if 'rep' in parameters['convergence_scheme']:
logger.info(
'Setting max number of energy repetitions (reqrep) = {0}'.
format(parameters['reqrep']))
else:
parameters['reqrep'] = int(parameters['reqrep'])
if 'tolerance' not in parameters:
parameters['tolerance'] = 0.001
if rank == 0:
if 'rep' in parameters['convergence_scheme']:
logger.info(
'Setting energy tolerance (tolerance) = {0}'.format(
parameters['tolerance']))
else:
parameters['tolerance'] = float(parameters['tolerance'])
if 'predator' not in parameters:
parameters['predator'] = 'mutation_dups'
if rank == 0:
logger.info('Setting predator = {0}'.format(
parameters['predator']))
else:
parameters['predator'] = str(parameters['predator']).lower()
if 'adaptbegin' not in parameters:
parameters['adaptbegin'] = 0.75
if rank == 0:
if parameters['predator'] == 'adapting':
logger.info(
'Setting adaptation predator to begin (adaptbegin) at genrep*{0}'
.format(parameters['adaptbegin']))
else:
parameters['adaptbegin'] = float(parameters['adaptbegin'])
if 'adaptmultiplier' not in parameters:
parameters['adaptmultiplier'] = 3.0
if rank == 0:
if parameters['predator'] == 'adapting':
logger.info(
'Setting adaptation predator multiplier (adaptmultiplier) = {0}'
.format(parameters['adaptmulitplier']))
else:
parameters['adaptmultiplier'] = float(parameters['adaptmultiplier'])
if 'demin' not in parameters:
parameters['demin'] = 0.005
if rank == 0:
logger.info(
'Setting cutoff convergence energy (demin) = {0}'.format(
parameters['demin']))
else:
parameters['demin'] = float(parameters['demin'])
return parameters