def addAdsorbate(constraint="1"):
constrnts = AND(COMPLETED, RELAX, SURFACE, SYMMETRIC(False), constraint)
ads = {'H': ['O1']}
output = db.query(['fwid', 'params_json', 'finaltraj_pckl'], constrnts)
question = 'Are you sure you want to add adsorbates to %d slabs?' % len(
output)
if ask(question):
for fw, paramStr, ftraj in output:
params = json.loads(paramStr)
newsurf = surfFuncs.adsorbedSurface(
ftraj, json.loads(params['facet_json']), ads)
if jobs.Job(params).spinpol():
newsurf.set_initial_magnetic_moments([
3 if e in misc.magElems else 0
for e in newsurf.get_chemical_symbols()
])
ase.visualize.view(newsurf)
params['name'] += '_' + printAds(ads)
params['surfparent'] = fw
params['inittraj_pckl'] = pickle.dumps(newsurf)
params['adsorbates_json'] = json.dumps(ads)
job = jobs.Job(params)
if job.new():
viz.view(newsurf)
question = 'Does this structure look right?\n' + abbreviateDict(
params)
if ask(question):
job.check()
job.submit()
misc.launch()
def calculateBulkModulus(constraint="1"):
cons = AND(LATTICEOPT,
constraint) # necessary for bulkmodulus calc to be valid
output = db.query([
'fwid', 'storage_directory', 'job_name', 'structure_ksb',
'bulkvacancy_ksb', 'bulkscale_ksb', 'system_type_ksb',
'planewave_cutoff', 'xc', 'kptden_ksb', 'psp_ksb', 'dwrat_ksb',
'econv_ksb', 'mixing_ksb', 'nmix_ksb', 'maxstep_ksb', 'nbands_ksb',
'sigma_ksb', 'fmax_ksb', 'dftcode'
], cons)
existing = manage.listOfIncompleteJobStrs()
question = 'Are you sure you want to calculate bulk modulus for %d structures?' % len(
output)
if ask(question):
newjbs = []
for fwid, stor_dir, name, structure, bvj, bsj, kind, pw, xc, kptden, psp, dwrat, econv, mixing, nmix, maxstep, nbands, sigma, fmax, dftcode in output:
params = {
'jobkind': 'bulkmod',
'strain': 0.03,
'inittraj_pckl':
misc.restoreMagmom(misc.storageDir2pckl(stor_dir)),
'name': name + '_bulkmod',
'structure': structure,
'bulkvacancy_json': bvj,
'bulkscale_json': bsj,
'kind': kind,
'dftcode': dftcode,
'pw': pw,
'xc': xc,
'kptden': kptden,
'psp': psp,
'dwrat': dwrat,
'econv': econv,
'mixing': mixing,
'nmix': nmix,
'maxstep': maxstep,
'nbands': nbands,
'sigma': sigma,
'fmax': fmax,
'parent': fwid
}
job = jobs.Job(params)
if job.new(existing): newjbs.append(job)
if ask("launch %d new jobs?" % len(newjbs)):
for j in newjbs:
j.submit()
misc.launch()
def getXCcontribs(constraint="1"):
"""NOT YET DEBUGGED"""
cons = AND(GPAW, RELAXORLAT, constraint)
output = db.query([
'fwid', 'storage_directory', 'job_name', 'system_type_ksb',
'planewave_cutoff', 'xc', 'kptden_ksb', 'psp_ksb', 'dwrat_ksb',
'econv_ksb', 'mixing_ksb', 'nmix_ksb', 'maxstep_ksb', 'nbands_ksb',
'sigma_ksb', 'fmax_ksb', 'dftcode'
], cons)
question = 'Are you sure you want to calculate XC contributions for %d structures?' % len(
output)
if ask(question):
newjbs = []
for fwid, stor_dir, name, kind, pw, xc, kptden, psp, dwrat, econv, mixing, nmix, maxstep, nbands, sigma, fmax, dftcode in output:
params = {
'jobkind': 'xc',
'inittraj_pckl': storageDir2pckl(stor_dir),
'name': name + '_xc',
'kind': kind,
'dftcode': dftcode,
'pw': pw,
'xc': xc,
'kptden': kptden,
'psp': psp,
'dwrat': dwrat,
'econv': econv,
'mixing': mixing,
'nmix': nmix,
'maxstep': maxstep,
'nbands': nbands,
'sigma': sigma,
'fmax': fmax,
'parent': fwid
}
job = jobs.Job(params)
if job.new() and ask('does this look right %s' % str(params)):
newjbs.append(job)
if ask("launch %d new jobs?" % len(newjbs)):
for j in newjbs:
j.submit()
misc.launch()
def addInterstitial(constraint="1", emttol=0.2, vischeck=True, limit=1):
"""
Specify an interstitial and the number (up to which) you want to add them to the cell
Filter duplicates via EMT-calculated energy (stipulate a tolerance)
Assume final structure is triclinic
"""
inter, num = 'H', 2
cons = AND(LATTICEOPT, QE, PW(500), KPTDEN(2), constraint)
output = db.query(
[
'fwid',
'storage_directory',
'job_name',
'system_type_ksb' # generic parameters
,
'planewave_cutoff',
'xc',
'kptden_ksb',
'psp_ksb',
'dwrat_ksb',
'econv_ksb',
'mixing_ksb',
'nmix_ksb',
'maxstep_ksb',
'nbands_ksb',
'sigma_ksb',
'fmax_ksb',
'dftcode' # calc parameters
,
'bulkvacancy_ksb',
'bulkscale_ksb'
],
cons,
limit=limit) # bulk parameters
existing = manage.listOfIncompleteJobStrs(
) # check currently running/fizzled jobs to avoid duplicates
if ask('Are you sure you want to create interstitials for %d structures?' %
len(output)):
newjbs, totJbs, totStruct = [], 0, 0 #initialize variables
for fwid, stor_dir, name, kind, pw, xc, kptden, psp, dwrat, econv, mixing, nmix, maxstep, nbands, sigma, fmax, dftcode, bv, bs in output:
params = {
'jobkind': 'vcrelax',
'kind': 'bulk',
'dftcode': dftcode,
'name': name,
'pw': pw,
'xc': xc,
'kptden': kptden,
'psp': psp,
'bulkparent': fwid,
'dwrat': dwrat,
'econv': econv,
'mixing': mixing,
'nmix': nmix,
'maxstep': maxstep,
'nbands': nbands,
'sigma': sigma,
'fmax': fmax,
'structure': 'triclinic',
'bulkvacancy_json': bv,
'bulkscale_json': bs
} #bulk
ftraj = misc.storageDir2pckl(stor_dir)
spnpl = any([
x > 0
for x in pickle.loads(ftraj).get_initial_magnetic_moments()
])
trajs = [[(pickle.loads(ftraj), '')]] # initial state
for i in range(num):
lastround = trajs[
-1] # for all structures with n - 1 interstitials (note: initially there is only one, with 0 interstitials)
trajs.append(
[]
) # initialize container for all new structures with n interstitials
for inputtraj, strname in lastround:
for newtraj, newname in interstitialFuncs.getInterstitials(
inputtraj, inter, spnpl):
trajs[-1].append(
(newtraj, strname + newname)
) # add all new trajs found for all previous structures (append suffix to name)
def modParams(
par, trj, nam
): # For a given new traj and name, create input parameters
p = copy.deepcopy(
par) # All parameters common to all of these jobs
p['name'] += nam # Unique job name
p['inittraj_pckl'] = pickle.dumps(trj) # Initial structure
return p
def getEMT(x):
a = pickle.loads(x['inittraj_pckl'])
a.set_calculator(emt.EMT())
return a.get_potential_energy()
onelevel = [
modParams(params, item[0], item[1]) for sublist in trajs[1:]
for item in sublist
] #collapse list of lists to a single list of input parameters
emtPairs = sorted(
zip(map(getEMT, onelevel),
onelevel)) # pairs of (Energy,Params) ordered by energy
emtCounter, jbs = 0, []
for e, p in emtPairs:
if e - emtCounter > emttol:
jbs.append(jobs.Job(p))
emtCounter = e
newjbs.extend([x for x in jbs if x.new(existing)])
totJbs += len(jbs)
totStruct += len(onelevel)
check = ask(
'Do you want to check %d/%d new jobs? (%d filtered by emt)' %
(len(newjbs), totStruct, totStruct - totJbs))
if not check and ask('Do you want to exit?'): return 0
for jb in newjbs:
question = 'Does this structure look right?\n' + misc.abbreviateDict(
jb.params)
if vischeck: viz.view(pickle.loads(jb.params['inittraj_pckl']))
if not vischeck or ask(question): jb.submit()
misc.launch()
def getBareSlab(constraint="1", check=True):
"""Create a bare slab for some set of relaxed bulk structures. Alter parameters below as necessary"""
cons = AND(RELAXORLAT, BULK, MBEEF, PAWPSP, BCC, PW(1500), KPTDEN(2),
constraint)
facet = [1, 0, 0]
xy = [1, 1]
layers = 6
constrained = 2
symmetric = 1
vacuum = 10
vacancies = []
output = db.query([
'fwid', 'storage_directory', 'job_name', 'structure_ksb',
'planewave_cutoff', 'xc', 'kptden_ksb', 'psp_ksb', 'dwrat_ksb',
'econv_ksb', 'mixing_ksb', 'nmix_ksb', 'maxstep_ksb', 'nbands_ksb',
'sigma_ksb', 'fmax_ksb', 'dftcode'
], cons)
question = 'Are you sure you want to create bare slabs for %d structures?' % len(
output)
if ask(question):
existing = manage.listOfIncompleteJobStrs()
newjbs = []
for fwid, stor_dir, name, structure, pw, xc, kptden, psp, dwrat, econv, mixing, nmix, maxstep, nbands, sigma, fmax, dftcode in output:
ftraj = misc.readOnSherOrSlac(stor_dir + 'raw.pckl')
surf, img = surfFuncs.bulk2surf(ftraj, facet, xy, layers,
constrained, symmetric, vacuum,
vacancies)
name += '_' + ','.join(map(str, facet)) + '_' + 'x'.join(
map(str, (xy + [layers])))
params = {
'jobkind': 'relax',
'inittraj_pckl': pickle.dumps(surf),
'name': name,
'kind': 'surface',
'dftcode': dftcode,
'structure': structure # don't need bulk vacancies/scale?
,
'pw': pw,
'xc': xc,
'kptden': kptden,
'psp': psp,
'bulkparent': fwid,
'dwrat': dwrat,
'econv': econv,
'mixing': mixing,
'nmix': nmix,
'maxstep': maxstep,
'nbands': nbands,
'sigma': sigma,
'fmax': fmax,
'sites_base64': img,
'facet_json': json.dumps(facet),
'xy_json': json.dumps(xy),
'layers': layers,
'constrained': constrained,
'symmetric': symmetric,
'vacuum': vacuum,
'vacancies_json': json.dumps(vacancies),
'adsorbates_json': json.dumps({})
}
job = jobs.Job(params)
if job.new(existing):
if check: viz.view(surf)
question = 'Does this structure look right?\n' + misc.abbreviateDict(
params)
if not check or ask(question): job.submit()
misc.launch()
def getVibs(constraint="1"):
"""
Launch vibration calculation for all latticeopt or (vc-)relax jobs that meet some (SQL) constraint in the shared database
Only (and all) nonmetal atoms will be vibrated (defined in misc.py)
"""
existing = manage.listOfIncompleteJobStrs()
cons = AND(
QE, RELAXORLAT, constraint
) #could we add a column to quickly filter things with nonmetals?
output = db.query(
[
'fwid',
'storage_directory',
'job_name',
'system_type_ksb' # generic parameters
,
'planewave_cutoff',
'xc',
'kptden_ksb',
'psp_ksb',
'dwrat_ksb',
'econv_ksb',
'mixing_ksb',
'nmix_ksb',
'maxstep_ksb',
'nbands_ksb',
'sigma_ksb',
'fmax_ksb',
'dftcode' # calc parameters
,
'structure_ksb',
'bulkvacancy_ksb',
'bulkscale_ksb' # possible bulk parameters
,
'facet_ksb',
'xy_ksb',
'layers_ksb',
'constrained_ksb',
'symmetric_ksb',
'vacuum_ksb',
'vacancies_ksb',
'adsorbates_ksb',
'sites_ksb'
],
cons) # possible surf parameters
question = 'Are you sure you want to calculate vibrations for %d structures?' % len(
output)
if ask(question):
newjbs = []
for fwid, stor_dir, name, kind, pw, xc, kptden, psp, dwrat, econv, mixing, nmix, maxstep, nbands, sigma, fmax, dftcode, structure, bv, bs, facet, xy, lay, const, sym, vac, vacan, ads, sites in output:
inittraj_pckl = storageDir2pckl(stor_dir)
atoms = pickle.loads(inittraj_pckl)
nonmetal_inds = [
i for i, x in enumerate(atoms.get_chemical_symbols())
if x in misc.nonmetalSymbs
]
if len(nonmetal_inds) > 0:
params = {
'jobkind': 'vib',
'inittraj_pckl': inittraj_pckl,
'name': name + '_vib',
'kind': kind,
'dftcode': dftcode,
'structure': structure,
'pw': pw,
'xc': xc,
'kptden': kptden,
'psp': psp,
'dwrat': dwrat,
'econv': econv,
'mixing': mixing,
'nmix': nmix,
'maxstep': maxstep,
'nbands': nbands,
'sigma': sigma,
'fmax': fmax,
'structure': structure,
'bulkvacancy_json': bv,
'bulkscale_json': bs,
'facet_json': facet,
'xy_json': xy,
'layers': lay,
'constrained': const,
'symmetric': sym,
'vacuum': vac,
'vacancies_json': vacan,
'adsorbates_json': ads,
'sites_base64': sites,
'parent': fwid,
'vibids_json': json.dumps(nonmetal_inds),
'delta': 0.04
}
job = jobs.Job(params)
if job.new() and ask('does this look right %s' % str(params)):
newjbs.append(job)
if ask("launch %d new jobs?" % len(newjbs)):
for j in newjbs:
j.submit(existing)
misc.launch()
def query(self, rowid):
return db.query(self.inputcols, 'jobid = %d' % rowid)[0]