def amp_jobs(fdata,ndata,ntrain,job, dstype, dslist, amp_pes, Lload_amp, HL, E_conv, f_conv_list, Lgraph, ncore, n_mol, Ltwinx):
### parameter file
outf = "hyperparam_" + job + ".dat" # to write input info
total_images = get_total_image(fdata,ndata) # can read extxyz, OUTCAR,
ntotal = len(total_images)
print(f"dstype {dstype} dslist {dslist} with total {ntotal} data")
tr_images, te_images = data_selection(total_images, dstype, dslist, job)
if job == 'tr':
ntrain = len(tr_images) # ntrain is redefined in case of job==train
ntest = len(te_images)
print(f"{ntest} test images")
### Force input in training
f_conv = f_conv_list[0]
f_coeff = 0.1 # default
if len(f_conv_list) == 2:
f_coeff = f_conv_list[1]
### AMP running parts
if re.search("pr", job):
y=[]
for mol in total_images:
y.append(mol.get_potential_energy())
if fdata.endswith('extxyz'):
mplot_nvector([],y,fdata.split(".")[0],'sample','E(eV)')
elif fdata == "OUTCAR":
mplot_nvector([],y,Xtitle='sample',Ytitle='E(eV)')
### JOB == TRAINING
elif re.search("tr",job):
pwd = os.getcwd()
### prevent running tr in tr-directory
amp_pot = None
if os.path.isfile(amp_pes):
if not Lload_amp:
print(f"There is {amp_pes}, can't run training")
sys.exit(10)
else:
amp_pot = amp_pes
print("data training:total sets %d/%d" % (ntrain, ntotal))
f_write(outf, HL, E_conv, f_conv, f_coeff, ntotal, dstype, dslist)
calc_train_images(tr_images, HL, E_conv, f_conv, f_coeff, ncore, amp_pot=amp_pot)
### Test after training:: not run in QSUB
server = socket.gethostname()
if server == 'chi' or server == 'login':
outf = "runamp_te.dat" # to write input info
title, suptitle = get_title(fdata, HL, E_conv,f_conv, f_coeff, ntrain, ntest)
print("data test:train sets %d/%d" % (ntest, ntrain))
rmserr, max_res = calc_test_images(te_images, amp_pes, title, suptitle,ncore,Lgraph,Ltwinx=Ltwinx)
f_write(outf, HL, E_conv, f_conv, f_coeff, ntotal,dstype, dslist, err=rmserr, max_res=max_res)
### JOB == TEST
elif re.search("te",job):
title, suptitle = get_title( fdata, HL, E_conv, f_conv,f_coeff, ntrain, ntest)
print("data test:train sets %d/%d" % (ntest, ntrain))
rmserr, max_res = calc_test_images(te_images, amp_pes, title, suptitle,ncore,Lgraph,nmol=n_mol,Ltwinx=Ltwinx)
f_write(outf, HL, E_conv, f_conv,f_coeff, ntotal,dstype, dslist, err=rmserr, max_res=max_res)
return
def amp_jobs(fdata,ndata,job, dstype, dslist, amp_pes, HL, E_conv, f_conv, Lgraph, ncore, n_mol, Ltwinx):
outf = "runamp_" + job + ".dat" # to write input info
if ndata:
if len(ndata) == 1:
ini = 0
fin = ndata[0]
elif len(ndata) == 2:
ini = ndata[0]
fin = ndata[1]
ntotal = fin - ini
st = '%d:%d' % (ini, fin)
else:
st = ':'
ntotal = len(fdata)
#print(st)
total_images = ase.io.read(fdata, index=st) # can read extxyz, OUTCAR,
ntotal = len(total_images)
print(f"dstype {dstype} dslist {dslist}")
tr_images, te_images = data_selection(total_images, dstype, dslist, job)
ntrain = len(tr_images)
ntest = len(te_images)
if re.search("pr", job):
y=[]
for mol in total_images:
y.append(mol.get_potential_energy())
if fdata.endswith('extxyz'):
mplot_nvector([],y,fdata.split(".")[0],'sample','E(eV)')
elif fdata == "OUTCAR":
mplot_nvector([],y,Xtitle='sample',Ytitle='E(eV)')
### JOB == TRAINING
elif re.search("tr",job):
print("data training:total sets %d/%d" % (ntrain, ntotal))
f_write(outf, HL, E_conv, f_conv, ntotal, dstype, dslist)
calc_train_images(tr_images, HL, E_conv, f_conv, ncore)
### test after training:: Do not turn on in qsub
server = socket.gethostname()
if server == 'chi' or server == 'login':
outf = "runamp_te.dat" # to write input info
title, suptitle = get_title(fdata, HL, E_conv,f_conv, ntrain, ntest)
print("data test:train sets %d/%d" % (ntest, ntrain))
rmserr, max_res = calc_test_images(te_images, amp_pes, title, suptitle,Lgraph,ncore,Ltwinx=Ltwinx)
f_write(outf, HL, E_conv, f_conv, ntotal,dstype, dslist, err=rmserr, max_res=max_res)
### JOB == TEST
elif re.search("te",job):
title, suptitle = get_title( fdata, HL, E_conv, f_conv, ntrain, ntest)
print("data test:train sets %d/%d" % (ntest, ntrain))
rmserr, max_res = calc_test_images(te_images, amp_pes, title, suptitle,Lgraph,ncore,nmol=n_mol,Ltwinx=Ltwinx)
f_write(outf, HL, E_conv, f_conv, ntotal,dstype, dslist, err=rmserr, max_res=max_res)
elif re.search('md',job):
# use first geometry
atoms = ase.io.read(fdata, index='0')
run_md(atoms)
return
def amp_train(ntotal, images, amp_pes, HL, E_conv, f_conv, Lgraph, ncore,
n_mol, Ltwinx):
#total_images = ase.io.read(fdata, index=':') # can read extxyz, OUTCAR,
#images_sets = Images(total_images, nsets=data_int)
print("data training:total sets %d/%d" % (len(images), ntotal))
calc_train_images(images, HL, E_conv, f_conv, ncore)
### test after training:: Do not turn on in qsub
title, suptitle = get_title(fdata, HL, E_conv, f_conv, len(total_images),
len(images))
print("data test:total sets %d/%d" % (len(images), len(total_images)))
rmserr, max_res = calc_test_images(job,
images,
amp_pes,
title,
suptitle,
Lgraph,
ncore,
Ltwinx=Ltwinx)
f_write('CONTCAR', HL, E_conv, f_conv, rmserr, max_res, job)
return
def amp_jobs(fdata, job, data_int, amp_pes, HL, E_conv, f_conv, Lgraph, ncore,
n_mol, Ltwinx):
total_images = ase.io.read(fdata, index=':') # can read extxyz, OUTCAR,
images_sets = Images(total_images, nsets=data_int)
#if not os.path.isfile(amp_pes):
if re.search("pr", job):
y = []
for mol in total_images:
y.append(mol.get_potential_energy())
if fdata.endswith('extxyz'):
mplot_nvector([], y, fdata.split(".")[0], 'sample', 'E(eV)')
elif fdata == "OUTCAR":
mplot_nvector([], y, Xtitle='sample', Ytitle='E(eV)')
### JOB == TRAINING
elif re.search("tr", job):
if isinstance(data_int, int):
images = images_sets.get_training_images()
else:
d_list = data_int[:2]
images = images_sets.get_training_images(d_list=d_list)
print("data training:total sets %d/%d" %
(len(images), len(total_images)))
calc_train_images(images, HL, E_conv, f_conv, ncore)
### test after training:: Do not turn on in qsub
server = socket.gethostname()
if server == 'chi':
if isinstance(data_int, int):
images = images_sets.get_test_images()
else:
if len(data_int) >= 3:
if len(data_int) == 3:
d_list = data_int[1:]
else:
d_list = data_int[2:]
images = images_sets.get_test_images(d_list=d_list)
else:
print("There is no test set region in -di ")
title, suptitle = get_title(fdata, HL, E_conv, f_conv,
len(total_images), len(images))
print("data test:total sets %d/%d" %
(len(images), len(total_images)))
rmserr, max_res = calc_test_images(job,
images,
amp_pes,
title,
suptitle,
Lgraph,
ncore,
Ltwinx=Ltwinx)
f_write(fdata, HL, E_conv, f_conv, rmserr, max_res, job)
### JOB == TEST
elif re.search("te", job):
if isinstance(data_int, int):
if data_int == 0:
images = total_images
else:
images = images_sets.get_test_images()
### for data interval
else:
images = images_sets.get_test_images(d_list=data_int)
title, suptitle = get_title(fdata, HL, E_conv, f_conv,
len(total_images), len(images))
print("data test:total sets %d/%d" % (len(images), len(total_images)))
rmserr, max_res = calc_test_images(job,
images,
amp_pes,
title,
suptitle,
Lgraph,
ncore,
nmol=n_mol,
Ltwinx=Ltwinx)
f_write(fdata, HL, E_conv, f_conv, rmserr, max_res, job)
elif re.search('md', job):
# use first geometry
atoms = ase.io.read(fdata, index='0')
run_md(atoms)
return