def opt(T=350):
A = read('packed.gen')
# A = press_mol(A)
write_gulp_in(A, runword='opti conv qiterative ', T=T, lib='reax')
print('\n- running gulp optimize ...')
system('gulp<inp-gulp>gulp.out')
xyztotraj('his.xyz', mode='w')
def nvt_wt(T=350,
time_step=0.1,
tot_step=100,
gen='packed.gen',
mode='w',
wt=25):
A = read(gen, index=-1)
write_gulp_in(A,
runword='md conv qiterative ',
T=T,
time_step=time_step,
tot_step=tot_step,
lib='reax')
print('\n- running gulp nvt ...')
system('nohup gulp<inp-gulp>gulp.out 2>&1 &')
thread = popen('ps -aux | grep "gulp"')
lines = thread.readlines()
time.sleep(wt)
for line in lines:
l = line.split()
if l[7] == 'R+' or l[7] == 'R':
system('kill %s' % l[1])
xyztotraj('his.xyz', mode=mode)
def npt(T=350, time_step=0.1, tot_step=10.0):
A = read('packed.gen')
write_gulp_in(A,
runword='md conp',
T=T,
time_step=time_step,
tot_step=tot_step,
lib='reax')
system('gulp<inp-gulp>gulp.out')
xyztotraj('his.xyz')
def nvt(T=350,time_step=0.1,tot_step=5000,gen='siesta.traj',index=-1,mode='w'):
''' a gulp MD run '''
A = read(gen,index=index)
write_gulp_in(A,runword='md qiterative conv',
T=T,
time_step=time_step,
tot_step=tot_step,
lib='reax')
print('\n- running gulp nvt ...')
system('gulp<inp-gulp>gulp.out')
xyztotraj('his.xyz',mode=mode)
def nvt(T=350, time_step=0.1, tot_step=100, gen='packed.gen', mode='w'):
A = read(gen, index=-1)
write_gulp_in(A,
runword='md qiterative conv',
T=T,
time_step=time_step,
tot_step=tot_step,
lib='reax')
print('\n- running gulp nvt ...')
system('gulp<inp-gulp>gulp.out')
xyztotraj('his.xyz', mode=mode)
def debug_h(direcs={'ch4': '/home/feng/siesta/train/ch4'},
gulp_cmd='/home/feng/gulp/Src/gulp<inp-gulp >gulp.out'):
for key in direcs:
mol = key
rn = ReaxFF(libfile='ffield',
direcs=direcs,
dft='siesta',
rc_scale='none',
optword='all',
batch_size=1,
sort=False,
pkl=True,
interactive=True)
molecules = rn.initialize()
rn.session(learning_rate=1.0 - 4, method='AdamOptimizer')
hblab = rn.lk.hblab
bdlab = rn.lk.bdlab
rbd = rn.get_value(rn.rbd)
rhb = rn.get_value(rn.rhb)
rik = rn.lk.rik
rij = rn.lk.rij
fhb = rn.get_value(rn.fhb)
frhb = rn.get_value(rn.frhb)
bohb = rn.get_value(rn.BOhb)
exphb1 = rn.get_value(rn.exphb1)
exphb2 = rn.get_value(rn.exphb2)
sin4 = rn.get_value(rn.sin4)
hbthe = rn.get_value(rn.hbthe)
ehb = rn.get_value(rn.EHB)
cell = rn.cell[mol]
A = Atoms(symbols=molecules[mol].atom_name,
positions=molecules[mol].x[0],
cell=cell,
pbc=(1, 1, 1))
write_gulp_in(A, runword='gradient nosymmetry conv qite verb')
# system('/home/feng/gulp/gulp-5.0/Src/gulp<inp-gulp >gulp.out')
system(gulp_cmd)
# system('/home/gfeng/gulp/gulp-5.0/Src/gulp<inp-gulp >gulp.out')
atom_name = molecules[mol].atom_name
fg = open('gulp.out', 'r')
for line in fg.readlines():
if line.find('- ehb:') >= 0:
l = line.split()
i = int(l[2]) - 1
j = int(l[3]) - 1
k = int(l[4]) - 1
hb = atom_name[i] + '-' + atom_name[j] + '-' + atom_name[k]
hbk = [mol, i, j, k]
bd = atom_name[i] + '-' + atom_name[j]
if not bd in rn.bonds:
bd = atom_name[j] + '-' + atom_name[i]
bdk = [mol, i, j]
bdkr = [mol, j, i]
if bdk in bdlab[bd]:
nbd = bdlab[bd].index(bdk)
elif bdkr in bdlab[bd]:
nbd = bdlab[bd].index(bdkr)
find = False
if hbk in hblab[hb]:
# nb = hblab[hb].index(hbk)
# print('------------------------------------')
nbs = []
for nb, bb in enumerate(hblab[hb]):
if bb == hbk:
nbs.append(nb)
find = True
if find:
ib = 0
for nb in nbs:
if abs(rhb[hb][nb][0] - float(l[6])) < 0.00001:
print('- ReaxFF %d %s:' % (ib, hb),
'rbd: %10.6f' % rbd[bd][nbd][0],
'rhb: %10.6f' % rhb[hb][nb][0],
'exphb1: %10.6f' % exphb1[hb][nb][0],
'bohb: %10.6f' % bohb[hb][nb][0],
'exphb2: %10.6f' % exphb2[hb][nb][0],
'sin4: %10.6f' % sin4[hb][nb][0],
'hbthe: %10.6f' % hbthe[hb][nb][0],
'ehb: %10.6f' % ehb[hb][nb][0],
'rik: %10.6f' % rik[hb][nb][0])
print('- GULP %d %s:' % (ib, hb),
'rbd: %10.6f' % float(l[5]),
'rhb: %10.6f' % float(l[6]),
'exphb1: %10.6f' % float(l[9]),
'bohb: %10.6f' % float(l[13]),
'exphb2: %10.6f' % float(l[10]),
'sin4: %10.6f' % float(l[11]),
'hbthe: %10.6f' % float(l[15]),
'ehb: %10.6f' % float(l[12]),
'rik: %10.6f' % float(l[14]))
ib += 1
else:
print('N.F-GULP %s:' % hb, 'rbd: %10.6f' % float(l[5]),
'rhb: %10.6f' % float(l[6]), 'fhb: %10.6f' % float(l[7]),
'exphb1: %10.6f' % float(l[9]),
'bohb: %10.6f' % float(l[13]),
'exphb2: %10.6f' % float(l[10]),
'sin4: %10.6f' % float(l[11]),
'ehb: %10.6f' % float(l[12]))
fg.close()
def debug_ang(direcs={'cho-4': '/home/feng/siesta/cho4'},
gulp_cmd='/home/feng/gulp/gulp-5.0/Src/gulp<inp-gulp >gulp.out'):
for key in direcs:
mol = key
rn = ReaxFF(libfile='ffield',
direcs=direcs,
dft='siesta',
optword='nocoul',
batch_size=1,
sort=False,
pkl=True,
interactive=True)
molecules = rn.initialize()
rn.session(learning_rate=1.0 - 4, method='AdamOptimizer')
f7 = rn.get_value(rn.f_7)
f8 = rn.get_value(rn.f_8)
expang = rn.get_value(rn.expang)
eang = rn.get_value(rn.EANG)
theta = rn.get_value(rn.theta)
theta0 = rn.get_value(rn.theta0)
SBO = rn.get_value(rn.SBO)
sbo = rn.get_value(rn.sbo)
pbo = rn.get_value(rn.pbo)
rnlp = rn.get_value(rn.rnlp)
D_ang = rn.get_value(rn.D_ang)
Delta = rn.get_value(rn.Delta)
atom_name = molecules[mol].atom_name
p = rn.p_
nang = rn.nang
anglab = rn.lk.anglab
angs = rn.angs
atom_lab = rn.lk.atom_lab
cell = rn.cell[mol]
# print(eang)
# print(nang)
A = Atoms(symbols=molecules[mol].atom_name,
positions=molecules[mol].x[0],
cell=cell,
pbc=(1, 1, 1))
write_gulp_in(A, runword='gradient nosymmetry conv qite verb')
system(gulp_cmd)
fg = open('gulp.out', 'r')
angfind = {}
for atn in rn.angs:
angfind[atn] = []
for line in fg.readlines():
if line.find('- eval:') >= 0:
l = line.split()
i = int(l[2]) - 1
j = int(l[3]) - 1
k = int(l[4]) - 1
atn = atom_name[i] + '-' + atom_name[j] + '-' + atom_name[k]
atnk = [mol, i, j, k]
if not atn in angs:
atn = atom_name[k] + '-' + atom_name[j] + '-' + atom_name[i]
atnkr = [mol, k, j, i]
find = False
if atnk in anglab[atn]:
na = anglab[atn].index(atnk)
find = True
elif atnkr in anglab[atn]:
na = anglab[atn].index(atnkr)
find = True
aj = atom_lab.index([mol, j])
if find:
angfind[atn].append(na)
print('- ReaxFF %s:' % atn, na, atnk, 'sbo:', sbo[atn][na][0],
'theta:', theta[atn][na][0], 'theta0:',
theta0[atn][na][0], 'expang:', expang[atn][na][0],
'eang:', eang[atn][na][0])
print('- GULP %s:' % atn, na, atnk, 'sbo:', l[11], 'theta:',
l[9], 'theta0:', l[10], 'expang:', l[6], 'eang:', l[5])
else:
print('-NF GULP %s:' % atn, na, 'sbo:', l[11], 'theta:', l[9],
'theta0:', l[10], 'expang:', l[6], 'eang:', l[5])
fg.close()
print('\n- angles not find\n')
for atn in rn.angs:
for na in range(rn.nang[atn]):
if not na in angfind[atn]:
al = anglab[atn][na]
print('- ReaxFF %s:' % atn, na, al, 'sbo:', sbo[atn][na][0],
'theta0:', theta0[atn][na][0], 'expang:',
expang[atn][na][0], 'eang:', eang[atn][na][0])
def debug_pen(mol='ch4', direcs={'ch4': '/home/feng/siesta/train/ch4'}):
rn = ReaxFF(libfile='ffield',
direcs=direcs,
dft='siesta',
rc_scale='none',
optword='all',
batch_size=1,
sort=False,
pkl=True,
interactive=True)
molecules = rn.initialize()
rn.session(learning_rate=1.0 - 4, method='AdamOptimizer')
# ebond = rn.get_value(rn.ebond[mol])
bo = rn.get_value(rn.bo)
bo0 = rn.get_value(rn.bo0)
bop = rn.get_value(rn.bop)
rbd = rn.get_value(rn.rbd)
bodiv1 = rn.get_value(rn.bodiv1)
bop_pi = rn.get_value(rn.bop_pi)
f = rn.get_value(rn.F)
f11 = rn.get_value(rn.F_11)
f12 = rn.get_value(rn.F_12)
f45 = rn.get_value(rn.F_45)
f4 = rn.get_value(rn.f_4)
f5 = rn.get_value(rn.f_5)
dboci = rn.get_value(rn.Di_boc)
dbocj = rn.get_value(rn.Dj_boc)
Dp = rn.get_value(rn.Dp)
BOP = rn.get_value(rn.BOP)
# print('- shape of BOP: ',BOP.shape)
nbd = rn.nbd
bdlab = rn.lk.bdlab
bonds = rn.bonds
atom_name = molecules[mol].atom_name
p = rn.p_
nang = rn.nang
anglab = rn.lk.anglab
angs = rn.angs
cell = rn.cell[mol]
fbo = open('bo.txt', 'w')
bd = 'C-C'
# for bd in bonds:
if nbd[bd] > 0:
# print('- shape of new style: ',bo[bd].shape)
print('\n- bd: %s \n' % bd, file=fbo)
for nb in range(nbd[bd]):
print('- ',
bd,
'r:',
rbd[bd][nb][0],
'BOp:',
bop[bd][nb][0],
'F:',
f[bd][nb][0],
'F_11:',
f11[bd][nb][0],
'F_12:',
f12[bd][nb][0],
'F_45:',
f45[bd][nb][0],
'F_4:',
f4[bd][nb][0],
'F_5:',
f5[bd][nb][0],
'Dboci:',
dboci[bd][nb][0],
'Dbocj:',
dbocj[bd][nb][0],
'BO:',
bo0[bd][nb][0],
file=fbo)
fbo.close()
ff = open('f.txt', 'w')
f1 = rn.get_value(rn.f_1)
f2 = rn.get_value(rn.f_2)
f3 = rn.get_value(rn.f_3)
dexpf3 = rn.get_value(rn.dexpf3)
dexpf3t = rn.get_value(rn.dexpf3t)
f3log = rn.get_value(rn.f3log)
# for bd in bonds:
if nbd[bd] > 0:
print('\n- bd: %s \n' % bd, file=ff)
for nb in range(nbd[bd]):
print('- ',
bd,
'r:',
rbd[bd][nb][0],
'f_1:',
f1[bd][nb][0],
'f_2:',
f2[bd][nb][0],
'f_3:',
f3[bd][nb][0],
'dexpf3:',
dexpf3[bd][nb][0],
'dexpf3t:',
dexpf3[bd][nb][0],
'f3log:',
f3log[bd][nb][0],
file=ff)
ff.close()
fa = open('ang.txt', 'w')
eang = rn.get_value(rn.EANG)
expang = rn.get_value(rn.expang)
f7 = rn.get_value(rn.f_7)
f8 = rn.get_value(rn.f_8)
thet = rn.get_value(rn.thet)
for a in angs:
if nang[a] > 0:
print('\n- a: %s \n' % bd, file=fa)
for na in range(nang[a]):
iat = anglab[a][na][1]
jat = anglab[a][na][2]
kat = anglab[a][na][3]
print('- ',
a,
'thet:',
thet[a][na][0],
'f7: ',
f7[a][na][0],
'f8: ',
f8[a][na][0],
'expang: ',
expang[a][na][0],
'eang: ',
eang[a][na][0],
file=fa)
fa.close()
f9 = rn.get_value(rn.f_9)
epen = rn.get_value(rn.EPEN)
fijk = rn.get_value(rn.fijk)
bo = rn.get_value(rn.bo)
A = Atoms(symbols=molecules[mol].atom_name,
positions=molecules[mol].x[0],
cell=cell,
pbc=(1, 1, 1))
write_gulp_in(A, runword='gradient nosymmetry conv qite verb')
system('/home/feng/gulp/gulp-5.0/Src/gulp<inp-gulp >gulp.out')
fg = open('gulp.out', 'r')
for line in fg.readlines():
if line.find('- epen:') >= 0:
l = line.split()
i = int(l[2]) - 1
j = int(l[3]) - 1
k = int(l[4]) - 1
atn = atom_name[i] + '-' + atom_name[j] + '-' + atom_name[k]
atnk = [mol, i, j, k]
if not atn in angs:
atn = atom_name[k] + '-' + atom_name[j] + '-' + atom_name[i]
atnkr = [mol, k, j, i]
find = False
if atnk in anglab[atn]:
na = anglab[atn].index(atnk)
find = True
elif atnkr in anglab[atn]:
na = anglab[atn].index(atnkr)
find = True
if find:
print('- ReaxFF %s:' % atn, 'f9:', f9[atn][na][0], 'epen:',
epen[atn][na][0], fijk[atn][na][0])
print('- GULP %s:' % atn, 'f9:', l[9], 'epen:', l[5], l[6])
else:
bd1 = atom_name[i] + '-' + atom_name[j]
bd1k = [mol, i, j]
if not bd1 in bonds:
bd1 = atom_name[j] + '-' + atom_name[i]
bd1k = [mol, j, i]
bd2 = atom_name[i] + '-' + atom_name[k]
bd2k = [mol, i, k]
if not bd2 in bonds:
bd2 = atom_name[k] + '-' + atom_name[i]
bd2k = [mol, k, i]
if bd1k in bdlab[bd1] and bd2k in bdlab[bd2]:
nb1 = bdlab[bd1].index(bd1k)
nb2 = bdlab[bd2].index(bd2k)
boij = bo[bd1][nb1][0]
boik = bo[bd2][nb2][0]
print('- GULP %s:' % atn, l[9], l[5], l[6], 'bo:', bd1,
boij, l[10], bd2, boik, l[11])
else:
print('- GULP %s:' % atn, l[9], l[5], l[6], 'bo:', bd1,
'none', l[10], bd2, 'none', l[11])
fg.close()
def debug_be(direcs={'ch4': '/home/feng/siesta/train/ch4'}):
for key in direcs:
mol = key
rn = ReaxFF(libfile='ffield',
direcs=direcs,
dft='siesta',
rc_scale='none',
clip_op=False,
optword='all',
batch_size=1,
sort=False,
pkl=True,
interactive=True)
molecules = rn.initialize()
rn.session(learning_rate=1.0 - 4, method='AdamOptimizer')
bdlab = rn.lk.bdlab
rbd = rn.get_value(rn.rbd)
ebond = rn.get_value(rn.EBD)
sieng = rn.get_value(rn.sieng)
pieng = rn.get_value(rn.pieng)
ppeng = rn.get_value(rn.ppeng)
bosi = rn.get_value(rn.bosi)
bopi = rn.get_value(rn.bopi)
bopp = rn.get_value(rn.bopp)
cell = rn.cell[mol]
A = Atoms(symbols=molecules[mol].atom_name,
positions=molecules[mol].x[0],
cell=cell,
pbc=(1, 1, 1))
write_gulp_in(A, runword='gradient nosymmetry conv qite verb')
system('/home/feng/gulp/Src/gulp<inp-gulp >gulp.out')
atom_name = molecules[mol].atom_name
fg = open('gulp.out', 'r')
for line in fg.readlines():
if line.find('- ebond:') >= 0:
l = line.split()
i = int(l[2]) - 1
j = int(l[3]) - 1
bn = atom_name[i] + '-' + atom_name[j]
if not bn in rn.bonds:
bn = atom_name[j] + '-' + atom_name[i]
bnk = [mol, i, j]
bnkr = [mol, j, i]
find = False
if bnk in bdlab[bn]:
nb = bdlab[bn].index(bnk)
find = True
elif bnkr in bdlab[bn]:
nb = bdlab[bn].index(bnkr)
find = True
if find:
if abs(ebond[bn][nb][0] - float(l[4])) > 0.001:
print('- ReaxFF %s:' % bn, 'rbd:', rbd[bn][nb][0],
'ebond:', ebond[bn][nb][0], 'sieng:',
sieng[bn][nb][0], 'pieng:', pieng[bn][nb][0],
'ppeng:', ppeng[bn][nb][0], 'bosi:', bosi[bn][nb][0],
'bopi:', bopi[bn][nb][0], 'bopp:', bopp[bn][nb][0])
print('- GULP %s:' % bn, 'rbd:', l[4], 'ebond:', l[5],
'sieng:', l[6], 'pieng:', l[7], 'ppeng:', l[8],
'bosi:', l[9], 'bopi:', l[10], 'bopp:', l[11])
fg.close()
def debug_eu(direcs={'ch4': '/home/feng/siesta/train/ch4'},
gulp_cmd='/home/feng/gulp/gulp-5.0/Src/gulp<inp-gulp >gulp.out'):
for key in direcs:
mol = key
rn = ReaxFF(libfile='ffield',
direcs=direcs,
dft='siesta',
rc_scale='none',
clip_op=False,
optword='all',
batch_size=1,
sort=False,
pkl=True,
interactive=True)
molecules = rn.initialize()
rn.session(learning_rate=1.0 - 4, method='AdamOptimizer')
atlab = rn.lk.atlab
rbd = rn.get_value(rn.rbd)
eunder = rn.get_value(rn.EUN)
D_lp = rn.get_value(rn.Delta_lpcorr)
Delta_lp = rn.get_value(rn.Delta_lp)
DPI = rn.get_value(rn.Dpi)
eu1 = rn.get_value(rn.eu1)
eu2 = rn.get_value(rn.eu2)
expeu2 = rn.get_value(rn.expeu2)
p = rn.get_value(rn.p)
cell = rn.cell[mol]
atom_name = molecules[mol].atom_name
A = Atoms(symbols=molecules[mol].atom_name,
positions=molecules[mol].x[0],
cell=cell,
pbc=(1, 1, 1))
write_gulp_in(A, runword='gradient nosymmetry conv qite verb')
system(gulp_cmd)
fg = open('gulp.out', 'r')
for line in fg.readlines():
if line.find('- eunder:') >= 0:
l = line.split()
i = int(l[2]) - 1
an = atom_name[i]
ank = [mol, i]
find = False
if ank in atlab[an]:
na = atlab[an].index(ank)
find = True
if find:
# if abs(ebond[bn][nb][0]-float(l[4]))>0.001:
print('- ReaxFF %d %s:' % (na, an),
'eunder: %10.6f' % eunder[an][na][0],
'eu1: %10.6f' % eu1[an][na][0],
'expeu2: %10.6f' % expeu2[an][na][0],
'ovun2: %10.6f' % p['ovun2_' + an],
'Delta_lp: %10.6f' % Delta_lp[an][na][0],
'DPI: %10.6f' % DPI[an][na][0],
'Delta_lpcorr: %10.6f' % D_lp[an][na][0])
print('- GULP %d %s:' % (na, an),
'eunder: %10.6f' % float(l[3]),
'eu1: %10.6f' % float(l[5]),
'expeu2: %10.6f' % float(l[7]),
'ovun2: %10.6f' % float(l[8]),
'Delta_lp: %10.6f' % float(l[10]),
'DPI: %10.6f' % float(l[11]),
'Delta_lpcorr: %10.6f' % float(l[9]))
fg.close()
def debug_bo(direcs={'ch4': '/home/feng/siesta/train/ch4'},
gulp_cmd='/home/feng/gulp/gulp-5.0/Src/gulp<inp-gulp >gulp.out'):
for key in direcs:
mol = key
rn = ReaxFF(libfile='ffield',
direcs=direcs,
dft='siesta',
optword='nocoul',
batch_size=1,
sort=False,
pkl=True,
interactive=True)
molecules = rn.initialize()
rn.session(learning_rate=1.0 - 4, method='AdamOptimizer')
bdlab = rn.lk.bdlab
bosi = rn.get_value(rn.bosi)
bo = rn.get_value(rn.bo)
bo0 = rn.get_value(rn.bo0)
bop = rn.get_value(rn.bop)
rbd = rn.get_value(rn.rbd)
bop_si = rn.get_value(rn.bop_si)
bop_pi = rn.get_value(rn.bop_pi)
bop_pp = rn.get_value(rn.bop_pp)
f = rn.get_value(rn.F)
f11 = rn.get_value(rn.F_11)
f12 = rn.get_value(rn.F_12)
f45 = rn.get_value(rn.F_45)
f4 = rn.get_value(rn.f_4)
f5 = rn.get_value(rn.f_5)
ebond = rn.get_value(rn.ebond)
cell = rn.cell[mol]
A = Atoms(symbols=molecules[mol].atom_name,
positions=molecules[mol].x[0],
cell=cell,
pbc=(1, 1, 1))
write_gulp_in(A, runword='gradient nosymmetry conv qite verb')
system(gulp_cmd)
atom_name = molecules[mol].atom_name
fg = open('gulp.out', 'r')
for line in fg.readlines():
if line.find('- bosi:') >= 0:
l = line.split()
i = int(l[2]) - 1
j = int(l[3]) - 1
bn = atom_name[i] + '-' + atom_name[j]
if not bn in rn.bonds:
bn = atom_name[j] + '-' + atom_name[i]
bnk = [mol, i, j]
bnkr = [mol, j, i]
find = False
if bnk in bdlab[bn]:
nb = bdlab[bn].index(bnk)
find = True
elif bnkr in bdlab[bn]:
nb = bdlab[bn].index(bnkr)
find = True
if find:
# if abs(rbd[bn][nb][0]-float(l[4]))>0.0001:
print('- ReaxFF %s:' % bn, 'rbd:', rbd[bn][nb][0], 'bop_si:',
bop_si[bn][nb][0], 'bop_pi:', bop_pi[bn][nb][0],
'bop_pp:', bop_pp[bn][nb][0])
print('- GULP %s:' % bn, 'rbd:', l[4], 'bop_si:', l[5],
'bop_pi:', l[6], 'bop_pp:', l[7])
else:
print('- GULP %s:' % bn, 'rbd:', l[4], 'bop_si:', l[5],
'bop_pi:', l[6], 'bop_pp:', l[7])
fg.close()
e_,eb_,el_,eo_,eu_,ea_,ep_,etc_,et_,ef_,ev_,ehb_,ecl_,esl_= \
get_reax_energy(fo='gulp.out')
print('- ebond - IRFF %f GULP %f.' % (ebond[mol], eb_))
def debug_v(direcs={'ch4': '/home/feng/siesta/train/ch4'},
gulp_cmd='/home/feng/gulp/Src/gulp<inp-gulp >gulp.out'):
for key in direcs:
mol = key
rn = ReaxFF(libfile='ffield',
direcs=direcs,
dft='siesta',
optword='nocoul',
batch_size=1,
sort=False,
pkl=True,
interactive=True)
molecules = rn.initialize()
rn.session(learning_rate=1.0 - 4, method='AdamOptimizer')
vlab = rn.lk.vlab
rv = rn.get_value(rn.rv)
expvdw1 = rn.get_value(rn.expvdw1)
expvdw2 = rn.get_value(rn.expvdw2)
evdw = rn.get_value(rn.EVDW)
f13 = rn.get_value(rn.f_13)
cell = rn.cell[mol]
A = Atoms(symbols=molecules[mol].atom_name,
positions=molecules[mol].x[0],
cell=cell,
pbc=(1, 1, 1))
write_gulp_in(A, runword='gradient nosymmetry conv qite verb')
system(gulp_cmd)
atom_name = molecules[mol].atom_name
fg = open('gulp.out', 'r')
for line in fg.readlines():
if line.find('- evdw:') >= 0:
l = line.split()
i = int(l[2]) - 1
j = int(l[3]) - 1
vb = atom_name[i] + '-' + atom_name[j]
vbk = [mol, i, j]
vbkr = [mol, j, i]
if not vb in rn.bonds:
vb = atom_name[j] + '-' + atom_name[i]
find = False
if vbk in vlab[vb]:
# nb = hblab[hb].index(hbk)
# print('------------------------------------')
nbs = []
for nb, bb in enumerate(vlab[vb]):
if bb == vbk:
nbs.append(nb)
find = True
elif vbkr in vlab[vb]:
# nb = hblab[hb].index(hbk)
# print('------------------------------------')
nbs = []
for nb, bb in enumerate(vlab[vb]):
if bb == vbkr:
nbs.append(nb)
find = True
if find:
ib = 0
for nb in nbs:
if abs(rv[vb][nb][0] - float(l[4])) < 0.00001:
# if abs(evdw[vb][nb][0]-float(l[5]))>0.0001:
print('- ReaxFF %d %s:' % (ib, vb),
'rv: %10.6f' % rv[vb][nb][0],
'evdw: %10.6f' % evdw[vb][nb][0])
print('- GULP %d %s:' % (ib, vb),
'rv: %10.6f' % float(l[4]),
'evdw: %10.6f' % float(l[5]))
ib += 1
else:
print('- N.F GULP %s:' % vb, 'rv: %10.6f' % float(l[4]),
'evdw: %10.6f' % float(l[5]))
fg.close()