def opt(struct, calc):
opt = GaussianOptimizer(struct, calc)
print("\nBeginning optimization...")
opt.run(opt=['tight','maxcycles=100'])
print("Optimization completed!")
return struct
def test_optimizer(atoms):
pos = atoms.positions.copy()
atoms.calc = get_calc(label='opt', scf='qc')
opt_gauss = GaussianOptimizer(atoms)
opt_gauss.run(fmax='tight')
e_gaussopt = read('opt.log', index=-1).get_potential_energy()
atoms.positions[:] = pos
atoms.calc.set_label('sp')
opt_ase = LBFGS(atoms, trajectory='ase_opt.traj')
opt_ase.run(fmax=1e-2)
e_aseopt = atoms.get_potential_energy()
assert e_gaussopt - e_aseopt == pytest.approx(0., abs=1e-3)
def test_irc(atoms):
calc_ts = get_calc(label='ts', chk='ts.chk')
ts = GaussianOptimizer(atoms, calc_ts)
ts.run(opt='calcall,ts,noeigentest')
tspos = atoms.positions.copy()
atoms.calc = get_calc(label='sp', chk='sp.chk', freq='')
e_ts = atoms.get_potential_energy()
calc_irc_for = get_calc(label='irc_for', oldchk='sp', chk='irc_for.chk')
irc_for = GaussianIRC(atoms, calc_irc_for)
irc_for.run(direction='forward', irc='rcfc')
e_for = read('irc_for.log', index=-1).get_potential_energy()
atoms.positions[:] = tspos
calc_irc_rev = get_calc(label='irc_rev', oldchk='sp', chk='irc_rev.chk')
irc_rev = GaussianIRC(atoms, calc_irc_rev)
irc_rev.run(direction='reverse', irc='rcfc')
e_rev = read('irc_rev.log', index=-1).get_potential_energy()
assert e_ts - e_for == pytest.approx(1.282, abs=1e-3)
assert e_ts - e_rev == pytest.approx(0.201, abs=1e-3)