def test_multiple_paramaters(self):
"""
This test makes sure that input files with multi-parameter keywords
and route cards with multiple lines can be parsed accurately.
"""
filepath = os.path.join(test_dir, "l-cysteine.inp")
route = {
"test": None,
"integral": {
"grid": "UltraFine"
},
"opt": {
"Z-Matrix": None,
"maxcycles": "80",
"tight": None
},
}
gin = GaussianInput.from_file(filepath)
self.assertEqual(gin.dieze_tag, "#n")
self.assertEqual(gin.functional, "B3LYP")
self.assertEqual(gin.basis_set, "6-31+G**")
self.assertEqual(gin.route_parameters, route)
self.assertEqual(gin.title, "L-cysteine neutral")
self.assertEqual(gin.charge, 0)
self.assertEqual(gin.spin_multiplicity, 1)
def test_from_file(self):
filepath = os.path.join(test_dir, "MethylPyrrolidine_drawn.gjf")
gau = GaussianInput.from_file(filepath)
self.assertEqual(gau.molecule.composition.formula, "H11 C5 N1")
self.assertIn("opt", gau.route_parameters)
self.assertEqual(gau.route_parameters["geom"], "connectivity")
self.assertEqual(gau.functional, "b3lyp")
self.assertEqual(gau.basis_set, "6-311+g(d,p)")
filepath = os.path.join(test_dir, "g305_hb.txt")
with open(filepath) as f:
txt = f.read()
toks = txt.split("--link1--")
for i, t in enumerate(toks):
lines = t.strip().split("\n")
lines = [l.strip() for l in lines]
gau = GaussianInput.from_string("\n".join(lines))
self.assertIsNotNone(gau.molecule)
if i == 0:
mol = gau.molecule
ans = """Full Formula (H4 O2)
Reduced Formula: H2O
Charge = 0, Spin Mult = 1
Sites (6)
0 O 0.000000 0.000000 0.000000
1 O 0.000000 0.000000 2.912902
2 H 0.892596 0.000000 -0.373266
3 H 0.143970 0.000219 0.964351
4 H -0.582554 0.765401 3.042783
5 H -0.580711 -0.766761 3.043012"""
self.assertEqual(str(mol), ans)
def test_from_file(self):
filepath = os.path.join(test_dir, 'MethylPyrrolidine_drawn.gjf')
gau = GaussianInput.from_file(filepath)
self.assertEqual(gau.molecule.composition.formula, "H11 C5 N1")
self.assertIn("opt", gau.route_parameters)
self.assertEqual(gau.route_parameters["geom"], "connectivity")
self.assertEqual(gau.functional, "b3lyp")
self.assertEqual(gau.basis_set, "6-311+g(d,p)")
filepath = os.path.join(test_dir, "g305_hb.txt")
with open(filepath) as f:
txt = f.read()
toks = txt.split("--link1--")
for i, t in enumerate(toks):
lines = t.strip().split("\n")
lines = [l.strip() for l in lines]
gau = GaussianInput.from_string("\n".join(lines))
self.assertIsNotNone(gau.molecule)
if i == 0:
mol = gau.molecule
ans = """Full Formula (H4 O2)
Reduced Formula: H2O
Charge = 0, Spin Mult = 1
Sites (6)
0 O 0.000000 0.000000 0.000000
1 O 0.000000 0.000000 2.912902
2 H 0.892596 0.000000 -0.373266
3 H 0.143970 0.000219 0.964351
4 H -0.582554 0.765401 3.042783
5 H -0.580711 -0.766761 3.043012"""
self.assertEqual(str(mol), ans)
def read_mol(filename):
"""
Reads a molecule based on file extension. For example, anything ending in
a "xyz" is assumed to be a XYZ file. Supported formats include xyz,
gaussian input (gjf|g03|g09|com|inp), Gaussian output (.out|and
pymatgen's JSON serialized molecules. Using openbabel,
many more extensions are supported but requires openbabel to be installed.
Args:
filename (str): A filename to read from.
Returns:
A Molecule object.
"""
fname = os.path.basename(filename)
if fnmatch(fname.lower(), "*.xyz*"):
return XYZ.from_file(filename).molecule
elif any([fnmatch(fname.lower(), "*.{}*".format(r)) for r in ["gjf", "g03", "g09", "com", "inp"]]):
return GaussianInput.from_file(filename).molecule
elif any([fnmatch(fname.lower(), "*.{}*".format(r)) for r in ["out", "lis", "log"]]):
return GaussianOutput(filename).final_structure
elif fnmatch(fname, "*.json*") or fnmatch(fname, "*.mson*"):
with zopen(filename) as f:
s = json.load(f, cls=MontyDecoder)
if type(s) != Molecule:
raise IOError("File does not contain a valid serialized " "molecule")
return s
else:
m = re.search("\.(pdb|mol|mdl|sdf|sd|ml2|sy2|mol2|cml|mrv)", filename.lower())
if m:
return BabelMolAdaptor.from_file(filename, m.group(1)).pymatgen_mol
raise ValueError("Unrecognized file extension!")
def better_guess(cls,
gaussian_cmd,
input_file,
output_file,
stderr_file='stderr.txt',
backup=True,
cart_coords=True):
orig_input = GaussianInput.from_file(input_file)
yield (GaussianJob(gaussian_cmd=gaussian_cmd,
input_file=input_file,
output_file=output_file,
stderr_file=stderr_file,
suffix='.guess1',
backup=backup))
if GaussianErrorHandler.activate_better_guess:
# TODO: check why it comes here only if the lower job is not
# failing and not in the else condition
# continue only if other corrections are invalid or failed
lower_output = GaussianOutput(output_file)
if len(lower_output.errors) == 0:
# if the calculation at the lower level of theory succeeded
if not filter(os.listdir('.'), '*.[Cc][Hh][Kk]'):
raise FileNotFoundError(
'Missing checkpoint file. Required '
'to read initial guesses')
gin = GaussianInput(
mol=None,
charge=orig_input.charge,
spin_multiplicity=orig_input.spin_multiplicity,
title=orig_input.title,
functional=orig_input.functional,
basis_set=orig_input.basis_set,
route_parameters=lower_output.route_parameters,
input_parameters=orig_input.input_parameters,
link0_parameters=orig_input.link0_parameters,
dieze_tag=orig_input.dieze_tag,
gen_basis=orig_input.gen_basis)
gin.route_parameters['Guess'] = 'Read'
gin.route_parameters['Geom'] = 'Checkpoint'
gin.write_file(input_file, cart_coords=cart_coords)
yield (GaussianJob(gaussian_cmd=gaussian_cmd,
input_file=input_file,
output_file=output_file,
stderr_file=stderr_file,
suffix='.guess2',
backup=backup))
else:
logger.info('Failed to generate a better initial guess')
else:
logger.info('Calculation completed normally without having to '
'generate a better initial guess')
def test_multiple_paramaters(self):
"""
This test makes sure that input files with multi-parameter keywords
and route cards with multiple lines can be parsed accurately.
"""
filepath = os.path.join(test_dir, "l-cysteine.inp")
route = {"test": None, "integral": {"grid": "UltraFine"},
"opt": {"Z-Matrix": None, "maxcycles": "80", "tight": None}}
gin = GaussianInput.from_file(filepath)
self.assertEqual(gin.dieze_tag, "#n")
self.assertEqual(gin.functional, "B3LYP")
self.assertEqual(gin.basis_set, "6-31+G**")
self.assertEqual(gin.route_parameters, route)
self.assertEqual(gin.title, "L-cysteine neutral")
self.assertEqual(gin.charge, 0)
self.assertEqual(gin.spin_multiplicity, 1)
def read_mol(filename):
"""
Reads a molecule based on file extension. For example, anything ending in
a "xyz" is assumed to be a XYZ file. Supported formats include xyz,
gaussian input (gjf|g03|g09|com|inp), Gaussian output (.out|and
pymatgen's JSON serialized molecules. Using openbabel,
many more extensions are supported but requires openbabel to be installed.
Args:
filename (str): A filename to read from.
Returns:
A Molecule object.
"""
fname = os.path.basename(filename)
if fnmatch(fname.lower(), "*.xyz*"):
return XYZ.from_file(filename).molecule
elif any([
fnmatch(fname.lower(), "*.{}*".format(r))
for r in ["gjf", "g03", "g09", "com", "inp"]
]):
return GaussianInput.from_file(filename).molecule
elif any([
fnmatch(fname.lower(), "*.{}*".format(r))
for r in ["out", "lis", "log"]
]):
return GaussianOutput(filename).final_structure
elif fnmatch(fname, "*.json*") or fnmatch(fname, "*.mson*"):
with zopen(filename) as f:
s = json.load(f, cls=MontyDecoder)
if type(s) != Molecule:
raise IOError("File does not contain a valid serialized "
"molecule")
return s
else:
m = re.search("\.(pdb|mol|mdl|sdf|sd|ml2|sy2|mol2|cml|mrv)",
filename.lower())
if m:
return BabelMolAdaptor.from_file(filename, m.group(1)).pymatgen_mol
raise ValueError("Unrecognized file extension!")
def calculate_properties(rin, directory):
with cd(directory):
# found that the relax often crashed or didn't finish, so will restart
# the calculation in these cases
try:
rout = GaussianOutput('relax.log')
if not rout.properly_terminated:
logging.info('restarting {}'.format(directory))
route = rout.route
route['integral'] = '(acc2e=12)'
rout.to_input('relax.com',
cart_coords=True,
route_parameters=route)
os.system('g09 < relax.com > relax.log')
except IOError:
# relaxation hasn't been run yet
logging.info('started processing {}'.format(directory))
rin.write_file('relax.com', cart_coords=True)
os.system('g09 < relax.com > relax.log')
except IndexError, AttributeError:
# the relax calculation used the wrong header, fix this and restart
rin = GaussianInput.from_file('relax.com')
rin.route_parameters['integral'] = '(acc2e=12)'
rin.write_file('relax.com', cart_coords=True)
os.system('g09 < relax.com > relax.log')
rout = GaussianOutput('relax.log')
if not rout.properly_terminated:
logging.error(
'{} relaxation did not terminate correctly'.format(directory))
return 0
# do the TD-DFT calculation
tdin = GaussianInput(rout.final_structure,
charge=0,
title=rin.title,
functional=functional,
spin_multiplicity=1,
basis_set=basis_set,
dieze_tag=dieze_tag,
link0_parameters=link0,
route_parameters=td_params)
td_done = False
if os.path.isfile('td.log'):
tdout = GaussianOutput('td.log')
if tdout.properly_terminated:
td_done = True
if not td_done:
tdin.write_file('td.com', cart_coords=True)
os.system('g09 < td.com > td.log')
tdout = GaussianOutput('td.log')
if not tdout.properly_terminated:
logging.error(
'{} TD-DFT did not terminate correctly'.format(directory))
return 0
# do the TD-DFT calculation w. Tamm-Dancoff approx
tdain = GaussianInput(rout.final_structure,
charge=0,
title=rin.title,
spin_multiplicity=1,
functional=functional,
basis_set=basis_set,
dieze_tag=dieze_tag,
link0_parameters=link0,
route_parameters=tda_params)
tda_done = False
if os.path.isfile('tda.log'):
tdaout = GaussianOutput('tda.log')
if tdaout.properly_terminated:
tda_done = True
if not tda_done:
tdain.write_file('tda.com', cart_coords=True)
os.system('g09 < tda.com > tda.log')
tdaout = GaussianOutput('tda.log')
if not tdaout.properly_terminated:
logging.error(
'{} TDA-DFT did not terminate correctly'.format(directory))
return 0
# add the dummy atoms for the NICS(1)_zz calculations
mol_nics = rout.final_structure
mol_nics = add_dummy_atoms(mol_nics, six_mem_a)
mol_nics = add_dummy_atoms(mol_nics, six_mem_b)
mol_nics = add_dummy_atoms(mol_nics, five_mem_a)
mol_nics = add_dummy_atoms(mol_nics, five_mem_b)
# run NICS on the ground state and triplet state
nicssin = GaussianInput(mol_nics,
charge=0,
title=rin.title,
spin_multiplicity=1,
functional=functional,
basis_set=basis_set,
dieze_tag=dieze_tag,
link0_parameters=link0,
route_parameters=nmr_params)
nicssin.write_file('nics_singlet.com', cart_coords=True)
# work around as pymatgen does not allow Bq as an element
os.system("sed -i 's/X-Bq0+/Bq/g' nics_singlet.com")
os.system('g09 < nics_singlet.com > nics_singlet.log')
# can't have NICS job completion check due to above mentioned pmg bug
nicstin = GaussianInput(mol_nics,
charge=0,
title=rin.title,
spin_multiplicity=3,
functional=functional,
basis_set=basis_set,
dieze_tag=dieze_tag,
link0_parameters=link0,
route_parameters=nmr_params)
nicstin.write_file('nics_triplet.com', cart_coords=True)
os.system("sed -i 's/X-Bq0+/Bq/g' nics_triplet.com")
os.system('g09 < nics_triplet.com > nics_triplet.log')
'title': ginp.title, 'index': index})
############################################
# generate substituted pyridine structures #
############################################
# need to consider that nx positions 2 and 3 are incompatible
# with substitutional sites z and y, respectively and ny
# position 3 is incompatible with substitional site x
#
# the other important note is that we need to substitute the atoms from the
# highest site_id downwards, otherwise the order and index of the atoms will
# change
gin = GaussianInput.from_file(os.path.join('..', 'templates',
'input-template.com'))
gin.route_parameters['integral'] = '(acc2e=12)' # hack to get round pmg bug
print "generating pyridine substituted structures..."
for nx, ny, x_sub, y_sub, z_sub in itertools.product(nx_sites, ny_sites,
subs, subs, subs):
mol = copy.deepcopy(gin)
# if there are any clashes then skip the structure
if ((x_sub[0] and ny[0] == 3) or (y_sub[0] and nx[0] == 3) or
(z_sub[0] and nx[0] == 2)):
continue
# make a list of the substitutions and sort from highest to lowest site_id
# this list includes the H atoms next to the N we introduce in to the rings
sub_list = []