def read_nwchem(filename):
"""Method to read geometry from a nwchem output
"""
f = filename
if isinstance(filename, str):
f = open(filename)
lines = f.readlines()
i = 0
while i < len(lines):
if lines[i].find('XYZ format geometry') >= 0:
natoms = int(lines[i + 2].split()[0])
string = ''
for j in range(2, natoms + 4):
xyzstring = lines[i + j]
symbol = xyzstring.split()[0].strip()
# replace bq ghost with X: MDTMP can we do better?
if symbol.startswith('bq'):
xyzstring = xyzstring.replace(symbol, 'X')
string += xyzstring
atoms = read_xyz(StringIO(string))
i += natoms + 4
else:
i += 1
if type(filename) == str:
f.close()
return atoms
def read_nwchem(filename):
"""Method to read geometry from a nwchem output
"""
from ase import Atoms, Atom
if isinstance(filename, str):
f = open(filename)
lines = f.readlines()
i = 0
while i < len(lines):
if lines[i].find('XYZ format geometry') >=0:
natoms = int(lines[i + 2].split()[0])
string = ''
for j in range(2, natoms + 4):
xyzstring = lines[i + j]
symbol = xyzstring.split()[0].strip()
# replace bq ghost with X: MDTMP can we do better?
if symbol.startswith('bq'):
xyzstring = xyzstring.replace(symbol, 'X')
string += xyzstring
atoms = read_xyz(StringIO(string))
i += natoms + 4
else:
i += 1
if type(filename) == str:
f.close()
return atoms
def read_nwchem_input(filename):
"""Method to read geometry from an NWChem input file."""
f = filename
if isinstance(filename, str):
f = open(filename)
lines = f.readlines()
# Find geometry region of input file.
stopline = 0
for index, line in enumerate(lines):
if line.startswith('geometry'):
startline = index + 1
stopline = -1
elif (line.startswith('end') and stopline == -1):
stopline = index
# Format and send to read_xyz.
xyz_text = '%i\n' % (stopline - startline)
xyz_text += ' geometry\n'
for line in lines[startline:stopline]:
xyz_text += line
atoms = read_xyz(StringIO(xyz_text))
atoms.set_cell((0., 0., 0.)) # no unit cell defined
if type(filename) == str:
f.close()
return atoms
def read_orca_input(filename):
"""Method to read geometry from an ORCA input file."""
f = filename
if isinstance(filename, str):
f = open(filename)
lines = f.readlines()
# Find geometry region of input file.
done = False
i = 0
xyzstring = ''
for l, line in enumerate(lines):
if line.find('xyz') > -1 and line.find('*') > -1:
while not done:
i += 1
if not (lines[l + i].find('*') > -1):
xyzstring += lines[l + i]
sym = lines[l + i].strip().split()[0]
else:
done = True
if done:
break
xyzstring = str(i - 1) + '\n\n' + xyzstring
atoms = read_xyz(StringIO(xyzstring))
if type(filename) == str:
f.close()
return atoms
def evaluate_molecules(molecules: List[str],
b3lyp_energies: List[float]) -> List[float]:
"""Compute the atomization energy of molecules
Args:
molecules ([str]): XYZ-format molecular structures. Assumed to be
fully-relaxed
b3lyp_energies ([float]): B3LYP total energies of structures
Returns:
([float]): Estimated G4MP2 atomization energies of molecules
"""
# Convert the molecules to atoms objects
atoms = [next(read_xyz(StringIO(x))) for x in molecules]
# Generate the local environment for each atom
conv = AtomsConverter(environment)
inputs = [conv.convert_atoms(atom) for atom in atoms]
# Add the b3lyp_energies to each atom object
for i, e in zip(inputs, b3lyp_energies):
i['u0'] = torch.Tensor(np.expand_dims(e, 0))
# Execute in batches
results = []
for i in inputs:
outputs = model(i)
results.append(np.squeeze(outputs['y'].cpu().data.numpy()))
# Return atomization energy
return [
compute_atomization_energy(a, e, 'g4mp2')
for a, e in zip(atoms, results)
]
def read_nwchem_input(filename):
"""Method to read geometry from an NWChem input file."""
f = filename
if isinstance(filename, str):
f = open(filename)
lines = f.readlines()
# Find geometry region of input file.
stopline = 0
for index, line in enumerate(lines):
if line.startswith('geometry'):
startline = index + 1
stopline = -1
elif (line.startswith('end') and stopline == -1):
stopline = index
# Format and send to read_xyz.
xyz_text = '%i\n' % (stopline - startline)
xyz_text += ' geometry\n'
for line in lines[startline:stopline]:
xyz_text += line
atoms = read_xyz(StringIO(xyz_text))
atoms.set_cell((0., 0., 0.)) # no unit cell defined
if type(filename) == str:
f.close()
return atoms
def read_orca(filename):
"""Method to read geometry from a ORCA output
"""
f = filename
if isinstance(filename, str):
f = open(filename)
lines = f.readlines()
done = False
i = 0
xyzstring = ''
for l, line in enumerate(lines):
if line.find('CARTESIAN COORDINATES (ANGSTROEM)') >= 0:
i += 1
while not done:
i += 1
if not lines[l + i] == '\n':
xyzstring += lines[l + i]
sym = lines[l + i].strip().split()[0]
else:
done = True
if done:
break
xyzstring = str(i - 2) + '\n\n' + xyzstring
atoms = read_xyz(StringIO(xyzstring))
if type(filename) == str:
f.close()
return atoms
def evaluate_molecules(molecules: List[str],
b3lyp_energies: List[float]) -> List[float]:
"""Compute the atomization energy of molecules
Args:
molecules ([str]): XYZ-format molecular structures. Assumed to be
fully-relaxed
b3lyp_energies ([float]): B3LYP total energies of structures
Returns:
([float]): Estimated G4MP2 atomization energies of molecules
"""
# Convert all of the molecules to the qml representation
compnds = [Compound(StringIO(x)) for x in molecules]
# Compute the atomization energy for each compound
b3lyp_atom = [
compute_atomization_energy(next(read_xyz(StringIO(x))), u0, 'b3lyp')
for x, u0 in zip(molecules, b3lyp_energies)
]
# Compute the representaiton for each compound
def compute_rep(x):
"""Generates representation and returns the values"""
x.generate_fchl_representation(max_size)
return x.representation
reps = np.array(list(map(compute_rep, compnds)))
# Compute the delta between B3LYP and G4MP2
delta = model.predict(reps)
# Return the sum of the two
return np.add(b3lyp_atom, delta)
def _load_molecule(self, molecule_path):
"""
Load molecule from file (can handle all ase formats).
Args:
molecule_path (str): Path to molecular geometry
"""
file_format = os.path.splitext(molecule_path)[-1]
if file_format == 'xyz':
self.molecule = read_xyz(molecule_path)
else:
self.molecule = read(molecule_path)
def acetaldehyde():
return next(
read_xyz(
StringIO("""7
H4 C2 O1
C -0.002945 1.509914 0.008673
C 0.026083 0.003276 -0.037459
O 0.942288 -0.655070 -0.456826
H 0.922788 1.926342 -0.391466
H -0.862015 1.878525 -0.564795
H -0.150506 1.843934 1.042891
H -0.894430 -0.486434 0.357749""")))
def main(filename, source, x=1, y=1, z=1):
temp = list(xyz.read_xyz(filename))[0]
if source == "NA":
temp.cell = Cell([[x, 0, 0], [0, y, 0], [0, 0, z]])
else:
try:
temp.cell = vasp.read_vasp(source).cell
except:
raise Exception("source is neither valid POSCAR nor NA")
vasp.write_vasp("../POSCAR_ec", temp, sort=True, vasp5=True)
def evaluate_schnet(models: List[Union[TorchMessage, torch.nn.Module, Path]],
molecules: List[str],
property_name: str,
batch_size: int = 64,
device: str = 'cpu') -> np.ndarray:
"""Run inference for a machine learning model
Args:
models: List of models to evaluate. Either a SchNet model or
the bytes corresponding to a serialized model
molecules: XYZ-format structures of molecules to be evaluate
property_name: Name of the property being predicted
batch_size: Number of molecules to evaluate per batch
device: Device on which to run the computation
"""
# Make sure the models are converted to Torch models
if isinstance(models[0], TorchMessage):
models = [m.get_model(device) for m in models]
elif isinstance(models[0], (Path, str)):
models = [torch.load(m, map_location='cpu')
for m in models] # Load to main memory first
# Make the dataset
with TemporaryDirectory() as td:
# Convert the molecules to ase.Atoms objects
atoms = [next(read_xyz(StringIO(x), slice(None))) for x in molecules]
# Save the data to an ASE Atoms database
run_file = os.path.join(td, 'run_data.db')
db = AtomsData(run_file, available_properties=[])
db.add_systems(atoms, [{} for _ in atoms])
# Build the data loader
loader = AtomsLoader(db, batch_size=batch_size)
# Run the models
y_preds = []
for model in models:
y_pred = []
model.to(device) # Move the model to the device
for batch in loader:
# Push the batch to the device
batch = {k: v.to(device) for k, v in batch.items()}
# Run it and save results
pred = model(batch)
y_pred.append(pred[property_name].detach().cpu().numpy())
y_preds.append(np.squeeze(np.concatenate(y_pred)))
return np.vstack(y_preds).T
def read_centers(filename=None):
# read
atoms = list(read_xyz(filename))[0]
symbols = atoms.get_chemical_symbols()
positions = atoms.get_positions()
# split X (wannier centers) and atoms.
# map( (filter( (lambda x: x[0]=='X'), zip(symbols,positions))))
# wannier_centers= [xpos[1] for xpos in zip(symbols,positions) if xpos[0] ]
wannier_centers = [
xpos[1] for xpos in zip(symbols, positions) if xpos[0] == 'X'
]
asymbols = [xpos[0] for xpos in zip(symbols, positions) if xpos[0] != 'X']
aposes = [xpos[1] for xpos in zip(symbols, positions) if xpos[0] != 'X']
cell = atoms.get_cell()
atoms = Atoms(symbols=asymbols, positions=aposes, cell=cell)
return wannier_centers, atoms
def make_schnetpack_data(dataset,
dbpath,
properties,
xyz_col='xyz',
conformers=None,
overwrite=True):
"""Convert a Pandas dictionary to a SchNet database
Args:
dataset (pd.DataFrame): Dataset to convert
dbpath (string): Path to database to be saved
properties ([string]): List of properties to include in the dataset
conformers (str): Name of column with conformers as xyz
xyz_col (string): Name of the column with the XYZ data
overwrite (True): Whether to overwrite the database
"""
# If needed, delete the previous database
if os.path.exists(dbpath) and overwrite:
os.unlink(dbpath)
# Convert all entries to ase.Atoms objects
atoms = dataset[xyz_col].apply(lambda x: read_xyz(StringIO(x)).__next__())
# Every column besides the training set will be a property
prop_cols = set(properties).difference([xyz_col])
property_list = [
dict(zip(prop_cols, [np.atleast_1d(row[p]) for p in prop_cols]))
for i, row in dataset.iterrows()
]
# Add conformers to the property list, but it isn't a required property when loading entries
if conformers is not None:
for d, c in zip(property_list, dataset[conformers]):
d['conformers'] = np.atleast_1d(c)
# Initialize the object
db = AtomsData(dbpath,
required_properties=properties,
conformers=conformers is not None)
# Add every system to the db object
db.add_systems(atoms, property_list)
return db
def get_initial_structure(smiles: str) -> Tuple[Atoms, pybel.Molecule]:
"""Generate an initial guess for a molecular structure
Args:
smiles: SMILES string
Returns:
Generate an Atoms object
"""
# Make the 3D structure
mol = pybel.readstring("smi", smiles)
mol.make3D()
# Convert it to ASE
atoms = next(read_xyz(StringIO(mol.write('xyz')), slice(None)))
atoms.charge = mol.charge
atoms.set_initial_charges([a.formalcharge for a in mol.atoms])
return atoms, mol
def run_model(model, data, xyz_col, additional_cols=None, progbar=True):
"""Runs a SchNetPack model on the column of a dataframe containing XYZ files
Args:
model (AtomisticModel): Model to be evaluated
data (DataFrame): Data to be evaluated
xyz_col (string): Column containing the XYZ data
additional_cols ([string]): Any other columns to add to the input (e.g., B3LYP results)
progbar (boolean): Whether to display a progress bar
Returns:
(ndarray) Predictions from the model
"""
# Get default value for additional_cols
if additional_cols is None:
additional_cols = []
# Make the tool to convert ase.Atoms to SchNet inputs
c = AtomsConverter()
results = []
for xyz, more_data in tqdm(list(
zip(data[xyz_col], data[additional_cols].values)),
disable=not progbar,
leave=False):
# Convert the XYZ file to an ASE object
atoms = next(read_xyz(StringIO(xyz)))
# Generate it in the input format needed
inputs = c.convert_atoms(atoms)
# Add in the additional columns
for i, col in enumerate(additional_cols):
inputs[col] = torch.Tensor(np.expand_dims(more_data[i], 0))
# Run it through the model
outputs = model(inputs)
# Get the value in numpy format
results.append(np.squeeze(outputs['y'].cpu().data.numpy()))
return np.array(results)
def read_gamess_us(filename):
"""Method to read geometry from a GAMESS-US output
"""
f = filename
if isinstance(filename, str):
f = open(filename)
lines = f.readlines()
done = False
i = 0
xyzstring = ''
for l, line in enumerate(lines):
if line.find('CHARGE X Y Z'
) >= 0:
while not done:
i += 1
if not lines[l + i] == '\n':
sym = lines[l + i].strip().split()[0]
for c in range(2, 5):
pos = map(float, lines[l + i].strip().split()[2:5])
xyzstring += sym + ' ' + str(pos[0] * Bohr) + ' ' + str(
pos[1] * Bohr) + ' ' + str(pos[2] * Bohr) + '\n'
else:
done = True
if done:
break
xyzstring = str(i - 1) + '\n\n' + xyzstring
atoms = read_xyz(StringIO(xyzstring))
if type(filename) == str:
f.close()
return atoms
def read_gamess_us_input(filename):
"""Method to read geometry from an GAMESS-US input file."""
f = filename
if isinstance(filename, str):
f = open(filename)
lines = f.readlines()
# Find geometry region of input file.
done = False
i = 0
xyzstring = ''
for l, line in enumerate(lines):
if line.find(' $data') > -1:
i += 2
while not done:
i += 1
if not (lines[l + i].find('$end') > -1):
sym = lines[l + i].strip().split()[0]
pos = map(float, lines[l + i].strip().split()[2:5])
xyzstring += sym + ' ' + str(pos[0]) + ' ' + str(
pos[1]) + ' ' + str(pos[2]) + '\n'
else:
done = True
if done:
break
xyzstring = str(i - 3) + '\n\n' + xyzstring
atoms = read_xyz(StringIO(xyzstring))
if type(filename) == str:
f.close()
return atoms
def get_initial_structure(smiles: str) -> Tuple[Atoms, Dict[int, Set[int]]]:
"""Generate an initial guess for a molecular structure
Args:
smiles: SMILES string
Returns:
An ASE atoms object, bond graph
"""
# Make the 3D structure
mol = pybel.readstring("smi", smiles)
mol.make3D()
# Convert it to ASE
atoms = next(read_xyz(StringIO(mol.write('xyz')), slice(None)))
# Get the bonding graph
g = nx.Graph()
g.add_nodes_from(range(len(mol.atoms)))
for bond in OBMolBondIter(mol.OBMol):
g.add_edge(bond.GetBeginAtomIdx() - 1, bond.GetEndAtomIdx() - 1, data={"rotor": bond.IsRotor()})
return atoms, g
def read_gamess_us(filename):
"""Method to read geometry from a GAMESS-US output
"""
f = filename
if isinstance(filename, str):
f = open(filename)
lines = f.readlines()
done = False
i = 0
xyzstring = ''
for l, line in enumerate(lines):
if line.find('CHARGE X Y Z') >= 0:
while not done:
i += 1
if not lines[l+i] == '\n':
sym = lines[l+i].strip().split()[0]
for c in range(2,5):
pos = map(float, lines[l+i].strip().split()[2:5])
xyzstring += sym + ' ' + str(pos[0]*Bohr)+ ' ' + str(pos[1]*Bohr)+ ' ' + str(pos[2]*Bohr)+'\n'
else:
done = True
if done:
break
xyzstring = str(i-1) + '\n\n' + xyzstring
atoms = read_xyz(StringIO(xyzstring))
if type(filename) == str:
f.close()
return atoms
def read_gamess_us_input(filename):
"""Method to read geometry from an GAMESS-US input file."""
f = filename
if isinstance(filename, str):
f = open(filename)
lines = f.readlines()
# Find geometry region of input file.
done = False
i = 0
xyzstring = ''
for l, line in enumerate(lines):
if line.find(' $data') > -1:
i += 2
while not done:
i += 1
if not (lines[l+i].find('$end') > -1):
sym = lines[l+i].strip().split()[0]
pos = map(float, lines[l+i].strip().split()[2:5])
xyzstring += sym + ' ' + str(pos[0])+ ' ' + str(pos[1])+ ' ' + str(pos[2])+'\n'
else:
done = True
if done:
break
xyzstring = str(i-3) + '\n\n' + xyzstring
atoms = read_xyz(StringIO(xyzstring))
if type(filename) == str:
f.close()
return atoms
def read(filename, index=-1, format=None):
"""Read Atoms object(s) from file.
filename: str
Name of the file to read from.
index: int or slice
If the file contains several configurations, the last configuration
will be returned by default. Use index=n to get configuration
number n (counting from zero).
format: str
Used to specify the file-format. If not given, the
file-format will be guessed by the *filetype* function. If
it's 'babel', will try to use the OpenBabel library.
Known formats:
========================= ===========
format short name
========================= ===========
GPAW restart-file gpw
Dacapo netCDF output file dacapo
Old ASE netCDF trajectory nc
Virtual Nano Lab file vnl
ASE pickle trajectory traj
GPAW text output gpaw-text
CUBE file cube
XCrySDen Structure File xsf
Dacapo text output dacapo-text
XYZ-file xyz
VASP POSCAR/CONTCAR file vasp
Protein Data Bank pdb
VTK XML Image Data vti
VTK XML Structured Grid vts
VTK XML Unstructured Grid vtu
========================= ===========
"""
p = filename.rfind('@')
if p != -1:
try:
index = string2index(filename[p + 1:])
except ValueError:
pass
else:
filename = filename[:p]
if format is None:
format = filetype(filename)
if format.startswith('gpw'):
import gpaw
r = gpaw.io.open(filename, 'r')
positions = r.get('CartesianPositions') * Bohr
numbers = r.get('AtomicNumbers')
cell = r.get('UnitCell') * Bohr
pbc = r.get('BoundaryConditions')
tags = r.get('Tags')
magmoms = r.get('MagneticMoments')
atoms = Atoms(positions=positions,
numbers=numbers,
cell=cell,
pbc=pbc)
if tags.any():
atoms.set_tags(tags)
if magmoms.any():
atoms.set_initial_magnetic_moments(magmoms)
return atoms
if format == 'xyz':
from ase.io.xyz import read_xyz
return read_xyz(filename, index)
if format == 'traj':
from ase.io.trajectory import read_trajectory
return read_trajectory(filename, index)
if format == 'cube':
from ase.io.cube import read_cube
return read_cube(filename, index)
if format == 'nc':
from ase.io.netcdf import read_netcdf
return read_netcdf(filename, index)
if format == 'gpaw-text':
from ase.io.gpawtext import read_gpaw_text
return read_gpaw_text(filename, index)
if format == 'dacapo-text':
from ase.io.dacapo import read_dacapo_text
return read_dacapo_text(filename)
if format == 'dacapo':
from ase.io.dacapo import read_dacapo
return read_dacapo(filename)
if format == 'xsf':
from ase.io.xsf import read_xsf
return read_xsf(filename, index)
if format == 'vasp':
from ase.io.vasp import read_vasp
return read_vasp(filename)
if format == 'mol':
from ase.io.mol import read_mol
return read_mol(filename)
if format == 'pdb':
from ase.io.pdb import read_pdb
return read_pdb(filename)
if format == 'cif':
from ase.io.cif import read_cif
return read_cif(filename)
if format == 'babel':
from ase.io.babel import read_babel
return read_babel(filename, index=index)
if format == 'vti':
from ase.io.vtkxml import read_vti
return read_vti(filename)
if format == 'vts':
from ase.io.vtkxml import read_vts
return read_vts(filename)
if format == 'vtu':
from ase.io.vtkxml import read_vtu
return read_vtu(filename)
if format == 'iwm':
from ase.io.iwm import read_iwm
return read_iwm(filename)
if format == 'Cmdft':
from ase.io.cmdft import read_I_info
return read_I_info(filename)
raise RuntimeError('That can *not* happen!')
# Get the system information
host_info = get_platform_info()
# Set the random seed
np.random.seed(args.random)
rng = np.random.RandomState(args.random)
# Download the QM9 dataset and get the molecule of interest
qm9_path = get_qm9_path()
with gzip.open(qm9_path, 'rt') as fp:
for _, d in zip(range(args.mol), fp):
pass
mol_info = json.loads(d)
# Parse the molecule coordinates into an ASE object
atoms = next(read_xyz(StringIO(mol_info['xyz'])))
# Open an experiment directory
start_time = datetime.utcnow()
out_dir = os.path.join('runs', f'{start_time.strftime("%d%b%y-%H%M%S")}')
os.makedirs(out_dir)
# Save the parameters and host information
with open(os.path.join(out_dir, 'run_params.json'), 'w') as fp:
json.dump(run_params, fp, indent=2)
with open(os.path.join(out_dir, 'host_info.json'), 'w') as fp:
json.dump(host_info, fp, indent=2)
# Initialize the ASE calculator
calc = Psi4(memory='500MB', **_fidelity[args.fidelity])
"-f",
"--format",
action="store",
type="string",
default="xyz",
help="format of the output file: POSCAR, (xyz in preparation)")
(options, args) = parser.parse_args()
print_debug = False
if (num < 2):
parser.print_help()
else:
file_xyz = read_xyz(input_file, slice(0, None, 1))
natoms = file_xyz[0].get_number_of_atoms()
# if possible read unit-cell
if (os.path.isfile(options.cell)):
cell = [[], [], []]
f = open(options.cell, "r")
ls = f.read().splitlines()
for i in range(3):
l = ls[i].split()
cell[i] = [float(l[0]), float(l[1]), float(l[2])]
for step in file_xyz:
step.set_cell(cell)
step.set_pbc([True, True, True])
# --------------- Atoms -----------------
butadiene = """10
C 3.649801161546418 5.442281389577507 3.863313703750026
C 5.051651240044169 5.368220758269772 4.162165876906096
C 5.750174626862403 4.162261915959347 4.240449977068684
C 7.150130182125531 4.155384186721486 4.537328602062397
H 3.218154657585170 4.565210696328925 3.522601038049320
H 3.077656122062729 6.375092902842770 3.826039498180272
H 5.478464901706067 6.370680001794822 4.422235395756437
H 5.320549047980879 3.220584852467720 3.974551561510350
H 7.723359150977955 3.224855971783890 4.574146712279462
H 7.580803493981530 5.034479218283977 4.877211530909463
"""
h = 0.3
atoms = Cluster(read_xyz(StringIO.StringIO(butadiene)))
atoms.minimal_box(3., h)
atoms.set_calculator(GPAW(h=h))
if 0:
dyn = FIRE(atoms)
dyn.run(fmax=0.05)
atoms.write('butadiene.xyz')
vibname = 'fcvib'
vib = Vibrations(atoms, name=vibname)
vib.run()
# Modul
a = FranckCondon(atoms, vibname, minfreq=250)
# excited state forces
from ase.io.xyz import read_xyz
from gpaw import GPAW
from gpaw.mixer import Mixer
from gpaw import ConvergenceError
from gpaw.mpi import rank
from gpaw.eigensolvers.rmm_diis import RMM_DIIS
from gpaw import setup_paths
# Use setups from the $PWD and $PWD/.. first
setup_paths.insert(0, '.')
setup_paths.insert(0, '../')
atoms = read_xyz('../Au102_revised.xyz')
prefix = 'Au_cluster'
L = 32.0
atoms.set_cell((L,L,L),scale_atoms=False)
atoms.center()
atoms.set_pbc(1)
r = [1, 1, 1]
atoms = atoms.repeat(r)
n = [240 * ri for ri in r]
# nbands (>=1683) is the number of bands per cluster
nbands = 3*6*6*16 # 1728
for ri in r: nbands = nbands*ri
mixer = Mixer(beta=0.1, nmaxold=5, weight=100.0)
# the next three lines decrease memory usage
es = RMM_DIIS(keep_htpsit=False)
butadiene = """10
C 3.649801161546418 5.442281389577507 3.863313703750026
C 5.051651240044169 5.368220758269772 4.162165876906096
C 5.750174626862403 4.162261915959347 4.240449977068684
C 7.150130182125531 4.155384186721486 4.537328602062397
H 3.218154657585170 4.565210696328925 3.522601038049320
H 3.077656122062729 6.375092902842770 3.826039498180272
H 5.478464901706067 6.370680001794822 4.422235395756437
H 5.320549047980879 3.220584852467720 3.974551561510350
H 7.723359150977955 3.224855971783890 4.574146712279462
H 7.580803493981530 5.034479218283977 4.877211530909463
"""
h = 0.3
atoms = Cluster(read_xyz(StringIO.StringIO(butadiene)))
atoms.minimal_box(3.0, h)
atoms.set_calculator(GPAW(h=h))
if 0:
dyn = FIRE(atoms)
dyn.run(fmax=0.05)
atoms.write("butadiene.xyz")
vibname = "fcvib"
vib = Vibrations(atoms, name=vibname)
vib.run()
# Modul
a = FranckCondon(atoms, vibname, minfreq=250)
# excited state forces
def train_schnet(
model: Union[TorchMessage, torch.nn.Module, Path],
database: Dict[str, float],
num_epochs: int,
reset_weights: bool = True,
property_name: str = 'output',
test_set: Optional[List[str]] = None,
device: str = 'cpu',
batch_size: int = 32,
validation_split: float = 0.1,
bootstrap: bool = False,
random_state: int = 1,
learning_rate: float = 1e-3,
patience: int = None,
timeout: float = None
) -> Union[Tuple[TorchMessage, pd.DataFrame], Tuple[TorchMessage, pd.DataFrame,
List[float]]]:
"""Train a SchNet model
Args:
model: Model to be retrained
database: Mapping of XYZ format structure to property
num_epochs: Number of training epochs
property_name: Name of the property being predicted
reset_weights: Whether to re-initialize weights before training, or start training from previous
test_set: Hold-out set. If provided, function will return the performance of the model on those weights
device: Device (e.g., 'cuda', 'cpu') used for training
batch_size: Batch size during training
validation_split: Fraction to training set to use for the validation loss
bootstrap: Whether to take a bootstrap sample of the training set before training
random_state: Random seed used for generating validation set and bootstrap sampling
learning_rate: Initial learning rate for optimizer
patience: Patience until learning rate is lowered. Default: epochs / 8
timeout: Maximum training time in seconds
Returns:
- model: Retrained model
- history: Training history
- test_pred: Predictions on ``test_set``, if provided
"""
# Make sure the models are converted to Torch models
if isinstance(model, TorchMessage):
model = model.get_model(device)
elif isinstance(model, (Path, str)):
model = torch.load(model,
map_location='cpu') # Load to main memory first
# If desired, re-initialize weights
if reset_weights:
for module in model.modules():
if hasattr(module, 'reset_parameters'):
module.reset_parameters()
# Separate the database into molecules and properties
xyz, y = zip(*database.items())
xyz = np.array(xyz)
y = np.array(y)
# Convert the xyz files to ase Atoms
atoms = np.array([next(read_xyz(StringIO(x), slice(None))) for x in xyz])
# Make the training and validation splits
rng = np.random.RandomState(random_state)
train_split = rng.rand(len(xyz)) > validation_split
train_X = atoms[train_split]
train_y = y[train_split]
valid_X = atoms[~train_split]
valid_y = y[~train_split]
# Perform a bootstrap sample of the training data
if bootstrap:
sample = rng.choice(len(train_X), size=(len(train_X), ), replace=True)
train_X = train_X[sample]
train_y = train_y[sample]
# Start the training process
with TemporaryDirectory() as td:
# Save the data to an ASE Atoms database
train_file = os.path.join(td, 'train_data.db')
db = AtomsData(train_file, available_properties=[property_name])
db.add_systems(train_X, [{property_name: i} for i in train_y])
train_loader = AtomsLoader(db, batch_size=batch_size, shuffle=True)
valid_file = os.path.join(td, 'valid_data.db')
db = AtomsData(valid_file, available_properties=[property_name])
db.add_systems(valid_X, [{property_name: i} for i in valid_y])
valid_loader = AtomsLoader(db, batch_size=batch_size)
# Make the trainer
opt = optim.Adam(model.parameters(), lr=learning_rate)
loss = trn.build_mse_loss(['delta'])
metrics = [spk.metrics.MeanSquaredError('delta')]
if patience is None:
patience = num_epochs // 8
hooks = [
trn.CSVHook(log_path=td, metrics=metrics),
trn.ReduceLROnPlateauHook(opt,
patience=patience,
factor=0.8,
min_lr=1e-6,
stop_after_min=True)
]
if timeout is not None:
hooks.append(TimeoutHook(timeout))
trainer = trn.Trainer(
model_path=td,
model=model,
hooks=hooks,
loss_fn=loss,
optimizer=opt,
train_loader=train_loader,
validation_loader=valid_loader,
checkpoint_interval=num_epochs + 1 # Turns off checkpointing
)
trainer.train(device, n_epochs=num_epochs)
# Load in the best model
model = torch.load(os.path.join(td, 'best_model'))
# If desired, report the performance on a test set
test_pred = None
if test_set is not None:
test_pred = evaluate_schnet([model],
test_set,
property_name=property_name,
batch_size=batch_size,
device=device)
# Move the model off of the GPU to save memory
if 'cuda' in device:
model.to('cpu')
# Load in the training results
train_results = pd.read_csv(os.path.join(td, 'log.csv'))
# Return the results
if test_pred is None:
return TorchMessage(model), train_results
else:
return TorchMessage(model), train_results, test_pred[:, 0].tolist()
def read(filename, index=-1, format=None):
"""Read Atoms object(s) from file.
filename: str
Name of the file to read from.
index: int or slice
If the file contains several configurations, the last configuration
will be returned by default. Use index=n to get configuration
number n (counting from zero).
format: str
Used to specify the file-format. If not given, the
file-format will be guessed by the *filetype* function.
Known formats:
========================= ===========
format short name
========================= ===========
GPAW restart-file gpw
Dacapo netCDF output file dacapo
Old ASE netCDF trajectory nc
Virtual Nano Lab file vnl
ASE pickle trajectory traj
ASE bundle trajectory bundle
GPAW text output gpaw-text
CUBE file cube
XCrySDen Structure File xsf
Dacapo text output dacapo-text
XYZ-file xyz
VASP POSCAR/CONTCAR file vasp
VASP OUTCAR file vasp_out
SIESTA STRUCT file struct_out
ABINIT input file abinit
V_Sim ascii file v_sim
Protein Data Bank pdb
CIF-file cif
FHI-aims geometry file aims
FHI-aims output file aims_out
VTK XML Image Data vti
VTK XML Structured Grid vts
VTK XML Unstructured Grid vtu
TURBOMOLE coord file tmol
TURBOMOLE gradient file tmol-gradient
exciting input exi
AtomEye configuration cfg
WIEN2k structure file struct
DftbPlus input file dftb
CASTEP geom file cell
CASTEP output file castep
CASTEP trajectory file geom
ETSF format etsf.nc
DFTBPlus GEN format gen
CMR db/cmr-file db
CMR db/cmr-file cmr
LAMMPS dump file lammps
========================= ===========
"""
if isinstance(filename, str):
p = filename.rfind('@')
if p != -1:
try:
index = string2index(filename[p + 1:])
except ValueError:
pass
else:
filename = filename[:p]
if isinstance(index, str):
index = string2index(index)
if format is None:
format = filetype(filename)
if format.startswith('gpw'):
import gpaw
r = gpaw.io.open(filename, 'r')
positions = r.get('CartesianPositions') * Bohr
numbers = r.get('AtomicNumbers')
cell = r.get('UnitCell') * Bohr
pbc = r.get('BoundaryConditions')
tags = r.get('Tags')
magmoms = r.get('MagneticMoments')
energy = r.get('PotentialEnergy') * Hartree
if r.has_array('CartesianForces'):
forces = r.get('CartesianForces') * Hartree / Bohr
else:
forces = None
atoms = Atoms(positions=positions, numbers=numbers, cell=cell, pbc=pbc)
if tags.any():
atoms.set_tags(tags)
if magmoms.any():
atoms.set_initial_magnetic_moments(magmoms)
else:
magmoms = None
atoms.calc = SinglePointCalculator(energy, forces, None, magmoms,
atoms)
return atoms
if format == 'castep':
from ase.io.castep import read_castep
return read_castep(filename, index)
if format == 'castep_cell':
import ase.io.castep
return ase.io.castep.read_cell(filename, index)
if format == 'castep_geom':
import ase.io.castep
return ase.io.castep.read_geom(filename, index)
if format == 'exi':
from ase.io.exciting import read_exciting
return read_exciting(filename, index)
if format == 'xyz':
from ase.io.xyz import read_xyz
return read_xyz(filename, index)
if format == 'traj':
from ase.io.trajectory import read_trajectory
return read_trajectory(filename, index)
if format == 'bundle':
from ase.io.bundletrajectory import read_bundletrajectory
return read_bundletrajectory(filename, index)
if format == 'cube':
from ase.io.cube import read_cube
return read_cube(filename, index)
if format == 'nc':
from ase.io.netcdf import read_netcdf
return read_netcdf(filename, index)
if format == 'gpaw-text':
from ase.io.gpawtext import read_gpaw_text
return read_gpaw_text(filename, index)
if format == 'dacapo-text':
from ase.io.dacapo import read_dacapo_text
return read_dacapo_text(filename)
if format == 'dacapo':
from ase.io.dacapo import read_dacapo
return read_dacapo(filename)
if format == 'xsf':
from ase.io.xsf import read_xsf
return read_xsf(filename, index)
if format == 'vasp':
from ase.io.vasp import read_vasp
return read_vasp(filename)
if format == 'vasp_out':
from ase.io.vasp import read_vasp_out
return read_vasp_out(filename, index)
if format == 'abinit':
from ase.io.abinit import read_abinit
return read_abinit(filename)
if format == 'v_sim':
from ase.io.v_sim import read_v_sim
return read_v_sim(filename)
if format == 'mol':
from ase.io.mol import read_mol
return read_mol(filename)
if format == 'pdb':
from ase.io.pdb import read_pdb
return read_pdb(filename, index)
if format == 'cif':
from ase.io.cif import read_cif
return read_cif(filename, index)
if format == 'struct':
from ase.io.wien2k import read_struct
return read_struct(filename)
if format == 'struct_out':
from ase.io.siesta import read_struct
return read_struct(filename)
if format == 'vti':
from ase.io.vtkxml import read_vti
return read_vti(filename)
if format == 'vts':
from ase.io.vtkxml import read_vts
return read_vts(filename)
if format == 'vtu':
from ase.io.vtkxml import read_vtu
return read_vtu(filename)
if format == 'aims':
from ase.io.aims import read_aims
return read_aims(filename)
if format == 'aims_out':
from ase.io.aims import read_aims_output
return read_aims_output(filename, index)
if format == 'iwm':
from ase.io.iwm import read_iwm
return read_iwm(filename)
if format == 'Cmdft':
from ase.io.cmdft import read_I_info
return read_I_info(filename)
if format == 'tmol':
from ase.io.turbomole import read_turbomole
return read_turbomole(filename)
if format == 'tmol-gradient':
from ase.io.turbomole import read_turbomole_gradient
return read_turbomole_gradient(filename)
if format == 'cfg':
from ase.io.cfg import read_cfg
return read_cfg(filename)
if format == 'dftb':
from ase.io.dftb import read_dftb
return read_dftb(filename)
if format == 'sdf':
from ase.io.sdf import read_sdf
return read_sdf(filename)
if format == 'etsf':
from ase.io.etsf import ETSFReader
return ETSFReader(filename).read_atoms()
if format == 'gen':
from ase.io.gen import read_gen
return read_gen(filename)
if format == 'db':
from ase.io.cmr_io import read_db
return read_db(filename, index)
if format == 'lammps':
from ase.io.lammps import read_lammps_dump
return read_lammps_dump(filename, index)
raise RuntimeError('File format descriptor ' + format + ' not recognized!')
def read(filename, index=-1, format=None):
"""Read Atoms object(s) from file.
filename: str
Name of the file to read from.
index: int or slice
If the file contains several configurations, the last configuration
will be returned by default. Use index=n to get configuration
number n (counting from zero).
format: str
Used to specify the file-format. If not given, the
file-format will be guessed by the *filetype* function.
Known formats:
========================= ===========
format short name
========================= ===========
GPAW restart-file gpw
Dacapo netCDF output file dacapo
Old ASE netCDF trajectory nc
Virtual Nano Lab file vnl
ASE pickle trajectory traj
ASE bundle trajectory bundle
GPAW text output gpaw-text
CUBE file cube
XCrySDen Structure File xsf
Dacapo text output dacapo-text
XYZ-file xyz
VASP POSCAR/CONTCAR file vasp
VASP OUTCAR file vasp_out
SIESTA STRUCT file struct_out
ABINIT input file abinit
V_Sim ascii file v_sim
Protein Data Bank pdb
CIF-file cif
FHI-aims geometry file aims
FHI-aims output file aims_out
VTK XML Image Data vti
VTK XML Structured Grid vts
VTK XML Unstructured Grid vtu
TURBOMOLE coord file tmol
TURBOMOLE gradient file tmol-gradient
exciting input exi
AtomEye configuration cfg
WIEN2k structure file struct
DftbPlus input file dftb
CASTEP geom file cell
CASTEP output file castep
CASTEP trajectory file geom
ETSF format etsf.nc
DFTBPlus GEN format gen
CMR db/cmr-file db
CMR db/cmr-file cmr
LAMMPS dump file lammps
========================= ===========
"""
if isinstance(filename, str):
p = filename.rfind('@')
if p != -1:
try:
index = string2index(filename[p + 1:])
except ValueError:
pass
else:
filename = filename[:p]
if isinstance(index, str):
index = string2index(index)
if format is None:
format = filetype(filename)
if format.startswith('gpw'):
import gpaw
r = gpaw.io.open(filename, 'r')
positions = r.get('CartesianPositions') * Bohr
numbers = r.get('AtomicNumbers')
cell = r.get('UnitCell') * Bohr
pbc = r.get('BoundaryConditions')
tags = r.get('Tags')
magmoms = r.get('MagneticMoments')
energy = r.get('PotentialEnergy') * Hartree
if r.has_array('CartesianForces'):
forces = r.get('CartesianForces') * Hartree / Bohr
else:
forces = None
atoms = Atoms(positions=positions,
numbers=numbers,
cell=cell,
pbc=pbc)
if tags.any():
atoms.set_tags(tags)
if magmoms.any():
atoms.set_initial_magnetic_moments(magmoms)
else:
magmoms = None
atoms.calc = SinglePointCalculator(energy, forces, None, magmoms,
atoms)
return atoms
if format == 'castep':
from ase.io.castep import read_castep
return read_castep(filename, index)
if format == 'castep_cell':
import ase.io.castep
return ase.io.castep.read_cell(filename, index)
if format == 'castep_geom':
import ase.io.castep
return ase.io.castep.read_geom(filename, index)
if format == 'exi':
from ase.io.exciting import read_exciting
return read_exciting(filename, index)
if format == 'xyz':
from ase.io.xyz import read_xyz
return read_xyz(filename, index)
if format == 'traj':
from ase.io.trajectory import read_trajectory
return read_trajectory(filename, index)
if format == 'bundle':
from ase.io.bundletrajectory import read_bundletrajectory
return read_bundletrajectory(filename, index)
if format == 'cube':
from ase.io.cube import read_cube
return read_cube(filename, index)
if format == 'nc':
from ase.io.netcdf import read_netcdf
return read_netcdf(filename, index)
if format == 'gpaw-text':
from ase.io.gpawtext import read_gpaw_text
return read_gpaw_text(filename, index)
if format == 'dacapo-text':
from ase.io.dacapo import read_dacapo_text
return read_dacapo_text(filename)
if format == 'dacapo':
from ase.io.dacapo import read_dacapo
return read_dacapo(filename)
if format == 'xsf':
from ase.io.xsf import read_xsf
return read_xsf(filename, index)
if format == 'vasp':
from ase.io.vasp import read_vasp
return read_vasp(filename)
if format == 'vasp_out':
from ase.io.vasp import read_vasp_out
return read_vasp_out(filename, index)
if format == 'abinit':
from ase.io.abinit import read_abinit
return read_abinit(filename)
if format == 'v_sim':
from ase.io.v_sim import read_v_sim
return read_v_sim(filename)
if format == 'mol':
from ase.io.mol import read_mol
return read_mol(filename)
if format == 'pdb':
from ase.io.pdb import read_pdb
return read_pdb(filename, index)
if format == 'cif':
from ase.io.cif import read_cif
return read_cif(filename, index)
if format == 'struct':
from ase.io.wien2k import read_struct
return read_struct(filename)
if format == 'struct_out':
from ase.io.siesta import read_struct
return read_struct(filename)
if format == 'vti':
from ase.io.vtkxml import read_vti
return read_vti(filename)
if format == 'vts':
from ase.io.vtkxml import read_vts
return read_vts(filename)
if format == 'vtu':
from ase.io.vtkxml import read_vtu
return read_vtu(filename)
if format == 'aims':
from ase.io.aims import read_aims
return read_aims(filename)
if format == 'aims_out':
from ase.io.aims import read_aims_output
return read_aims_output(filename, index)
if format == 'iwm':
from ase.io.iwm import read_iwm
return read_iwm(filename)
if format == 'Cmdft':
from ase.io.cmdft import read_I_info
return read_I_info(filename)
if format == 'tmol':
from ase.io.turbomole import read_turbomole
return read_turbomole(filename)
if format == 'tmol-gradient':
from ase.io.turbomole import read_turbomole_gradient
return read_turbomole_gradient(filename)
if format == 'cfg':
from ase.io.cfg import read_cfg
return read_cfg(filename)
if format == 'dftb':
from ase.io.dftb import read_dftb
return read_dftb(filename)
if format == 'sdf':
from ase.io.sdf import read_sdf
return read_sdf(filename)
if format == 'etsf':
from ase.io.etsf import ETSFReader
return ETSFReader(filename).read_atoms()
if format == 'gen':
from ase.io.gen import read_gen
return read_gen(filename)
if format == 'db':
from ase.io.cmr_io import read_db
return read_db(filename, index)
if format == 'lammps':
from ase.io.lammps import read_lammps_dump
return read_lammps_dump(filename, index)
raise RuntimeError('File format descriptor '+format+' not recognized!')
from ase.io.xyz import read_xyz
from gpaw import GPAW
from gpaw.mixer import Mixer
from gpaw import ConvergenceError
from gpaw.mpi import rank
from gpaw.eigensolvers.rmm_diis_old import RMM_DIIS
from gpaw import setup_paths
# Use setups from the $PWD and $PWD/.. first
setup_paths.insert(0, '.')
setup_paths.insert(0, '../')
atoms = read_xyz('../Au102_revised.xyz')
prefix = 'Au_cluster'
L = 32.0
atoms.set_cell((L,L,L),scale_atoms=False)
atoms.center()
atoms.set_pbc(1)
r = [1, 1, 1]
atoms = atoms.repeat(r)
n = [240 * ri for ri in r]
# nbands (>=1683) is the number of bands per cluster
nbands = 3*6*6*16 # 1728
for ri in r: nbands = nbands*ri
mixer = Mixer(beta=0.1, nmaxold=5, weight=100.0)
# the next three lines decrease memory usage
es = RMM_DIIS(keep_htpsit=False)
natoms = 0
if(num < 2):
parser.print_help()
else:
# >>>>>>>>>>>>>>>>>>>>> READ GEOMETRY <<<<<<<<<<<<<<<<<<<<
if(iformat == "geometry.in"):
atoms = read_aims(ifile)
# elif(iformat == "cube"):
# atoms = read_cube(sys.argv[num-1])
# elif(iformat == "xsf"):
# atoms = read_xsf(sys.argv[num-1],read_data=True)
elif(iformat == "POSCAR"):
atoms = read_vasp(ifile)
elif(iformat == "xyz"):
atoms = read_xyz(ifile)
cfile = options.xyzcell
ThereIsCellFile = os.path.isfile(cfile)
if(ThereIsCellFile):
cell = [[],[],[]]
f = open(cfile, "r")
ls = f.read().splitlines()
for i in range(3):
l = ls[i].split()
cell[i] = [float(l[0]), float(l[1]), float(l[2])]
atoms.set_cell(cell)
atoms.set_pbc([True,True,True])
if(options.vector == True):
a = options.atoms
