def read_pdb(fileobj, index=-1):
"""Read PDB files.
The format is assumed to follow the description given in
http://www.wwpdb.org/documentation/format32/sect9.html."""
if isinstance(fileobj, str):
fileobj = open(fileobj)
images = []
atoms = Atoms()
for line in fileobj.readlines():
if line.startswith('ATOM') or line.startswith('HETATM'):
try:
# Atom name is arbitrary and does not necessarily contain the element symbol.
# The specification requires the element symbol to be in columns 77+78.
symbol = line[76:78].strip().lower().capitalize()
words = line[30:55].split()
position = np.array([float(words[0]),
float(words[1]),
float(words[2])])
atoms.append(Atom(symbol, position))
except:
pass
if line.startswith('ENDMDL'):
images.append(atoms)
atoms = Atoms()
if len(images) == 0:
images.append(atoms)
return images[index]
def read_pdb(fileobj, index=-1):
"""Read PDB files.
The format is assumed to follow the description given in
http://www.wwpdb.org/documentation/format32/sect9.html."""
if isinstance(fileobj, str):
fileobj = open(fileobj)
atoms = Atoms()
for line in fileobj.readlines():
if line.startswith('ATOM'):
try:
symbol = line[12:16].strip()
# we assume that the second character is a label
# in case that it is upper case
if len(symbol) > 1 and symbol[1].isupper():
symbol = symbol[0]
words = line[30:55].split()
position = np.array([float(words[0]),
float(words[1]),
float(words[2])])
atoms.append(Atom(symbol, position))
except:
pass
return atoms
def read_pdb(fileobj, index=-1):
"""Read PDB files.
The format is assumed to follow the description given in
http://www.wwpdb.org/documentation/format32/sect9.html."""
if isinstance(fileobj, str):
fileobj = open(fileobj)
images = []
atoms = Atoms()
for line in fileobj.readlines():
if line.startswith('ATOM') or line.startswith('HETATM'):
try:
# Atom name is arbitrary and does not necessarily contain the element symbol.
# The specification requires the element symbol to be in columns 77+78.
symbol = line[76:78].strip().lower().capitalize()
words = line[30:55].split()
position = np.array([float(words[0]),
float(words[1]),
float(words[2])])
atoms.append(Atom(symbol, position))
except:
pass
if line.startswith('ENDMDL'):
images.append(atoms)
atoms = Atoms()
if len(images) == 0:
images.append(atoms)
return images[index]
def read_pdb(fileobj, index=-1):
"""Read PDB files.
The format is assumed to follow the description given in
http://www.wwpdb.org/documentation/format32/sect9.html."""
if isinstance(fileobj, str):
fileobj = open(fileobj)
atoms = Atoms()
for line in fileobj.readlines():
if line.startswith('ATOM') or line.startswith('HETATM'):
try:
symbol = line[12:16].strip()
# we assume that the second character is a label
# in case that it is upper case
if len(symbol) > 1 and symbol[1].isupper():
symbol = symbol[0]
words = line[30:55].split()
position = np.array(
[float(words[0]),
float(words[1]),
float(words[2])])
atoms.append(Atom(symbol, position))
except:
pass
return atoms
def colored(self, elements):
res = Atoms()
res.set_cell(self.get_cell())
for atom in self:
elem = self.types[atom.tag]
if elem in elements:
elem = elements[elem]
res.append(Atom(elem, atom.position))
return res
def read_gpaw_text(fileobj, index=-1):
if isinstance(fileobj, str):
fileobj = open(fileobj)
lines = fileobj.readlines()
images = []
while True:
try:
i = lines.index('Unit Cell:\n')
cell = [float(line.split()[2]) for line in lines[i + 3:i + 6]]
pbc = [line.split()[1] == 'yes' for line in lines[i + 3:i + 6]]
except ValueError:
pass
try:
i = lines.index('Positions:\n')
except ValueError:
break
atoms = Atoms(cell=cell, pbc=pbc)
for line in lines[i + 1:]:
words = line.split()
if len(words) != 5:
break
n, symbol, x, y, z = words
symbol = symbol.split('.')[0]
atoms.append(Atom(symbol, [float(x), float(y), float(z)]))
lines = lines[i + 5:]
try:
i = lines.index('-------------------------\n')
except ValueError:
e = None
else:
line = lines[i + 9]
assert line.startswith('Zero Kelvin:')
e = float(line.split()[-1])
try:
ii = lines.index('Forces in eV/Ang:\n')
except ValueError:
f = None
else:
f = []
for i in range(ii + 1, ii + 1 + len(atoms)):
try:
x, y, z = lines[i].split()[-3:]
f.append((float(x), float(y), float(z)))
except (ValueError, IndexError), m:
raise IOError('Malformed GPAW log file: %s' % m)
if len(images) > 0 and e is None:
break
if e is not None or f is not None:
atoms.set_calculator(SinglePointCalculator(e, f, None, None, atoms)) ### Fixme magmoms
images.append(atoms)
lines = lines[i:]
def colored(self, elements):
res = Atoms()
res.set_cell(self.get_cell())
for atom in self:
elem = self.types[atom.tag]
if elem in elements:
elem = elements[elem]
res.append(Atom(elem, atom.position))
return res
def build_atoms(positions, method, cell, alat):
"""Creates the atoms for a quantum espresso in file."""
if method != "crystal":
raise NotImplementedError("Only supported for crystal method of " "ATOMIC_POSITIONS, not %s." % method)
atoms = Atoms()
for el, (x, y, z) in positions:
atoms.append(Atom(el, (x, y, z)))
cell *= f2f(alat) * units.Bohr
atoms.set_cell(cell, scale_atoms=True)
return atoms
def build_atoms(positions, method, cell, alat):
"""Creates the atoms for a quantum espresso in file."""
if method != 'crystal':
raise NotImplementedError('Only supported for crystal method of '
'ATOMIC_POSITIONS, not %s.' % method)
atoms = Atoms()
for el, (x, y, z) in positions:
atoms.append(Atom(el, (x, y, z)))
cell *= f2f(alat) * units.Bohr
atoms.set_cell(cell, scale_atoms=True)
return atoms
def make_atoms(index, lines, key, cell):
"""Scan through lines to get the atomic positions."""
atoms = Atoms()
if key == 'Cartesian axes':
for line in lines[index + 3:]:
entries = line.split()
if len(entries) == 0:
break
symbol = entries[1][:-1]
x = float(entries[6])
y = float(entries[7])
z = float(entries[8])
atoms.append(Atom(symbol, (x, y, z)))
atoms.set_cell(cell)
elif key == 'ATOMIC_POSITIONS (crystal)':
for line in lines[index + 1:]:
entries = line.split()
if len(entries) == 0 or (entries[0] == 'End'):
break
symbol = entries[0][:-1]
x = float(entries[1])
y = float(entries[2])
z = float(entries[3])
atoms.append(Atom(symbol, (x, y, z)))
atoms.set_cell(cell, scale_atoms=True)
# Energy is located after positions.
energylines = [
number for number, line in enumerate(lines)
if ('!' in line and 'total energy' in line)
]
energyline = min([n for n in energylines if n > index])
energy = float(lines[energyline].split()[-2]) * units.Ry
# Forces are located after positions.
forces = np.zeros((len(atoms), 3))
forcelines = [
number for number, line in enumerate(lines)
if 'Forces acting on atoms (Ry/au):' in line
]
forceline = min([n for n in forcelines if n > index])
for line in lines[forceline + 4:]:
words = line.split()
if len(words) == 0:
break
fx = float(words[-3])
fy = float(words[-2])
fz = float(words[-1])
atom_number = int(words[1]) - 1
forces[atom_number] = (fx, fy, fz)
forces *= units.Ry / units.Bohr
calc = SinglePointCalculator(atoms, energy=energy, forces=forces)
atoms.set_calculator(calc)
return atoms
def read_cmdft(fileobj):
lines = fileobj.readlines()
del lines[0]
finished = False
s = Atoms()
while not finished:
w = lines.pop(0).split()
if w[0].startswith('"'):
position = Bohr * np.array([float(w[3]), float(w[4]), float(w[5])])
s.append(Atom(w[0].replace('"', ''), position))
else:
finished = True
yield s
def read_cif(fileobj, index=-1):
if isinstance(fileobj, str):
fileobj = open(fileobj)
def search_key(fobj, key):
for line in fobj:
if key in line:
return line
return None
def get_key(fobj, key, pos=1):
line = search_key(fobj, key)
if line:
return float(line.split()[pos].split('(')[0])
return None
a = get_key(fileobj, '_cell_length_a')
b = get_key(fileobj, '_cell_length_b')
c = get_key(fileobj, '_cell_length_c')
alpha = pi * get_key(fileobj, '_cell_angle_alpha') / 180
beta = pi * get_key(fileobj, '_cell_angle_beta') / 180
gamma = pi * get_key(fileobj, '_cell_angle_gamma') / 180
va = a * np.array([1, 0, 0])
vb = b * np.array([cos(gamma), sin(gamma), 0])
cx = cos(beta)
cy = (cos(alpha) - cos(beta) * cos(gamma)) / sin(gamma)
cz = sqrt(1. - cx*cx - cy*cy)
vc = c * np.array([cx, cy, cz])
cell = np.array([va, vb, vc])
atoms = Atoms(cell=cell)
read = False
for line in fileobj:
if not read:
if '_atom_site_disorder_group' in line:
read = True
else:
word = line.split()
if len(word) < 5:
break
symbol = word[1]
pos = (float(word[2].split('(')[0]) * va +
float(word[3].split('(')[0]) * vb +
float(word[4].split('(')[0]) * vc )
atoms.append(Atom(symbol, pos))
return atoms
def make_atoms(index, lines, key, cell):
"""Scan through lines to get the atomic positions."""
atoms = Atoms()
if key == 'Cartesian axes':
for line in lines[index + 3:]:
entries = line.split()
if len(entries) == 0:
break
symbol = entries[1][:-1]
x = float(entries[6])
y = float(entries[7])
z = float(entries[8])
atoms.append(Atom(symbol, (x, y, z)))
atoms.set_cell(cell)
elif key == 'ATOMIC_POSITIONS (crystal)':
for line in lines[index + 1:]:
entries = line.split()
if len(entries) == 0 or (entries[0] == 'End'):
break
symbol = entries[0][:-1]
x = float(entries[1])
y = float(entries[2])
z = float(entries[3])
atoms.append(Atom(symbol, (x, y, z)))
atoms.set_cell(cell, scale_atoms=True)
# Energy is located after positions.
energylines = [number for number, line in enumerate(lines) if
('!' in line and 'total energy' in line)]
energyline = min([n for n in energylines if n > index])
energy = float(lines[energyline].split()[-2]) * units.Ry
# Forces are located after positions.
forces = np.zeros((len(atoms), 3))
forcelines = [number for number, line in enumerate(lines) if
'Forces acting on atoms (Ry/au):' in line]
forceline = min([n for n in forcelines if n > index])
for line in lines[forceline + 4:]:
words = line.split()
if len(words) == 0:
break
fx = float(words[-3])
fy = float(words[-2])
fz = float(words[-1])
atom_number = int(words[1]) - 1
forces[atom_number] = (fx, fy, fz)
forces *= units.Ry / units.Bohr
calc = SinglePointCalculator(atoms, energy=energy, forces=forces)
atoms.set_calculator(calc)
return atoms
def read_I_info(fileobj, index=-1):
if isinstance(fileobj, str):
fileobj = open(fileobj)
lines = fileobj.readlines()
del lines[0]
finished = False
s = Atoms()
while not finished:
w = lines.pop(0).split()
if w[0].startswith('"'):
s.append(Atom(w[0].replace('"',''),
[float(w[3]), float(w[4]), float(w[5])]))
else:
finished = True
return s
def read_proteindatabank(fileobj, index=-1):
"""Read PDB files."""
if isinstance(fileobj, basestring):
fileobj = open(fileobj)
images = []
orig = np.identity(3)
trans = np.zeros(3)
atoms = Atoms()
for line in fileobj.readlines():
if line.startswith('CRYST1'):
cellpar = [float(word) for word in line[6:54].split()]
atoms.set_cell(cellpar_to_cell(cellpar))
atoms.pbc = True
for c in range(3):
if line.startswith('ORIGX' + '123'[c]):
pars = [float(word) for word in line[10:55].split()]
orig[c] = pars[:3]
trans[c] = pars[3]
if line.startswith('ATOM') or line.startswith('HETATM'):
try:
# Atom name is arbitrary and does not necessarily
# contain the element symbol. The specification
# requires the element symbol to be in columns 77+78.
symbol = line[76:78].strip().lower().capitalize()
words = line[30:55].split()
position = np.array(
[float(words[0]),
float(words[1]),
float(words[2])])
position = np.dot(orig, position) + trans
atoms.append(Atom(symbol, position))
except Exception as ex:
warnings.warn(
'Discarding atom when reading PDB file: {}'.format(ex))
if line.startswith('ENDMDL'):
images.append(atoms)
atoms = Atoms()
if len(images) == 0:
images.append(atoms)
return images[index]
def read_proteindatabank(fileobj, index=-1):
"""Read PDB files.
The format is assumed to follow the description given in
http://www.wwpdb.org/documentation/format32/sect8.html and
http://www.wwpdb.org/documentation/format32/sect9.html."""
if isinstance(fileobj, str):
fileobj = open(fileobj)
images = []
orig = np.identity(3)
trans = np.zeros(3)
atoms = Atoms()
for line in fileobj.readlines():
if line.startswith('CRYST1'):
cellpar = [float(word) for word in line[6:54].split()]
atoms.set_cell(cellpar_to_cell(cellpar))
for c in range(3):
if line.startswith('ORIGX' + '123'[c]):
pars = [float(word) for word in line[10:55].split()]
orig[c] = pars[:3]
trans[c] = pars[3]
if line.startswith('ATOM') or line.startswith('HETATM'):
try:
# Atom name is arbitrary and does not necessarily
# contain the element symbol. The specification
# requires the element symbol to be in columns 77+78.
symbol = line[76:78].strip().lower().capitalize()
words = line[30:55].split()
position = np.array([float(words[0]),
float(words[1]),
float(words[2])])
position = np.dot(orig, position) + trans
atoms.append(Atom(symbol, position))
except:
pass
if line.startswith('ENDMDL'):
images.append(atoms)
atoms = Atoms()
if len(images) == 0:
images.append(atoms)
return images[index]
def read_pdb(fileobj, index=-1):
"""Read PDB files.
The format is assumed to follow the description given in
http://www.wwpdb.org/documentation/format32/sect8.html and
http://www.wwpdb.org/documentation/format32/sect9.html."""
if isinstance(fileobj, str):
fileobj = open(fileobj)
images = []
orig = np.identity(3)
trans = np.zeros(3)
atoms = Atoms()
for line in fileobj.readlines():
if line.startswith('CRYST1'):
cellpar = [float(word) for word in line[6:54].split()]
atoms.set_cell(cellpar_to_cell(cellpar))
for c in range(3):
if line.startswith('ORIGX' + '123'[c]):
pars = [float(word) for word in line[10:55].split()]
orig[c] = pars[:3]
trans[c] = pars[3]
if line.startswith('ATOM') or line.startswith('HETATM'):
try:
# Atom name is arbitrary and does not necessarily contain the element symbol.
# The specification requires the element symbol to be in columns 77+78.
symbol = line[76:78].strip().lower().capitalize()
words = line[30:55].split()
position = np.array(
[float(words[0]),
float(words[1]),
float(words[2])])
position = np.dot(orig, position) + trans
atoms.append(Atom(symbol, position))
except:
pass
if line.startswith('ENDMDL'):
images.append(atoms)
atoms = Atoms()
if len(images) == 0:
images.append(atoms)
return images[index]
def read_I_info(fileobj, index=-1):
if isinstance(fileobj, str):
fileobj = open(fileobj)
lines = fileobj.readlines()
del lines[0]
finished = False
s = Atoms()
while not finished:
w = lines.pop(0).split()
if w[0].startswith('"'):
position = Bohr * np.array([float(w[3]), float(w[4]), float(w[5])])
s.append(Atom(w[0].replace('"',''), position))
else:
finished = True
return s
def read_proteindatabank(fileobj, index=-1):
"""Read PDB files."""
if isinstance(fileobj, basestring):
fileobj = open(fileobj)
images = []
orig = np.identity(3)
trans = np.zeros(3)
atoms = Atoms()
for line in fileobj.readlines():
if line.startswith('CRYST1'):
cellpar = [float(word) for word in line[6:54].split()]
atoms.set_cell(cellpar_to_cell(cellpar))
atoms.pbc = True
for c in range(3):
if line.startswith('ORIGX' + '123'[c]):
pars = [float(word) for word in line[10:55].split()]
orig[c] = pars[:3]
trans[c] = pars[3]
if line.startswith('ATOM') or line.startswith('HETATM'):
try:
# Atom name is arbitrary and does not necessarily
# contain the element symbol. The specification
# requires the element symbol to be in columns 77+78.
symbol = line[76:78].strip().lower().capitalize()
words = line[30:55].split()
position = np.array([float(words[0]),
float(words[1]),
float(words[2])])
position = np.dot(orig, position) + trans
atoms.append(Atom(symbol, position))
except Exception as ex:
warnings.warn('Discarding atom when reading PDB file: {}'
.format(ex))
if line.startswith('ENDMDL'):
images.append(atoms)
atoms = Atoms()
if len(images) == 0:
images.append(atoms)
return images[index]
def convertXYZtoASE(filename):
with open(filename,'r') as f:
# Read number of atoms (line 4)
n_atoms = string.atoi(string.split(f.readline())[0])
# Read remaining lines
atoms = Atoms()
for line in f.readlines():
if len(line)>5: # Ignore blank lines
data = string.split(line)
if len(data)==4:
symbol = data[0]
position = np.array([string.atof(data[1+j]) for j in range(3)])
atom = Atom(symbol=symbol, position=position)
atoms.append(atom)
if len(atoms) != n_atoms:
print 'SiestaIO.ReadXYZFile: Inconstency in %s detected!' %filename
return atoms
def xyz_to_nuclei_fod(ase_atoms):
# this function is from the pyflosic flosic.py routine
# get the number of atoms + fod
syssize = len(ase_atoms)
# Determine, if fod is included or not.
# A fod is included, if it is no closer then fodeps to another fod.
fodeps = 0.1
included = []
nombernotincluded = 0
dist = -1.0
for i in range(0, syssize):
# Cores are always included.
if ase_atoms[i].symbol != 'X' and ase_atoms[i].symbol != 'He':
included.append(1)
if ase_atoms[i].symbol == 'X' or ase_atoms[i].symbol == 'He':
distold = fodeps
for j in range(0, i):
dist = np.sqrt(
(ase_atoms[i].position[0] - ase_atoms[j].position[0])**2 +
(ase_atoms[i].position[1] - ase_atoms[j].position[1])**2 +
(ase_atoms[i].position[2] - ase_atoms[j].position[2])**2)
if dist < distold and included[j] == 1:
distold = dist
if distold < fodeps:
included.append(0)
nombernotincluded += 1
else:
included.append(1)
# process fods and cores separetely.
tmp = []
nuclei = Atoms([])
fod1 = Atoms([])
fod2 = Atoms([])
nrofnuclei = 0
for i in range(0, syssize):
if ase_atoms[i].symbol != 'X':
nrofnuclei = nrofnuclei + 1
if ase_atoms[i].symbol == 'X':
break
for i in range(0, syssize):
if i < nrofnuclei:
tmp.append(ase_atoms[i].symbol + ' ' +
str(ase_atoms[i].position[0]) + ' ' +
str(ase_atoms[i].position[1]) + ' ' +
str(ase_atoms[i].position[2]) + '\n')
nuclei.append(ase_atoms[i])
elif ase_atoms[i].symbol == 'X':
fod1.append(ase_atoms[i])
if included[i] == 1:
tmp.append('ghost1' + ' ' + str(ase_atoms[i].position[0]) +
' ' + str(ase_atoms[i].position[1]) + ' ' +
str(ase_atoms[i].position[2]) + '\n')
elif ase_atoms[i].symbol == 'He':
fod2.append(ase_atoms[i])
if included[i] == 1:
tmp.append('ghost2' + ' ' + str(ase_atoms[i].position[0]) +
' ' + str(ase_atoms[i].position[1]) + ' ' +
str(ase_atoms[i].position[2]) + '\n')
geo = ''.join(tmp)
return geo, nuclei, fod1, fod2, included
def get_atoms_from_xml_output(filename, schema=None, output=None):
"""
Returns an Atoms object constructed from an XML QE file (according to the schema).
:param filename: Name of the XML file to read from or a file descriptor.
:param schema: Optional XML Schema file to use (for default schema is \
found using the schemaLocation attribute of the XML root).
:param output: Optional dictionary containing the output tree of the XML file. \
If provided skips file access and builds Atoms object directly from output dictionary.
:return: An Atoms object.
"""
if output is None:
output = xmlschema.to_dict(filename,
schema=schema,
path="./qes:espresso/output")
a1 = np.array(output["atomic_structure"]["cell"]["a1"])
a2 = np.array(output["atomic_structure"]["cell"]["a2"])
a3 = np.array(output["atomic_structure"]["cell"]["a3"])
a_p = (output["atomic_structure"]["atomic_positions"]["atom"])
atoms = Atoms()
# First define the unit cell from a1, a2, a3 and alat
cell = np.zeros((3, 3))
cell[0] = a1
cell[1] = a2
cell[2] = a3
atoms.set_cell(cell)
# Now the atoms in the unit cell
for atomx in a_p:
# TODO: extend to all possible cases the symbol splitting (for now, only numbering up to 9 work). Not a very common case...
symbol = split_atomic_symbol(atomx['@name'])[0]
x = float(atomx['$'][0])
y = float(atomx['$'][1])
z = float(atomx['$'][2])
atoms.append(Atom(symbol, (x, y, z)))
return atoms
def read_struct(fname):
"""Read a siesta struct file"""
from ase.atoms import Atoms, Atom
f = open(fname, 'r')
cell = []
for i in range(3):
cell.append([float(x) for x in f.readline().split()])
natoms = int(f.readline())
atoms = Atoms()
for atom in f:
Z, pos_x, pos_y, pos_z = atom.split()[1:]
atoms.append(Atom(int(Z), position = (float(pos_x), float(pos_y), float(pos_z))))
if len(atoms) != natoms:
raise IOError('Badly structured input file')
atoms.set_cell(cell, scale_atoms = True)
return atoms
def read_struct_out(fname):
"""Read a siesta struct file"""
from ase.atoms import Atoms, Atom
f = fname
cell = []
for i in range(3):
cell.append([float(x) for x in f.readline().split()])
natoms = int(f.readline())
atoms = Atoms()
for atom in f:
Z, pos_x, pos_y, pos_z = atom.split()[1:]
atoms.append(Atom(int(Z),
position=(float(pos_x), float(pos_y), float(pos_z))))
if len(atoms) != natoms:
raise IOError('Badly structured input file')
atoms.set_cell(cell, scale_atoms=True)
return atoms
def get_on_opt(f_nuc,f_fod1,f_fod2):
nuc = read(f_nuc)
fod1 = read(f_fod1)
fod2 = read(f_fod2)
geo,nuclei,init_fod1,init_fod2,included =xyz_to_nuclei_fod(nuc)
#write('final0.xyz',fod1,'xyz')
#write('final1.xyz',fod2,'xyz')
#print(nuclei.get_positions())
#print(fod1.get_positions())
comb = Atoms([])
for n in nuclei:
comb.append(n)
for f1 in fod1:
comb.append(f1)
for f2 in fod2:
comb.append(f2)
write('final_ON.xyz',comb,'xyz')
def read_gpaw_text(fileobj, index=-1):
if isinstance(fileobj, str):
fileobj = open(fileobj)
def index_startswith(lines, string):
for i, line in enumerate(lines):
if line.startswith(string):
return i
raise ValueError
lines = fileobj.readlines()
images = []
while True:
try:
i = lines.index('Unit Cell:\n')
except ValueError:
pass
else:
cell = []
pbc = []
for line in lines[i + 3:i + 6]:
words = line.split()
if len(words) == 5: # old format
cell.append(float(words[2]))
pbc.append(words[1] == 'yes')
else: # new format with GUC
cell.append([float(word) for word in words[3:6]])
pbc.append(words[2] == 'yes')
try:
i = lines.index('Positions:\n')
except ValueError:
break
atoms = Atoms(cell=cell, pbc=pbc)
for line in lines[i + 1:]:
words = line.split()
if len(words) != 5:
break
n, symbol, x, y, z = words
symbol = symbol.split('.')[0]
atoms.append(Atom(symbol, [float(x), float(y), float(z)]))
lines = lines[i + 5:]
try:
i = lines.index('-------------------------\n')
except ValueError:
e = None
else:
line = lines[i + 9]
assert line.startswith('Zero Kelvin:')
e = float(line.split()[-1])
try:
ii = index_startswith(lines, 'Fermi Level:')
except ValueError:
eFermi = None
else:
try:
eFermi = float(lines[ii].split()[2])
except ValueError: # we have two Fermi levels
fields = lines[ii].split()
def strip(string):
for rubbish in '[],':
string = string.replace(rubbish, '')
return string
eFermi = [float(strip(fields[2])), float(strip(fields[3]))]
# read Eigenvalues and occupations
ii1 = ii2 = 1e32
try:
ii1 = index_startswith(lines, ' Band Eigenvalues Occupancy')
except ValueError:
pass
try:
ii2 = index_startswith(lines, ' Band Eigenvalues Occupancy')
except ValueError:
pass
ii = min(ii1, ii2)
if ii == 1e32:
kpts = None
else:
ii += 1
words = lines[ii].split()
vals = []
while (len(words) > 2):
vals.append([float(word) for word in words])
ii += 1
words = lines[ii].split()
vals = np.array(vals).transpose()
kpts = [SinglePointKPoint(0, 0)]
kpts[0].eps_n = vals[1]
kpts[0].f_n = vals[2]
if vals.shape[0] > 3:
kpts.append(SinglePointKPoint(0, 1))
kpts[1].eps_n = vals[3]
kpts[1].f_n = vals[4]
# read charge
try:
ii = index_startswith(lines, 'Total Charge:')
except ValueError:
q = None
else:
q = float(lines[ii].split()[2])
# read dipole moment
try:
ii = index_startswith(lines, 'Dipole Moment:')
except ValueError:
dipole = None
else:
line = lines[ii].replace(']', '').replace('[', '')
dipole = np.array([float(c) for c in line.split()[-3:]])
try:
ii = index_startswith(lines, 'Local Magnetic Moments')
except ValueError:
magmoms = None
else:
magmoms = []
for i in range(ii + 1, ii + 1 + len(atoms)):
iii, magmom = lines[i].split()[:2]
magmoms.append(float(magmom))
try:
ii = lines.index('Forces in eV/Ang:\n')
except ValueError:
f = None
else:
f = []
for i in range(ii + 1, ii + 1 + len(atoms)):
try:
x, y, z = lines[i].split()[-3:]
f.append((float(x), float(y), float(z)))
except (ValueError, IndexError), m:
raise IOError('Malformed GPAW log file: %s' % m)
if len(images) > 0 and e is None:
break
if e is not None or f is not None:
calc = SinglePointDFTCalculator(e, f, None, magmoms, atoms, eFermi)
if kpts is not None:
calc.kpts = kpts
if dipole is not None:
calc.set_dipole_moment(dipole)
atoms.set_calculator(calc)
if q is not None and len(atoms) > 0:
n = len(atoms)
atoms.set_charges([q / n] * n)
images.append(atoms)
lines = lines[i:]
def xyz_to_nuclei_fod(ase_atoms):
# Get the number of atoms and fods.
syssize = len(ase_atoms)
# First the preparations for enabling the usage of GHOST atoms are done.
# We need to determine if a FOD is included in the creation of the GHOST atoms or
# not. If two GHOST atoms are created very close to each other (or other atoms) this
# will result in the DFT calculation crashing. A FOD is included if it is no closer
# then fodeps to another FOD or a nuclei.
fodeps = 0.1
included = []
numbernotincluded = 0
dist = -1.0
# Iterate over FODs and nuclei positions.
for i in range(0,syssize):
# Nuclei are always included, obviously. NOTE: FODs for spin0 are given with an
# X, the for spin1 with an He symbol.
if ase_atoms[i].symbol != 'X' and ase_atoms[i].symbol != 'He':
included.append(1)
if ase_atoms[i].symbol == 'X' or ase_atoms[i].symbol == 'He':
# For every FOD the distance to every included FOD and the nuclei is calculated
# in order to determine whether or not it has to be included.
distold = fodeps
for j in range(0,i):
dist = np.sqrt((ase_atoms[i].position[0]-ase_atoms[j].position[0])**2+(ase_atoms[i].position[1]-ase_atoms[j].position[1])**2+(ase_atoms[i].position[2]-ase_atoms[j].position[2])**2)
if dist < distold and included[j]==1:
distold = dist
# If the smallest distance is smaller than fodeps, the FOD will not be included.
# Elsewise it is included.
if distold < fodeps:
included.append(0)
numbernotincluded += 1
else:
included.append(1)
# Now the actual splitting is done.
# These arrays will hold nuclei and FODs (for each spin channel separately).
nuclei = Atoms([])
fod1 = Atoms([])
fod2 = Atoms([])
nrofnuclei = 0
# tmp will be used to create the list that can be used to enable GHOST atoms.
tmp = []
# Determine where to split.
for i in range(0,syssize):
if ase_atoms[i].symbol != 'X':
nrofnuclei = nrofnuclei + 1
# If the first FOD is found, nuclei assigning will be done.
if ase_atoms[i].symbol == 'X':
break
# Split everything.
for i in range(0,syssize):
# Assign the nuclei.
if i < nrofnuclei:
tmp.append(ase_atoms[i].symbol+' '+str(ase_atoms[i].position[0])+' '+str(ase_atoms[i].position[1])+' '+str(ase_atoms[i].position[2])+'\n')
nuclei.append(ase_atoms[i])
# Assing FODs for spin0.
elif ase_atoms[i].symbol == 'X':
fod1.append(ase_atoms[i])
if included[i] == 1:
tmp.append('ghost1'+' '+str(ase_atoms[i].position[0])+' '+str(ase_atoms[i].position[1])+' '+str(ase_atoms[i].position[2])+'\n')
# Assign FODs for spin1.
elif ase_atoms[i].symbol == 'He':
fod2.append(ase_atoms[i])
if included[i] == 1:
tmp.append('ghost2'+' '+str(ase_atoms[i].position[0])+' '+str(ase_atoms[i].position[1])+' '+str(ase_atoms[i].position[2])+'\n')
# geo holds both nuclei and GHOST atoms at FOD positions.
geo = ''.join(tmp)
# geo can now be used for calculation with GHOST atoms, nuclei and fod1/fod2 for
# every other calculation.
return geo,nuclei,fod1,fod2,included
def xyz_to_database(xyzfile,
num_examples=None,
xyz_format='xyz',
verbose=True,
unit_to_ev=None,
restart=False):
"""
Convert the xyz file to an `ase.db.core.Database`.
Args:
xyzfile: a `str` as the file to parse.
num_examples: a `int` as the maximum number of examples to parse. If None,
all examples in the given file will be saved.
xyz_format: a `str` representing the format of the given xyz file.
verbose: a `bool` indicating whether we should log the parsing progress.
unit_to_ev: a `float` as the unit for converting energies to eV. Defaults
to None so that default units will be used.
restart: a `bool`. If True, the database will be re-built even if already
existed.
Returns:
database: an `ase.db.core.Database`.
auxdict: a `dict` as the auxiliary dict for this database.
"""
if xyz_format.lower() == 'ase':
formatter = _ase_xyz
else:
formatter = _xyz
dbfile = "{}.db".format(splitext(xyzfile)[0])
if isfile(dbfile):
if restart:
remove(dbfile)
else:
auxdict = _load_auxiliary_dict(dbfile)
return connect(name=dbfile), auxdict
unit = unit_to_ev or formatter.default_unit
parse_forces = formatter.parse_forces
count = 0
ai = 0
natoms = 0
stage = 0
atoms = None
num_examples = num_examples or 0
natoms_counter = Counter()
y_min = np.inf
y_max = -np.inf
max_occurs = Counter()
database = connect(name=dbfile)
tic = time.time()
if verbose:
sys.stdout.write("Extract cartesian coordinates ...\n")
with open(xyzfile) as f:
for line in f:
if num_examples and count == num_examples:
break
line = line.strip()
if line == "":
continue
if stage == 0:
if line.isdigit():
natoms = int(line)
atoms = Atoms(calculator=ProvidedCalculator())
if parse_forces:
atoms.info['provided_forces'] = np.zeros((natoms, 3))
natoms_counter[natoms] += 1
stage += 1
elif stage == 1:
m = formatter.energy_patt.search(line)
if m:
if xyz_format.lower() == 'extxyz':
energy = float(m.group(3)) * unit
elif xyz_format.lower() == 'ase':
energy = float(m.group(2)) * unit
atoms.set_cell(
np.reshape([float(x) for x in m.group(1).split()],
(3, 3)))
atoms.set_pbc([
True if x == "T" else False
for x in m.group(3).split()
])
else:
energy = float(m.group(1)) * unit
atoms.info['provided_energy'] = energy
y_min = min(y_min, energy)
y_max = max(y_max, energy)
stage += 1
elif stage == 2:
m = formatter.string_patt.search(line)
if m:
atoms.append(
Atom(symbol=m.group(1),
position=[float(v) for v in m.groups()[1:4]]))
if parse_forces:
atoms.info['provided_forces'][ai, :] = [
float(v) * unit for v in m.groups()[4:7]
]
ai += 1
if ai == natoms:
atoms.calc.calculate()
database.write(atoms)
counter = Counter(atoms.get_chemical_symbols())
for symbol, n in counter.items():
max_occurs[symbol] = max(max_occurs[symbol], n)
ai = 0
stage = 0
count += 1
if verbose and count % 1000 == 0:
sys.stdout.write(
"\rProgress: {:7d} / {:7d} | Speed = {:.1f}".
format(count, num_examples,
count / (time.time() - tic)))
if verbose:
print("")
print("Total time: %.3f s\n" % (time.time() - tic))
# Dump the auxiliary dict.
auxdict = {
"max_occurs": dict(max_occurs),
"y_range": [y_min, y_max],
"natoms_counter": dict(natoms_counter)
}
_save_auxiliary_dict(dbfile, auxdict)
return database, auxdict
def lewis2xyz(lewis,f='SMILES',aromatic=False,dim=2,addH=True):
# Input
# lewis ... smiles format string
# aromatic ... is the system aromatic
# dim ... 2d == 2 and 3d == 3
# Output
# Bonding information
# Lewis pictures
# sdf and xyz files
if f == 'pdb':
m = Chem.MolFromPDBFile(lewis)
if f == 'xyz':
struct = read(lewis)
write('lewis.pdb',struct,'proteindatabank')
m = Chem.MolFromPDBFile(lewis)
if f == 'SMILES':
m = Chem.MolFromSmiles(lewis)
# Change armatic bonds to alternate single and double bonds
if aromatic == True:
if m.GetBondWithIdx(0).GetIsAromatic() == True:
Chem.Kekulize(m)
if addH == True:
m=Chem.AddHs(m)
if dim == 1 or dim == 2:
AllChem.Compute2DCoords(m)
if dim == 3:
AllChem.EmbedMolecule(m)
AllChem.UFFOptimizeMolecule(m)
fods = Atoms()
# core electrons
for atom in m.GetAtoms():
idx = atom.GetIdx()
# atomic number
N = m.GetAtomWithIdx(idx).GetAtomicNum()
# number of valence electrons
V = m.GetAtomWithIdx(idx).GetTotalValence()
# number of radicals
R = atom.GetNumRadicalElectrons()
# formal charge of the atom
F = atom.GetFormalCharge()
# number of core electrons
C = N-V-R-F
# lone pairs
#L = 0
#bonds = atom.GetBonds()
#for b in range(len(bonds)):
# bond_type = bonds[b].GetBondType()
# if str(bond_type) == 'SINGLE':
# L = L + 1
# if str(bond_type) == 'DOUBLE':
# L = L + 2
# if str(bond_type) == 'TRIPLE':
# L = L + 3
#print(V-L)
if C > 0:
pos = m.GetConformer().GetAtomPosition(idx)
offset = 0.002
if C == 1:
fods.append(Atom('X',[pos[0],pos[1],pos[2]]))
if C == 2:
fods.append(Atom('X',[pos[0],pos[1],pos[2]]))
fods.append(Atom('He',[pos[0]+offset,pos[1]+offset,pos[2]+offset]))
if C == 3:
fods.append(Atom('X',[pos[0],pos[1],pos[2]]))
fods.append(Atom('He',[pos[0]+offset,pos[1]+offset,pos[2]+offset]))
fods.append(Atom('X',[pos[0]+offset,pos[1]+offset,pos[2]+4*offset]))
if C == 4:
fods.append(Atom('X',[pos[0],pos[1],pos[2]]))
fods.append(Atom('He',[pos[0]+offset,pos[1]+offset,pos[2]+offset]))
fods.append(Atom('X',[pos[0]+offset,pos[1]+offset,pos[2]+40*offset]))
fods.append(Atom('He',[pos[0]+offset,pos[1]+offset,pos[2]-40*offset]))
if C == 5:
fods.append(Atom('X',[pos[0],pos[1],pos[2]]))
fods.append(Atom('He',[pos[0]+offset,pos[1]+offset,pos[2]+offset]))
fods.append(Atom('X',[pos[0]+offset,pos[1]+offset,pos[2]+40*offset]))
fods.append(Atom('He',[pos[0]+offset,pos[1]+offset,pos[2]-40*offset]))
fods.append(Atom('X',[pos[0]+40*offset,pos[1]+40*offset,pos[2]+offset]))
if C == 6:
fods.append(Atom('X',[pos[0],pos[1],pos[2]]))
fods.append(Atom('He',[pos[0]+offset,pos[1]+offset,pos[2]+offset]))
fods.append(Atom('X',[pos[0]+40*offset,pos[1]+40*offset,pos[2]+40*offset]))
fods.append(Atom('He',[pos[0]+offset,pos[1]-40*offset,pos[2]-40*offset]))
fods.append(Atom('X',[pos[0]-40*offset,pos[1]+40*offset,pos[2]-40*offset]))
fods.append(Atom('He',[pos[0]-40*offset,pos[1]-40*offset,pos[2]+40*offset]))
if C == 7:
fods.append(Atom('X',[pos[0],pos[1],pos[2]]))
fods.append(Atom('He',[pos[0]+offset,pos[1]+offset,pos[2]+offset]))
fods.append(Atom('X',[pos[0]+40*offset,pos[1]+40*offset,pos[2]+40*offset]))
fods.append(Atom('He',[pos[0]+offset,pos[1]-40*offset,pos[2]-40*offset]))
fods.append(Atom('X',[pos[0]-40*offset,pos[1]+40*offset,pos[2]-40*offset]))
fods.append(Atom('He',[pos[0]-40*offset,pos[1]-40*offset,pos[2]+40*offset]))
fods.append(Atom('He',[pos[0]+42*offset,pos[1]+42*offset,pos[2]+42*offset]))
if C == 8:
fods.append(Atom('X',[pos[0],pos[1],pos[2]]))
fods.append(Atom('He',[pos[0]+offset,pos[1]+offset,pos[2]+offset]))
fods.append(Atom('X',[pos[0]+40*offset,pos[1]+40*offset,pos[2]+40*offset]))
fods.append(Atom('He',[pos[0]+40*offset,pos[1]-40*offset,pos[2]-40*offset]))
fods.append(Atom('X',[pos[0]-40*offset,pos[1]+40*offset,pos[2]-40*offset]))
fods.append(Atom('He',[pos[0]-40*offset,pos[1]-40*offset,pos[2]+40*offset]))
fods.append(Atom('He',[pos[0]+42*offset,pos[1]+42*offset,pos[2]+42*offset]))
fods.append(Atom('X',[pos[0]+42*offset,pos[1]-42*offset,pos[2]-42*offset]))
# find radicals
for atom in m.GetAtoms():
rad = atom.GetNumRadicalElectrons()
#print(rad)
if rad == 1:
idx = atom.GetIdx()
pos = m.GetConformer().GetAtomPosition(idx)
bonds = atom.GetBonds()
vec_BA_x = 0
vec_BA_y = 0
vec_BA_z = 0
BA = []
for b in range(len(bonds)):
ai = bonds[b].GetBeginAtomIdx()
aj = bonds[b].GetEndAtomIdx()
bond_type = bonds[b].GetBondType()
# Positions
pos_ai = m.GetConformer().GetAtomPosition(ai)
pos_aj = m.GetConformer().GetAtomPosition(aj)
# adding bonding vectors for 3d
vec_BA_x = vec_BA_x + (pos_aj[0]-pos_ai[0])
vec_BA_y = vec_BA_y + (pos_aj[1]-pos_ai[1])
vec_BA_z = vec_BA_z + (pos_aj[2]-pos_ai[2])
# save bonding vectors for 2d
BA.append(np.array([(pos_aj[0]-pos_ai[0]),(pos_aj[1]-pos_ai[1]),(pos_aj[2]-pos_ai[2])]))
if dim == 1:
fods.append(Atom('X',[-1*BA[0][0],-1*BA[0][1],-1*BA[0][2]]))
if dim == 2:
vec = np.cross(BA[0],BA[1])
norm = np.sqrt(vec[0]**2 + vec[1]**2 + vec[2]**2)
pos = vec/norm
fods.append(Atom('X',[pos[0],pos[1],pos[2]]))
if dim == 3:
norm = np.sqrt(vec_BA_x**2 + vec_BA_y**2 + vec_BA_z**2)
pos = np.array([-1*vec_BA_x,-1*vec_BA_y,-1*vec_BA_z])/norm
fods.append(Atom('X',[pos[0],pos[1],pos[2]]))
bonds = m.GetBonds()
for b in range(len(bonds)):
ai = bonds[b].GetBeginAtomIdx()
aj = bonds[b].GetEndAtomIdx()
bond_type = bonds[b].GetBondType()
# Positions
pos_ai = m.GetConformer().GetAtomPosition(ai)
pos_aj = m.GetConformer().GetAtomPosition(aj)
# Bondlength
#print('%0.5f' % np.linalg.norm([pos_ai[0]-pos_aj[0],pos_ai[1]-pos_aj[1],pos_ai[2]-pos_aj[2]]))
# Middle of the vector
MP = [(pos_ai[0]+pos_aj[0])/2.,(pos_ai[1]+pos_aj[1])/2.,(pos_ai[2]+pos_aj[2])/2.]
#print('%0.5f %0.5f %0.5f' %(MP[0],MP[1],MP[2]))
# Print positions
#print('%0.5f %0.5f %0.5f' %(pos_ai[0],pos_ai[1],pos_ai[2]))
#print('%0.5f %0.5f %0.5f' %(pos_aj[0],pos_aj[1],pos_aj[2]))
for spin in ['He','X']:
if spin == 'He':
offset = 0.000
if spin == 'X':
#offset = 0.002
offset = 0.02
if str(bond_type) == 'SINGLE':
#print('%s %0.5f %0.5f %0.5f' %(spin,MP[0]+offset,MP[1]+offset,MP[2]+offset))
fods.append(Atom(spin,[MP[0]+offset,MP[1]+offset,MP[2]+offset]))
if str(bond_type) == 'DOUBLE':
#print('%s %0.5f %0.5f %0.5f' %(spin,MP[0]+offset,MP[1]+offset,MP[2]+0.5+offset))
#print('%s %0.5f %0.5f %0.5f' %(spin,MP[0]+offset,MP[1]+offset,MP[2]-0.5+offset))
fods.append(Atom(spin,[MP[0]+offset,MP[1]+offset,MP[2]+0.5+offset]))
fods.append(Atom(spin,[MP[0]+offset,MP[1]+offset,MP[2]-0.5+offset]))
if str(bond_type) == 'TRIPLE':
#print('%s %0.5f %0.5f %0.5f' %(spin,MP[0]+offset,MP[1]+offset,MP[2]+0.5+offset))
#print('%s %0.5f %0.5f %0.5f' %(spin,MP[0]+offset,MP[1]+offset,MP[2]-0.5+offset))
#print('%s %0.5f %0.5f %0.5f' %(spin,MP[0]-0.5+offset,MP[1]+offset,MP[2]+offset))
fods.append(Atom(spin,[MP[0]+offset,MP[1]+offset,MP[2]+0.5+offset]))
fods.append(Atom(spin,[MP[0]+offset,MP[1]+offset,MP[2]-0.5+offset]))
fods.append(Atom(spin,[MP[0]-0.5+offset,MP[1]+offset,MP[2]+offset]))
w = Chem.SDWriter('%s.sdf' % 'pylewis')
w.write(m)
w.flush()
# sdf2xyz
struct = read('%s.sdf' % 'pylewis')
struct_fods = Atoms()
struct_fods.extend(struct)
struct_fods.extend(fods)
#print(struct.get_chemical_symbols())
write('%s.xyz' % 'pylewis',struct,'xyz')
write('%s_fods.xyz' % 'pylewis',struct_fods,'xyz')
def read_gpaw_text(fileobj, index=-1):
if isinstance(fileobj, str):
fileobj = open(fileobj)
def index_startswith(lines, string):
for i, line in enumerate(lines):
if line.startswith(string):
return i
raise ValueError
lines = fileobj.readlines()
images = []
while True:
try:
i = lines.index('Unit Cell:\n')
except ValueError:
pass
else:
cell = []
pbc = []
for line in lines[i + 3:i + 6]:
words = line.split()
if len(words) == 5: # old format
cell.append(float(words[2]))
pbc.append(words[1] == 'yes')
else: # new format with GUC
cell.append([float(word) for word in words[3:6]])
pbc.append(words[2] == 'yes')
try:
i = lines.index('Positions:\n')
except ValueError:
break
atoms = Atoms(cell=cell, pbc=pbc)
for line in lines[i + 1:]:
words = line.split()
if len(words) != 5:
break
n, symbol, x, y, z = words
symbol = symbol.split('.')[0]
atoms.append(Atom(symbol, [float(x), float(y), float(z)]))
lines = lines[i + 5:]
try:
i = lines.index('-------------------------\n')
except ValueError:
e = None
else:
line = lines[i + 9]
assert line.startswith('Zero Kelvin:')
e = float(line.split()[-1])
try:
ii = index_startswith(lines, 'Fermi Level:')
except ValueError:
eFermi = None
else:
try:
eFermi = float(lines[ii].split()[2])
except ValueError: # we have two Fermi levels
fields = lines[ii].split()
def strip(string):
for rubbish in '[],':
string = string.replace(rubbish, '')
return string
eFermi = [float(strip(fields[2])),
float(strip(fields[3])) ]
# read Eigenvalues and occupations
ii1 = ii2 = 1e32
try:
ii1 = index_startswith(lines, ' Band Eigenvalues Occupancy')
except ValueError:
pass
try:
ii2 = index_startswith(lines, ' Band Eigenvalues Occupancy')
except ValueError:
pass
ii = min(ii1, ii2)
if ii == 1e32:
kpts = None
else:
ii += 1
words = lines[ii].split()
vals = []
while(len(words) > 2):
vals.append([float(word) for word in words])
ii += 1
words = lines[ii].split()
vals = np.array(vals).transpose()
kpts = [SinglePointKPoint(0, 0)]
kpts[0].eps_n = vals[1]
kpts[0].f_n = vals[2]
if vals.shape[0] > 3:
kpts.append(SinglePointKPoint(0, 1))
kpts[1].eps_n = vals[3]
kpts[1].f_n = vals[4]
# read charge
try:
ii = index_startswith(lines, 'Total Charge:')
except ValueError:
q = None
else:
q = float(lines[ii].split()[2])
try:
ii = index_startswith(lines, 'Local Magnetic Moments')
except ValueError:
magmoms = None
else:
magmoms = []
for i in range(ii + 1, ii + 1 + len(atoms)):
iii, magmom = lines[i].split()[:2]
magmoms.append(float(magmom))
try:
ii = lines.index('Forces in eV/Ang:\n')
except ValueError:
f = None
else:
f = []
for i in range(ii + 1, ii + 1 + len(atoms)):
try:
x, y, z = lines[i].split()[-3:]
f.append((float(x), float(y), float(z)))
except (ValueError, IndexError), m:
raise IOError('Malformed GPAW log file: %s' % m)
if len(images) > 0 and e is None:
break
if e is not None or f is not None:
calc = SinglePointDFTCalculator(e, f, None, magmoms, atoms, eFermi)
if kpts is not None:
calc.kpts = kpts
atoms.set_calculator(calc)
if q is not None and len(atoms) > 0:
n = len(atoms)
atoms.set_charges([q / n] * n)
images.append(atoms)
lines = lines[i:]
def make_Td(names, sizes):
#work out the expansion factor
factor = sum(sizes)
#factor = 3.0
Td = Atoms()
Td.append(Atom('F', position=(0, 0, 0)))
Td.append(Atom('N', position=(0, 0, factor)))
#Atom 3
Td.append(Atom('N', position=(1.0, 0, 0.0)))
Td.set_angle([1, 0, 2], 1.91062939)
Td.set_distance(0, 2, factor, fix=0)
#Atom 4
Td.append(Atom('N', position=(1.0, 0, 0.0)))
Td.set_distance(0, 3, factor, fix=0)
Td.set_angle([1, 0, 3], 1.91062939, mask=None)
Td.set_dihedral([2, 1, 0, 3], 2.0943951, mask=None)
#print Td.get_angle([1,0,2])
#Atom 5
Td.append(Atom('N', position=(1.0, 0, 0)))
Td.set_angle([1, 0, 4], 1.91062939)
Td.set_dihedral([2, 1, 0, 4], -2.0943951)
Td.set_distance(0, 4, factor, fix=0)
#find side centroids
tmp1 = Td[1:4]
coside = tmp1.get_center_of_mass()
Td.append(Atom('C', position=coside))
#
tmp1 = Td[2:5]
coside = tmp1.get_center_of_mass()
Td.append(Atom('C', position=coside))
#
tmp1 = Td[3:5]
tmp1.append(Td[1])
coside = tmp1.get_center_of_mass()
Td.append(Atom('C', position=coside))
#
tmp1 = Td[1:3]
tmp1.append(Td[4])
coside = tmp1.get_center_of_mass()
Td.append(Atom('C', position=coside))
write('td.xyz', Td)
# Now we make tags etc
## C -> triangles (faces), N => triangle caps
eps = 0.05
model_molecule = {}
n_obj = -1 #initialise to -1 such that it can be indexed at start of add
n_tags = 0
#First detect global bonding
cov_rad = []
for atom in Td:
#cov_rad.append(covalent_radii[atom.number])
cov_rad.append(factor / 2)
nlist = NeighborList(cov_rad,
skin=0.1,
self_interaction=False,
bothways=True)
nlist.build(Td)
nbond = 1
bond_matrix = np.zeros((len(Td), len(Td)))
pbc_bond_matrix = np.zeros((len(Td), len(Td)))
bond_dict = {}
for atom in Td:
print "---------------------"
indices, offsets = nlist.get_neighbors(atom.index)
for index, offset in zip(indices, offsets):
print atom.index, index, atom.symbol, Td[
index].symbol, offset, Td.get_distance(atom.index,
index,
mic=True)
if atom.index < index:
this_bond = [atom.index, index]
for o in offset:
this_bond.append(o)
bond = tuple(this_bond)
print bond
bond_dict[bond] = nbond
#Now we need to find the same bond the other direction to get the offsets right
indices2, offsets2 = nlist.get_neighbors(index)
for i2, o2 in zip(indices2, offsets2):
if i2 == atom.index:
#print "sum of offsets = ", offset, o2, sum(offset + o2)
if sum(offset + o2) == 0: #same bond
this_bond = [index, atom.index]
for o in o2:
this_bond.append(o)
bond = tuple(this_bond)
print bond
bond_dict[bond] = nbond
nbond += 1
print nbond
for k, v in bond_dict.items():
print k, v
#Now we look for all the things with unit*factor bondlengths
#Start with N (rectangles)
for atom in Td:
if atom.symbol == 'N':
print '======================================='
print 'N Atom ', atom.index
n_obj += 1
model_molecule[n_obj] = Atoms()
model_molecule[n_obj] += atom
model_molecule[n_obj][0].original_index = atom.index
model_molecule[n_obj][0].symbol = 'F'
indices, offsets = nlist.get_neighbors(atom.index)
#Just checking the symbols here
print indices
print offsets
symbols = ([Td[index].symbol for index in indices])
symbol_string = ''.join(
sorted([Td[index].symbol for index in indices]))
#print symbol_string
#for i,o in zip(indices, offsets):
# print i,o
for index, offset in zip(indices, offsets):
print atom.index, index, offset
this_bond = [atom.index, index]
for o in offset:
this_bond.append(o)
bond = tuple(this_bond)
print bond_dict[bond]
if Td[index].symbol == 'C':
#By definition, we can't be going over a periodic boundary (no pbc)
model_molecule[n_obj] += Td[index]
model_molecule[n_obj][-1].tag = bond_dict[bond]
model_molecule[n_obj].positions[-1] = Td.positions[index]
n_centers = n_obj
##Now we do the C (triangles)
for atom in Td:
if atom.symbol == 'C':
#print'======================================='
#print 'N Atom ',atom.index, " finding squares1"
n_obj += 1
model_molecule[n_obj] = Atoms()
model_molecule[n_obj] += atom
indices, offsets = nlist.get_neighbors(atom.index)
for index, offset in zip(indices, offsets):
print index, offset, Td[index].symbol
this_bond = [atom.index, index]
for o in offset:
this_bond.append(o)
bond = tuple(this_bond)
#if not bond_dict.has_key(bond):
#Then
if Td[index].symbol == 'N':
#By definition, we can't be going over a periodic boundary (no pbc)
model_molecule[n_obj] += Td[index]
model_molecule[n_obj][-1].tag = bond_dict[bond]
model_molecule[n_obj].positions[-1] = Td.positions[index]
f = open('Td.model', 'w')
g = open('control-mofgen.txt', 'w')
#Just for checking, now lets gather everything in the model_molecule dict into one big thing and print it
test_mol = Atoms()
for obj in model_molecule:
test_mol += model_molecule[obj]
write('test_model.xyz', test_mol)
#print n_centers, n_model, n_obj
#Headers and cell
f.write('%-20s %-3d\n' % ('Number of objects =', n_obj + 1))
f.write('%-20s\n' % ('build = systre'))
f.write('%5s\n' % ('Cell:'))
f.write('%8.3f %8.3f %8.3f \n' % (0.0, 0.0, 0.0))
f.write('%8.3f %8.3f %8.3f \n' % (0.0, 0.0, 0.0))
f.write('%8.3f %8.3f %8.3f \n' % (0.0, 0.0, 0.0))
g.write('%-20s\n' % ('model = Td'))
for obj in xrange(n_centers + 1):
f.write('\n%-8s %-3d\n' % ('Center: ', obj + 1))
f.write('%-3d\n' % (len(model_molecule[obj])))
f.write('%-20s\n' % ('type = triangle'))
#process positions to make it a bit more ideal
for atom in model_molecule[obj]:
#if atom.symbol == 'C':
if atom.index != 0:
model_molecule[obj].set_distance(atom.index, 0, 1.0, fix=1)
for atom in model_molecule[obj]:
(x, y, z) = atom.position
#print atom.symbol, atom.position, atom.tag
if atom.tag:
f.write('%-2s %15.8f %15.8f %15.8f %-4s\n' %
('X', x, y, z, atom.tag))
#f.write('%-2s %15.8f %15.8f %15.8f %-4s\n' % (atom.symbol, x, y, z, atom.tag))
else:
f.write('%-2s %15.8f %15.8f %15.8f\n' % ('Q', x, y, z))
#f.write('%-2s %15.8f %15.8f %15.8f\n' % (atom.symbol, x, y, z))
g.write('%-9s %-50s\n' % ('center =', names[0]))
for obj in xrange(n_centers + 1, n_obj + 1):
f.write('\n%-8s %-3d\n' % ('Linker: ', obj - n_centers))
f.write('%-3d\n' % (len(model_molecule[obj])))
f.write('%-20s\n' % ('type = triangle'))
#process positions to make it a bit more ideal
for atom in model_molecule[obj]:
#if atom.symbol == 'C':
if atom.index != 0:
model_molecule[obj].set_distance(atom.index, 0, 1.0, fix=1)
for atom in model_molecule[obj]:
(x, y, z) = atom.position
#print atom.symbol, atom.position, atom.tag
if atom.tag:
f.write('%-2s %15.8f %15.8f %15.8f %-4s\n' %
('X', x, y, z, atom.tag))
#f.write('%-2s %15.8f %15.8f %15.8f %-4s\n' % (atom.symbol, x, y, z, atom.tag))
else:
f.write('%-2s %15.8f %15.8f %15.8f\n' % ('Q', x, y, z))
#f.write('%-2s %15.8f %15.8f %15.8f\n' % (atom.symbol, x, y, z))
g.write('%-9s %-50s\n' % ('linker =', names[1]))
test_mol = Atoms()
for obj in model_molecule:
test_mol += model_molecule[obj]
write('test_model2.xyz', test_mol)
def read_gpaw_text(fileobj, index=-1):
if isinstance(fileobj, str):
fileobj = open(fileobj)
def index_startswith(lines, string):
for i, line in enumerate(lines):
if line.startswith(string):
return i
raise ValueError
lines = fileobj.readlines()
images = []
while True:
try:
i = lines.index('Unit Cell:\n')
except ValueError:
pass
else:
cell = []
pbc = []
for line in lines[i + 3:i + 6]:
words = line.split()
if len(words) == 5: # old format
cell.append(float(words[2]))
pbc.append(words[1] == 'yes')
else: # new format with GUC
cell.append([float(word) for word in words[3:6]])
pbc.append(words[2] == 'yes')
try:
i = lines.index('Positions:\n')
except ValueError:
break
atoms = Atoms(cell=cell, pbc=pbc)
for line in lines[i + 1:]:
words = line.split()
if len(words) != 5:
break
n, symbol, x, y, z = words
symbol = symbol.split('.')[0]
atoms.append(Atom(symbol, [float(x), float(y), float(z)]))
lines = lines[i + 5:]
try:
i = lines.index('-------------------------\n')
except ValueError:
e = None
else:
line = lines[i + 9]
assert line.startswith('Zero Kelvin:')
e = float(line.split()[-1])
try:
ii = index_startswith(lines, 'Fermi Level:')
except ValueError:
eFermi = None
else:
try:
eFermi = float(lines[ii].split()[2])
except ValueError: # we have two Fermi levels
fields = lines[ii].split()
def strip(string):
for rubbish in '[],':
string = string.replace(rubbish, '')
return string
eFermi = [float(strip(fields[2])), float(strip(fields[3]))]
try:
ii = index_startswith(lines, 'Total Charge:')
except ValueError:
q = None
else:
q = float(lines[ii].split()[2])
try:
ii = index_startswith(lines, 'Local Magnetic Moments')
except ValueError:
magmoms = None
else:
magmoms = []
for i in range(ii + 1, ii + 1 + len(atoms)):
iii, magmom = lines[i].split()[:2]
magmoms.append(float(magmom))
try:
ii = lines.index('Forces in eV/Ang:\n')
except ValueError:
f = None
else:
f = []
for i in range(ii + 1, ii + 1 + len(atoms)):
try:
x, y, z = lines[i].split()[-3:]
f.append((float(x), float(y), float(z)))
except (ValueError, IndexError), m:
raise IOError('Malformed GPAW log file: %s' % m)
if len(images) > 0 and e is None:
break
if e is not None or f is not None:
calc = SinglePointDFTCalculator(e, f, None, magmoms, atoms, eFermi)
atoms.set_calculator(calc)
if q is not None:
n = len(atoms)
atoms.set_charges([q / n] * n)
images.append(atoms)
lines = lines[i:]
def make_Td(names,sizes):
#work out the expansion factor
factor = sum(sizes)
#factor = 3.0
Td = Atoms()
Td.append(Atom('F', position=(0,0,0)))
Td.append(Atom('N', position=(0,0,factor)))
#Atom 3
Td.append(Atom('N', position=(1.0,0,0.0)))
Td.set_angle([1,0,2],1.91062939)
Td.set_distance(0,2, factor, fix=0)
#Atom 4
Td.append(Atom('N', position=(1.0,0,0.0)))
Td.set_distance(0,3, factor, fix=0)
Td.set_angle([1,0,3],1.91062939, mask=None)
Td.set_dihedral([2,1,0,3],2.0943951, mask=None)
#print Td.get_angle([1,0,2])
#Atom 5
Td.append(Atom('N', position=(1.0,0,0)))
Td.set_angle([1,0,4],1.91062939)
Td.set_dihedral([2,1,0,4],-2.0943951)
Td.set_distance(0,4, factor, fix=0)
#find side centroids
tmp1 = Td[1:4]
coside = tmp1.get_center_of_mass()
Td.append(Atom('C', position = coside))
#
tmp1 = Td[2:5]
coside = tmp1.get_center_of_mass()
Td.append(Atom('C', position = coside))
#
tmp1 = Td[3:5]
tmp1.append(Td[1])
coside = tmp1.get_center_of_mass()
Td.append(Atom('C', position = coside))
#
tmp1 = Td[1:3]
tmp1.append(Td[4])
coside = tmp1.get_center_of_mass()
Td.append(Atom('C', position = coside))
write('td.xyz',Td)
# Now we make tags etc
## C -> triangles (faces), N => triangle caps
eps = 0.05
model_molecule={}
n_obj = -1 #initialise to -1 such that it can be indexed at start of add
n_tags = 0
#First detect global bonding
cov_rad=[]
for atom in Td:
#cov_rad.append(covalent_radii[atom.number])
cov_rad.append(factor / 2)
nlist = NeighborList(cov_rad,skin=0.1,self_interaction=False,bothways=True)
nlist.build(Td)
nbond = 1
bond_matrix = np.zeros( (len(Td),len(Td)) )
pbc_bond_matrix = np.zeros( (len(Td),len(Td)) )
bond_dict = {}
for atom in Td:
print "---------------------"
indices, offsets = nlist.get_neighbors(atom.index)
for index,offset in zip(indices,offsets):
print atom.index, index, atom.symbol, Td[index].symbol, offset, Td.get_distance(atom.index,index,mic=True)
if atom.index < index:
this_bond = [atom.index, index]
for o in offset:
this_bond.append(o)
bond = tuple(this_bond)
print bond
bond_dict[bond] = nbond
#Now we need to find the same bond the other direction to get the offsets right
indices2,offsets2 = nlist.get_neighbors(index)
for i2,o2 in zip(indices2,offsets2):
if i2 == atom.index:
#print "sum of offsets = ", offset, o2, sum(offset + o2)
if sum(offset + o2) == 0: #same bond
this_bond = [index, atom.index]
for o in o2:
this_bond.append(o)
bond = tuple(this_bond)
print bond
bond_dict[bond] = nbond
nbond +=1
print nbond
for k,v in bond_dict.items():
print k,v
#Now we look for all the things with unit*factor bondlengths
#Start with N (rectangles)
for atom in Td:
if atom.symbol == 'N':
print'======================================='
print 'N Atom ',atom.index
n_obj+=1
model_molecule[n_obj] = Atoms()
model_molecule[n_obj] += atom
model_molecule[n_obj][0].original_index = atom.index
model_molecule[n_obj][0].symbol = 'F'
indices, offsets = nlist.get_neighbors(atom.index)
#Just checking the symbols here
print indices
print offsets
symbols = ([Td[index].symbol for index in indices])
symbol_string = ''.join(sorted([Td[index].symbol for index in indices]))
#print symbol_string
#for i,o in zip(indices, offsets):
# print i,o
for index,offset in zip(indices,offsets):
print atom.index, index, offset
this_bond = [atom.index, index]
for o in offset:
this_bond.append(o)
bond = tuple(this_bond)
print bond_dict[bond]
if Td[index].symbol == 'C':
#By definition, we can't be going over a periodic boundary (no pbc)
model_molecule[n_obj] += Td[index]
model_molecule[n_obj][-1].tag = bond_dict[bond]
model_molecule[n_obj].positions[-1] = Td.positions[index]
n_centers = n_obj
##Now we do the C (triangles)
for atom in Td:
if atom.symbol == 'C':
#print'======================================='
#print 'N Atom ',atom.index, " finding squares1"
n_obj+=1
model_molecule[n_obj] = Atoms()
model_molecule[n_obj] += atom
indices, offsets = nlist.get_neighbors(atom.index)
for index,offset in zip(indices,offsets):
print index, offset, Td[index].symbol
this_bond = [atom.index, index]
for o in offset:
this_bond.append(o)
bond = tuple(this_bond)
#if not bond_dict.has_key(bond):
#Then
if Td[index].symbol == 'N':
#By definition, we can't be going over a periodic boundary (no pbc)
model_molecule[n_obj] += Td[index]
model_molecule[n_obj][-1].tag = bond_dict[bond]
model_molecule[n_obj].positions[-1] = Td.positions[index]
f = open('Td.model','w')
g = open('control-mofgen.txt','w')
#Just for checking, now lets gather everything in the model_molecule dict into one big thing and print it
test_mol = Atoms()
for obj in model_molecule:
test_mol += model_molecule[obj]
write('test_model.xyz',test_mol)
#print n_centers, n_model, n_obj
#Headers and cell
f.write('%-20s %-3d\n' %('Number of objects =',n_obj+1))
f.write('%-20s\n' %('build = systre'))
f.write('%5s\n' %('Cell:'))
f.write('%8.3f %8.3f %8.3f \n' % (0.0, 0.0, 0.0))
f.write('%8.3f %8.3f %8.3f \n' % (0.0, 0.0, 0.0))
f.write('%8.3f %8.3f %8.3f \n' % (0.0, 0.0, 0.0))
g.write('%-20s\n' %('model = Td'))
for obj in xrange(n_centers+1):
f.write('\n%-8s %-3d\n' %('Center: ', obj+1))
f.write('%-3d\n' %(len(model_molecule[obj])))
f.write('%-20s\n' %('type = triangle'))
#process positions to make it a bit more ideal
for atom in model_molecule[obj]:
#if atom.symbol == 'C':
if atom.index != 0:
model_molecule[obj].set_distance(atom.index,0,1.0,fix=1)
for atom in model_molecule[obj]:
(x,y,z) = atom.position
#print atom.symbol, atom.position, atom.tag
if atom.tag:
f.write('%-2s %15.8f %15.8f %15.8f %-4s\n' % ('X', x, y, z, atom.tag))
#f.write('%-2s %15.8f %15.8f %15.8f %-4s\n' % (atom.symbol, x, y, z, atom.tag))
else:
f.write('%-2s %15.8f %15.8f %15.8f\n' % ('Q', x, y, z))
#f.write('%-2s %15.8f %15.8f %15.8f\n' % (atom.symbol, x, y, z))
g.write('%-9s %-50s\n' %('center =', names[0]))
for obj in xrange(n_centers+1,n_obj+1):
f.write('\n%-8s %-3d\n' %('Linker: ', obj-n_centers))
f.write('%-3d\n' %(len(model_molecule[obj])))
f.write('%-20s\n' %('type = triangle'))
#process positions to make it a bit more ideal
for atom in model_molecule[obj]:
#if atom.symbol == 'C':
if atom.index != 0:
model_molecule[obj].set_distance(atom.index,0,1.0,fix=1)
for atom in model_molecule[obj]:
(x,y,z) = atom.position
#print atom.symbol, atom.position, atom.tag
if atom.tag:
f.write('%-2s %15.8f %15.8f %15.8f %-4s\n' % ('X', x, y, z, atom.tag))
#f.write('%-2s %15.8f %15.8f %15.8f %-4s\n' % (atom.symbol, x, y, z, atom.tag))
else:
f.write('%-2s %15.8f %15.8f %15.8f\n' % ('Q', x, y, z))
#f.write('%-2s %15.8f %15.8f %15.8f\n' % (atom.symbol, x, y, z))
g.write('%-9s %-50s\n' %('linker =', names[1]))
test_mol = Atoms()
for obj in model_molecule:
test_mol += model_molecule[obj]
write('test_model2.xyz',test_mol)
def read_turbomole_gradient(filename='gradient', index=-1):
""" Method to read turbomole gradient file """
if isinstance(filename, str):
f = open(filename)
# read entire file
lines = [x.strip() for x in f.readlines()]
# find $grad section
start = end = -1
for i, line in enumerate(lines):
if not line.startswith('$'):
continue
if line.split()[0] == '$grad':
start = i
elif start >= 0:
end = i
break
if end <= start:
raise RuntimeError('File %s does not contain a valid \'$grad\' section' % (filename))
def formatError():
raise RuntimeError('Data format in file %s does not correspond to known Turbomole gradient format' % (filename))
# trim lines to $grad
del lines[:start+1]
del lines[end-1-start:]
# Interpret $grad section
from ase import Atoms, Atom
from ase.calculators.singlepoint import SinglePointCalculator
from ase.units import Bohr
images = []
while len(lines): # loop over optimization cycles
# header line
# cycle = 1 SCF energy = -267.6666811409 |dE/dxyz| = 0.157112
fields = lines[0].split('=')
try:
cycle = int(fields[1].split()[0])
energy = float(fields[2].split()[0])
gradient = float(fields[3].split()[0])
except (IndexError, ValueError):
formatError()
# coordinates/gradient
atoms = Atoms()
forces = []
for line in lines[1:]:
fields = line.split()
if len(fields) == 4: # coordinates
# 0.00000000000000 0.00000000000000 0.00000000000000 c
try:
symbol = fields[3].lower().capitalize()
position = tuple([bohr2angstrom(float(x)) for x in fields[0:3] ])
except ValueError:
formatError()
atoms.append(Atom(symbol, position))
elif len(fields) == 3: # gradients
# -.51654903354681D-07 -.51654903206651D-07 0.51654903169644D-07
try:
grad = [float(x.replace('D', 'E')) * Bohr for x in fields[0:3] ]
except ValueError:
formatError()
forces.append(grad)
else: # next cycle
break
# calculator
calc = SinglePointCalculator(atoms, energy=energy, forces=forces)
atoms.set_calculator(calc)
# save frame
images.append(atoms)
# delete this frame from data to be handled
del lines[:2*len(atoms)+1]
return images[index]
def read_proteindatabank(fileobj, index=-1, read_arrays=True):
"""Read PDB files."""
if isinstance(fileobj, basestring):
fileobj = open(fileobj)
images = []
orig = np.identity(3)
trans = np.zeros(3)
atoms = Atoms()
occ = []
bfactor = []
for line in fileobj.readlines():
if line.startswith('CRYST1'):
cellpar = [
float(line[6:15]), # a
float(line[15:24]), # b
float(line[24:33]), # c
float(line[33:40]), # alpha
float(line[40:47]), # beta
float(line[47:54])
] # gamma
atoms.set_cell(cellpar_to_cell(cellpar))
atoms.pbc = True
for c in range(3):
if line.startswith('ORIGX' + '123'[c]):
orig[c] = [
float(line[10:20]),
float(line[20:30]),
float(line[30:40])
]
trans[c] = float(line[45:55])
if line.startswith('ATOM') or line.startswith('HETATM'):
try:
# Atom name is arbitrary and does not necessarily
# contain the element symbol. The specification
# requires the element symbol to be in columns 77+78.
# Fall back to Atom name for files that do not follow
# the spec, e.g. packmol.
try:
symbol = label_to_symbol(line[76:78].strip())
except (KeyError, IndexError):
symbol = label_to_symbol(line[12:16].strip())
# Don't use split() in case there are no spaces
position = np.array([
float(line[30:38]), # x
float(line[38:46]), # y
float(line[46:54])
]) # z
try:
occ.append(float(line[54:60]))
bfactor.append(float(line[60:66]))
except (IndexError, ValueError):
pass
position = np.dot(orig, position) + trans
atoms.append(Atom(symbol, position))
except Exception as ex:
warnings.warn(
'Discarding atom when reading PDB file: {}\n{}'.format(
line.strip(), ex))
if line.startswith('END'):
# End of configuration reached
# According to the latest PDB file format (v3.30),
# this line should start with 'ENDMDL' (not 'END'),
# but in this way PDB trajectories from e.g. CP2K
# are supported (also VMD supports this format).
if read_arrays and len(occ) == len(atoms):
atoms.set_array('occupancy', np.array(occ))
if read_arrays and len(bfactor) == len(atoms):
atoms.set_array('bfactor', np.array(bfactor))
images.append(atoms)
atoms = Atoms()
occ = []
bfactor = []
if len(images) == 0:
# Single configuration with no 'END' or 'ENDMDL'
if read_arrays and len(occ) == len(atoms):
atoms.set_array('occupancy', np.array(occ))
if read_arrays and len(bfactor) == len(atoms):
atoms.set_array('bfactor', np.array(bfactor))
images.append(atoms)
return images[index]
def make_soc(names,sizes):
#work out the expansion factor
factor = sum(sizes) #
a = 2.8284 * factor
soc = crystal(['C', 'In'], [(0.25, 0, 0), (0.25, 0.25, 0.25)], spacegroup=229, #229
cellpar=[a, a, a, 90, 90, 90])
#Use this purely to get the COM wrt In only
shadow_soc = crystal(['C', 'In'], [(0.25, 0, 0), (0.25, 0.25, 0.25)], spacegroup=229,
cellpar=[a, a, a, 90, 90, 90])
del shadow_soc[[atom.index for atom in shadow_soc if atom.symbol=='C']]
coIn = Atom('X', position = shadow_soc.get_center_of_mass())
#shadow_soc += coIn
#write('In_only.xyz',shadow_soc)
del shadow_soc[[atom.index for atom in shadow_soc if atom.symbol=='In']]
#print "shadow_soc",shadow_soc
#Now let's sort our In atoms into two sets of diagonals
first_In = 0
for atom in soc:
if atom.symbol !='In':
first_In +=1
else:
break
#print first_In
intra_In_distances = {}
for atom in soc:
if atom.symbol == 'In':
this_dist = soc.get_distance(first_In,atom.index)
intra_In_distances[atom.index] = this_dist
#print intra_In_distances
sorted_IID = sorted(intra_In_distances, key=intra_In_distances.get)
#print sorted_IID
#Now assign into two sets of diagonals
cw_set = [first_In,sorted_IID[4],sorted_IID[5],sorted_IID[6]]
acw_set = [sorted_IID[1],sorted_IID[2],sorted_IID[3],sorted_IID[7]]
#write('test.xyz',soc)
eps = 0.01
model_molecule={}
n_obj = -1 #initialise to -1 such that it can be indexed at start of add
n_tags = 0
#First detect global bonding
cov_rad=[]
for atom in soc:
#cov_rad.append(covalent_radii[atom.number])
cov_rad.append(factor / 2)
nlist = NeighborList(cov_rad,self_interaction=False,bothways=True)
nlist.build(soc)
#To sort out tags, we need to label each bond
nbond = 1
bond_matrix = np.zeros( (len(soc),len(soc)) )
for atom in soc:
indices, offsets = nlist.get_neighbors(atom.index)
for index in indices:
if atom.index < index:
bond_matrix[atom.index,index] = nbond
bond_matrix[index,atom.index] = nbond
print atom.index, index, nbond
nbond +=1
print nbond
print bond_matrix
#Now we look for all the things with 2unit bondlengths
#Starting with In (6 connected things)
for atom in soc:
if atom.symbol == 'In':
print'======================================='
print 'In Atom ',atom.index
n_obj+=1
model_molecule[n_obj] = Atoms()
model_molecule[n_obj] += atom
model_molecule[n_obj][0].original_index = atom.index
#Find the oxygens with three In bound to it
indices, offsets = nlist.get_neighbors(atom.index)
symbols = ([soc[index].symbol for index in indices])
symbol_string = ''.join(sorted([soc[index].symbol for index in indices]))
print symbol_string
#for i,o in zip(indices, offsets):
# print i,o
for index,offset in zip(indices,offsets):
if soc[index].symbol == 'C':
dist_mic = soc.get_distance(atom.index,index,mic=True)
dist_no_mic = soc.get_distance(atom.index,index,mic=False)
print index, dist_no_mic
if (abs(dist_mic - dist_no_mic) <= eps) and (abs(dist_mic - factor) < eps): #Cell expands by the factor. Default systre bondlength is 1.0
model_molecule[n_obj] += soc[index]
model_molecule[n_obj][-1].tag = bond_matrix[atom.index,index]
elif abs(dist_mic - factor) < eps:
#If we're going over a periodic boundary, we need to negate the tag
#print "Tag, ", soc[index].tag, " goes over pbc"
#soc[index].tag = -(soc[index].tag)
model_molecule[n_obj] += soc[index]
model_molecule[n_obj][-1].tag = -bond_matrix[atom.index,index]
print model_molecule[n_obj].positions[-1]
model_molecule[n_obj].positions[-1] = soc.positions[index] + np.dot(offset, soc.get_cell())
print model_molecule[n_obj].positions[-1]
model_molecule[n_obj] += coIn
#print model_molecule[n_obj].original_indices
n_centers = n_obj
#Assigning In atoms to 'sets'
intra_In_distances = {}
for atom in soc:
if atom.symbol == 'In':
this_dist = soc.get_distance(first_In,atom.index)
intra_In_distances[atom.index] = this_dist
print intra_In_distances
sorted_IID = sorted(intra_In_distances, key=intra_In_distances.get)
print sorted_IID
#Now assign into two sets of diagonals
cw_set = [first_In,sorted_IID[4],sorted_IID[5],sorted_IID[6]]
acw_set = [sorted_IID[1],sorted_IID[2],sorted_IID[3],sorted_IID[7]]
angle = 30.0 * pi /180.0
##Now we need to rotate these carbons around
for obj in range(n_centers+1):
#First work out the two sets of C atoms
print "Looking at model object, ",obj,model_molecule[obj][0].original_index
cx_distances = {}
for atom in model_molecule[obj]:
if atom.symbol == 'C':
cx_distances[atom.index] = model_molecule[obj].get_distance(atom.index,-1)
sorted_CXD = sorted(cx_distances, key = cx_distances.get)
inner_set = [sorted_CXD[0],sorted_CXD[1],sorted_CXD[2]]
outer_set = [sorted_CXD[3],sorted_CXD[4],sorted_CXD[5]]
rotation_axis = model_molecule[obj][0].position - model_molecule[obj][-1].position #In-X
#Now copy each set of C and rotate, then replace the original
inner_mol = Atoms()
for a in inner_set:
this_atom = Atom('N',position=model_molecule[obj][a].position, tag=model_molecule[obj][a].tag)
inner_mol.append(this_atom)
outer_mol = Atoms()
for a in outer_set:
this_atom = Atom('N',position=model_molecule[obj][a].position, tag=model_molecule[obj][a].tag)
outer_mol.append(this_atom)
if model_molecule[obj][0].original_index in cw_set:
#inner_mol goes cw
inner_mol.rotate(rotation_axis,angle,center='COM')
outer_mol.rotate(rotation_axis,-angle,center='COM')
elif model_molecule[obj][0].original_index in acw_set:
#inner_mol goes acw
inner_mol.rotate(rotation_axis,-angle,center='COM')
outer_mol.rotate(rotation_axis,angle,center='COM')
model_molecule[obj] += inner_mol
model_molecule[obj] += outer_mol
soc += inner_mol
soc += outer_mol
del model_molecule[obj][[atom.index for atom in model_molecule[obj] if atom.symbol=='C' or atom.symbol == 'X']]
print "=============="
#Now look for C (4 connected things)
for atom in soc:
if atom.symbol == 'C':
print'======================================='
print 'C Atom ',atom.index
n_obj+=1
model_molecule[n_obj] = Atoms()
model_molecule[n_obj] += atom
#Find the oxygens with three In bound to it
indices, offsets = nlist.get_neighbors(atom.index)
symbols = ([soc[index].symbol for index in indices])
symbol_string = ''.join(sorted([soc[index].symbol for index in indices]))
print symbol_string
#for i,o in zip(indices, offsets):
# print i,o
for index,offset in zip(indices,offsets):
if soc[index].symbol == 'In':
dist_mic = soc.get_distance(atom.index,index,mic=True)
dist_no_mic = soc.get_distance(atom.index,index,mic=False)
print index, dist_no_mic
print soc[index].original_index
if (abs(dist_mic - dist_no_mic) <= eps) and (abs(dist_mic - factor) < eps):
this_tag = bond_matrix[atom.index,index]
print "Searching for tag ",this_tag
for atom2 in soc:
if atom2.symbol == 'N' and abs(atom2.tag) == this_tag:
model_molecule[n_obj] += Atom('O', position=atom2.position)
model_molecule[n_obj][-1].tag = this_tag
#model_molecule[n_obj] += soc[index]
#model_molecule[n_obj][-1].tag = bond_matrix[atom.index,index]
elif abs(dist_mic - factor) < eps:
#If we're going over a periodic boundary, we need to mark the tag
#print "Tag, ", soc[index].tag, " goes over pbc"
this_tag = bond_matrix[atom.index,index]
for atom2 in soc:
if atom2.symbol == 'N' and abs(atom2.tag) == this_tag:
model_molecule[n_obj] += Atom('O', position=atom2.position)
model_molecule[n_obj][-1].tag = -this_tag
model_molecule[n_obj].positions[-1] = soc.positions[atom2.index] + np.dot(offset, soc.get_cell())
#model_molecule[n_obj] += soc[index]
#model_molecule[n_obj][-1].tag = -bond_matrix[atom.index,index]
#print model_molecule[n_obj].positions[-1]
#model_molecule[n_obj].positions[-1] = soc.positions[index] + np.dot(offset, soc.get_cell())
for obj in xrange(n_centers+1,n_obj+1):
tmp_mol = Atoms()
tmp_mol = model_molecule[obj].copy()
del tmp_mol[[atom.index for atom in tmp_mol if atom.symbol!='O']]
#process positions correct COM
model_molecule[obj][0].position = tmp_mol.get_center_of_mass()
#Now most of the tags have switched
for obj in xrange(n_centers+1):
#process positions to make it a bit more ideal
for atom in model_molecule[obj]:
if atom.symbol == 'C' or atom.symbol == 'N':
model_molecule[obj].set_distance(atom.index,0,1.0,fix=1)
shadow_soc += Atom('N', position=atom.position,tag=atom.tag) #Keep these tags
for obj in xrange(n_centers+1,n_obj+1):
#process positions to make it a bit more ideal
for atom in model_molecule[obj]:
if atom.symbol == 'O':
model_molecule[obj].set_distance(atom.index,0,1.0,fix=1)
shadow_soc += Atom('O', position=atom.position)
#Now find the nearest N
rON={}
for atom2 in shadow_soc:
if atom2.symbol == 'N':
this_dist = shadow_soc.get_distance(-1,atom2.index,mic=True)
rON[atom2.index] = this_dist
print atom.index, atom2.index, this_dist
min_index = min(rON, key=rON.get)
atom.tag = shadow_soc[min_index].tag
shadow_soc[-1].tag = shadow_soc[min_index].tag
#write('shadow_soc.xyz',shadow_soc)
#Tags get horribly messed up
t = open('tags_check','w')
for obj in xrange(n_obj+1):
#process positions to make it a bit more ideal
for atom in model_molecule[obj]:
(x,y,z) = atom.position
#print atom.symbol, atom.position, atom.tag
if atom.tag:
t.write('%-2s %15.8f %15.8f %15.8f %-4s\n' % (atom.symbol, x, y, z, atom.tag))
else:
t.write('%-2s %15.8f %15.8f %15.8f\n' % (atom.symbol, x, y, z))
f = open('soc.model','w')
g = open('control-mofgen.txt','w')
#Just for checking, now lets gather everything in the model_molecule dict into one big thing and print it
#test_mol = Atoms()
#for obj in model_molecule:
# test_mol += model_molecule[obj]
#
#write('test_model.xyz',test_mol)
#print n_centers, n_model, n_obj
#Headers and cell
f.write('%-20s %-3d\n' %('Number of objects =',n_obj+1))
f.write('%-20s\n' %('build = systre'))
f.write('%5s\n' %('Cell:'))
f.write('%8.3f %8.3f %8.3f \n' %
(soc.get_cell()[0][0],
soc.get_cell()[0][1],
soc.get_cell()[0][2]))
f.write('%8.3f %8.3f %8.3f \n' %
(soc.get_cell()[1][0],
soc.get_cell()[1][1],
soc.get_cell()[1][2]))
f.write('%8.3f %8.3f %8.3f \n' %
(soc.get_cell()[2][0],
soc.get_cell()[2][1],
soc.get_cell()[2][2]))
g.write('%-20s\n' %('model = soc'))
for obj in xrange(n_centers+1):
f.write('\n%-8s %-3d\n' %('Centre: ', obj+1))
f.write('%-3d\n' %(len(model_molecule[obj])))
f.write('%-20s\n' %('type = tri_prism'))
#process positions to make it a bit more ideal
#for atom in model_molecule[obj]:
# if atom.symbol == 'C' or atom.symbol == 'N':
# pass#model_molecule[obj].set_distance(atom.index,0,1.0,fix=1)
for atom in model_molecule[obj]:
(x,y,z) = atom.position
#print atom.symbol, atom.position, atom.tag
if atom.tag:
f.write('%-2s %15.8f %15.8f %15.8f %-4s\n' % ('X', x, y, z, atom.tag))
else:
f.write('%-2s %15.8f %15.8f %15.8f\n' % ('Q', x, y, z))
g.write('%-9s %-50s\n' %('center =', names[0]))
for obj in xrange(n_centers+1,n_obj+1):
f.write('\n%-8s %-3d\n' %('Linker: ', obj-n_centers))
f.write('%-3d\n' %(len(model_molecule[obj])))
f.write('%-20s\n' %('type = rectangle'))
#process positions to make it a bit more ideal
#for atom in model_molecule[obj]:
# if atom.symbol == 'In':
# pass#model_molecule[obj].set_distance(atom.index,0,1.0,fix=1)
for atom in model_molecule[obj]:
(x,y,z) = atom.position
#print atom.symbol, atom.position, atom.tag
if atom.tag:
f.write('%-2s %15.8f %15.8f %15.8f %-4s\n' % ('X', x, y, z, atom.tag))
else:
f.write('%-2s %15.8f %15.8f %15.8f\n' % ('Q', x, y, z))
g.write('%-9s %-50s\n' %('linker =', names[1]))
test_mol = Atoms()
for obj in model_molecule:
test_mol += model_molecule[obj]
write('test_model2.xyz',test_mol)
write('test_model2.cif',soc)
def read_turbomole_gradient(filename='gradient', index=-1):
""" Method to read turbomole gradient file """
if isinstance(filename, str):
f = open(filename)
# read entire file
lines = [x.strip() for x in f.readlines()]
# find $grad section
start = end = -1
for i, line in enumerate(lines):
if not line.startswith('$'):
continue
if line.split()[0] == '$grad':
start = i
elif start >= 0:
end = i
break
if end <= start:
raise RuntimeError(
'File %s does not contain a valid \'$grad\' section' % (filename))
def formatError():
raise RuntimeError(
'Data format in file %s does not correspond to known Turbomole gradient format'
% (filename))
# trim lines to $grad
del lines[:start + 1]
del lines[end - 1 - start:]
# Interpret $grad section
from ase import Atoms, Atom
from ase.calculators.singlepoint import SinglePointCalculator
from ase.units import Bohr
images = []
while len(lines): # loop over optimization cycles
# header line
# cycle = 1 SCF energy = -267.6666811409 |dE/dxyz| = 0.157112
fields = lines[0].split('=')
try:
cycle = int(fields[1].split()[0])
energy = float(fields[2].split()[0])
gradient = float(fields[3].split()[0])
except (IndexError, ValueError):
formatError()
# coordinates/gradient
atoms = Atoms()
forces = []
for line in lines[1:]:
fields = line.split()
if len(fields) == 4: # coordinates
# 0.00000000000000 0.00000000000000 0.00000000000000 c
try:
symbol = fields[3].lower().capitalize()
position = tuple(
[bohr2angstrom(float(x)) for x in fields[0:3]])
except ValueError:
formatError()
atoms.append(Atom(symbol, position))
elif len(fields) == 3: # gradients
# -.51654903354681D-07 -.51654903206651D-07 0.51654903169644D-07
try:
grad = [
float(x.replace('D', 'E')) * Bohr for x in fields[0:3]
]
except ValueError:
formatError()
forces.append(grad)
else: # next cycle
break
# calculator
calc = SinglePointCalculator(energy, forces, None, None, atoms)
atoms.set_calculator(calc)
# save frame
images.append(atoms)
# delete this frame from data to be handled
del lines[:2 * len(atoms) + 1]
return images[index]
atoms = read(glob.glob('final.xyz')[0])
atoms.set_cell(np.eye(3) * l)
print atoms
up = Atoms()
up.set_cell(atoms.get_cell())
up_target = 'X'
dn = Atoms()
dn.set_cell(atoms.get_cell())
dn_target = 'He'
nuclei = Atoms()
nuclei.set_cell(atoms.get_cell())
for a in range(len(atoms)):
if atoms[a].symbol == up_target:
up.append(atoms[a])
if atoms[a].symbol == dn_target:
dn.append(atoms[a])
if atoms[a].symbol != up_target and atoms[a].symbol != dn_target:
nuclei.append(atoms[a])
data_up = get_rdf(up, l, nmax)
data_dn = get_rdf(dn, l, nmax)
data_nuclei = get_rdf(nuclei, l, nmax)
data_all = get_rdf(atoms, l, nmax)
# data_up[0]/max(data_up[0])*len(up)
# data_dn[0]/max(data_dn[0])*len(dn)
bar(data_up[1], data_up[0], label='up', width=0.02)
bar(data_dn[1], data_dn[0], label='dn', width=0.01)
bar(data_nuclei[1], data_nuclei[0], label='nuclei', width=0.02)
#bar(data_all[1],data_all[0],label='all',width=0.05)
