def get_KPPRA(self):
# Open up the OUTCAR
with open(self.outcar) as fp:
#store the number of atoms and number of irreducible K-points
for line in fp:
if "number of ions NIONS =" in line:
words = line.split()
NI = int(words[11])
elif "k-points NKPTS =" in line:
words = line.split()
NIRK = float(words[3])
#check if the number of k-points was reduced by VASP if so, sum all the k-points weight
if "irreducible" in open(self.outcar).read():
fp.seek(0)
for line in fp:
#sum all the k-points weight
if "Coordinates Weight" in line:
NK = 0
counter = 0
for line in fp:
if counter == NIRK:
break
NK += float(line.split()[3])
counter += 1
return Value(scalars=[Scalar(value=NI * NK)])
#if k-points were not reduced KPPRA equals the number of atoms * number of irreducible k-points
else:
return Value(scalars=[Scalar(value=NI * NIRK)])
def get_KPPRA(self):
# Open up the OUTCAR
with open(os.path.join(self._directory, 'OUTCAR')) as fp:
#store the number of atoms and number of irreducible K-points
for line in fp:
if "NIONS" in line:
words = line.split()
NI = int(words[11])
elif "NKPTS" in line:
words = line.split()
NIRK = float(words[3])
#check if the number of k-points was reduced by VASP if so, sum all the k-points weight
if "irreducible" in open(os.path.join(self._directory,
'OUTCAR')).read():
fp.seek(0)
for line in fp:
#sum all the k-points weight
if "Coordinates Weight" in line:
NK = 0
counter = 0
for line in fp:
if counter == NIRK:
break
NK += float(line.split()[3])
counter += 1
return Value(scalars=[Scalar(value=NI * NK)])
#if k-points were not reduced KPPRA equals the number of atoms * number of irreducible k-points
else:
return Value(scalars=[Scalar(value=NI * NIRK)])
# Error handling: NKPTS or NIONS not found
raise Exception('NIONS, irredicuble or Coordinates not found')
def get_dos(self):
file_path = os.path.join(self._directory, 'DOSCAR')
if not os.path.isfile(file_path):
return None
#open DOSCAR
with open(os.path.join(self._directory, 'DOSCAR')) as fp:
for i in range(6):
l = fp.readline()
n_step = int(l.split()[2])
energy = []
dos = []
for i in range(n_step):
l = fp.readline().split()
e = float(l.pop(0))
energy.append(Scalar(value=e))
dens = 0
for j in range(int(len(l) / 2)):
dens += float(l[j])
dos.append(Scalar(value=dens))
# Convert to property
return Property(scalars=dos,
units='number of states per unit cell',
conditions=Value(name='energy',
scalars=energy,
units='eV'))
def test_unambig():
"""Test that properties are mirrored in a top level dic"""
pif = System()
pif.properties = [
Property(name="foo", scalars=[Scalar(value=1.0)]),
Property(name="bar", scalars=[Scalar(value=2.0)])
]
r = ReadView(pif)
assert r["foo"].scalars[0].value == 1.0
assert r["bar"].scalars[0].value == 2.0
def get_forces(self):
self.atoms = read_vasp_out(os.path.join(self._directory, 'OUTCAR'))
forces_raw = self.atoms.get_calculator().results['forces'].tolist()
forces_wrapped = [[Scalar(value=x) for x in y] for y in forces_raw]
positions_raw = self.atoms.positions.tolist()
positions_wrapped = [[Scalar(value=x) for x in y]
for y in positions_raw]
return Property(vectors=forces_wrapped,
conditions=Value(name="positions",
vectors=positions_wrapped))
def test_condition_elevation():
"""Test that read views elevate conditions"""
condition = Value(name="spam", scalars=[Scalar(value="eggs")])
pif = System(properties=[
Property(
name="foo", scalars=[Scalar(value="bar")], conditions=[condition])
])
r = ReadView(pif)
assert r["foo"].scalars[0].value == "bar", "Didn't elevate property key"
assert r["spam"].scalars[0].value == "eggs", "Didn't elevate condition key"
def test_read_view():
"""Test that properties are passed through to the readview"""
pif = System()
pif.uid = "10245"
pif.names = ["example", "ex"]
pif.properties = [
Property(name="foo", scalars=[Scalar(value=1.0)]),
Property(name="bar", scalars=[Scalar(value=2.0)])
]
r = ReadView(pif)
assert r.uid == pif.uid
assert r.names == pif.names
assert r.properties["foo"].scalars[0].value == 1.0
assert r.properties["bar"].scalars[0].value == 2.0
def test_update_conflict():
"""Test that update works when not extending and a field is redefined"""
more_pif = System(
properties=[
Property(name="band gap", scalars=[Scalar(value=1.3)]),
Property(name="phase", scalars=[Scalar(value="gas")]),
]
)
combined = update(test_pif, more_pif)
expected = {"band gap": 1.3, "phase": "gas"}
assert set(x.name for x in combined.properties) == set(expected.keys())
assert set(x.scalars[0].value for x in combined.properties) == set(expected.values())
for prop in combined.properties:
assert expected[prop.name] == prop.scalars[0].value
def test_update_extend():
"""Test that update works when extending with no dups"""
more_pif = System(
properties=[
Property(name="band gap", scalars=[Scalar(value=1.3)]),
Property(name="phase", scalars=[Scalar(value="gas")]),
]
)
combined = update(test_pif, more_pif, extend=True)
expected = {"band gap": 1.3, "phase": "gas", "foo": "bar", "spam": 2.7}
assert set(x.name for x in combined.properties) == set(expected.keys())
assert set(x.scalars[0].value for x in combined.properties) == set(expected.values())
for prop in combined.properties:
assert expected[prop.name] == prop.scalars[0].value
def test_nested_read_view():
"""Test that nested Pios (system here) are recursively processed"""
pif = System()
pif.uid = "10245"
pif.properties = [
Property(name="foo", scalars=[Scalar(value=1.0)]),
Property(name="bar", scalars=[Scalar(value=2.0)])
]
pif2 = System(sub_systems=[pif])
r = ReadView(pif2)
assert r.sub_systems["10245"].uid == pif.uid
assert r["10245"].uid == pif.uid
assert r.sub_systems["10245"].properties["foo"].scalars[0].value == 1.0
assert r.sub_systems["10245"].properties["bar"].scalars[0].value == 2.0
assert r["foo"].scalars[0].value == 1.0
assert r["bar"].scalars[0].value == 2.0
def get_stresses(self):
#Check if ISIF = 0 is used
if "ISIF = 0" in open(os.path.join(self._directory,
'OUTCAR')).read():
return None
#Check if ISIF = 1 is used
elif "ISIF = 1" in open(os.path.join(self._directory,
'OUTCAR')).read():
return None
else:
#scan file in reverse to have the final pressure
for line in open(os.path.join(self._directory,
'OUTCAR')).readlines():
if "in kB" in line:
words = line.split()
XX = float(words[2])
YY = float(words[3])
ZZ = float(words[4])
XY = float(words[5])
YZ = float(words[6])
ZX = float(words[7])
matrix = [[XX, XY, ZX], [XY, YY, YZ], [ZX, YZ, ZZ]]
wrapped = [[Scalar(value=x) for x in y] for y in matrix]
return Property(matrices=[wrapped], units='kbar')
# Error handling: "in kB" not found
raise Exception('in kB not found')
def test_update_conflicts():
"""Test that update works when there are conflicts"""
conflict_pif = System(
properties=[
Property(name="band gap", scalars=[Scalar(value=1.3)]),
Property(name="phase", scalars=[Scalar(value="gas")]),
Property(name="spam", scalars=[Scalar(value=2.0)]),
]
)
combined = update(test_pif, conflict_pif, extend=True)
expected = {"band gap": 1.3, "phase": "gas", "foo": "bar", "spam": 2.0}
assert set(x.name for x in combined.properties) == set(expected.keys())
assert set(x.scalars[0].value for x in combined.properties) == set(expected.values())
for prop in combined.properties:
assert expected[prop.name] == prop.scalars[0].value
def test_multiple_instances():
"""Test that keys that show up in different places with the same value are kept"""
pif = System()
pif.uid = "10245"
pif.properties = [
Property(name="foo", scalars=[Scalar(value=1.0)]),
Property(name="bar", scalars=[Scalar(value=2.0)])
]
pif2 = System(
sub_systems=[
pif,
],
properties=[Property(name="bar", scalars=[Scalar(value=2.0)])])
r = ReadView(pif2)
assert r.properties["bar"].scalars[0].value == 2.0
assert r.sub_systems["10245"].properties["bar"].scalars[0].value == 2.0
assert r["bar"].scalars[0].value == 2.0
def get_total_energy(self):
with open(self.outcar) as fp:
last_energy = None
for line in fp:
if line.startswith(' free energy TOTEN'):
last_energy = float(line.split()[4])
if last_energy is None:
return None
return Property(scalars=[Scalar(value=last_energy)], units='eV')
def get_xc_functional(self):
# Open up the OUTCAR
with open(self.outcar) as fp:
# Look for TITEL
for line in fp:
if "TITEL" in line:
words = line.split()
return Value(scalars=[Scalar(value=words[2])])
def get_xc_functional(self):
# Open up the OUTCAR
with open(os.path.join(self._directory, 'OUTCAR')) as fp:
# Look for TITEL
for line in fp:
if "TITEL" in line:
words = line.split()
return Value(scalars=[Scalar(value=words[2])])
break
def get_band_gap(self):
"""Get the bandgap, either from the EIGENVAL or DOSCAR files"""
if self.outcar is not None and self.eignval is not None:
bandgap = VaspParser._get_bandgap_eigenval(self.eignval,
self.outcar)
elif self.doscar is not None:
bandgap = VaspParser._get_bandgap_doscar(self.doscar)
else:
return None
return Property(scalars=[Scalar(value=round(bandgap, 3))], units='eV')
def test_method_software():
"""Testing that method and software names are elevated"""
method = Method(name="spam", software=[Software(name="magic")])
pif = System(properties=[
Property(name="foo", scalars=[Scalar(value="bar")], methods=[method])
])
r = ReadView(pif)
assert r["foo"].scalars[0].value == "bar", "Didn't elevate property key"
assert "spam" in r.keys(), "Expected spam in keys"
assert "magic" in r.keys(), "Expected magic in keys"
def test_ambiguity():
"""Test that ambiguous keys are removed from the top level dict"""
pif = System()
pif.uid = "10245"
pif.properties = [
Property(name="foo", scalars=[Scalar(value=1.0)]),
Property(name="bar", scalars=[Scalar(value=2.0)])
]
pif2 = System(
sub_systems=[
pif,
],
properties=[Property(name="foo", scalars=[Scalar(value=10.0)])])
r = ReadView(pif2)
assert r.properties["foo"].scalars[0].value == 10.0
assert r.sub_systems["10245"].properties["foo"].scalars[0].value == 1.0
assert "foo" not in r.keys()
assert r.sub_systems["10245"]["foo"].scalars[0].value == 1.0
assert r["bar"].scalars[0].value == 2.0
def get_pp_name(self):
# Open up the OUTCAR
with open(self.outcar) as fp:
#initialize empty list to store pseudopotentials
pp = []
# Look for TITEL
for line in fp:
if "TITEL" in line:
words = line.split()
pp.append(words[3])
return Value(vectors=[[Scalar(value=x) for x in pp]])
def get_cutoff_energy(self):
# Open up the OUTCAR
with open(os.path.join(self._directory, 'OUTCAR'), 'r') as fp:
# Look for ENCUT
for line in fp:
if "ENCUT" in line:
words = line.split()
return Value(scalars=[Scalar(value=float(words[2]))],
units=words[3])
# Error handling: ENCUT not found
raise Exception('ENCUT not found')
def get_total_magnetization(self):
if self.outcar is None:
return None
parser = OutcarParser()
with open(self.outcar, "r") as fp:
matches = list(
filter(lambda x: "total magnetization" in x,
parser.parse(fp.readlines())))
if len(matches) == 0:
return None
total_magnetization = matches[-1]["total magnetization"]
return Property(scalars=[Scalar(value=total_magnetization)],
units="Bohr")
def get_band_gap(self):
"""Get the bandgap, either from the EIGENVAL or DOSCAR files"""
doscar_path = os.path.join(self._directory, 'DOSCAR')
outcar_path = os.path.join(self._directory, 'OUTCAR')
eigenval_path = os.path.join(self._directory, 'EIGENVAL')
if os.path.isfile(outcar_path) and os.path.isfile(eigenval_path):
bandgap = VaspParser._get_bandgap_eigenval(eigenval_path,
outcar_path)
elif os.path.isfile(doscar_path):
bandgap = VaspParser._get_bandgap_doscar(doscar_path)
else:
return None
return Property(scalars=[Scalar(value=round(bandgap, 3))], units='eV')
def get_pp_name(self):
# Open up the OUTCAR
with open(os.path.join(self._directory, 'OUTCAR')) as fp:
#initialize empty list to store pseudopotentials
pp = []
# Look for TITEL
for line in fp:
if "TITEL" in line:
words = line.split()
pp.append(words[3])
return Value(vectors=[[Scalar(value=x) for x in pp]])
# Error handling: TITEL not found
raise Exception('TITEL not found')
def get_pressure(self):
#define pressure dictionnary because since when is kB = kbar? Come on VASP people
pressure_dict = {'kB': 'kbar'}
#Check if ISIF = 0 is used
if "ISIF = 0" in open(self.outcar).read():
#if ISIF = 0 is used, print this crap
return None
#if ISIF is not 0 then extract pressure and units
else:
#scan file in reverse to have the final pressure
for line in reversed(open(self.outcar).readlines()):
if "external pressure" in line:
words = line.split()
return Property(scalars=[Scalar(value=float(words[3]))],
units=pressure_dict[words[4]])
break
def get_stresses(self):
#Check if ISIF = 0 is used
if "ISIF = 0" in open(self.outcar).read():
return None
#Check if ISIF = 1 is used
elif "ISIF = 1" in open(self.outcar).read():
return None
else:
#scan file in reverse to have the final pressure
for line in open(self.outcar).readlines():
if "in kB" in line:
words = line.split()
XX = float(words[2])
YY = float(words[3])
ZZ = float(words[4])
XY = float(words[5])
YZ = float(words[6])
ZX = float(words[7])
matrix = [[XX, XY, ZX], [XY, YY, YZ], [ZX, YZ, ZZ]]
wrapped = [[Scalar(value=x) for x in y] for y in matrix]
return Property(matrices=[wrapped], units='kbar')
def get_vdW_settings(self):
#define the name of the vdW methods in function of their keyword
vdW_dict = {
'BO': 'optPBE-vdW',
'MK': 'optB88-vdW',
'ML': 'optB86b-vdW',
'RE': 'vdW-DF',
'OR': 'Klimes-Bowler-Michaelides'
}
#Open up the OUTCAR
with open(self.outcar) as fp:
#Check if vdW is used
if "LUSE_VDW" in open(self.outcar).read():
#if vdW is used, get its keyword
for line in fp:
if "GGA =" in line:
words = line.split()
return Value(
scalars=[Scalar(value=vdW_dict[words[2]])])
#if vdW is not used, return None
else:
return None
def make_pif(df):
"""
Extracts information from Pandas Dataframe with GDB-9 molecule data
to create a PIF object containing metadata and structural information
:param: df - Pandas Dataframe object
:return: PIF object
"""
pif_data = ChemicalSystem()
pif_data.contacts = [
Person(
name=Name(given='Lawrence', family='Crosby'),
email='*****@*****.**',
orcid='0000-0001-7644-3762'
)
]
pif_data.references = [Reference(doi='10.1038/sdata.2014.22')]
pif_data.licenses = [
License(
name='CC0 1.0',
description='Creative Commons Public Domain Dedication',
url='https://creativecommons.org/publicdomain/zero/1.0/'
)
]
software_list = [
Software(name='Corina', version='3.491 2013', producer='Altamira LLC'),
Software(name='MOPAC', version='13.136L 2012', producer='CAChe Research LLC')
]
pif_data.preparation = [ProcessStep(software=software_list)]
pif_data.chemical_formula = df.iloc[-1, 0].split('/')[1] # extract chem formula from InChI
# set from last row, first element in DataFrame
pif_data.uid = df.iloc[-1, 1] # use B3LYP InChI as uid
# set from last row, second element in DataFrame
ids = []
gdb9_id = Id(name='GDB9 Id', value=df.iloc[1, 1])
ids.append(gdb9_id)
inchi_id_corina = Id(name='InChI Corina', value=df.iloc[-1, 0])
ids.append(inchi_id_corina) # add Corina InChI to id list
smiles_id_gdb17 = Id(name='SMILES GDB-17', value=df.iloc[-2, 0])
ids.append(smiles_id_gdb17)# add SMILES GDB-17 to id list
smiles_id_b3lyp = Id(name='SMILES B3LYP', value=df.iloc[-2, 1])
ids.append(smiles_id_b3lyp)# add SMILES B3YLP to id list
pif_data.ids = ids
properties = []
vib_freqs = Property(name='Harmonic Vibrational Frequencies', units='cm-1',
dataType='COMPUTATIONAL')
vib_freqs.scalars = [Scalar(value=x) for x in df.iloc[-3, :]]
# set vibrational frequencies using 3rd from last row in DataFrame
properties.append(vib_freqs)
num_atoms = Property(name='Number of Atoms', dataType='COMPUTATIONAL')
n_atoms = df.iloc[0, 0]
# get n_atoms from first element of DataFrame
num_atoms.scalars = [Scalar(value=n_atoms)]
properties.append(num_atoms)
atoms = Property(name='Atoms', dataType='COMPUTATIONAL')
elems = [df.iloc[a, 0] for a in range(2, int(n_atoms)+2)]
# set elements using 1st column starting from 2nd row of DateFrame
atoms.scalars = [Scalar(elem) for elem in elems]
properties.append(atoms)
atomic_positions = Property(name='Atomic Positions',
dataType='COMPUTATIONAL', units='angstrom')
x_coords = [df.iloc[a, 1] for a in range(2, int(n_atoms)+2)]
y_coords = [df.iloc[a, 2] for a in range(2, int(n_atoms)+2)]
z_coords = [df.iloc[a, 3] for a in range(2, int(n_atoms)+2)]
# set coordinates using 2nd, 3rd, and 4th columns starting from 2nd row of DateFrame
atomic_positions.vectors = [
[
Scalar(x_coords[i]),
Scalar(y_coords[i]),
Scalar(z_coords[i])
]
for i in range(int(n_atoms))
]
properties.append(atomic_positions)
partial_charges = Property(name='Partial Charge',
dataType='COMPUTATIONAL', units='e')
charges = [df.iloc[a, 4] for a in range(2, int(n_atoms)+2)]
# set charges using 5th columns starting from 2nd row of DateFrame
partial_charges.scalars = [Scalar(charge) for charge in charges]
properties.append(partial_charges)
rot_a = Property(
name='Rotational Constant A',
dataType='COMPUTATIONAL',
units='GHz',
scalars=Scalar(value=df.iloc[1, 2])
)
properties.append(rot_a)
rot_b = Property(
name='Rotational Constant B',
dataType='COMPUTATIONAL',
units='GHz',
scalars=Scalar(value=df.iloc[1, 3])
)
properties.append(rot_b)
rot_c = Property(
name='Rotational Constant C',
dataType='COMPUTATIONAL',
units='GHz',
scalars=Scalar(value=df.iloc[1, 4])
)
properties.append(rot_b)
mu = Property(
name='Dipole Moment',
dataType='COMPUTATIONAL',
units='D',
scalars=Scalar(value=df.iloc[1, 5])
)
properties.append(mu)
alpha = Property(
name='Isotropic Polarizability',
dataType='COMPUTATIONAL',
units='angstrom^3',
scalars=Scalar(value=df.iloc[1, 6])
)
properties.append(alpha)
e_homo = Property(
name='Energy of H**O',
dataType='COMPUTATIONAL',
units='Hartree',
scalars=Scalar(value=df.iloc[1, 7])
)
properties.append(e_homo)
e_lumo = Property(
name='Energy of LUMO',
dataType='COMPUTATIONAL',
units='Hartree',
scalars=Scalar(value=df.iloc[1, 8])
)
properties.append(e_lumo)
e_gap = Property(
name='Energy Gap',
dataType='COMPUTATIONAL',
units='Hartree',
scalars=Scalar(value=df.iloc[1, 9])
)
properties.append(e_gap)
r2 = Property(
name='Electronic Spatial Extent',
dataType='COMPUTATIONAL',
units='angstrom^2',
scalars=Scalar(value=df.iloc[1, 10])
)
properties.append(r2)
zpve = Property(
name='Zero Point Vibrational Energy',
dataType='COMPUTATIONAL',
units='Hartree',
scalars=Scalar(value=df.iloc[1, 11])
)
properties.append(zpve)
u_0 = Property(
name='Internal Energy at OK',
dataType='COMPUTATIONAL',
units='Hartree',
scalars=Scalar(value=df.iloc[1, 12])
)
properties.append(u_0)
u = Property(
name='Internal Energy at 298K',
dataType='COMPUTATIONAL',
units='Hartree',
scalars=Scalar(value=df.iloc[1, 13])
)
properties.append(u)
h = Property(
name='Enthalpy at 298K',
dataType='COMPUTATIONAL',
units='Hartree',
scalars=Scalar(value=df.iloc[1, 14])
)
properties.append(h)
g = Property(
name='Free Energy at 298K',
dataType='COMPUTATIONAL',
units='Hartree',
scalars=Scalar(value=df.iloc[1, 15])
)
properties.append(g)
c_v = Property(
name='Heat Capacity at 298K',
dataType='COMPUTATIONAL',
units='cal/mol/K',
scalars=Scalar(value=df.iloc[1, 16])
)
properties.append(c_v)
pif_data.properties = properties
return pif_data
from pypif_sdk.func import replace_by_key, copy
from pypif_sdk.accessor import get_property_by_name
from pypif.obj import System, Property, Scalar, Value, FileReference
test_pif = System(names=["methane"],
properties=[
Property(name="foo", scalars=[Scalar(value="bar")]),
Property(name="spam",
units="eV",
scalars=[Scalar(value=2.7)],
conditions=[
Value(name="tomato",
units="eV",
scalars=[Scalar(value=1.0)])
]),
Property(
name="image",
files=[FileReference(relative_path="/tmp/file.png")])
])
def test_simple_replace():
"""Test replace a single field with default arguments"""
prop = get_property_by_name(copy(test_pif), "image")
assert prop.files[
0].relative_path == "/tmp/file.png", "Didn't shorten file name"
new_pif = replace_by_key(test_pif, "relative_path",
{"/tmp/file.png": "file.png"})
prop = get_property_by_name(new_pif, "image")
assert prop.files[
0].relative_path == "file.png", "Didn't shorten file name"
from pypif_sdk.func import update
from pypif.obj import System, Property, Scalar
test_pif = System(
properties=[
Property(name="foo", scalars=[Scalar(value="bar")]),
Property(name="spam", scalars=[Scalar(value=2.7)]),
]
)
def test_update_no_conflicts():
"""Test that update works when there are no conflicts"""
no_conflict_pif = System(
names=["gas", "methane"]
)
combined = update(test_pif, no_conflict_pif)
expected = {"foo": "bar", "spam": 2.7}
assert set(x.name for x in combined.properties) == set(expected.keys())
assert set(x.scalars[0].value for x in combined.properties) == set(expected.values())
for prop in combined.properties:
assert expected[prop.name] == prop.scalars[0].value
assert combined.names == ["gas", "methane"]
def test_update_conflict():
"""Test that update works when not extending and a field is redefined"""
more_pif = System(