def extract(s, force_recalc):
if (not s.has(MSDiagonal.default_name + '_evals_hsort')
or force_recalc):
MSDiagonal.get(s)
ms_evals = s.get_array(MSDiagonal.default_name + '_evals_hsort')
return _asymmetry(ms_evals)
def extract(s, force_recalc):
if (not s.has(EFGDiagonal.default_name + '_evals_hsort')
or force_recalc):
EFGDiagonal.get(s)
efg_evals = s.get_array(EFGDiagonal.default_name + '_evals_hsort')
return _asymmetry(efg_evals)
def extract(s, sel_i, sel_j, tag, isotopes, isotope_list, self_coupling,
force_recalc):
jDiagProp = JCDiagonal(sel_i=sel_i, sel_j=sel_j, isotopes=isotopes,
tag=tag, isotope_list=isotope_list,
self_coupling=self_coupling,
force_recalc=force_recalc)
jc_dict = jDiagProp(s)
jc_keys = jc_dict.keys()
jc_evals = [jc_dict[ij]['evals'] for ij in jc_keys]
jc_evals = _haeb_sort(jc_evals)
jc_skew = _asymmetry(jc_evals)
return dict(zip(jc_dict.keys(), jc_skew))
def nics_buildup(args=None):
parser = ap.ArgumentParser()
parser.add_argument('seedname', type=str, default=None)
parser.add_argument('-R', type=float, default=10, help="Max radius")
parser.add_argument('-minR', type=float, default=0, help="Min radius")
parser.add_argument('-n', type=int, default=100, help="Number of points")
parser.add_argument('-log',
action='store_true',
default=False,
help="Use a logarithmic grid")
parser.add_argument('-cfile',
type=str,
default='current.dat',
help="Name of current file")
parser.add_argument('-out',
type=str,
default=None,
help=("Prefix for output file names "
"(default is the seedname)"))
parser.add_argument('-csv',
action='store_true',
default=False,
help="Output NICS tensors in CSV format")
args = parser.parse_args()
cfile = CurrentFile(args.cfile)
cell = io.read(args.seedname + '.cell').get_cell()
nfile = nicsfile(args.seedname + '.nicslist')
try:
fracpoints = np.dot(nfile.nics_points_frac, cell)
except TypeError:
fracpoints = None
abspoints = nfile.nics_points_abs
ncomp = NicsCompute(cfile, cell)
allpoints = {'frac': fracpoints, 'abs': abspoints}
outname = args.seedname if args.out is None else args.out
outnics = open(outname + '_nics.txt', 'w')
if (args.csv):
outcsv = open(outname + '_nics.csv', 'w')
csvwriter = csv.writer(outcsv)
for ptype, plist in allpoints.items():
if plist is None:
continue
for i, p in enumerate(plist):
nics = ncomp.get_nics(p)
rrange, nbuild = ncomp.get_nics_buildup(p,
Rmax=args.R,
n=args.n,
Rmin=args.minR,
is_log=args.log)
# Print output
if (args.csv):
csvwriter.writerow([
i + 1,
np.trace(nics.nics) / 3.0,
np.trace(nics.nics_plus_chi) / 3.0
])
outnics.write('Point {0}_{1}:\n'.format(ptype, i + 1))
outnics.write('Coordinates: {0} {1} {2}\n'.format(*p))
outnics.write('NICS isotropy: {0} ppm\n'.format(
np.trace(nics.nics) / 3.0))
outnics.write('NICS+chi isotropy: {0} ppm\n'.format(
np.trace(nics.nics_plus_chi) / 3.0))
outnics.write('NICS tensor:\n{0}\n'.format(nics.nics))
outnics.write('NICS+chi tensor:\n{0}\n'.format(nics.nics_plus_chi))
outnics.write('\n------\n')
# For buildup, let's diagonalize them
all_evals = np.array(
[np.linalg.eigh((nb + nb.T) / 2.0)[0] for nb in nbuild])
all_evals = _haeb_sort(all_evals)
iso = np.average(all_evals, axis=1)
aniso = _anisotropy(all_evals)
asymm = _asymmetry(all_evals)
np.savetxt(outname + '_{0}_{1}_nicsbuild.dat'.format(ptype, i + 1),
np.array([rrange, iso, aniso, asymm]).T)
outnics.close()
if (args.csv): outcsv.close()
def asymmetry(self):
if self._asymmetry is None:
self._asymmetry = _asymmetry(self._haeb_evals[None, :])[0]
return self._asymmetry