# This is horribly slow, and for very skewed cells, n has to become very large.
result = None
n = 2
for i0 in range(-n, n+1):
for i1 in range(-n, n+1):
for i2 in range(-n, n+1):
rp = r+np.dot(a0*H.T, [i0,i1,i2])
d = np.linalg.norm(rp)
if (result is None) or (result[1] > d):
result = (rp, d)
return result[0]
poscar_filename = 'Ni_fcc_unit.gga.relaxed.vasp'
cell_poscar = Poscar()
cell_poscar.read(poscar_filename)
cell_poscar.normalize_h_matrix()
cell_sc = make_super_cell(structure=cell_poscar,sc=[3,3,3])
_positions = [a.position for a in cell_sc.atomic_basis]
_n_atoms = len(_positions)
min_distances = []
for i in range(1,_n_atoms):
r1 = np.array(_positions[0])
r2 = np.array(_positions[i])
r = r2-r1
r_min = brute_minimum_image_convention(r=r,H=cell_sc.H,a0=cell_sc.a0)
d = np.linalg.norm(r_min)
min_distances.append(d*cell_poscar.a0*cell_poscar.a1)
min_distances = np.array(min_distances)
distances = np.linspace(
np.min(min_distances),
np.max(min_distances),
import argparse
from pypospack.io.vasp import Poscar
_description = 'Normalizes the H-matrix'
_parser = argparse.ArgumentParser(description=_description)
_parser.add_argument('--in', action='store', dest='filename_in', type=str)
_parser.add_argument('--out', action='store', dest='filename_out', type=str)
_args = _parser.parse_args()
_filename_in = _args.filename_in
_filename_out = _args.filename_out
print('filename_in:{}'.format(_filename_in))
_poscar = Poscar()
_poscar.read(_filename_in)
print('a0:{}'.format(_poscar.a0))
print('H:\n{}'.format(_poscar.H))
print('filename_out:{}'.format(_filename_out))
_poscar.normalize_h_matrix()
print(80 * '-')
print('a0:{}'.format(_poscar.a0))
print('H:\n{}'.format(_poscar.H))
_poscar.write(_filename_out)