# This Source Code Form is subject to the terms of the Mozilla Public
# License, v. 2.0. If a copy of the MPL was not distributed with this
# file, You can obtain one at https://mozilla.org/MPL/2.0/.
from numbers import Integral
import numpy as np
from sisl import Grid
from sisl._internal import set_module
from .._help import grid_reduce_indices
from ..sile import add_sile, sile_fh_open
from .car import carSileVASP
from .sile import SileVASP
__all__ = ["chgSileVASP"]
@set_module("sisl.io.vasp")
class chgSileVASP(carSileVASP):
"""Charge density plus geometry
This file-object handles the charge-density from VASP
"""
[docs] @sile_fh_open(True)
def read_grid(self, index=0, dtype=np.float64, **kwargs):
"""Reads the charge density from the file and returns with a grid (plus geometry)
Parameters
----------
index : int or array_like, optional
the index of the grid to read. For a spin-polarized VASP calculation 0 and 1 are
allowed, UP/DOWN. For non-collinear 0, 1, 2 or 3 is allowed which equals,
TOTAL, x, y, z charge density with the Cartesian directions equal to the charge
magnetization. For array-like they refer to the fractional
contributions for each corresponding index.
dtype : numpy.dtype, optional
grid stored dtype
spin : optional
same as `index` argument. `spin` argument has precedence.
Returns
-------
Grid : charge density grid with associated geometry
"""
index = kwargs.get("spin", index)
geom = self.read_geometry()
V = geom.lattice.volume
rl = self.readline
# Now we are past the cell and geometry
# We can now read the size of CHGCAR
rl()
nx, ny, nz = list(map(int, rl().split()))
n = nx * ny * nz
is_index = True
if isinstance(index, Integral):
max_index = index + 1
else:
is_index = False
max_index = len(index)
vals = []
vext = vals.extend
i = 0
while i < n * max_index:
vext(rl().split())
i = len(vals)
if i % n == 0 and i < n * max_index:
# Each time a new spin-index is present, we need to read the coordinates
j = 0
while j < geom.na:
j += len(rl().split())
# one line of nx, ny, nz
rl()
# Cut size before proceeding (otherwise it *may* fail)
vals = np.array(vals).astype(dtype, copy=False)
if is_index:
val = vals[n * index : n * (index + 1)].reshape(nz, ny, nx)
else:
vals = vals[: n * max_index].reshape(-1, nz, ny, nx)
val = grid_reduce_indices(vals, index, axis=0)
del vals
# Make it C-ordered with nx, ny, nz
val = np.swapaxes(val, 0, 2) / V
# Create the grid with data
# Since we populate the grid data afterwards there
# is no need to create a bigger grid than necessary.
grid = Grid([1, 1, 1], dtype=dtype, geometry=geom)
grid.grid = val
return grid
# CHG has low-precision, so the user should prefer CHGCAR
add_sile("CHG", chgSileVASP, gzip=True)
add_sile("CHGCAR", chgSileVASP, gzip=True)