Source code for nanotools.totalenergy

# -*- coding: utf-8 -*-
"""
Created on 2020-05-11

@author: Vincent Michaud-Rioux
"""

from nanotools.system import (
    System,
    calc_magnetic_moments_collinear,
    calc_magnetic_moments_non_collinear,
)
from nanotools.base import Base, Quantity
from nanotools.energy import Energy
from nanotools.io.calculators import solve_generic
from nanotools.solver import Solver
from nanotools.utils import read_field, dict_converter
import attr
import matplotlib.pyplot as plt
import numpy as np
import scipy
import scipy.signal


[docs] @attr.s class TotalEnergy(Base): """``TotalEnergy`` class. Examples:: a = 2.818 # lattice constant (ang) cell = Cell(avec=[[0.,a,a],[a,0.,a],[a,a,0.]], resolution=0.12) fxyz = [[0.00,0.00,0.00],[0.25,0.25,0.25]] atoms = Atoms(fractional_positions=fxyz, formula="GaAs") sys = System(cell=cell, atoms=atoms) sys.kpoint.set_grid([5,5,5]) calc = TotalEnergy(sys) calc.solve() print(calc.energy.etot) Attributes: system: Object containing system related parameters. energy: Object containing the total energy and its derivatives (force, stress, etc.). solver: Object containing solver related parameters. """ # input is dictionary with default constructor system: System = attr.ib( converter=lambda d: dict_converter(d, System), validator=attr.validators.instance_of(System), ) # optional energy: Energy = attr.ib( factory=Energy, converter=lambda d: dict_converter(d, Energy), validator=attr.validators.instance_of(Energy), ) solver: Solver = attr.ib( factory=Solver, converter=lambda d: dict_converter(d, Solver), validator=attr.validators.instance_of(Solver), ) classname: str = attr.ib() @classname.default def _classname_default_value(self): return self.__class__.__name__ def __attrs_post_init__(self): # atomic system if self.system is None: raise Exception('"system" parameters not found in input file.') self.solver.eig.set_target_irange( ispin=self.system.hamiltonian.ispin, valence_charge=self.system.atoms.valence_charge, n_orb=self.system.atoms.get_number_of_orbitals(), )
[docs] def get_total_magnetic_moment(self): """Returns the total magnetic moment.""" return self.system.atoms.get_total_magnetic_moment()
[docs] def get_magnetic_moments(self): """Returns the magnetic moments.""" magnetic_moments = self.system.atoms.get_magnetic_moments() if magnetic_moments is None: natm = self.system.atoms.get_number_of_atoms() if self.system.hamiltonian.ispin == 1: magnetic_moments = np.zeros((natm)) elif self.system.hamiltonian.ispin == 2: denpath = self.solver.restart.densityPath magnetic_moments = calc_magnetic_moments_collinear( self.system.cell, self.system.atoms, denpath ) elif self.system.hamiltonian.ispin == 4: denpath = self.solver.restart.densityPath magnetic_moments = calc_magnetic_moments_non_collinear( self.system.cell, self.system.atoms, denpath ) self.system.atoms.set_magnetic_moments(magnetic_moments) return magnetic_moments
[docs] def get_number_of_atoms(self): """Returns the number of atoms.""" return self.system.atoms.get_number_of_atoms()
[docs] def get_potential(self, name, units="si"): """Returns the number of atoms. Args: name (string): Path to the HDF5 file containing the potential data. units (string, optional): Unit of energy for the potential. """ filename = self.solver.restart.densityPath if name == "electrostatic": vna = read_field(filename, "potential/neutralatom") vdh = read_field(filename, "potential/deltahartree") v = vna + vdh elif name == "deltahartree": v = read_field(filename, "potential/deltahartree") elif name == "effective": v = read_field(filename, "potential/effective") elif name == "external": v = read_field(filename, "potential/external") elif name == "neutralatom": v = read_field(filename, "potential/neutralatom") elif name == "xc": v = read_field(filename, "potential/xc") else: raise ValueError("No potential %s." % (name)) if units == "atomic": v.ito("hartree") elif units != "si": raise ValueError("Invalid units %s." % (units)) return v
[docs] def get_total_energy(self): """Returns the total energy.""" return self.energy.etot
[docs] def get_total_energy_per_atom(self): """Returns the total energy per atom.""" n = self.get_number_of_atoms() etot = self.get_total_energy() return etot / n
[docs] def get_average_field(self, field, axis): """Loads field from ``self.solver.restart.densityPath`` and compute the average along a given axis. Args: field (str/NDArray): str: HDF5 path. For example, "potential/effective". NDArray: 3d array. axis (int): Axis remaining after averaging. The value should be between 0 and 2. Returns: NDArray: Average field. """ if axis not in [0, 1, 2]: raise ValueError("Axis must be in [0,1,2].") avg = list(range(3)) avg.pop(axis) if type(field) == str: field = self.get_field(field) return np.mean(field.m, axis=tuple(avg))
[docs] def get_field(self, field): """Loads field from ``self.solver.restart.densityPath``.""" filename = self.solver.restart.densityPath field = read_field(filename, field) return field
[docs] def plot_field(self, field, axis, filename=None, show=True): """Plots a field averaged along a given axis. Args: field (str): Path in the HDF5 filed pointed to by self.solver.restart.densityPath. axis (int): Axis remaining after averaging. The value should be between 0 and 2. filename (str, optional): If not None, then the figure is saved to filename. show (bool, optional): If True block and show figure. If False, do not show figure. Returns: fig (:obj:`matplotlib.figure.Figure`): A figure handle. """ if axis not in [0, 1, 2]: raise Exception("axis must take the value 0, 1 or 2.") if isinstance(field, np.ndarray): vavg = field elif isinstance(field, Quantity): vavg = field.m else: vavg = self.get_average_field(field, axis) x = self.system.cell.get_grid(axis=axis) lx = self.system.cell.get_length(axis=axis) fig = plt.figure() plt.plot(x, vavg) plt.xticks(np.arange(7) / 6 * lx.m) plt.grid(axis="x") plt.xlabel(f"position ({self.system.cell.avec.u})") plt.ylabel("field average (eV)") if show: plt.show() if filename is not None: fig.savefig(filename) return fig
[docs] def smooth_field(self, field, axis, width=2.0, shape="erf"): """Smooths a field by circular convolution. Args: field (str/NDArray): str: Path in the HDF5 filed pointed to by self.solver.restart.density. For example, "potential/effective". NDArray: 3d array. axis (int): Axis remaining after averaging. The value should be between 0 and 2. width (float): Unit system ('atomic' or 'si') """ field = self.get_average_field(field, axis) field3 = np.concatenate((field, field, field)) x = self.system.cell.get_grid(axis=axis) dx = self.system.cell.get_dx(axis) lx = self.system.cell.get_length(axis=axis) x3 = np.concatenate((x - lx, x, x + lx)) if shape == "n-points": cn = int(width // dx.m) # half-stencil width cn += cn % 2 cn = int(cn // 2) conv = np.zeros_like(x) conv[0 : cn + 1] = 1.0 / (2 * cn + 1) conv[-1 : -cn - 1 : -1] = 1.0 / (2 * cn + 1) conv = scipy.linalg.circulant(conv) sfield = field for i in np.arange(4): sfield = np.matmul(conv, sfield) else: pulse = 0.5 * (scipy.special.erf((x3.m - width / 2.0)) + 1.0) - 0.5 * ( scipy.special.erf((x3.m + width / 2.0)) + 1.0 ) pulse = pulse / np.sum(pulse) / dx sfield = field for i in np.arange(2): sfield = ( scipy.signal.convolve(np.tile(sfield, 3), pulse, mode="full") * dx ) sfield = sfield[2 * x.size : 3 * x.size] return sfield
[docs] def solve(self, input="nano_scf_in", output="nano_scf_out"): """ Performs a self-consistent calculation to obtain the ground state total energy. Args: input (str): The object is saved to an input file ``input`` to be read by the Fortran program. output (str): The results (with various extensions) are moved to files ``output`` and the results are loaded to the object. """ output = solve_generic(self, "scf", input, output) self._update(output + ".json") if self.energy.include_dftd3: self.energy.set_dftd3_forces(self.system) self.write(output + ".json")
[docs] def supercell(self, T): """ Replaces the underlying ``System`` object with a supercell ``System`` object. Args: T (NDArray): A (3x3) linear transformation. Returns: """ self.system.supercell(T=T) self.solver.restart.clear_paths() self.solver.eig.set_target_irange( ispin=self.system.hamiltonian.ispin, valence_charge=self.system.atoms.valence_charge, n_orb=self.system.atoms.get_number_of_orbitals(), )