1. k-point sampling
K-point sampling
이번장에서는 NanoCore를 이용해 그래핀 모델을 생성하고, k-point sampling 과정을 통해 전체 에너지가 수렴하는 k-point를 구할 것이다.
from NanoCore import *
import numpy as np
# Modeling
atoms = carbonlab.grp(0,0)
# Number of system's atoms
num_atoms = len(atoms._atoms)
# Test for supercell
kpoint_sample = [1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25]
Energy = [] # given energy
for i in kpoint_sample:
sim = s2.Siesta(atoms)
sim.set_option('CGsteps', 0)
sim.set_option('kgrid', [i, i, 1])
sim.set_option('kshift', [0.5,0.5,0.0]) # set k shift from gamma
sim.set_option('MixingWt', 0.10) # adjust mixing weight (density)
sim.set_option('BasisSize', 'SZ') # adjust basis size
sim.run(mpi=1, nproc=8)
energy = s2.get_total_energy()
Energy.append(energy)
f = open('Ediff.txt', 'w')
for i in range(len(Energy)+1):
if i == 0:
f.write('k-points : %02d %02d %02d, Energy = %9.5f, Ediff= %s' % (2*i+1, 2*i+1, 1, Energy[i], 'Not_intial_difference'))
f.write('\n')
elif i < len(Energy):
f.write('k-points : %02d %02d %02d, Energy = %9.5f, Ediff=%9.5f' % (2*i+1, 2*i+1, 1, Energy[i], Energy[i]-Energy[i-1]))
f.write('\n')
elif i == len(Energy):
f.write('over')
f.write('\n')
else:
pass
f.close()
k-points : 01 01 01, Energy = -321.36050, Ediff= Not_intial_difference
k-points : 03 03 01, Energy = -322.42307, Ediff= -1.06257
k-points : 05 05 01, Energy = -322.33655, Ediff= 0.08653
k-points : 07 07 01, Energy = -322.32165, Ediff= 0.01490
k-points : 09 09 01, Energy = -322.32810, Ediff= -0.00645
k-points : 11 11 01, Energy = -322.32556, Ediff= 0.00253
k-points : 13 13 01, Energy = -322.32485, Ediff= 0.00072
k-points : 15 15 01, Energy = -322.32595, Ediff= -0.00110
k-points : 17 17 01, Energy = -322.32532, Ediff= 0.00063
k-points : 19 19 01, Energy = -322.32518, Ediff= 0.00014
k-points : 21 21 01, Energy = -322.32555, Ediff= -0.00036
k-points : 23 23 01, Energy = -322.32530, Ediff= 0.00024
k-points : 25 25 01, Energy = -322.32526, Ediff= 0.00005
over
import os, sys
def file_len(filename):
with open(filename) as f:
for i, l in enumerate(f):
pass
return i+1
tlines = file_len('RUN.fdf')
struct_info1 = []
struct_info2 = []
f = open('RUN.fdf', 'r')
for i in range(tlines):
line = f.readline()
words = line.split()
struct_info1.append(line)
struct_info2.append(words)
f.close()
# Get line DM.Tolerence
for i in range(len(struct_info1)):
if len(struct_info2[i]) >= 2:
if (struct_info2[i][0]) =='DM.Tolerance':
startnum = i
else:
pass
tolerance = []
tolerance = struct_info2[startnum][1]
Tol = int(tolerance[4])
tole=1.00000
for i in range(Tol):
tole=tole*0.1
# Checking Energy difference and find proper k-point
tline2 = file_len('Ediff.txt')
struct_info3 = []
struct_info4 = []
f = open('Ediff.txt', 'r')
for i in range(tline2):
line = f.readline()
words = line.split()
struct_info3.append(line)
struct_info4.append(words)
f.close()
#---------------------------------------------------------
for i in range(len(struct_info3)):
if str(struct_info3[i].strip()) == 'over':
blacknum = i+1
else:
pass
for j in range(1, blacknum-1):
if abs(float(struct_info4[j][9])) <= tole:
print('kpoint= %s' % struct_info4[j][2])
kpoint=2*j+1
else:
print('kpoint is not converage')
print('Converge kpoint = %d' % kpoint)
결과정리
plt.plot(kpoint_sample, Energy)
plt.xlabel("Kpoint")
plt.ylabel("Energy (eV)")
plt.savefig('k-test.png', dpi=300, bbox_inches ="tight")
plt.show()
