On Sat, May 15, 2021 at 7:27 PM Anton Akhmerov <anton.akhmerov+kd@gmail.com> wrote:

Dear 张金龙,

Note the axis of the plot reads 1e-9+1, so the plot limits go from 1 - 0.2e-9 to 1 + 1e-9.

I still can't figure out the following correspondence: 1e-9+1 <---> 1 - 0.2e-9 to 1 + 1e-9. Any more hints? HY

This how matplotlib treats plots that are approximately constant.

Hope this helps, Anton

On Sat, 15 May 2021 at 13:24, 张金龙 <jlzhang1996@163.com> wrote:

...

Hello

I felt confusing, when I use pyplot.plt. The data in figure is different from the printed.

There are the codes:

# ================================================================

# Physics background

# ------------------

# Kwant features highlighted

# --------------------------

# - Numpy matrices as values in Builder

import kwant

import numpy as np

# For plotting

import matplotlib.pyplot as plt

# For matrix support

import tinyarray

import kwant.qsymm

# define Pauli-matrices for convenience

s_0 = tinyarray.array([[1, 0], [0, 1]])

s_x = tinyarray.array([[0, 1], [1, 0]])

s_y = tinyarray.array([[0, -1j], [1j, 0]])

s_z = tinyarray.array([[1, 0], [0, -1]])

mz=0 # AFM

mu=0.0 # stargged chemical potential

n_theta=12

def make_system(constant=1, W=30, L=30, mz=0, m_in=0, mu=0):

lat = kwant.lattice.honeycomb(a=constant,norbs=2, name=['a', 'b'])

# a, b = lat.sublattices

sym = kwant.TranslationalSymmetry(lat.vec((1, 0)))

syst = kwant.Builder()

def onsite_a(site,theta):

return mz*s_z+m_in*(np.cos(theta)*s_x+np.sin(theta)*s_y)+mu*s_0

def onsite_b(site,theta):

return -mz*s_z+m_in*(np.cos(theta)*s_x+np.sin(theta)*s_y)-mu*s_0

syst[(lat.a(x, y) for x in range(L) for y in range(W))] = onsite_a

syst[(lat.b(x, y) for x in range(L) for y in range(W))] = onsite_b

syst[lat.neighbors()] = s_0

# define lead

left_sym=kwant.TranslationalSymmetry(lat.vec((1, 0)))

left_lead=kwant.Builder(left_sym)

left_lead[(lat.a(0, y) for y in range(W))] = onsite_a

left_lead[(lat.b(0, y) for y in range(W))] = onsite_b

left_lead[lat.neighbors()] = s_0

syst.attach_lead(left_lead)

syst.attach_lead(left_lead.reversed())

return syst

def plot_conductance(syst, energy, theta_list):

# Compute conductance

data = np.zeros(n_theta)

num_i=-1

for theta_i in theta_list:

num_i=num_i+1

smatrix = kwant.smatrix(syst, energy, params=dict(theta=theta_i))

data[num_i]=smatrix.transmission(1, 0)

plt.figure()

# x=np.linspace(-1,1,n_theta)

# data_i=np.linspace(0,0,n_theta)

# print(x,data_i)

# plt.plot(x,data_i,'ow')

print(data)

plt.plot(theta_list, data)

plt.xlabel(r'$\theta/\pi$')

plt.ylabel("conductance [e^2/h]")

plt.savefig('tran_PHE.jpg')

#plt.show()

def main():

syst = make_system(m_in=0.0001)

# Check that the system looks as intended.

kwant.plot(syst,file='transport_lattice.pdf')

syst = syst.finalized()

theta_num=np.linspace(-np.pi,np.pi,num=n_theta)

plot_conductance(syst, energy=0.0001,

theta_list=theta_num)

if __name__ == '__main__':

main()

The result would be printed in terminal: [1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.].

However, the figure plotted is wrong.

--

Best Wishes Jinlong Zhang jlzhang@163.com

-- Assoc. Prof. Hongyi Zhao <hongyi.zhao@gmail.com> Theory and Simulation of Materials Hebei Vocational University of Technology and Engineering NO. 552 North Gangtie Road, Xingtai, China

Thank you for your answer. I think that it is the default option in matplotlib.pyplot, but I'm not used to this label. The plot ylimits go from 1e-8+1 to 9e-8+1. However, the scale of y-axis is 1、2、3、4、5、6、7、8、9, meanwhile label the 1e-8+1 in the top over y-axis. So I felt confusing whether the data is 1e-8+1、2e-8+2、3e-8+3、4e-8+4、...9e-8+9 or 1e-8+1、2e-8+1、3e-8+1、4e-8+1、...9e-8+1. With the suggestion of Pr. Anton Akhmerov, I try this codes: ####################### import numpy as np import matplotlib.pyplot as plt y=1+np.linspace(1,9,100)*1e-8 x=np.linspace(1,9,100) plt.plot(x,y) plt.show() ####################### The latter one is correct: 1e-8+1、2e-8+1、3e-8+1、4e-8+1、...9e-8+1. Sorry about that I don't know how to attach the pic in the community. jinlong zhang