python 中绘制的功率谱的 y 值存在问题

Question

I have got some data and I used MATLAB to plot the power spectrum. But I struggled to translate the same instructions to Python and obtain the exact plot. The shape of the plot obtained in Python seems to be fine except that the y-values are positive for some reason?, Can anyone please tell me where I went wrong. Below, I have attached the instructions in MATLAB and python along with the two plots obtained.

MATLAB:                                                    
N=length(signal);
xdft = fft(signal);
xdft = xdft(1:N/2+1);
psdx=(1/(fs*N))*abs(xdft).^2;
psdx(2:end-1) = 2*psdx(2:end-1);
freq = 0:fs/N:fs/2;
plot(freq, 10*log10(psdx))

Python:
N = len(signal)
psdx = pow(np.fft.rfft(signal),2)
psdx[1:len(psdx)-1] = 2*psdx[1:len(psdx)-1]
f = np.arrange(0, (fs/2) + (1/N), fs/N)
plt.plot(f, (10*np.log10(psdx)))
plt.show() 

Signal can be found using: https://gofile.io/d/tlJm7F/edit Note that you need to load/read the file first

Answer 1

from a quick glance, I think you are not dividing psdx by fs*N from your matlab code. That could be the reason. I would print values of psdx in matlab and check if it matches with the python.

Answer 2

I just used MATLAB code as it is without using numpy.fft.rfft function. Also fs I assumed to be equal to 120. Please change as your code wants it to be.

The plot looks like below, which to me is closer to your MATLAB code.

The code is here.

import matplotlib.pyplot as plt
import numpy as np
with open("./FileName2.txt","r") as FileHandle:
    Signal = FileHandle.readlines()
signal = np.array([i.rstrip("\n") for i in Signal])
fs = 120
N = len(signal)
xdft = np.fft.fft(signal)
xdft = xdft[:int(N/2)+1]
psdx=(1/(fs*N))*np.abs(xdft)**2
f = np.arange(0, (fs/2) + (1/N), fs/N)
plt.plot(f, (10*np.log10(psdx)))
plt.show()