scipy Fast fourier transform doesn't recognize the signal - python

i'm trying to get the frequency of a signal via fourier transform but it's not able to recognize it (sets the peak to f=0). Maybe something is wrong in my code (FULL reprudible code at the end of the page):
PF = fft.fft(Y[0,:])/Npoints #/Npoints to get the true amplitudes
ZF = fft.fft(Y[1,:])/Npoints
freq = fft.fftfreq(Npoints,deltaT)
PF = fft.fftshift(PF) #change of ordering so that the frequencies are increasing
ZF = fft.fftshift(ZF)
freq = fft.fftshift(freq)
plt.plot(freq, np.abs(PF))
plt.show()
plt.plot(T,Y[0,:])
plt.show()
where Npoints is the number of intervals (points) and deltaT is the time spacing of the intervals. You can see that the peak is at f=0
I show also a plot of Y[0,:] (my signal) over time where it's clear that the signal has a characteristic frequency
FULL REPRUDICIBLE CODE
import numpy as np
import matplotlib.pyplot as plt
#numerical integration
from scipy.integrate import solve_ivp
import scipy.fft as fft
r=0.5
g=0.4
e=0.6
H=0.6
m=0.15
#define a vector of K between 0 and 4 with 50 componets
K=np.arange(0.1,4,0.4)
tsteps=np.arange(7200,10000,5)
Npoints=len(tsteps)
deltaT=2800/Npoints #sample spacing
for k in K :
i=0
def RmAmodel(t,y):
return [r*y[0]*(1-y[0]/k)-g*y[0]/(y[0]+H)*y[1], e*g*y[0]/(y[1]+H)*y[1]-m*y[1]]
sol = solve_ivp(RmAmodel, [0,10000], [3,3], t_eval=tsteps) #t_eval specify the points where the solution is desired
T=sol.t
Y=sol.y
vk=[]
for i in range(Npoints):
vk.append(k)
XYZ=[vk,Y[0,:],Y[1,:]]
#check periodicity over P and Z with fourier transform
#try Fourier analysis just for the last value of K
PF = fft.fft(Y[0,:])/Npoints #/Npoints to get the true amplitudes
ZF = fft.fft(Y[1,:])/Npoints
freq = fft.fftfreq(Npoints,deltaT)
PF = fft.fftshift(PF) #change of ordering so that the frequencies are increasing
ZF = fft.fftshift(ZF)
freq = fft.fftshift(freq)
plt.plot(T,Y[0,:])
plt.show()
plt.plot(freq, np.abs(PF))
plt.show()

I can't pinpoint where the problem is. It looks like there is some problem in the fft code. Anyway, I have little time so I will just put a sample code I made before. You can use it as reference or copy-paste it. It should work.
import numpy as np
import matplotlib.pyplot as plt
from scipy.fft import fft, fftfreq
fs = 1000 #sampling frequency
T = 1/fs #sampling period
N = int((1 / T) + 1) #number of sample points for 1 second
t = np.linspace(0, 1, N) #time array
pi = np.pi
sig1 = 1 * np.sin(2*pi*10*t)
sig2 = 2 * np.sin(2*pi*30*t)
sig3 = 3 * np.sin(2*pi*50*t)
#generate signal
signal = sig1 + sig2 + sig3
#plot signal
plt.plot(t, signal)
plt.show()
signal_fft = fft(signal) #getting fft
f2 = np.abs(signal_fft / N) #full spectrum
f1 = f2[:N//2] #half spectrum
f1[1:] = 2*f1[1:] #actual amplitude
freq = fs * np.linspace(0,N/2,int(N/2)) / N #frequency array
#plot fft result
plt.plot(freq, f1)
plt.xlim(0,100)
plt.show()

Related

Signal processing with Fourier transform

I need to process a periodic signal and obtain its frequency.
This can be easily done with scipy.fft and it works fine.
However, I have a particular signal that is not strictly periodic. The period changes slightly but randomly.
Applying the fft to this signal is quite hard. Many peaks are obtained and I cannot extrapolate only the range of frequencies that are near the (1/period) I am interested in.
How can I do this?
This is a simple code snippet of what I am doing:
df = pd.read_csv('data.txt', header=None)
x = np.asarray(df.iloc[:, 0])
y = np.asarray(df.iloc[:, 1])
yf = fft(y)
xf = fftfreq(len(y))
plt.plot(xf, abs(yf))
You can find an example of such signal at the following GitHub repository, inside the README file: https://github.com/crazydecibel/Stack-Overflow-question
What about taking the weighted average of frequency of top energy bins?
import numpy as np
import matplotlib.pyplot as plt
your_data = '15.042,...' # your github data
t, y = np.array([list(map(float, row.split(','))) for row in your_data.split()]).T
# here is the solution in terms of t and y.
Y = np.fft.fftshift(np.fft.fft(y))
Ts = (t[-1] - t[0]) / (len(t) + 1) # sampling period
Fs = 1 / Ts # sampling frequency
f = np.fft.fftshift(np.fft.fftfreq(len(y))) * Fs
# get top 5%
top = np.argsort(abs(Y))[-10:]
plt.subplot(211)
plt.stem(f, abs(Y), 'o')
plt.stem(f[top], abs(Y[top]), '+r')
plt.xlim(-0.05 * Fs, 0.05 * Fs)
fc = np.sum(abs(f[top]*Y[top]**2))/np.sum(abs(Y[top])**2)
plt.subplot(212)
plt.plot(t, y)
plt.plot(t, np.sin(t*(2*np.pi*fc)))

Python, FFT and DC offset analysis

I test an FFT on a square signal (100Hz, 0V-5V) of 50% duty cycle and i don't understand why my DC offset is huge ?
In theory it should be 2.5V ?
Thanks,
Best regards.
The fundamental is ok and others harmonics are correct.
square signal 100Hz, TTL compatible 0V-5V, 50% duty cycle
FFT, DC offset problem, fundamental ok, harmonics ok
from scipy.fftpack import fft
from scipy import signal
import matplotlib.pyplot as plt
import numpy as np
#
# configuration
# time analyse = L * (1/Fsample)
#
L = 512 # lenght buffer
Fsample = 2000 # frequency sample
Fsignal = 100 # frequency
#********************************
# perdiode sample
Tsample = 1.0/Fsample
# time vector, start = 0s, stop = 0.1024s, step = (0.1024s / (1/Fe))
t = np.linspace(0.0, L*Tsample, L)
# signal definition, DC offet = 2.5V, Amplitude = 2.5V
signal = 2.5 + 2.5*signal.square(2 * np.pi * Fsignal * t, 0.5)
# plot time signal
plt.subplot(2,1,1)
plt.plot(t, signal)
# fft of time signal
yf = fft(signal)
# time vector of fft
xf = np.linspace(0.0, 1.0/(2.0*Tsample), L//2)
# plot spectral
plt.subplot(2,1,2)
plt.plot(xf, 2.0/L * np.abs(yf[0:L//2]))
On the last line, the normalization factor was wrong.
It was not 2/L but 1/L.
The correct normalization factor plt.plot(xf, 1.0/L * np.abs(yf[0:L//2]))
The code work fine now !
FFT correct amplitude

How to get a range of frequency bins for PSD

I am computing PSD of a signal, and I want the power from frequency range 0Hz to 20Hz. This is what i tried using linspace
df = pd.read_csv(path)
df = pd.DataFrame(df)
x = np.linspace(0, 20, 41)
dt = x[1] - x[0]
fs = 1 / dt
f,P = signal.welch(df, fs=5, nperseg=30, noverlap=5,axis=0)
Here, I get 6 frequency components from 0Hz to 2.5Hz, but I want to compute the power for frequency range till 20Hz
Can anyone help me here to do the same.
The PSD only goes from 0 to fs/2, you should specify the correct sample frequency in the fs parameter, 1024 in your case.
This example illustrates how to get the PSD for a sinusoidal signal:
import numpy as np
import matplotlib.pyplot as plt
from scipy.signal import welch
Fs = 1024 # Hz
Ts = 1/Fs
time = np.arange(0, 2, Ts) # 2 seconds
freqs = [20, 50, 100] # frequencies in Hz
x = np.zeros(len(time))
for f in freqs:
x += np.sin(2 * np.pi * f * time)
plt.plot(x)
f, P = welch(x, fs=Fs)
plt.figure()
plt.stem(f, P)

Filtering signal frequency in Python

I tried to filter some signal with fft.
The signal I am working on is quite complicated and im not really experienced in this topic.
That's why I created a simple sin wave 3Hz and tried to cut off the 3 Hz.
and so far, so good
import numpy as np
import matplotlib.pyplot as plt
from scipy.fftpack import fftfreq, irfft, rfft
t = np.linspace(0, 2*np.pi, 1000, endpoint=True)
f = 3.0 # Frequency in Hz
A = 100.0 # Amplitude in Unit
s = A * np.sin(2*np.pi*f*t) # Signal
dt = t[1] - t[0] # Sample Time
W = fftfreq(s.size, d=dt)
f_signal = rfft(s)
cut_f_signal = f_signal.copy()
cut_f_signal[(np.abs(W)>3)] = 0 # cut signal above 3Hz
cs = irfft(cut_f_signal)
fig = plt.figure(figsize=(10,5))
plt.plot(s)
plt.plot(cs)
What i expected
What i got
I don't really know where the noise is coming from.
I think it is some basic stuff, but i dont get it.
Can someone explain to to me?
Edit
Just further information
Frequency
yf = fft(s)
N = s.size
xf = np.linspace(0, fa/2, N/2, endpoint=True)
fig, ax = plt.subplots()
ax.plot(xf,(2.0/N * np.abs(yf[:N//2])))
plt.xlabel('Frequency ($Hz$)')
plt.ylabel('Amplitude ($Unit$)')
plt.show()
You could change the way you create your signal and use a sample frequency:
fs = 1000
t = np.linspace(0, 1000 / fs, 1000, endpoint=False) # 1000 samples
f = 3.0 # Frequency in Hz
A = 100.0 # Amplitude in Unit
s = A * np.sin(2*np.pi*f*t) # Signal
dt = 1/fs
And here the whole code:
import numpy as np
import matplotlib.pyplot as plt
from scipy.fftpack import fftfreq, irfft, rfft
fs = 1000
t = np.linspace(0, 1000 / fs, 1000, endpoint=False)
f = 3.0 # Frequency in Hz
A = 100.0 # Amplitude in Unit
s = A * np.sin(2*np.pi*f*t) # Signal
dt = 1/fs
W = fftfreq(s.size, d=dt)
f_signal = rfft(s)
cut_f_signal = f_signal.copy()
cut_f_signal[(np.abs(W)>3)] = 0 # cut signal above 3Hz
cs = irfft(cut_f_signal)
fig = plt.figure(figsize=(10,5))
plt.plot(s)
plt.plot(cs)
And with f = 3.0 Hz and (np.abs(W) >= 3):
And with f = 1.0 Hz:
Just some additional information about why A. As solution works better than yours:
A. A's model doesn't include any non-integer frequencies in its Solution and after filtering out the higher frequencies the result looks like:
1.8691714842589136e-12 * exp(2*pi*n*t*0.0)
1.033507502555532e-12 * exp(2*pi*n*t*1.0)
2.439774536202658e-12 * exp(2*pi*n*t*2.0)
-8.346741339115191e-13 * exp(2*pi*n*t*3.0)
-5.817427588021649e-15 * exp(2*pi*n*t*-3.0)
4.476938066992472e-14 * exp(2*pi*n*t*-2.0)
-3.8680170177940454e-13 * exp(2*pi*n*t*-1.0)
while your solution includes components like:
...
177.05936105690256 * exp(2*pi*n*t*1.5899578814880346)
339.28717376420747 * exp(2*pi*n*t*1.7489536696368382)
219.76658524130005 * exp(2*pi*n*t*1.9079494577856417)
352.1094590251063 * exp(2*pi*n*t*2.0669452459344453)
267.23939871205346 * exp(2*pi*n*t*2.2259410340832484)
368.3230130593005 * exp(2*pi*n*t*2.384936822232052)
321.0888818355804 * exp(2*pi*n*t*2.5439326103808555)
...
Please refer to this question regarding possible side effects of zeroing FFT bins out.

Strange FFT output python

I am trying to take the FFT and plot it. Problem is, my code works for small frequencies (like 50) but doesn't work for the bigger frequencies I need. What is going on with my code?! I expect to see a spike at the frequency of the sine wave I input, but the spike is at different frequencies depending on the sample spacing I use.
bins = 600
ss = 2048
freq = 44100
centerfreq = freq*bins/ss
# Number of samplepoints
N = ss
# sample spacing
T = 1 / 800.
x = np.linspace(0.0, N*T, N)
y = sin(2*np.pi*centerfreq*x)
yf = fft(y)
xf = np.linspace(0.0, 1.0/(2.0*T), N/2)
plt.plot(xf, 2.0/N * np.abs(yf[0:N/2]), 'r')
The code is right, you need to brush up your Fourier Theory and Nyquist Sampling Theorem and make sure the numbers make sense. The problem is with your x-axis scale. The plot function plots the first item in x with the first item in y, if x is not scaled up to your expectations, you are in for a surprise. You also see this if you plot a sinusoidal signal (sine wave) and expect 'degrees' and you get radians for instance. Its your duty to scale it up well so that it lines up to your expectation.
Refer to this SO answer https://stackoverflow.com/a/25735436/2061422.
from scipy import *
from numpy import *
from pylab import * # imports for me to get going
bins = 600
ss = 2048
freq = 44100
centerfreq = freq*bins/ss
print centerfreq
# Number of samplepoints
N = ss
# sample spacing
T = 1. / freq # i have decreased the spacing considerably
x = np.linspace(0.0, N*T, N)
sample_spacing = x[1] - x[0] # but this is the real sample spacing
y = sin(2*np.pi*centerfreq*x)
yf = fft(y)
xf = np.linspace(0.0, 1.0/(2.0*T), N/2)
freqs = np.fft.fftfreq(len(y), sample_spacing) # read the manual on this fella.
plt.plot(freqs[:N/2], 1.0/N * np.abs(yf[0:N/2]), 'r')
plt.grid()
plt.show()

Categories

Resources