How to Plot fft of ascii values in MATLAB? [closed

问题:

Closed. This question does not meet Stack Overflow guidelines. It is not currently accepting answers.

---

Want to improve this question? Update the question so it's on-topic for Stack Overflow.

Closed 6 years ago.

I have a ascii file containing 2048 x and y value pairs. I just want to know how to plot fft of y in MATLAB. I am writing following MATLAB code but could not be able to find appropriate result.

How can I do this? This is what I have tried:

I = load('data1.asc');

for i = 1:2048
    y = I(:,2);
end

plot(x)

Fs = 40000;                    
T = 1/Fs;                   
L = 2000;     
NFFT = 2^nextpow2(L);
Y = abs(fft(y,NFFT))/L;
f = Fs/2*linspace(0,1,NFFT/2+1);

figure, plot(f,2*abs(Y(1:NFFT/2+1))) 
axis([0 40000 0 40])
xlabel('Frequency (Hz)')
ylabel('|Y(f)|')

回答1:

Instead of diving straight into MATLAB's FFT routine you should instead consider using the periodogram function. When people say "FFT" they usual mean PSD or periodogram, i.e. a plot of power spectral density using a suitably windowed and FFTed sample. The periodogram function in MATLAB takes care of all the details of this for you, i.e. applying a window function, calculating the FFT, deriving magnitude from FFT output, appropriate scaling of axes, and even plotting if required.

Note: periodogram is in the MATLAB Signal Processing Toolbox - if you do not have access to this then you can either consider using Octave (free MATLAB clone) which has a periodogram clone, otherwise you would need to put the various building blocks together yourself:

• window function
• FFT
• calculate magnitude
• take scare of scaling of freq and magnitude values
• plot PSD

回答2:

The fft part of the code looks good to me. However, this bit doesn't make much sense:

for i = 1:2048
    y = I(:,2);
end

What are you trying to do here? You're not using the loop index i at all in the for loop.

Also, I assume y is of length 2000, can you confirm? Otherwise L = 2000 should be changed to L = length(y). Similarly, I assume that the sampling frequency of the data is 40kHz, otherwise Fs = 40000 is not correct.

EDIT following discussion in comments:

With the data that you have provided, I get the same results. The only thing I did is exclude the last data point from the analysis when it drops to zero. The way you read the data still doesn't make sense to me. Note I am using Octave, not MATLAB, but the code should give the same results in MATLAB.

load('ascii_value.txt')
y = ascii_value(1:end-1,2);
plot(y)
L=length(y);
Fs = 40000;
T = 1/Fs;
NFFT = 2^nextpow2(L);
Y = abs(fft(y,NFFT))/L;
f = Fs/2*linspace(0,1,NFFT/2+1);
figure, plot(f,2*abs(Y(1:NFFT/2+1)))
axis([0 40000 0 40])
xlabel('Frequency (Hz)')
ylabel('|Y(f)|')

The signal looks like this:

and the FFT like this:

Note: if you are sampling at 40 kHz, your FFT can only go up to 20kHz (Nyquist frequency).