Convert Time Domain Signal Data into Frequency Domain

Hi, I have dataset that contains time stamp and the acceleration data of three axes( x, y, and z) then i the data has been stored in an array called Segments This is an example of my data

 Time stamp    X           Y                Z
  0  1.0195313  0.16088867  -0.26391602
0  1.0976563  0.17456055  -0.33447266
0  1.2556152  0.24926758  -0.3774414
0  1.2314453  0.517334  -0.25732422
0  1.0212402  0.5761719  -0.09277344
0  1.0727539  0.51000977  0.007324219
0  1.1694336  0.32885742  -0.017822266
0  1.1247559  0.22924805  -0.10595703
0  1.0339355  0.27905273  -0.13623047
0  0.8273926  0.24560547  -0.080566406
0  0.75097656  0.25390625  -0.018310547
0  0.67626953  0.3046875  0.051757813
0  0.7282715  0.33764648  0.08911133
0  0.8227539  0.34545898  0.08325195
0  0.87939453  0.3413086  0.08300781
0  1.0527344  0.34326172  0.106933594
0  1.1125488  0.32128906  0.10058594
0  1.1477051  0.26416016  0.06347656
0  1.2807617  0.2680664  0.015136719
0  1.2800293  0.2668457  -0.006347656
1  0.7998047  0.2043457  -0.22460938
1  0.89868164  0.16308594  -0.18432617
1  0.9472656  0.107177734  -0.19458008
1  1.0109863  0.08178711  -0.20800781
1  1.0375977  0.111328125  -0.17626953
1  0.96313477  0.13378906  -0.13134766
1  1.0727539  0.15991211  -0.08935547
1  1.041748  0.2055664  -0.0637207
1  0.9003906  0.2607422  -0.013183594
1  0.94677734  0.23095703  0.025390625
1  1.0883789  0.24560547  0.021240234
1  1.1467285  0.22265625  -0.029785156
1  1.1020508  0.23388672  -0.08325195
1  0.88549805  0.22949219  -0.068603516
1  0.8017578  0.24829102  -0.03857422
1  0.83569336  0.27197266  -0.024902344
1  0.8642578  0.34106445  -0.037353516
1  0.91308594  0.3779297  -0.017333984
1  1.0136719  0.43603516  0.04321289
1  1.0756836  0.45166016  0.06567383
1  1.0771484  0.37670898  0.028076172

I have converted the time domain signal into frequency domain signal using *fft* function

I have sent the data of x-axis only into FFT function as illustrated in the following commands

    % FFTS = fft(X_Segments)
    plot(FFTS)

NOTE( X_Segments is an array that contains the acceleration data of X-axis only)

the graph was so strange!as shown in the pic

Can anyone help me? Thank you so much

[Image Analyst]

The variable type "dataset" is deprecated and it's recommended to use tables instead.

Is your variable Segments a table or a regular N-by-4 matrix (like a double)?

The fft() function takes a 1-D signal. Which column did you send into fft()?

[neamah al-naffakh]

hi mate thanks for your answer, actully, Segments contains the data of X,Y, and Z. that means it's contains 3 columns(x,y, and z). then i have send the data of x axis only into FFT function see the following commands that i have used

    % rr = fft(X_Segments)
plot(rr)

where X_Segments contains the accelration data of X axis only

[Star Strider]

The time has to be unique for each point, meaning that you have to know your sampling interval and sampling frequency. You cannot have an accurate fft if you don’t know your sampling frequency.

[Image Analyst]

So

X_Segments = Segments(:, 2);

right?

Then your code looks about right. I believe it will take only the real part for plotting though, and ignore the imaginary part. Perhaps you want pwelch() instead of fft(). Check it out.

[neamah al-naffakh]

thank you so much mate, see the following graph! I have got only one peak for 320 value of X data !

[Rehanos_boyzee]

Can you please share your code as im trying to do something similar

Answer 1

[Image Analyst]

Did I help you above such that you have no more questions?

[neamah al-naffakh]

iI am not sure to be honest, I am wondering if the graph is correct or not? because when I draw the data in time domain i got many peaks in the signal ! while I got only on in frequency domain signal?

[Image Analyst]

In general, you could have lots of peaks in both domains due to a variety of reasons.