I have the following code (a sample data file is attached).
load("sample_signal.mat");
whos

figure
subplot(2,1,1)
hold on; grid minor
plot(signals.signal_01.signal)
plot(signals.signal_02.signal)
plot(signals.signal_03.time, signals.signal_03.theta, ‘k’)
title("Signal");

subplot(2,1,2)
hold on; grid minor
fs = 1;
[freq, amplitude, phase] = generic_fft(signals.signal_01.signal, fs);
plot(freq, amplitude)
[freq, amplitude, phase] = generic_fft(signals.signal_02.signal, fs);
plot(freq, amplitude)
fs = 10;
[freq, amplitude, phase] = generic_fft(signals.signal_03.theta, fs);
plot(freq, amplitude, ‘k’)
xlim([0 0.5])
xlabel("$omega$ [rad/s]")
ylabel("fft")
title("FFT");

% fft function

function [freq, amplitude, phase] = generic_fft(signal, fs)
% Remove NaNs
signal = signal(~isnan(signal));
% Length of the signal
n = length(signal);
% Perform FFT
fft_values = fft(signal);
% Compute single-sided amplitude spectrum
amplitude = abs(fft_values / n); % Normalize by length
amplitude = amplitude(1:floor(n / 2) + 1); % Keep positive frequencies
amplitude(2:end-1) = 2 * amplitude(2:end-1);
% Compute single-sided phase spectrum
phase = angle(fft_values(1:floor(n / 2) + 1));
% Frequency vector
freq = (0:(n / 2)) * (fs / n);
end

Now, the signal plotted in black shows a clear, dominant frequency in the FFt, but the other two signals show additional "noise" (see, after 0.1 rad/s). Is it possible to maybe filter these out or only show the dominant frequency?
The code is rough and the data structure is also not ideal (I saved the data mid-simulation for this question 🙂 ). Also, any tips to improve the fft function is also very appreciated. TIA!I have the following code (a sample data file is attached).
load("sample_signal.mat");
whos

figure
subplot(2,1,1)
hold on; grid minor
plot(signals.signal_01.signal)
plot(signals.signal_02.signal)
plot(signals.signal_03.time, signals.signal_03.theta, ‘k’)
title("Signal");

subplot(2,1,2)
hold on; grid minor
fs = 1;
[freq, amplitude, phase] = generic_fft(signals.signal_01.signal, fs);
plot(freq, amplitude)
[freq, amplitude, phase] = generic_fft(signals.signal_02.signal, fs);
plot(freq, amplitude)
fs = 10;
[freq, amplitude, phase] = generic_fft(signals.signal_03.theta, fs);
plot(freq, amplitude, ‘k’)
xlim([0 0.5])
xlabel("$omega$ [rad/s]")
ylabel("fft")
title("FFT");

% fft function

function [freq, amplitude, phase] = generic_fft(signal, fs)
% Remove NaNs
signal = signal(~isnan(signal));
% Length of the signal
n = length(signal);
% Perform FFT
fft_values = fft(signal);
% Compute single-sided amplitude spectrum
amplitude = abs(fft_values / n); % Normalize by length
amplitude = amplitude(1:floor(n / 2) + 1); % Keep positive frequencies
amplitude(2:end-1) = 2 * amplitude(2:end-1);
% Compute single-sided phase spectrum
phase = angle(fft_values(1:floor(n / 2) + 1));
% Frequency vector
freq = (0:(n / 2)) * (fs / n);
end

Now, the signal plotted in black shows a clear, dominant frequency in the FFt, but the other two signals show additional "noise" (see, after 0.1 rad/s). Is it possible to maybe filter these out or only show the dominant frequency?
The code is rough and the data structure is also not ideal (I saved the data mid-simulation for this question 🙂 ). Also, any tips to improve the fft function is also very appreciated. TIA! I have the following code (a sample data file is attached).
load("sample_signal.mat");
whos

figure
subplot(2,1,1)
hold on; grid minor
plot(signals.signal_01.signal)
plot(signals.signal_02.signal)
plot(signals.signal_03.time, signals.signal_03.theta, ‘k’)
title("Signal");

subplot(2,1,2)
hold on; grid minor
fs = 1;
[freq, amplitude, phase] = generic_fft(signals.signal_01.signal, fs);
plot(freq, amplitude)
[freq, amplitude, phase] = generic_fft(signals.signal_02.signal, fs);
plot(freq, amplitude)
fs = 10;
[freq, amplitude, phase] = generic_fft(signals.signal_03.theta, fs);
plot(freq, amplitude, ‘k’)
xlim([0 0.5])
xlabel("$omega$ [rad/s]")
ylabel("fft")
title("FFT");

% fft function

function [freq, amplitude, phase] = generic_fft(signal, fs)
% Remove NaNs
signal = signal(~isnan(signal));
% Length of the signal
n = length(signal);
% Perform FFT
fft_values = fft(signal);
% Compute single-sided amplitude spectrum
amplitude = abs(fft_values / n); % Normalize by length
amplitude = amplitude(1:floor(n / 2) + 1); % Keep positive frequencies
amplitude(2:end-1) = 2 * amplitude(2:end-1);
% Compute single-sided phase spectrum
phase = angle(fft_values(1:floor(n / 2) + 1));
% Frequency vector
freq = (0:(n / 2)) * (fs / n);
end

Now, the signal plotted in black shows a clear, dominant frequency in the FFt, but the other two signals show additional "noise" (see, after 0.1 rad/s). Is it possible to maybe filter these out or only show the dominant frequency?
The code is rough and the data structure is also not ideal (I saved the data mid-simulation for this question 🙂 ). Also, any tips to improve the fft function is also very appreciated. TIA! fft, filter, plot MATLAB Answers — New Questions

​