frequency to time domain

I have magnitude and phase of signal in frequency domain and I need to have it in time domain. I know I have to use ifft function, but I heard about mirroring and concatenating these values and I'm kinda lost. I'd appreciate any help with that

I have tried doing something like this, but I'm clearly missing the point

%A - magnitude
z = A .* exp(i*phase);
X = ifft(z);
figure, plot(abs(X))
figure, plot(angle(X))

These are my magnitude and phase in frequency domain

Answer 1

[Star Strider]

First, the phase must be in radians, and the highest frequency in both plots must be the Nyquist frequency for any inversion to work.

Then, experiment with this:

Fs = 2*max(frequency_vector)
t = linspace(0, 2*numel(frequency_vector), 2*numel(frequency_vector)).'/Fs;     % Time Vector
z = [A(:) .* exp(1i*phase(:)); conj(A(:) .* exp(1i*phase(:)))];                 % Force Column Vectors & Concatenate
X = ifft(z);

figure
plot(t, real(X))

There should not be any imaginary values in ‘X’, however they could be present although extremely small, so this would be the approach I would take. Depending upon how ‘z’ was calculated, there could be a scaling issue of numel(t)/2 that it could be necessary to multiply by to recover ‘X’ correctly.

I cannot be certain that this will work with your data, however it should at least get you started.

[NASRIN AKTER]

Hello

I have 2 signals and between their frequency spectrum I was expecting for a shift in frequency at maximum amplitude. I got a shift but how do I prove that this frequency shift is not because of a phase shift in time domain? I am attaching an image of the signals.

Thanks in advance

[Star Strider]

Please post this as a new Question. It does not actually pertain to this problem, and there is not enuogh information. I will delete it and this Comment in a few hours.