How to design an arbitrary phase flat magnitude FIR filter?

I need to design an FIR filter with flat magnitude, but arbitrary phase. This seems like a pretty straight forward request, but I have searched through all of the FIR functions and cannot find anyting that will work. I most recently experimented with fdesign.arbmagnphase, but the results did not look correct. Plus, it gives me a design object when what I want are the FIR filter coefficients that I can export to a CSV file and load into a DSP.

If you'd like to try running the code, here's a link to download the phase variable below.

Fs = 192000;
magnitudeFlat = zeros(height(Fs),1);
phase = phaseA - phaseB; % The phase specification is based on the difference of two other responses.

% convert to complex vector
complexVector = @(mag_dB,phase_radians) [10.^(mag_dB/20) .* exp(1j*(phase_radians))];
Z = complex(magnitudeFlat,phaseDiff);

% design FIR filter
n = 2^12;
Nyq = Fs/2;
f = frequencyHz(1:Nyq) / Nyq; f(end)=1; % Normalize the last row to 1 since it's off by a tiny amount.
b = fir2(n,f,Z(1:Nyq)); % So far, this is the only FIR function I have found that will take in a complex array.

% plot filter kernel
semilogx(abs(fft(b))*2) % The resulting plot does not have flat magnitude.

Answer 1

[Nathan Lively]

User error. I created use function complexVector, but was calling complex. ¯\_(ツ)_/¯