I feel like I’m missing something. I have a length 1024 vector and want to take the FFT of that vector. I would like to take a length 2048 FFT of that vector. Is the proper solution to zero-pad the vector manually, and then take the FFT?
Depends on what you mean mathematically by “the length-2048 FFT” of a length-1024 vector; you are not using standard mathematical terms.
fft computes a discrete Fourier transform (DFT); what you do with it is up to you.