Below you will see a measurement of a human heart beat interval as a function of time. I take the Fourier transform of this data (but a much longer measurement, obviously). However, notice that the wave form is not symmetric on the upward path and the downward path. This is related to Respiratory Sinus Arrhythmia (RSA), which is perfectly health, although the exact relationship between RSA and the data points here escapes me right now. Anyway, if I take the Fourier Transform of this kind of data, what artifact will this asymmetry create? Suppose I don't want to look at the RSA but only the frequency. Will the RSA distort the central frequency of the Fourier transform? Could it create the bimodal peak in the frequency domain that I sometimes see? Should I correct the data and somehow even out this waveform before I take the transform?
Thanks.
