Hi all!

I want to measure the coherence of my input-output measurement data. The experiment was performed on a plate using an accelerometer and actuator. I took three data sets for each point on the plate (all with same time scale). The problem is that the sweep frequency duration for each data set is different. I want to know whether i can average the data in time domain without caring about the different sweep length or should i first take the fft of each measurement and then average in frequency domain. I want some suggestions on how to measure coherence from the data i have.

Thanks in anticipation.