Helper function for testing assumption of autocorrelation processing
test.sigma.assumption.Rd
The approach we're using is valid if autocorr (red lines) is below mean firing rate (blue line). If so, then the equation we're using to convert spiking autocorr to guassian sigma should work.