Auto-/Cross-correlation and Auto-/Cross-covariance #939
Description
I'm surprised by the functions autocor, autocov, corsscor and corsscov as the books I'm familiar with use the same definition of autocorrelation an auto-covariance as on Wikipedia and similarly for cross-correlation and cross-covariance.
In essence, the difference between covariance and correlation lies in the subtraction of the mean, while in the StatsBase functions the difference is in the normalization.
What's the reason for you choice? Is this common in some fields?