covariance and correlation

[rodinalex]

There appears to be an issue in the autocovariance and autocorrelation functions. After taking the dot product of the two vectors, the result is divided by the length of the total vector lx. Shouldn't the division be by (lx - lag) to correspond to the number of terms in the dot-sum since we are looking for the mean of the sum?

[andreasnoack]

Dup of #562