welford - Online Variance with Welfords method

Computes the variance of a dataset by incrementally adding values to an accumulator. Welfords method is more numerically stable than the standard method.

Welfords method goes back to a 1962 paper by B. P. Welford and is presented in Donald Knuth’s Art of Computer Programming, Vol 2.

This implemenation is modelled after the following:

• Accurately computing running variance
• Computing skewness and kurtosis in one pass

See also:

• WP: Online Algorithm
• WP: Parallel Algorithm