Transformations¶
Box Cox power transformation¶
The Box-Cox power transformation is a data transformation to achieve a better normality and to stabilize the variance. In MCRA, the transformation parameter p in (yp−1)/p is determined by maximizing the log-likelihood function
l(p)=−nslog[1nn∑i=1(y(p)i−¯y(p))2]+(p−1)n∑i=1logyi
where i indexes the n observations and
¯y(p)=1nn∑i=1y(p)i
is the average of the y(p)i (Box & Cox, 1964) [Box et al., 1964].