How does the Box-Cox Transformation in Matlab work?

What is the explicitly Log-Likelihood Function (LLF) maximized?

Are the variance and mean values computed from the data?

Answer 1

[Tushar Athawale]

Hi,

The log-likelihood function being maximized with respect to parameter λ for the Box-Cox transformation is following:

-(n/2)*log(σ(Y)^2)+(λ-1)*(∑log(y_i)), where Y represents the data transformed using the Box-Cox transformation and n represents the number of data points.

The equations for the Box-Cox transformation can be found in the following link: http://www.mathworks.com/help/finance/boxcox.html

You can also learn more about the implementation of the Box-Cox transformation in MATLAB by studying the code for 'boxcox.m':

>> edit boxcox.m