Is there a package implementing the Box and Cox (1964) method for transforming data?


I’m new to Julia and looking to convert some of my workflow. I was wondering if there is an implementation of the maximum likelihood-like approach of Box and Cox (1964)?

I may be missing something, but isn’t the Box-Cox transformation rather trivial to implement? Why do you need a package for this?

ML should be easy with a generic optimization library like Optim.jl. Eg see a recent thread.

Yes, it is relatively trivial to implement. I just thought I would ask given that it’s commonly implemented in statistical packages in most data science / statistical oriented languages.

I just created a package :wink: Not registered though.


Incidentally, I like your use of 𝐱 in code. It solves the eternal elt vs x vs xs dilemma.

1 Like

For those who are interested, I’ve updated the package:

  1. Ability to specify custom lambda parameter
  2. Ability to scale results with geometric mean for better interpretability with units
  3. Ability to switch between two different log-likelihood implementations
  4. Better README and doc strings