# On Box-Cox transform in regression models

A few days ago, a former student of mine, David, came back to me about Box-Cox tests in linear models. It made me look more carefully at the test, and I do not understand what is computed, to be honest. Let us start with something simple, like a linear simple regression, i.e.

Let us introduced – as suggested in Box & Cox (1964) – the following family of (power) transformations

on the variable of interest. Then assume that

As mentioned in Chapter 14 of Davidson & MacKinnon (1993) – in French – the log-likelihood of this model (assuming that observations are independent, with distribution ) can be written

We can then use profile-likelihood techniques (see here) to derive the optimal transformation.

This can be done in R extremely simply,

> library(MASS)
> boxcox(lm(dist~speed,data=cars),lambda=seq(0,1,by=.1))

we then get the following graph, If we look at the code of the function, it is based on the QR decomposition of the matrix (since we assume that  is a full-rank matrix). More precisely,  where  is a  matrix,  is a  orthonornal matrix, and is a  upper triangle matrix. It might be convenient to use this matrix since, for instance, .  Thus, we do have an upper triangle system of equations.

> X=lm(dist~speed,data=cars)$qr The code used to get the previous graph is (more or less) the following, > g=function(x,lambda){ + y=NA + if(lambda!=0){y=(x^lambda-1)/lambda} + if(lambda==0){y=log(x)} + return(y)} > n=nrow(cars) > X=lm(dist~speed,data=cars)$qr
> Y=cars$dist > logv=function(lambda){ + -n/2*log(sum(qr.resid(X, g(Y,lambda)/ + exp(mean(log(Y)))^(lambda-1))^2))} > L=seq(0,1,by=.05) > LV=Vectorize(logv)(L) > points(L,LV,pch=19,cex=.85,col="red") As we can see (with those red dots) we can reproduce the R graph. But it might not be consistent with other techniques (and functions described above). For instance, we can plot the profile likelihood function, > logv=function(lambda){ + s=summary(lm(g(dist,lambda)~speed, + data=cars))$sigma
+ e=lm(g(dist,lambda)~speed,data=cars)$residuals + -n/2*log(2 * pi)-n*log(s)-.5/s^2*(sum(e^2))+ + (lambda-1)*sum(log(Y)) + } > L=seq(0,1,by=.01) > LV=Vectorize(logv)(L) > plot(L,LV,type="l",ylab="") > (maxf=optimize(logv,0:1,maximum=TRUE))$maximum
 0.430591

$objective  -197.6966 > abline(v=maxf$maximum,lty=2) The good point is that the optimal value of  is the same as the one we got before. The only problem is that the -axis has a different scale. And using profile likelihood techniques to derive a confidence interval will give us different results (with a larger confidence interval than the one given by the standard function),

> ic=maxf$objective-qchisq(.95,1) > #install.packages("rootSolve") > library(rootSolve) > f=function(x)(logv(x)-ic) > (lower=uniroot(f, c(0,maxf$maximum))$root)  0.1383507 > (upper=uniroot(f, c(maxf$maximum,1))$root)  0.780573 > segments(lower,ic,upper,ic,lwd=2,col="red") Actually, it possible to rewrite the log-likelihood as (let us just get rid of the constant), where Here, it becomes > logv=function(lambda){ + e=lm(g(dist,lambda)~speed,data=cars)$residuals
+ elY=(exp(mean(log(Y))))
+ -n/2*log(sum((e/elY^lambda)^2))
+ }
>
> L=seq(0,1,by=.01)
> LV=Vectorize(logv)(L)
> plot(L,LV,type="l",ylab="")
> optimize(logv,0:1,maximum=TRUE)
$maximum  0.430591$objective
 -47.73436

with again the same optimal value for , and the same confidence interval, since the function is the same, up to some additive constant. So we have been able to derive the optimal transformation according to Box-Cox transformation, but so far, the confidence interval is not the same (it might come from the fact that here we substituted an estimator to the unknown parameter . ## 2 thoughts on “On Box-Cox transform in regression models”

1. Dr. Adnan Haider says:

Thank you for this post. Just a small correction. There is a small typo in confidence interval formula. In order to compute similar results with R, you may use:
ic=maxf\$objective-0.5*qchisq(.95,1)
with this revision:
Lower: # 0.2203471
Upper: # 0.6696117
Means, half of Chi-Sq value will produce similar results.

This site uses Akismet to reduce spam. Learn how your comment data is processed.