I am trying to do a Yeo-Johnson on my response variable. However, the boxcox lambda value and the Yeo-Johnson lambda value are not close to equal. In theory, they should be very close because I have very few negative values in my response variable.
I tested the type='BC' in the transfo() command to see if I would get the same lambda value from the BoxCox.lambda() command (these should be exactly equal). Do you know why these two lambda values are not equal?
library(cellWise)
library(forecast)
transfo(data$Response,type="BC",robust=FALSE,prestandardize=FALSE)$lambdahats
## these two are not equal, how do I get them to be equal?
BoxCox.lambda(data$Response)
I have to use the transfo package because it has the 'YJ' option and I have negative values in the response variable.