ARIMA

noetsi

Fortran must die
#1
I am confused about two piece of documentation for ARIMA in the forecast package.

lambda = model$lambda,
I am not sure what this does. The documentation says:
Box-Cox transformation parameter. If lambda="auto", then a transformation isautomatically selected using BoxCox.lambda. The transformation is ignored ifNULL. Otherwise, data transformed before model is estimated

The default is not 'auto' but I am not certain what it is, that is what it does or if the default simply does nothing. It is not clear to me, and this is important, how you would tell the software to log the data which is often done with financial data.

biasadj = FALSE,
When you do box cox logging etc you need to transform the data back. It is not clear to me how you do this. The documentation says (I don't understand the practical ramifications of doing this)

Use adjusted back-transformed mean for Box-Cox transformations. If transformed data is used to produce forecasts and fitted values, a regular back transformation will result in median forecasts. If biasadj is TRUE, an adjustment will be made to produce mean forecasts and fitted values.