Generic Box-Cox Analysis Function
The function is implemented as a generic function with methods for classes "FitAR", "Arima", "ts" and "numeric".
For lambda!=0, the Box-Cox transformation is of x is (x^lambda-1)/lambda. If the minimum data value is <= 0, a small positive constant, equal to the negative of the minimum plus 0.25, is added to all the data values.
BoxCox(object, ...)
object |
model object |
... |
optional arguments |
No value returned. Graphical output is produced as side-effect. The plot shows relative likelihood function as well as the MLE and a confidence interval.
The MASS package has a similar function boxcox but this is implemented only for regression and analysis of variance.
A.I. McLeod and Y. Zhang
## Not run: #takes a few seconds BoxCox(lynx) out<-FitAR(lynx, c(1,2,4,10,11), ARModel="ARp", MLEQ=FALSE) BoxCox(out) out<-FitAR(lynx, c(1,2,4,5,7,8,10,11,12)) BoxCox(out) ## End(Not run)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.