Problem with predict(armaFit()) and small series

[Matthieu Stigler]

hi

There is a problem with the predict method on armaFit()when the series 
is too small, I guess because of the pre-specified n.ahead and n.back 
values, see:

 >library(fArma)
 > l<-armaFit(~arma(1,1), lynx[1:20])
 > predict(l)
Error in `[.default`(data, (n - n.back + 1):n) :
   only 0's may be mixed with negative subscripts
 > predict(l, n.ahead=1)
Error in `[.default`(data, (n - n.back + 1):n) :
   only 0's may be mixed with negative subscripts
 > predict(l, n.ahead=1, n.back=1)
Error in `colnames<-`(`*tmp*`, value = c("Low 95", "Low 80", "Forecast",  :
   attempt to set colnames on object with less than two dimensions

Thanks!

Matthieu