I was sniffing in the fracdiff library (this is for fractionally integrated
ARMA processes; Haslett and Raftery 1989).
The documentation suggests that one tries the following simple example:
library(fracdiff)
ts.test <- fracdiff.sim( 5000, ar = .2, ma = -.4, d = .3)
fracdiff( ts.test$series, nar = length(ts.test$ar), nma = length(ts.test$ma))
When I run this, I get the following error:
R --vanilla < demo.R > demo.out
Warning message:
unable to compute correlation matrix in: switch(temp$info, warning("warning in gamma function"), warning("singular Hessian"),
This doesn't look nice. Should I be worried? Is this a bug report? Does
R have a system like Debian's `reportbug' for submitting bugs.
Bug report for fracdiff
2 messages · Ajay Shah, Uwe Ligges
Ajay Shah wrote:
I was sniffing in the fracdiff library (this is for fractionally integrated ARMA processes; Haslett and Raftery 1989). The documentation suggests that one tries the following simple example: library(fracdiff) ts.test <- fracdiff.sim( 5000, ar = .2, ma = -.4, d = .3) fracdiff( ts.test$series, nar = length(ts.test$ar), nma = length(ts.test$ma))
Works on Windows with R-1.8.1 with the recent version of fracdiff 1.1-1 compiled with gcc-3.3.1. (you haven't told us any version information).
When I run this, I get the following error:
R --vanilla < demo.R > demo.out
Warning message:
unable to compute correlation matrix in: switch(temp$info, warning("warning in gamma function"), warning("singular Hessian"),
This doesn't look nice. Should I be worried? Is this a bug report? Does
R have a system like Debian's `reportbug' for submitting bugs.
Yes, see ?bug.report, but for contributed packages you should contact the package maintainer directly, library(help = fracdiff) tells you who it is ... Uwe Ligges