search for: pred_garch

Hi, I had a quick question regarding the pred_garch function (as part of the tseries library, garch.c), and specifically: the function calculates the 'h' vector of conditional variances, h[n+1] if the data is genuine. I am very much a beginner, but from what I understand I would think that it is computed from the errors vector, not 'y...

Kurt Hornik wrote: > >>>>> Duncan Murdoch writes: > > > The problem is in the .C call to pred_garch, which is short by one > > parameter. In the garch.c file, the header is > > > void pred_garch (double *y, double *h, int *n, double *par, int *p, > > int *q, int *genuine) > > > but in garch.R, the call is coded as > > > pred <- .C("pred_garch&qu...

...month 01 day 30 language R The tseries package can be loaded successfully and most functions are running fine, but garch will cause a segmentation fault. I did a very simple debug, and found the problem lied in function call: .C( "pred_garch", ... ). BTW, my gcc version is 2.96. If I compile the R and all packages by using gcc 3.0.4, the problem persists. The following are the output when I am trying to run the example in garch help: > n <- 1100 > a <- c(0.1, 0.5, 0.2) # ARCH(2) coefficients > e <- rnorm(n) &...

Platform: Windows 98 R version: Version 1.6.1 (2002-11-01) Package 'tseries' version: 0.9-4 Built: R 1.6.0; Win32; Mon Oct 7 14:02:38 2002) (dependanted packages 'quadprog': Built: R 1.5.1; Win32; 2002-07-03 12:20:50) I've just upgraded to R 1.6.1. By typing library(tseries) example(garch) => I've got a crash with the following message: "RGUI a causé une

Platform: Windows 98 R version: Version 1.6.1 (2002-11-01) Package 'tseries' version: 0.9-4 Built: R 1.6.0; Win32; Mon Oct 7 14:02:38 2002) (dependanted packages 'quadprog': Built: R 1.5.1; Win32; 2002-07-03 12:20:50) I've just upgraded to R 1.6.1. By typing library(tseries) example(garch) => I've got a crash with the following message: "RGUI a causé une