Hello,
I write this message because I have a problem with cointegration and causality
tests on R.
I'm working with time series data. I use ucra and vars packages.To perform
cointegration and Granger causality tests, I respectively write :
sjv <- vardata[, c("upd", "nc", "r", "up", "op", "ur", "expl")]
sjv.vecm <- ca.jo(sjv, type="eigen", K=2, spec = "longrun", season =4)
summary(sjv.vecm)
reg1.caus <- VAR(vardata[, c("upd", "nc", "r", "up", "op", "ur", "expl")],
p = 2, type = "const")
causality(reg1.caus, cause = "up")
In both cases, I respectively obtain the following error messages:
Error is solve.default(M11) :
the system is numerically singular : conditionnement de la réciproque =
9.47575e-17
and
Error is solve.default(crossprod(as.matrix(Z))) :
the system is numerically singular : conditionnement de la réciproque =
7.87318e-22
What do these messages mean ? And how can I resolve this problem? I precise
that I use quarterly data and that when I try to directly estimate VAR model
using the command below I encounter no problem:
reg1 <- VAR(vardata[, c("upd", "nc", "r", "up", "op", "ur", "expl")],
p = 1, type ="const")
reg1
Thank you in advance for answer.
Axel
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
