| To: | <s-news@lists.biostat.wustl.edu> |
|---|---|
| Subject: | Problems with anova.gls when calling within a function |
| From: | "Steffen Barembruch" <steffen@barembruch.de> |
| Date: | Thu, 1 Feb 2007 19:30:12 +0100 |
| In-reply-to: | |
| References: | |
| Thread-index: | Acc0C7EsuM0kp7HIS56yba2edqrnBgD2A64gA5JcHeA= |
|
Hello!
I've got a problem with calling "anova.gls" within a function. Here is my
code, which I am executing within a function:
gls.lr <- gls(formula.fixed, data = "" na.action = ""
method="ML")
lme.lr <- lme(formula.fixed, data = "" random = formula.random, na.action= "" method="ML") LR.test <- anova.gls(gls.lr, lme.lr) (formula.random and formula.fixed have been assigned before, to be the
appropriate formulae)
I
get an error, that the object "lme.lr" and the object "gls.lr" are not
found.
I
think this is, because anova.gls looks for objects in the working directory and
not in the current frame. When I execute the same code at
the command line (not within a function), it does work.
And when I assign the objects to the working directory as following, it
also works
gls.lr <- gls(formula.fixed, data = ""
na.action = "" method="ML")
assign("gls.lr", gls.lr, where=1) lme.lr <- lme(formula.fixed, data =
"" random = formula.random, na.action= ""
method="ML")
assign("lme.temp1",
lme.temp1, where=1)
LR.test <-
anova.gls(gls.lr, lme.lr)
Does anyone
know anything about this? Why does this error occur? Do I have to assign the
objects to the working directory beforehand? Did I do anything
wrong
Surprisingly
this problem does not occur, when you call anova.gls with one parameter,
like
anova.gls(gls.lr)
Thanks a lot. I hope someone got some insight for me
Steffen
Barembruch |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | inconsistent results for survReg() vs censorReg(), Feng Gao |
|---|---|
| Next by Date: | Upcoming Course**** R/Splus Fundamentals and Programming Techniques**** In Washington DC, San Francisco and Princeton, elvis |
| Previous by Thread: | inconsistent results for survReg() vs censorReg(), Feng Gao |
| Next by Thread: | Upcoming Course**** R/Splus Fundamentals and Programming Techniques**** In Washington DC, San Francisco and Princeton, elvis |
| Indexes: | [Date] [Thread] [Top] [All Lists] |