Dear R help,
I've encountered a problem with survreg's anova(). I am currently
writing general code to fit a variety of models using different fitting
functions. Here's a simple example of what I'm trying to do:
---begin code---
# general function to analyse data
analyse.data <- function(formula, FUN, data, ...)
{
fit <- FUN(formula, data=data, ...)
anova(fit)
}
---end code---
In theory I should be able to call analyse.data() specifying FUN as I
please. In practice FUN=lm and FUN=glm work but FUN=survreg does not.
For example:
---begin code---
library(survival)
data <- data.frame(
y = rexp(100, rate=1),
status = rbinom(100,1,prob=0.8),
x = rnorm(100))
# analyse as lm
formula <- as.formula("y ~ x")
analyse.data(formula, FUN=lm, data=data)
# works!
# analyse as glm
formula <- as.formula("y ~ x")
analyse.data(formula, FUN=glm, data=data, family="Gamma")
# works!
# analyse as survival
formula <- as.formula("Surv(y, status) ~ x")
analyse.data(formula, survreg, data=data)
# Error in eval(expr, envir, enclos) : couldn't find function
"FUN"
# check survival works outside function
fit <- survreg(formula, data)
anova(fit)
# works!
---end code---
I suspect this is something to do with parent frames in anova.survreg().
However, although R mentions anova.survreg on traceback(), the code for
this function seems not to be viewable.
I couldn't find any other positings about this. Any suggestions?
Note: I am using R version 2.1.1 and "survival" package version 2.18.
Many thanks,
William
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-Dr William Valdar
++44 (0)1865 287 717
Wellcome Trust Centre valdar at well.ox.ac.uk
for Human Genetics, Oxford www.well.ox.ac.uk/~valdar
