how to write assignment form of function
Dear Professor Ripley, thank you for your answer. Adding a return value, as also Dimitris Rizopoulos suggested the function does what I need, that is to rename factor levels. I tried to look at levels<-.factor in R-devel but I have to admit that I do not know exactly where to look and searching I did not find it. For the moment my problem is solved and I interpret your hint that way that in the future levels<-.factor will not more drop all other attributes. Thanks again Heinz T??chler
At 15:18 10.08.2005 +0100, Prof Brian Ripley wrote:
On Wed, 10 Aug 2005, Heinz Tuechler wrote:
where can I find information about how to write an assigment form of a function?
In all good books on S programming, or by studying examples. But in this case the problem is actually about defining functions with the return value you expect.
For curiosity I tried to write a different form of the levels()-function, since the original method for factor deletes all other attributes of a
factor.
Of course, the simple method would be to use instead of levels(x) <- newlevels, attr(x, 'levels') <- newlevels.
And that would not do what the current function does, which is to merge levels as required. I suggest you look at levels<-.factor in R-devel, which does not drop attributes.
I tried the following:
## example
x <- factor(c(1,1,NA,2,3,4,4,4,1,2)); y <- x
attr(x, 'levels') <- c('a', 'b', 'c', 'd') # does what I want
x
[1] a a <NA> b c d d d a b
Levels: a b c d
'levels.simple<-' <- function (x, value)
{
attr(x, 'levels') <- value
}
This did not return anything! Try returning 'x'.
levels.simple(y) <- c('a', 'b', 'c', 'd') # does not what I want
y
[1] "a" "b" "c" "d"
-- Brian D. Ripley, ripley at stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UK Fax: +44 1865 272595