[R] model.frame: how does one use it?

[Deepayan Sarkar]

On 6/15/07, hadley wickham <h.wickham at gmail.com> wrote:
> On 6/15/07, Deepayan Sarkar <deepayan.sarkar at gmail.com> wrote:

[...]

> > By 'name', I mean a symbol, such as 'Age' or 'myweight'.  So
> > basically, everything is as you would expect if the name is visible in
> > data, but if not, the search starts in the environment of the formula,
> > not the environment where the function call is being made (which is
> > the standard evaulation behaviour).  This is a feature, not a bug
> > (things would be a lot more confusing if it were the other way round).
>
> Could you give an example?  It's always seemed confusing to me and I
> don't see why looking in the environment of the formula helps.

Good question. I remember being convinced that it was a good idea, but
no longer remember why. This is the best I can come up with right now:


## stupid function that fits a model on a random subsample

lmsub <- function(formula, data, p = 0.7)
{
    n <- nrow(data)
    newdata <- data[sample(round(n * p)), ]
    lm(formula, newdata)
}

mydata <- data.frame(x = 1:100, y = rnorm(100))

n <- 2

lmsub(y ~ poly(x, n), data = mydata)

I don't think the use of the name 'n' is unusual in either case, and
we definitely wouldn't want the one inside 'lmsub' to be used for the
formula.

-Deepayan