I rewrote scatter.smooth to handle missing values, but I have a question
about a move I had to make. Here's the code:
Mscatter.smooth<-function (x, y, span = 2/3, degree = 1, family =
c("symmetric",
"gaussian"), xlab = deparse(substitute(x)), ylab =
deparse(substitute(y)),
ylim = range(y, prediction$y), evaluation = 50, ...)
{
if (inherits(x, "formula")) {
if (length(x) < 3)
stop("need response in formula")
thiscall <- match.call()
thiscall$x <- x[[3]]
thiscall$y <- x[[2]]
return(invisible(eval(thiscall, sys.parent())))
}
##################
plot(x, y, xlab = xlab, ylab = ylab, ...)
i<-complete.cases(x,y)
x<-x[i]
y<-y[i]
##################
prediction <- loess.smooth(x, y, span, degree, family, evaluation)
lines(prediction)
invisible()
}
The changes I made are between the '###########'. The idea was only to
pass complete cases on to loess.smooth and thus avoice the NA error
you'd usually get by calling scatter.smooth with missing values.
I had to move the plot above the loess.smooth, whereas in the orginal
scatter.smooth is was just below to take adavantage of prediction$y to
set the ylim in the plotting function.
But if I try to plot after creating the complete cases, the names of the
x and y variables disappear and instead the plotting labels on the x and
y axis look like, for example, 'c(1,2,1.2,5.2, ...)', that is, the
vector of values being plotted.
Why does that happen?
Matt
--
William (Matt) Briggs
Assistant Professor of Biostatistics
Division of General Internal Medicine
Weill Medical College of Cornell University
525 East 68th St., Box #46
New York, NY 10021
Tel: 212-628-0128; Fax: 212-746-8965
http://drclamp.wmbriggs.com
Lazy evaluation. See V&R's S Programming for a good explanation. But, in brief, the default values for xlab and ylab are not evaluated until they are needed. If plot(..., xlab=...) appears after the assignments to x and y, then x and y are no longer the original argument expressions but numeric vectors. -- Bert Gunter Genentech Non-Clinical Statistics South San Francisco, CA "The business of the statistician is to catalyze the scientific learning process." - George E. P. Box> -----Original Message----- > From: r-help-bounces at stat.math.ethz.ch > [mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of William Briggs > Sent: Thursday, February 10, 2005 1:29 PM > To: r-help at stat.math.ethz.ch > Subject: [R] rewrite of scatter.smooth to handle NAs > > > I rewrote scatter.smooth to handle missing values, but I have > a question > about a move I had to make. Here's the code: > > Mscatter.smooth<-function (x, y, span = 2/3, degree = 1, family = > c("symmetric", > "gaussian"), xlab = deparse(substitute(x)), ylab = > deparse(substitute(y)), > ylim = range(y, prediction$y), evaluation = 50, ...) > { > if (inherits(x, "formula")) { > if (length(x) < 3) > stop("need response in formula") > thiscall <- match.call() > thiscall$x <- x[[3]] > thiscall$y <- x[[2]] > return(invisible(eval(thiscall, sys.parent()))) > } > ################## > plot(x, y, xlab = xlab, ylab = ylab, ...) > i<-complete.cases(x,y) > x<-x[i] > y<-y[i] > ################## > prediction <- loess.smooth(x, y, span, degree, family, > evaluation) > lines(prediction) > invisible() > } > > The changes I made are between the '###########'. The idea > was only to > pass complete cases on to loess.smooth and thus avoice the NA error > you'd usually get by calling scatter.smooth with missing values. > > I had to move the plot above the loess.smooth, whereas in the orginal > scatter.smooth is was just below to take adavantage of > prediction$y to > set the ylim in the plotting function. > > But if I try to plot after creating the complete cases, the > names of the > x and y variables disappear and instead the plotting labels > on the x and > y axis look like, for example, 'c(1,2,1.2,5.2, ...)', that is, the > vector of values being plotted. > > Why does that happen? > > Matt > -- > William (Matt) Briggs > Assistant Professor of Biostatistics > Division of General Internal Medicine > Weill Medical College of Cornell University > 525 East 68th St., Box #46 > New York, NY 10021 > Tel: 212-628-0128; Fax: 212-746-8965 > > http://drclamp.wmbriggs.com > > ______________________________________________ > R-help at stat.math.ethz.ch mailing list > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide! > http://www.R-project.org/posting-guide.html >
You need evaluate deparse(substitute(x)) _before_ you change x.
so have
xlab; ylab
early in the body of your function.
However, NEWS in R-devel says
o scatter.smooth() and loess.smooth() now handle missing values
in their inputs.
so why reinvent that wheel?
On Thu, 10 Feb 2005, William Briggs wrote:
>
> I rewrote scatter.smooth to handle missing values, but I have a question
> about a move I had to make. Here's the code:
>
> Mscatter.smooth<-function (x, y, span = 2/3, degree = 1, family =
> c("symmetric",
> "gaussian"), xlab = deparse(substitute(x)), ylab =
> deparse(substitute(y)),
> ylim = range(y, prediction$y), evaluation = 50, ...)
> {
> if (inherits(x, "formula")) {
> if (length(x) < 3)
> stop("need response in formula")
> thiscall <- match.call()
> thiscall$x <- x[[3]]
> thiscall$y <- x[[2]]
> return(invisible(eval(thiscall, sys.parent())))
> }
> ##################
> plot(x, y, xlab = xlab, ylab = ylab, ...)
> i<-complete.cases(x,y)
> x<-x[i]
> y<-y[i]
> ##################
> prediction <- loess.smooth(x, y, span, degree, family, evaluation)
> lines(prediction)
> invisible()
> }
>
> The changes I made are between the '###########'. The idea was
only to pass
> complete cases on to loess.smooth and thus avoice the NA error you'd
usually
> get by calling scatter.smooth with missing values.
>
> I had to move the plot above the loess.smooth, whereas in the orginal
> scatter.smooth is was just below to take adavantage of prediction$y to set
> the ylim in the plotting function.
>
> But if I try to plot after creating the complete cases, the names of the x
> and y variables disappear and instead the plotting labels on the x and y
axis
> look like, for example, 'c(1,2,1.2,5.2, ...)', that is, the vector
of values
> being plotted.
>
> Why does that happen?
Because it is documented that way!
--
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