R help - Jul 2010 - xyplot with all columns of a data.frame on a single plot

Hi, I have a data.frame with columns named X, D1, D2, D3

I know I can get a single plot with 3 curves by doing

xyplot(D1 + D2 + D3 ~ X, data)

but in some cases I might have columns D1 ...  D10.

Is there a way to plot all 10 columns without having to specify each
individual term?

(By analogy with formulae in lm, I thought, xyplot(. ~ X, data) would
work, but it didn't)

Thanks,

-- 
Rajarshi Guha
NIH Chemical Genomics Center

One method:

dd <- do.call(make.groups, mydata[,-1])
dd$X <- mydata$X
xyplot(data ~ X | which, dd)

Another method:

form <- paste(paste(colnames(mydata)[-1], collapse = " + "),
"~ X")
xyplot(as.formula(form), mydata)

Yet another method:

library(latticeExtra)
xyplot.list(mydata[,-1], FUN = function(z, ...) xyplot(z ~ mydata$X, ...))


On 27 July 2010 12:26, Rajarshi Guha <rajarshi.guha at gmail.com>
wrote:
> Hi, I have a data.frame with columns named X, D1, D2, D3
>
> I know I can get a single plot with 3 curves by doing
>
> xyplot(D1 + D2 + D3 ~ X, data)
>
> but in some cases I might have columns D1 ... ?D10.
>
> Is there a way to plot all 10 columns without having to specify each
> individual term?
>
> (By analogy with formulae in lm, I thought, xyplot(. ~ X, data) would
> work, but it didn't)
>
> Thanks,
>
> --
> Rajarshi Guha
> NIH Chemical Genomics Center
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>


-- 
Felix Andrews / ???
http://www.neurofractal.org/felix/

Hi:

Another approach might be to use the melt() function in package reshape
before creating the plot with xyplot, something along the lines of the
following:

library(reshape)
mdat <- melt(data, id = 'X')

This should create a data frame with three columns: X, variable (all the D*
names as factor levels) and value (stacked version of the D*s). Then use
something like

xyplot(value ~ X, data = mdat, groups = 'variable', ...)
xyplot(value ~ X | variable, data = mdat, ...)

One advantage of this approach is that you'll get the same structure out of
melt() no matter how many D* columns you have;
another is that the code block is small and relatively easy to remember six
months from now.  Here's a simple toy example:

library(reshape)
library(lattice)
d <- data.frame(x = 1:20, y1 = rnorm(20), y2 = rnorm(20), y3 = rnorm(20))

# Reshape the data:
m <- melt(d, id = 'x')

# xyplot with a basic legend

# melted data
xyplot(value ~ x, data = m, groups = variable,
    auto.key = list(space = 'right', points = TRUE, lines = FALSE))
# plot from the original data
xyplot(y1 + y2 + y3 ~ x, data = d,
    auto.key = list(space = 'right', points = TRUE, lines = FALSE))   #
identical except for y label

HTH,
Dennis

On Mon, Jul 26, 2010 at 7:26 PM, Rajarshi Guha
<rajarshi.guha@gmail.com>wrote:
> Hi, I have a data.frame with columns named X, D1, D2, D3
>
> I know I can get a single plot with 3 curves by doing
>
> xyplot(D1 + D2 + D3 ~ X, data)
>
> but in some cases I might have columns D1 ...  D10.
>
> Is there a way to plot all 10 columns without having to specify each
> individual term?
>
> (By analogy with formulae in lm, I thought, xyplot(. ~ X, data) would
> work, but it didn't)
>
> Thanks,
>
> --
> Rajarshi Guha
> NIH Chemical Genomics Center
>
> ______________________________________________
> R-help@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
	[[alternative HTML version deleted]]

One of the nice features of R's formula syntax is that
you can create a character string containing a formula,
and pass it to the formula() function.  For example:

xyplot(formula(paste(paste(paste('D',1:10,sep=''),collapse='+'),'X',sep='~')),data)

will do what you want.

 					- Phil Spector
 					 Statistical Computing Facility
 					 Department of Statistics
 					 UC Berkeley
 					 spector at stat.berkeley.edu


On Mon, 26 Jul 2010, Rajarshi Guha wrote:
> Hi, I have a data.frame with columns named X, D1, D2, D3
>
> I know I can get a single plot with 3 curves by doing
>
> xyplot(D1 + D2 + D3 ~ X, data)
>
> but in some cases I might have columns D1 ...  D10.
>
> Is there a way to plot all 10 columns without having to specify each
> individual term?
>
> (By analogy with formulae in lm, I thought, xyplot(. ~ X, data) would
> work, but it didn't)
>
> Thanks,
>
> -- 
> Rajarshi Guha
> NIH Chemical Genomics Center
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>