[R] shared axes in multipanel plot

[Gabor Grothendieck]

On Mon, Dec 14, 2009 at 11:30 AM, Jennifer Young
<Jennifer.Young at math.mcmaster.ca> wrote:
> splendid!
>
> This worked well, but there are two oddities that I can't resolve.
>
> 1. In the real data, the "baseline" is a cumulative probability plot (from
> simulations) rather than the straight line.  The panel.lines plots this
> curve, but seems to join the first and last points together.
> panel.points(x, baseline, type="l") did the same.
> I checked that the vector is indeed sorted properly, so I'm not sure why
> it should connect the first point to the last.

I can't reproduce the problem based on this description.

>
> 2. The screens are correctly labeled, but in the wrong order (left to
> right, top to bottom: 3,4,1,2). Is this easily corrected?

xyplot(..., as.table = TRUE) will give one reordering.

Another possibility is:
   plt <- xplot(...)
   plt[ix]
where ix is a permutation of 1:4

>
> I've been cowardly avoiding learning xyplot() so thanks for the jumpstart!
>
>> Try this using xyplot.zoo in the zoo package.  We define the baseline
>> and a panel function.   The panel function just performs the default
>> action to display the graphs and adds the baseline.   The screens
>> variable is 1,1,2,2,3,3,4,4.  We create a zoo object from dat and use
>> screens to name the columns according to their group.  Finally we call
>> xyplot.zoo passing it screens so that the successive columns go in the
>> indicated panels and also passing the other items.  See ?xyplot.zoo in
>> zoo and ?xyplot in lattice.
>>
>> library(zoo)
>> library(lattice)
>>
>> baseline <- 1:nrow(dat)/nrow(dat)
>> pnl <- function(x, ...) {
>>       panel.plot.default(x, ...)
>>       panel.lines(x, baseline, lwd = 2, col = grey(0.5))
>> }
>> nc <- ncol(dat)
>> screens <- rep(1:(nc/2), each = 2)
>> z <- zoo(dat)
>> colnames(z) <- paste("Group", screens)
>> xyplot(z, screens = screens , layout = c(2, 2), col = "black", lty =
>> 2, scales = list(y = list(relation = "same")), panel = pnl)
>>
>>
>> On Fri, Dec 11, 2009 at 10:02 AM, Jennifer Young
>> <Jennifer.Young at math.mcmaster.ca> wrote:
>>> Hello
>>>
>>> I've created a function to make a plot with multiple pannels from
>>> columns
>>> of data that are created in a previous function.  In the example below
>>> the
>>> number of columns is 8, giving 4 pannels, but in general it takes data
>>> with any number of columns and figures out a nice layout.
>>>
>>> The panels all have the same axes, and so I wonder what functions are
>>> avialable to create axes only on the left and bottom of the whole plot
>>> rather than each pannel.
>>> I'd really like a generic way to do this for any number of plots, but
>>> was
>>> even having trouble figuring out how to do it manually for this example;
>>> How are pannels referred to, in a layout context?
>>> That is, how do I say,
>>>
>>> if(current.pannel==4) {do stuff}
>>>
>>> Here's a simple version of the code.
>>>
>>> baseline <- (1:20)/20    #example data
>>> dat1 <- matrix(baseline,20,8)
>>> dat <- dat1+matrix(rnorm(20*8)/30, 20,8)
>>>
>>> nstrat <- ncol(dat)
>>> rows <- ceiling(nstrat/4)
>>> layout(matrix(1:(rows*2), rows, 2, T))
>>> par(oma=c(4,4,3,1))
>>> par(mar=c(1,1,0,1))
>>> for(i in which(1:nstrat%%2!=0)){
>>>    plot(baseline, type="l", col="grey", lwd=2,
>>>            xlab="", ylab="", ylim=c(0,1), xaxt='n', yaxt='n')
>>>    axis(1, labels=F); axis(2, labels=F)
>>>    points(dat[,i], type="l", lty=2)
>>>    points(dat[,i+1], type="l", lty=2)
>>> }
>>>
>>>
>>>
>>> Thank you muchly
>>> Jennifer Young
>>>
>>> PS: I am a subscriber, but can't for the life of me figure out how to
>>> send
>>> an email while logged in so that the moderators don't have to take the
>>> time to read it over.  I always get the "please wait while we check it
>>> over" email.  Likely I'm being dumb.
>>>
>>> ______________________________________________
>>> 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.
>>>
>>
>
>