[R] Background color in a grid plot seems to interfere with plot()

Marius Hofert marius.hofert at math.ethz.ch
Mon Sep 24 08:06:52 CEST 2012


Dear Paul,

Thank you for helping. This works great. 

I then tried to put in a grid (via grid()). Why does that fail? 

Cheers,

Marius


require(grid)
require(gridBase)

pdf(file="Rplot.pdf", width=8, height=8, onefile=FALSE)

## set up the grid layout
plot.new() # start (empty) new page with 'graphics'
gl <- grid.layout(5, 5, widths=unit(c(1.8, 8, 0.8, 8, 0.8), "cm"),
		  heights=unit(c(0.8, 8, 0.8, 8, 1.5), "cm"))
pushViewport(viewport(layout=gl))

## plot data
par. <- par(no.readonly=TRUE) # save plot settings
for(i in 1:2) { # rows
    i. <- if(i > 1) i+2 else i+1 # jumping over gaps
    for(j in 1:2) { # columns
        j. <- if(j > 1) j+2 else j+1 # jumping over gaps
        pushViewport(viewport(layout.pos.row=i., layout.pos.col=j.))
        grid.rect(gp=gpar(fill="gray90")) # background
        par(plt=gridPLT())
        ## plot
        par(new=TRUE) # always do this before each new 'graphics' plot
        grid(col=1)
        plot(1:10, 1:10, log="y", xlab="", ylab="",
             xaxt=if(i==2) "s" else "n", yaxt=if(j==1) "s" else "n")
        upViewport()
    }
}
par(par.)
dev.off()



Paul Murrell <paul at stat.auckland.ac.nz> writes:

> Hi
>
> On 24/09/12 09:36, Marius Hofert wrote:
>> Hi,
>>
>> Why does the upper left panel (in the plot below) not have a gray background?
>
> That is a symptom of the conflict that occurs when the 'graphics' package and
> the 'grid' package both try to initialise a new page.
> A good rule of thumb is to start a new page with 'graphics' first and THEN add
> grid' stuff ('grid' is better at sharing), so a minor adjustment to your code
> would be (#PAUL marks the changes) ...
>
>
> pdf(file="Rplot.pdf", width=8, height=8, onefile=FALSE)
> #PAUL
> # Start (empty) new page with 'graphics'
> plot.new()
> ## set up the grid layout
> gl <- grid.layout(5, 5, widths=unit(c(1.8, 8, 0.8, 8, 0.8), "cm"),
> 		  heights=unit(c(0.8, 8, 0.8, 8, 1.5), "cm"))
> pushViewport(viewport(layout=gl))
> ## plot data
> par. <- par(no.readonly=TRUE) # save plot settings
> for(i in 1:2) { # rows
>     i. <- if(i > 1) i+2 else i+1 # jumping over gaps
>     for(j in 1:2) { # columns
>         j. <- if(j > 1) j+2 else j+1 # jumping over gaps
>         pushViewport(viewport(layout.pos.row=i., layout.pos.col=j.))
>         grid.rect(gp=gpar(fill="gray90")) # background
>         par(plt=gridPLT())
>         ## plot
>         #PAUL
>         # ALWAYS do this before each new 'graphics' plot
>         par(new=TRUE)
>         plot(1:10, 1:10, log="y", xlab="", ylab="",
>              xaxt=if(i==2) "s" else "n", yaxt=if(j==1) "s" else "n")
>         upViewport()
>     }
> }
> par(par.)
> dev.off()
>
>
> Hope that helps.
>
> Paul
>
>> Cheers,
>>
>> Marius
>>
>>
>> require(grid)
>> require(gridBase)
>>
>> pdf(file="Rplot.pdf", width=8, height=8, onefile=FALSE)
>>
>> ## set up the grid layout
>> gl <- grid.layout(5, 5, widths=unit(c(1.8, 8, 0.8, 8, 0.8), "cm"),
>> 		  heights=unit(c(0.8, 8, 0.8, 8, 1.5), "cm"))
>> if(FALSE) grid.show.layout(gl)
>> pushViewport(viewport(layout=gl))
>>
>> ## plot data
>> par. <- par(no.readonly=TRUE) # save plot settings
>> for(i in 1:2) { # rows
>>      i. <- if(i > 1) i+2 else i+1 # jumping over gaps
>>      for(j in 1:2) { # columns
>>          j. <- if(j > 1) j+2 else j+1 # jumping over gaps
>>          pushViewport(viewport(layout.pos.row=i., layout.pos.col=j.))
>>          grid.rect(gp=gpar(fill="gray90")) # background
>>          par(plt=gridPLT())
>>          ## plot
>>          plot(1:10, 1:10, log="y", xlab="", ylab="",
>>               xaxt=if(i==2) "s" else "n", yaxt=if(j==1) "s" else "n")
>>          par(new=TRUE) # to be run after first plot
>>          upViewport()
>>      }
>> }
>> par(par.)
>> dev.off()
>>
>> ______________________________________________
>> 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.
>>




More information about the R-help mailing list