[R-SIG-Mac] clearing plot / over plotting of points

[Simon Urbanek]

Jochen,

On Apr 30, 2012, at 8:40 AM, jochen laubrock wrote:

> Hello,
> 
> points are not completely erased if overplotted in the background color, using points from base graphics. Instead, they leave a faint grey trace, both if running R.app and R in Terminal. Is this intended?

Yes, you may want to read up on "anti-aliasing". You can disable anti-aliasing (with antialias=F) but you get very ugly plots.


> What is the proper way to erase a plot if one does not want to create a new plot, which presumably has some overhead in an animation?
> 

You presume wrongly - it is in fact less efficient to try to erase a plot by painting over it and re-drawing a new one as opposed to drawing a new page, because you keep piling up the display list which forces R to redraw everything including over-painted parts of the plot.

If you want to do animation I hope you are aware of the crucial commands dev.hold()/dev.flush() designed for that purpose (and efficiency).

Cheers,
Simon


> Example,  using points:
> 
> A <- matrix(rnorm(1000), 500, 2)
> plot(range(A), range(A), type="n")
> points(A)
> points(A[1:250, 1:2], col="white", bg="white")
> 
> I'm posting to this list because I have not tested under a different OS.
> 
> Jochen
> 
>> sessionInfo()
> R version 2.15.0 (2012-03-30)
> Platform: x86_64-apple-darwin9.8.0/x86_64 (64-bit)
> 
> locale:
> [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
> 
> attached base packages:
> [1] stats     graphics  grDevices utils     datasets  methods   base     
> 
> other attached packages:
> [1] lattice_0.20-6 MASS_7.3-17   
> 
> loaded via a namespace (and not attached):
> [1] grid_2.15.0  tools_2.15.0
> 
> 
> 
> PS: Points *are* erased if using image
> 
> # using image
> A <- matrix(0, 100, 100)
> ix <- runif(1000, 0, 10000)
> A[ix] <- 1
> image(seq(0,1,length=100), seq(0,1,length=100), A, col=c("white", "black"))
> A[ix[1:500]] <- 0
> image(seq(0,1,length=100), seq(0,1,length=100), A, col=c("white", "black"))
> _______________________________________________
> R-SIG-Mac mailing list
> R-SIG-Mac at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-mac
> 
>