R devel - Aug 2002 - polygon() draws non-transparent border, erase.screen draws non-transparent border (PR#1881)

# polygon ignores requests to have its border transparent, look at
par(bg="transparent")
plot(c(0, 3), 0:1)
polygon(c(0, 1, 1, 0), c(0, 0, 1, 1), border=NA, col = 0)
polygon(c(1, 2, 2, 1), c(0, 0, 1, 1), border="transparent", col = 0)
polygon(c(2, 3, 3, 2), c(0, 0, 1, 1), border=0, col = 0)

# a quick fix for erase.screen() is the following
erase.screen <-
function (n = .split.cur.screen)
{
    if (!exists(".split.screens", envir = .GlobalEnv))
        return(FALSE)
    if (!(n %in% .split.valid.screens) && n != 0)
        stop("Invalid screen number\n")
    old <- par(usr = c(0, 1, 0, 1), mar = c(0, 0, 0, 0), fig = if (n >
        0)
        .split.screens[[n]]$fig
    else c(0, 1, 0, 1), xaxs = "i", yaxs = "i")
    on.exit(par(old))
    par(new = TRUE)
    plot.new()
    # bug in polygon, ignores border="transparent"
    # was: polygon(c(0, 1, 1, 0), c(0, 0, 1, 1), border=NA, col = 0)  
    rect(0, 0, 1, 1, col=0, border=0)
    par(new = TRUE)
    invisible()
}

as borders are overlapping between subscreens, erasing one subscreen will
set parts of borders of neighbor subscreens to background color. I guess this
is better than not beeing able to erase the borders at all. However, if border
shall not be erased have 
  rect(0, 0, 1, 1, col=0, border="tranparent")
instead. Alternatively we could introduce a new parameter border= to
erase.screen(), screen() and split.screen()

Best regards


Jens Oehlschlägel


> version
         _              
platform i386-pc-mingw32
arch     i386           
os       mingw32        
system   i386, mingw32  
status                  
major    1              
minor    5.1            
year     2002           
month    06             
day      17             
language R 


-- 




-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To:
r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

Strangely, it works without the first line, which is the default on a
windows() device.

There's a bug in GA_Polyline (not Polygon) stemming from the changes to
graphics device internals at 1.4.0, it seems.  It is transparent polylines
one cannot draw, and only on the windows() device.

You don't give an example for erase.screen: does this need a separate fix?

On Tue, 6 Aug 2002 oehl_list@gmx.de wrote:
> # polygon ignores requests to have its border transparent, look at
> par(bg="transparent")
> plot(c(0, 3), 0:1)
> polygon(c(0, 1, 1, 0), c(0, 0, 1, 1), border=NA, col = 0)
> polygon(c(1, 2, 2, 1), c(0, 0, 1, 1), border="transparent", col =
0)
> polygon(c(2, 3, 3, 2), c(0, 0, 1, 1), border=0, col = 0)
>
> # a quick fix for erase.screen() is the following
> erase.screen <-
> function (n = .split.cur.screen)
> {
>     if (!exists(".split.screens", envir = .GlobalEnv))
>         return(FALSE)
>     if (!(n %in% .split.valid.screens) && n != 0)
>         stop("Invalid screen number\n")
>     old <- par(usr = c(0, 1, 0, 1), mar = c(0, 0, 0, 0), fig = if (n
>
>         0)
>         .split.screens[[n]]$fig
>     else c(0, 1, 0, 1), xaxs = "i", yaxs = "i")
>     on.exit(par(old))
>     par(new = TRUE)
>     plot.new()
>     # bug in polygon, ignores border="transparent"
>     # was: polygon(c(0, 1, 1, 0), c(0, 0, 1, 1), border=NA, col = 0)
>     rect(0, 0, 1, 1, col=0, border=0)
>     par(new = TRUE)
>     invisible()
> }
>
> as borders are overlapping between subscreens, erasing one subscreen will
> set parts of borders of neighbor subscreens to background color. I guess
this
> is better than not beeing able to erase the borders at all. However, if
border
> shall not be erased have
>   rect(0, 0, 1, 1, col=0, border="tranparent")
> instead. Alternatively we could introduce a new parameter border= to
> erase.screen(), screen() and split.screen()
>
> Best regards
>
>
> Jens Oehlschlägel
>
>
>
> > version
>          _
> platform i386-pc-mingw32
> arch     i386
> os       mingw32
> system   i386, mingw32
> status
> major    1
> minor    5.1
> year     2002
> month    06
> day      17
> language R
>
>
> --
>
>
>
>
>
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
> r-devel mailing list -- Read
http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
> Send "info", "help", or "[un]subscribe"
> (in the "body", not the subject !)  To:
r-devel-request@stat.math.ethz.ch
>
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
>
-- 
Brian D. Ripley,                  ripley@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 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To:
r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._