FW: [R] Maps and plotting

[Roger Bivand]

On Thu, 14 Oct 2004, Shawn Way wrote:

> 
> Thanks for the help on the translucent dots.  What would be the best
> method for creating a map of the facility?  I looked into map* in the
> libraries and didn't find anything on creating the maps, just using
> them.
> 

Depends what you mean by "creating"? If you mean accessing map data, then 
you most likely will have to define for where (and maybe when), what 
scale, etc., and which format the map data files should be in. Like Prof. 
Ripley, I also thought your question was about translucency, that is using 
the alpha channel to darken overlapping symbols, which works very nicely 
for the PDF device in R 2.0.0.

> Thanks again... 
> 
> 
> Shawn Way, PE
> Engineering Manager
> sway at tanox.com
> 
> -----Original Message-----
> From: Prof Brian Ripley [mailto:ripley at stats.ox.ac.uk]
> Sent: Thursday, October 14, 2004 6:05 AM
> To: Barry Rowlingson
> Cc: Shawn Way; R-help at stat.math.ethz.ch
> Subject: Re: [R] Maps and plotting
> 
> On Thu, 14 Oct 2004, Barry Rowlingson wrote:
> 
> > Shawn Way wrote:
> > >  At our facility we have multiple sample points that are sampled on 
> > > any given day.  What I would like to do is create a map of the 
> > > facility with the sample points (and point labels) and when we have 
> > > out of specification results, place a transparent dot over the area
> on the map.
> > > As the number of OOS results builds up, I envision the dot getting 
> > > darker.
> > > 
> > 
> >   Over what timescale? This sounds like it could be an interactive, 
> > real-time on-line monitoring thing. Is it?
> > 
> >   In which case R's graphics devices might not be good enough, and 
> > you'd be better off using a TclTk graphics canvas.
> > 
> >   library(tcltk) and read the docs!
> > 
> >   Another idea, if all you are doing is updating a daily image, would 
> > be to use a language like Python, and the Python Imaging Library (PIL)
> 
> > to draw pretty graphs.
> > 
> >   I've done something similar that produces daily maps of disease 
> > incidence, but I used different size and colour circles and not 
> > transparency, so I just used base R graphics and produced a PNG file.
> > If I wanted transparency I'd probably use Python/PIL, which can handle
> 
> > alpha channels.
> 
> I think this may mean *translucent* dots.  R has been able to do
> transparent dots for a very long time, but PNG cannot handle
> translucency.  On devices (e.g. pdf) which can, you can do this as of R
> 2.0.0.
> 
> 

-- 
Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Breiviksveien 40, N-5045 Bergen,
Norway. voice: +47 55 95 93 55; fax +47 55 95 93 93
e-mail: Roger.Bivand at nhh.no