[R-sig-Geo] R and spatial data

Michael Sumner mdsumner at utas.edu.au
Thu Aug 12 13:27:47 CEST 2004


Hello, I am slowly coming up to speed with S4 classes and the approach
to spatial data in sp and pixmap.  I see the comments in
"Class-SpatialData.R" alluding to conversion between pixmap and
SpatialDataGrid, and I'm wondering what the master plans (if any)
are.  Is pixmap intended to extend class SpatialData?  Are there
bigger ideas afoot?  I'm thinking along the lines of Roger Bivand's
note from

Date: Mon, 28 Jun 2004 11:02:49 +0200 (CEST)
Subject: Re: [R-sig-Geo] Where is the Changelog of maptools?

about the intentions to include read shapefiles into an S4 class.  Is there
something I'm missing in terms of documented goals and plans?

Are there plans to add coordinate system metadata (i.e. CRS) to
pixmap?

I'm interested in learning more (and hopefully helping develop) the spatial
capability in R.  I have a particular (side) interest in running Manifold GIS
from R and vice versa, and I have some simple and ad hoc 'tight' and also
'loose' couplings between them.  See

This is an example of running Manifold directly from R, I transfer a
matrix from R to a surface in Manifold, with examples of setting the
projection and data type:
http://www.georeference.org/Forums/forums/thread-view.asp?tid=876&posts=1#M3655

This uses basic R export and Manifold scripting to pass matrix / surface data
with (limited for now) metadata:

http://www.georeference.org/Forums/forums/thread-view.asp?tid=789&posts=4

They are very ad hoc but I can see a lot of potential (I need to learn fast).

I'm still very new to the CVS concept, and even trawling package
sources for learning R code, so I won't be surprised if I've missed a lot
and am asking silly questions.

Cheers, Mike.




###############################################

Michael Sumner - PhD. candidate
Maths and Physics (ACE CRC & IASOS) and Zoology (AWRU)
University of Tasmania
Private Bag 77, Hobart, Tas 7001, Australia
Phone: 6226 1752

R tips
http://www.ku.edu/~pauljohn/R/Rtips.html

Matlab has two problems: 1) it's expensive, and 2) it costs a lot:

http://www.dspguru.com/sw/opendsp/mathclo2.htm




More information about the R-sig-Geo mailing list