8 messages · Agustin Lobo, G. Allegri, Barry Rowlingson +1 more
Hi! Does anyone has experience at displaying spatial R objects with QGis (perhaps using the python console + Rpy or RSpython)? I'm thinking on starting some testing and would like to know about previous experiences. Thanks
Dr. Agustin Lobo Institut de Ciencies de la Terra "Jaume Almera" (CSIC) LLuis Sole Sabaris s/n 08028 Barcelona Spain Tel. 34 934095410 Fax. 34 934110012 email: Agustin.Lobo at ija.csic.es http://www.ija.csic.es/gt/obster
Hi! Does anyone has experience at displaying spatial R objects with QGis (perhaps using the python console + Rpy or RSpython)? I'm thinking on starting some testing and would like to know about previous experiences.
I've done a bit of work on this and I think the current best option
would be to save the R Spatial objects as shapefiles and then load them
into QGis. For raster data I save as ESRI Ascii Grid format.
Here's some other ideas:
* write a data provider for QGis in Python that uses Rpy to fetch the
data from an R Spatial object stored in a .RData file.
Advantages: neat and well-integrated
Disadvantages: tricky. might need some changes to the QGis C++
codebase. Only works in QGis.
* write a GDAL/OGR driver for R Spatial objects.
Advantages: every project that uses GDAL/OGR can use it. Frank
Warmerdam would think it's a good idea.
Disadvantages: would have to be in C++, I'm not sure how to write
GDAL/OGR drivers that use external libraries (libR.so in this case).
I think writing the GDAL/OGR driver would be the right thing in this
case, but it would require someone with lots of R and C experience to
code it up. Start here: http://www.gdal.org/ogr/ogr_drivertut.html
Barry
An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20071204/1c3b3b81/attachment.pl>
That's what I'd use to write GDAL/OGR data to read into QGis, but if you want tighter coupling between QGis and R Data objects you need one of my other ideas... Barry
An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20071204/ab0642aa/attachment.pl>
An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-sig-geo/attachments/20071204/060a31cc/attachment.pl>
Ok, but I don't think you need a GDAL/OGR driver. Rgdal+sp do already the work. You need come code to couple the pieces, like the spgrass6 package [1] do for Grass.
So you are sitting in front of QGis and there's some R-spatial objects sitting in a .RData file. How do we get that into QGis? How about the user selects an option 'Load R-spatial Data Object', browses the .RData file and chooses the object, then Rpy uses RGdal converts it to a shapefile or raster, then QGis loads that in? Is that the solution you are thinking of? That's an easy solution, but it's a looser coupling than being able to read R-spatial objects directly, and it requires an extra file creation. Code-wise it's quite straightforward... I already have an R object browser written in PyQt and Rpy! Barry
Agustin Lobo wrote:
Hi! Does anyone has experience at displaying spatial R objects with QGis (perhaps using the python console + Rpy or RSpython)? I'm thinking on starting some testing and would like to know about previous experiences.
I've done a bit of work on this and I think the current best option would be to save the R Spatial objects as shapefiles and then load them into QGis. For raster data I save as ESRI Ascii Grid format. Here's some other ideas: * write a data provider for QGis in Python that uses Rpy to fetch the data from an R Spatial object stored in a .RData file.
For what (little?) it's worth, I have been doing this quite a bit, lately. I sometimes edit the objects in QGis, then reload them into R.
Advantages: neat and well-integrated
Disadvantages: tricky. might need some changes to the QGis C++
codebase. Only works in QGis.
* write a GDAL/OGR driver for R Spatial objects.
Advantages: every project that uses GDAL/OGR can use it. Frank
Warmerdam would think it's a good idea.
Disadvantages: would have to be in C++, I'm not sure how to write
GDAL/OGR drivers that use external libraries (libR.so in this case).
I think writing the GDAL/OGR driver would be the right thing in this
case, but it would require someone with lots of R and C experience to
code it up. Start here: http://www.gdal.org/ogr/ogr_drivertut.html
Barry
_______________________________________________ R-sig-Geo mailing list R-sig-Geo at stat.math.ethz.ch https://stat.ethz.ch/mailman/listinfo/r-sig-geo
--------------------------------- Don MacQueen Lawrence Livermore National Laboratory Livermore, CA, USA 925-423-1062 macq at llnl.gov