[R-sig-Geo] R Raster to ArcGIS

Roger Bivand Roger.Bivand at nhh.no
Tue Jul 29 21:38:53 CEST 2008


On Tue, 29 Jul 2008, Guy Serbin wrote:

> Hi list,
>
> I'm going to borrow what I just learned from my other thread.  I'd
> export your raster data as an ENVI file, which ArcGIS should be able
> to read with the ENVI Reader for ArcGIS installed.
>
> 1. Go to www.ittvis.com and download the appropriate version of the
> ENVI Reader for ArcGIS (1.1 for ArcGIS 8.3-9.1, 1.2 for ArcGIS 9.2).
> You won't need the ENVI reader for ArcGIS 9.3 as support for this
> format is in the upcoming version.
> 2. Get the caTools package for R and load it.
> 3. Export your data from R using write.ENVI
> 4. Read data into ArcGIS using usual means.


Seems like a long way round when readGDAL/writeGDAL in the rgdal package 
have plenty of drivers, such as GeoTiff. Depending on which application 
needs to be "in front", using rgdal and system() from R is very flexible.

I've found GDAL-generated GeoTiff files very easy to read into ENVI, 
harder to read into ArcGIS (may need statistics setting to get the 
symbology reasonable).

Roger

>
> You could also see if one of the other packages, e.g., RPyGeo, allows
> you to directly interface between R and ArcGIS for your purposes.
>
> Best of luck,
> Guy Serbin
>
> On Tue, Jul 29, 2008 at 2:54 PM, Ashton Shortridge <ashton at msu.edu> wrote:
>> Hi Matt (and list),
>>
>> ArcGIS still (as far as I know) doesn't work with rectangular cells. This
>> isn't a problem with the AsciitoGrid function, but I think is more
>> fundamentally a problem with Arc's raster data structure.
>>
>> The trick to getting around this may be to convert your matrix to a point
>> object - with x, y, and z values - in R, where x and y correspond to the
>> ground coordinates of each cell in the matrix. I think you would have to do
>> this with your own code, but if you've programmed with arrays before it would
>> not be difficult. Then interpolate these points to a square grid. The
>> interpolation could be done in R (using gstat routines for example) or within
>> a GIS (that is, write the xyz object to a text file, or perhaps straight to a
>> shapefile, import it as a vector point set to a GIS, and then convert that to
>> a raster.
>>
>> Hope this helps,
>>
>> Ashton
>>
>>
>> On Tuesday 29 July 2008, Matt.Farnsworth at aphis.usda.gov wrote:
>>> Dear list,
>>>
>>> I'm trying to bring an ASCII raster file into ArcGIS that resulted from
>>> calling the R function Tsp(), which is in the "fields" library, followed
>>> by the use of the function predict.surface() to generate the matrix of
>>> predicted values. These functions use many of the same inputs as the
>>> Krig() function in R. My problem is that the output is a symmetrical
>>> matrix of dimension NxN, however, because R outputs an NxN matrix it
>>> necessarily results in having a rectangular grid cell (e.g., x dimension
>>> is 50 and y dimension is 30) for any input data set whose spatial extent
>>> is not square.
>>>
>>> My question is; does anybody know how to import a rectangular grid from an
>>> ASCII raster file into ArcGIS using something other than the AsciiToGrid
>>> function in ArcGIS, which only allows for square grids?
>>>
>>> Thanks in advance for any help,
>>> Matt
>>>
>>> *************************************************
>>> Matt Farnsworth
>>> Disease Ecologist
>>> Spatial Epidemiology Unit
>>> Centers for Epidemiology and Animal Health
>>
>> --
>> Ashton Shortridge
>> Associate Professor                     ashton at msu.edu
>> Dept of Geography                       http://www.msu.edu/~ashton
>> 235 Geography Building          ph (517) 432-3561
>> Michigan State University               fx (517) 432-1671
>>
>> _______________________________________________
>> R-sig-Geo mailing list
>> R-sig-Geo at stat.math.ethz.ch
>> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>>
>
>
>
>

-- 
Roger Bivand
Economic Geography Section, Department of Economics, Norwegian School of
Economics and Business Administration, Helleveien 30, N-5045 Bergen,
Norway. voice: +47 55 95 93 55; fax +47 55 95 95 43
e-mail: Roger.Bivand at nhh.no




More information about the R-sig-Geo mailing list