Skip to content

WPS client implementation in R

4 messages · Dr. Benedikt Gräler, Barry Rowlingson, Javier Martínez-López

#
I'm wondering what's the state of WPS (Web Processing Service) client
implementations in R? All I can find are the abstract of a 2008 UseR! talk:

https://www.r-project.org/conferences/useR-2008/abstracts/Henneboehl+Pebesma.pdf

and what looks like code designed to call a custom WPS server in the GSIF
package:

https://github.com/cran/GSIF/blob/master/R/WPSworldgrids.R

for a server that doesn't seem to work at the moment (wps.worldgrids.org).

Any advance on those?

Barry
#
Dear Barry,

I am not quite sure what you are targetting at:
i) a solution to expose (geo) R functionality as a WPS on the web or
ii) a R package to ease the access to an exposed WPS.

Regarding i), there is WPS4R (as a follow-up on the UseR! abstract; 
available as part of the 52?North WPS bundle [1]) where annotated R 
scripts can automatically be wrapped in a WPS process and exposed on the 
web (a tutorial can be found here [2]).

In terms of ii), I am not aware of any existing solution, but would be 
interested to contribute; possibly also as a GSoC project (I'll sketch 
the idea on the 52N's GSoC 2018 project page [3]).

Cheers,

  Ben

[1] https://wiki.52north.org/Geostatistics/WPS4R
[2] 
https://wiki.52north.org/Geoprocessing/GeoprocessingTutorials#WPS4R_tutorials
[3] https://wiki.52north.org/Projects/GSoC2018ProjectIdeas
On 20/02/2018 11:26, Barry Rowlingson wrote:

  
    
#
On Tue, Feb 20, 2018 at 3:13 PM, Dr. Benedikt Gr?ler <b.graeler at 52north.org>
wrote:
I'm currently more interested in a WPS client for R, to call WPS services.

I think if I wanted to implement a WPS service in R I'd probably hook into
pywps, but that's just my python bias against a java solution! I admit the
52N solution looks pretty neat (annotate an R script and there you have it).

thanks

Barry
#
You can try OpenCPU (https://www.opencpu.org/). It works great!

Best,

Javier

On Tue, Feb 20, 2018 at 4:31 PM, Barry Rowlingson <
b.rowlingson at lancaster.ac.uk> wrote: