One assumption of sharing models is that we are able to create models using in this case R and afterwards are able to publish them. In this master thesis, the student will investigate how an executable R model can be deployed as Web Processing Services (together with all relevant input data and input constraints). The context we are looking at are where data are spatial and/or spatio-temporal and/or temporal, being served over OGC web services (SOS/WFS/WCS,…) and that there is an interest in re-executing the R model e.g. for a new geographical location or region, or for a new moment in time or over a new period of time. A challenge is to find a generic way to map input data specified as OGC web service requests to their respective R data structures, and to pass the R model output to the WPS process.

• Supervisor: Edzer Pebesma
• Contact Person (IFGI): ?
• Contact Person (INPE): ?

(added by pat; extended by edzer)