RCallerService is a microservice for running R-scripts in forecast models written in different other languages.
Based on https://github.com/jbytecode/rcaller/tree/master/RCaller library.
If you are running it without Docker, you should also install R itself. The Docker image can be found at https://hub.docker.com/r/kopilov/rcallerservice
For executing a script send it as a body of POST request to URL
/<result_type>?result=<result_name>[&timeout=<timeout_value_in_seconds>]
Supported result types:
* double_array
* text_array
* text
Output format: csv
Start the service (in Docker by docker run -it --net=host kopilov/rcallerservice:1.0.1
or directly on the system by java -jar rcallerservice.jar --port=8080
)
Prepare the script such as
v <- 1:9;
v * 10 -> u;
Put it in a file named like example.r
Run the query: curl --data-binary @example.r http://127.0.0.1:8080/double_array?result=u
The output should be 10.0;20.0;30.0;40.0;50.0;60.0;70.0;80.0;90.0