node-red-contrib-R-nodes

A set of nodes which allows you to use the statistical computing language R in node RED.

Installation

Add @inNETMonitoring:registry=https://npm.pkg.github.com/ to your .npmrc file.

Execute npm install @inNETMonitoring/node-red-contrib-r-nodes inside your Node RED root directory.

Setup your R-Server

Install R on the computer you'd like to use as the R server. It can be downloaded on the R Project Website.

After installation of R install all requirements/libraries required for the remote code execution. Start R and execute the following commands:

install.packages('RJSONIO')
install.packages('Rserve')

Start your R-Server

Start the R interpreter and run the following commands to start Rserve.

library(Rserve)
Rserve()

Available Nodes

R Server

This is a configuration node to set up a connection to a host running R and RServe.

R Script

The R Script node provides the possibility to write R code directly in node RED. When the node receives a msg its payload is copied to the R server and your script is executed afterwards. You can manipulate the msg.payload property with default R code. After the code is successfully executed, the payload property is copied back to msg and is emitted at the nodes output.

Sample R code

Here is some sample code on how to access the msg object in R:

data <- payload$data;
payload$mean <- mean(data);
payload$median <- median(data);

Caution:
The variable type is not converted by this node. You have to make sure, the data types of the members you want to access or manipulate are represented by the correct datatype (e.g. array of floats).

Example

A working example flow can be found on the nodeRED flow repository.