A simple little module for passing data from NodeJS to R (and back again).
Data passed from node is converted into a list and loaded into the R environment as the variable input
. No special syntax in R is needed. For better portability/reliability, it's recommended to load packages with needs
(comes packaged inside the module — no installation required).
npm install r-script
var R = require("r-script");
// example.js
var out = R("ex-sync.R")
.data("hello world", 20)
.callSync();
console.log(out);
// [ 'oedorlwlh l', 'oldlrhelwo ', 'erllol dhow', ' lwrellodoh', 'holdlerw ol',
// 'lrlewdhol o', 'lll wohdeor', 'hwrlledl oo', 'elrooh lwld', 'ewrlo lhdlo',
// 'hlloroelwd', 'h eodollwlr', 'wr ldleohlo', 'or ohldlwel', 'lohe lowlrd',
// 'rhdwoelllo ', 'owhorldell ', 'rlle ohdolw', 'rhlwolle od', 'woro helldl' ]
# ex-sync.R
needs(magrittr)
set.seed(512)
do.call(rep, input) %>%
strsplit(NULL) %>%
sapply(sample) %>%
apply(2, paste, collapse = "")
// example.js
var attitude = JSON.parse(
require("fs").readFileSync("example/attitude.json", "utf8"));
R("example/ex-async.R")
.data({df: attitude, nGroups: 3, fxn: "mean" })
.call(function(err, d) {
if (err) throw err;
console.log(d);
});
// [ { group: '(40,55]', rating: 46.7143, advance: 41.1429 },
// { group: '(55,70]', rating: 64.6154, advance: 41.9231 },
// { group: '(70,85]', rating: 77.2, advance: 45.5 } ]
# ex-async.R
needs(dplyr)
attach(input[[1]])
return("early returns are ignored")
cat("so are undirected calls to cat")
print("or print")
cat("unless directed to a file", file = "out.Rout")
# output of final expression is returned to node
df %>%
mutate(group = cut(rating, nGroups, ordered = T)) %>%
group_by(group) %>%
summarize_each(funs_(fxn)) %>%
select(group, rating, advance) %>%
mutate(group = as.character(group))
R(path)
Creates a new object that will source the R script specified by path.
R.data(...)
Adds data to the object and returns itself. You can give any number of arguments of different types.
R.call([options], callback)
Calls R. Any previously supplied data is stringified into JSON and passed to R, where it's converted into a list and loaded into the R environment as the variable input
. On completion, the callback is invoked with two arguments: any error and the output from R, parsed back into a Javascript object.
Additional arguments for the conversion from R to JSON can be specified as options (see documentation for toJSON
from the R package jsonlite
for defaults).
R.callSync([options])
The same as above, but calls R synchronously.