Executes a transducer to transform the observable sequence.

Transducers are composable algorithmic transformations. They are independent from the context of their input and output sources and specify only the essence of the transformation in terms of an individual element. Because transducers are decoupled from input or output sources, they can be used in many different processes such as Observable sequences. Transducers compose directly, without awareness of input or creation of intermediate aggregates.

Such examples of transducers libraries are transducers-js from Cognitect and transducers.js from James Long.

In order for this operator to work, it must the transducers library must follow the following contract:

return {
  init: function() {
    // Return the item
    return observer;
  step: function(obs, input) {
    // Process next item
    return obs.onNext(input);
  result: function(obs) {
    // Mark completion
    return obs.onCompleted();


  1. transducer (Transducer): A transducer to execute.


(Observable): An observable sequence that results from the transducer execution.


Below is an example using transducers-js.

function even (x) { return x % 2 === 0; }
function mul10(x) { return x * 10; }

var t = transducers

var source = Rx.Observable.range(1, 5)
  .transduce(t.comp(t.filter(even), t.map(mul10)));

var subscription = source.subscribe(
  function (x) {
    console.log('Next: %s', x);
  function (err) {
    console.log('Error: %s', err);
  function () {

// => Next: 20
// => Next: 40
// => Completed