What's difference between “reduce” and “scan”

问题

I'm studying RXJS and stuck with the problem: the same code with operators "reduce" and "scan" works in different ways, but I think that must return equal result. Example below. Please help.

const txtElement1 = document.getElementById('txt1');
const txtElement2 = document.getElementById('txt2');
const txtElement3 = document.getElementById('txt3');

// function return Observable
function get(array, initValue) {
  return Rx.Observable.create(observer => {
    let timer = initValue;

    array.forEach(item => {
      setTimeout(() => observer.next(item), timer);
      timer += 1000;
    });
  });
}

// 1) don't work with "reduce"
var stream1$ = get(['John', 'Ann', 'Bob'])
  .reduce(function(acc, x) {
    return acc + ` ${x}`;
  }, 'first - ');

stream1$.subscribe(text => txtElement1.innerHTML = text);

// 2)  the same code, but with "scan" - working
var stream2$ = get(['John', 'Ann', 'Bob'])
  .scan(function(acc, x) {
    return acc + ` ${x}`;
  }, 'second - ');

stream2$.subscribe(text => txtElement2.innerHTML = text);

// 3)  and the simple Observable with "reduce" - working
var stream3$ = Rx.Observable.from(['John', 'Ann', 'Bob'])
  .reduce(function(acc, x) {
    return acc + ` ${x}`;
  }, 'third - ');

stream3$.subscribe(text => txtElement3.innerHTML = text);

回答1:

From RxJS documentation,

Scan

apply a function to each item emitted by an Observable, sequentially, and emit each successive value

Reduce

apply a function to each item emitted by an Observable, sequentially, and emit the final value

Example codes

Scan

var source = Rx.Observable.range(1, 3)
.scan(
    function (acc, x) {
        return acc + x;
    });

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

For each value emitted by the Observable, scan emits corresponding output sequentially, So the Output will have 3 values for range 1 to 3, as follows

Output
Next: 1
Next: 3
Next: 6
Completed

Reduce

var source = Rx.Observable.range(1, 3)
    .reduce(function (acc, x) {
        return acc * x;
    }, 1)

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

Reduce function reduces the values from observables to a single value (final result) and emit. So the output will be as follows,

Next: 6
Completed

回答2:

Rx.scan is continuous whereas reduce is not continuous.

回答3:

there almost the same but scan emits each iteration. and reduce just emits a final results. i like scan when i want to actually see what reduce is doing. so sometimesi . use reduce and i cant visualize what reduce is actually doing. in this case i simple swap out reduce for scan since it will emit on each iteration. this way i can put in a log statement and see the results for each iteration.

来源：https://stackoverflow.com/questions/45350806/whats-difference-between-reduce-and-scan

标签

javascript

rxjs