asyncplify vs rx.net

Introduction
Machine
Results
Tables
Code used for the test
Conclusion

Introduction

Last year, I’ve worked on a projet requiring lots of calculations to calculate financial indicators. The amount of data to process was huge. I managed to get it worked using Rx.Net. Coming from a .Net/C# background, it was a natural fit.

After a year of working on nodejs and angularjs, I felt in love with javascript. I decided test rxjs but the performance was not convincing me. I decided to implement my own framework based on rxjs but with performance in mind.

The asyncplify framework was born.

Machine

Microsoft Windows 8.1 Pro 64Bit
Intel i7 2.20GHz
12 GB Memory
.Net Framework 4.5.1
nodejs 0.12.4

Results

results

Tables

	rx.net	rxjs	asyncplify
sum	1497ms	930ms	19ms
count	1360ms	1024ms	5ms
groupBy/flatMap	1491ms	1538ms	37ms
multiple sum/subscribe	21ms	961ms	3ms

Code used for the test

rx.net code

using System;
using System.Diagnostics;
using System.Reactive.Linq;

namespace ConsoleApplication1
{
    class Program
    {
        static void Main(string[] args)
        {
            var w = Stopwatch.StartNew();

            Observable
                .Range(0, 1000000)
                .Sum(v => (long)v)
                .Subscribe(x => Console.WriteLine("sum = {0} in {1}ms", x, w.ElapsedMilliseconds));

            w.Restart();

            Observable
                .Range(0, 1000000)
                .Count()
                .Subscribe(x => Console.WriteLine("count = {0} in {1}ms", x, w.ElapsedMilliseconds));

            w.Restart();

            Observable
                .Range(0, 1000000)
                .GroupBy(x => x % 10)
                .SelectMany(g => g)
                .Subscribe(_ => { }, () => Console.WriteLine("groupBy/flatMap in {0}ms", w.ElapsedMilliseconds));
                
            w.Restart();

            for (var i = 0; i < 1000; i++)
            {
                Observable
                    .Range(0, 10)
                    .Sum()
                    .Subscribe();
            }

            Console.WriteLine("multiple sum {0}ms", w.ElapsedMilliseconds);

            Console.ReadLine();
        }
    }
}

Used Rx.net version 2.2.5. I have compile this code using visual studio .net 2013 in release mode and start the program without a debugger attached.

rxjs code

var rx = require('rx').Observable;
var d = new Date();
	
rx
	.range(0, 1000000)
	.sum()
	.subscribe(function (x) { console.log('sum = %d in %dms', x, new Date() - d); });
	
d = new Date();
	
rx
	.range(0, 1000000)
	.count()
	.subscribe(function (x) { console.log('count = %d in %dms', x, new Date() - d); });
	
d = new Date();
	
rx
	.range(0, 1000000)
	.groupBy(function (x) { return x % 10; })
	.flatMap(function (g) { return g; })
	.subscribe(function () {}, function () {}, function () { console.log('groupBy/flatMap in %dms', new Date() - d); });
    
d = new Date();
	
for (var i = 0; i < 1000; i++) {
	rx
		.range(0, 1000)
		.sum()
		.subscribe();
}

console.log("multiple sum in %dms", new Date() - d);

used rx version 2.5.3

asyncplify code

var asyncplify = require('asyncplify');
var d = new Date();

asyncplify
	.range(1000000)
	.sum()
	.subscribe(function (x) { console.log('sum = %d in %dms', x, new Date() - d); });
	
d = new Date();
	
asyncplify
	.range(1000000)
	.count()
	.subscribe(function (x) { console.log('count = %d in %dms', x, new Date() - d); });
	
d = new Date();
	
asyncplify
	.range(1000000)
	.groupBy(function (x) { return x % 10; })
	.flatMap(function (g) { return g; })
	.subscribe({ end: function () { console.log('groupBy/flatMap in %dms', new Date() - d); } });
    
d = new Date();

for (var i = 0; i < 1000; i++) {
	asyncplify
		.range(10)
		.sum()
		.subscribe();
}

console.log("multiple sum in %dms", new Date() - d);

used asyncplify version 0.4.2

Conclusion

We can see that asyncplify is fast, faster than the 2 other rx framework. rxjs is in most case faster than rx.net but when using many rxjs subscription, rx.net is faster than rxjs. After seeing such positive result, I will continue to develop asyncplify.

God bless you… Dany