https://github.com/simple-statistics/stream-statistics

streaming statistical calculations for node
https://github.com/simple-statistics/stream-statistics

javascript statistics

Last synced: about 2 months ago
JSON representation

streaming statistical calculations for node

• Host: GitHub
• URL: https://github.com/simple-statistics/stream-statistics
• Owner: simple-statistics
• License: bsd-3-clause
• Archived: true
• Created: 2012-06-27T16:43:40.000Z (over 12 years ago)
• Default Branch: master
• Last Pushed: 2017-04-10T20:42:37.000Z (almost 8 years ago)
• Last Synced: 2024-10-03T00:24:18.295Z (3 months ago)
• Topics: javascript, statistics
• Language: JavaScript
• Homepage:
• Size: 171 KB
• Stars: 61
• Watchers: 6
• Forks: 4
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Changelog: CHANGELOG.md
  • License: LICENSE

Awesome Lists containing this project

README

        
[![Build Status](https://travis-ci.org/tmcw/stream-statistics.svg?branch=v0.1.0)](https://travis-ci.org/tmcw/stream-statistics)

![](https://farm9.staticflickr.com/8282/7711892138_6a4c08cd71_b.jpg)

## stream-statistics

### Install

    npm install stream-statistics

### API

This module exposes a single function that creates a stream. The stream

reads data, which it parses with `parseFloat()`, and computes statistics

on that data. When the input stream ends, `stream-statistics` emits the

`data` object.

The statistics object has the following members:

* `min`

* `max`

* `sum`

* `mean`

* `mode`

* `variance`

* `standard_deviation`

* `geometric_mean`

* `harmonic_mean`

`mode` computation expects the stream to deliver numbers in sorted lowest-first

order, and will return `undefined` if that expectation is not met.

### Use

```javascript

var streamStatistics = require('stream-statistics'),

    assert = require('assert');

function rangeStream(a, b) {

    var rs = new Readable({ objectMode: true });

    for (var i = 10; i < 1000; i++) { rs.push(i); }

    rs.push(null);

    return rs;

}

rangeStream(10, 1000).pipe(streamStatistics())

    .on('data', function(d) {

        assert.equal(d.min, 10);

    });

```

## cli

This also provides a binary, `sstatistics`, that you can get if you

`npm install -g` the library. Pipe numbers into it and it'll return

a basic set of stats about its input.

![](https://github.com/tmcw/stream-statistics/blob/master/screenshot.png?raw=true)

## [Documentation](https://github.com/tmcw/stream-statistics/wiki)

## See Also

* The sister project, [simple-statistics](https://github.com/tmcw/simple-statistics), that implements

  many of the same measures in straightforward and literate fashion

* [streaming-k-means](https://github.com/piatra/streaming-k-means) implements k-means clustering in an online fashion