Anscombe's Quartet

Anscombe's quartet.

Anscombe's quartet is a set of 4 datasets which all have nearly identical simple statistical properties but vary considerably when graphed. Anscombe created the datasets to demonstrate why graphical data exploration should precede statistical data analysis and to show the effect of outliers on statistical properties.

Installation

$ npm install datasets-anscombes-quartet

Usage

var data = require( 'datasets-anscombes-quartet' );

data

Anscombe's quartet comprises 4 individual datasets, where each individual dataset is an array of [x,y] tuples.

console.log( data );
/*
    [
        [
            [10,8.04],
            [8,6.95],
            ...
        ],
        [
            [10,9.14],
            [8,8.14],
            ...
        ],
        ...
    ]
*/

Examples

var toMatrix = require( 'compute-to-matrix' ),
    mean = require( 'compute-mean' ),
    variance = require( 'compute-variance' ),
    data = require( 'datasets-anscombes-quartet' );
 
var len = data.length,
    mats = new Array( len ),
    mu,
    s2,
    i;
 
// Convert the individual datasets to matrices...
for ( i = 0; i < len; i++ ) {
    mats[ i ] = toMatrix( data[ i ] );
}
/*
                [ xi0 yi0
                  xi1 yi1
                  xi2 yi2
    mats[ i ] =      .
                     .
                     .
                  xiN yiN ]
*/
 
// Calculate the means and variances along the rows for each matrix...
for ( i = 0; i < len; i++ ) {
    mu = mean( mats[ i ], {
        'dim': 1
    });
    /*
        [ E[x], E[y] ]
    */
 
    s2 = variance( mats[ i ], {
        'dim': 1
    });
    /*
        [ Var[x], Var[y] ]
    */
 
    console.log( 'Dataset %d: E[x] = %d, Var[x] = %d.', i+1, mu.get(0,0), s2.get(0,0) );
    console.log( 'Dataset %d: E[y] = %d, Var[y] = %d.\n', i+1, mu.get(0,1), s2.get(0,1) );
}

To run the example code from the top-level application directory,

$ node ./examples/index.js

Tests

Unit

Unit tests use the Mocha test framework with Chai assertions. To run the tests, execute the following command in the top-level application directory:

$ make test

All new feature development should have corresponding unit tests to validate correct functionality.

Test Coverage

This repository uses Istanbul as its code coverage tool. To generate a test coverage report, execute the following command in the top-level application directory:

$ make test-cov

Istanbul creates a ./reports/coverage directory. To access an HTML version of the report,

$ make view-cov

---

License

MIT license.

Copyright

Copyright © 2015. The Compute.io Authors.