198 lines
5.5 KiB
JavaScript
198 lines
5.5 KiB
JavaScript
|
/**
|
||
|
* @license Apache-2.0
|
||
|
*
|
||
|
* Copyright (c) 2018 The Stdlib Authors.
|
||
|
*
|
||
|
* Licensed under the Apache License, Version 2.0 (the "License");
|
||
|
* you may not use this file except in compliance with the License.
|
||
|
* You may obtain a copy of the License at
|
||
|
*
|
||
|
* http://www.apache.org/licenses/LICENSE-2.0
|
||
|
*
|
||
|
* Unless required by applicable law or agreed to in writing, software
|
||
|
* distributed under the License is distributed on an "AS IS" BASIS,
|
||
|
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||
|
* See the License for the specific language governing permissions and
|
||
|
* limitations under the License.
|
||
|
*/
|
||
|
|
||
|
'use strict';
|
||
|
|
||
|
// MODULES //
|
||
|
|
||
|
var isCollection = require( '@stdlib/assert/is-collection' );
|
||
|
var isPlainObject = require( '@stdlib/assert/is-plain-object' );
|
||
|
var setReadOnly = require( '@stdlib/utils/define-read-only-property' );
|
||
|
var countBy = require( '@stdlib/utils/count-by' );
|
||
|
var objectKeys = require( '@stdlib/utils/keys' );
|
||
|
var rank = require( './../../ranks' );
|
||
|
var pow = require( '@stdlib/math/base/special/pow' );
|
||
|
var chisqCDF = require( './../../base/dists/chisquare/cdf' );
|
||
|
var identity = require( '@stdlib/utils/identity-function' );
|
||
|
var incrspace = require( '@stdlib/array/incrspace' );
|
||
|
var validate = require( './validate.js' );
|
||
|
var print = require( './print.js' ); // eslint-disable-line stdlib/no-redeclare
|
||
|
|
||
|
|
||
|
// MAIN //
|
||
|
|
||
|
/**
|
||
|
* Computes the Kruskal-Wallis test for equality of medians.
|
||
|
*
|
||
|
* @param {...NumberArray} arguments - either two or more number arrays or a single numeric array if an array of group indicators is supplied as an option
|
||
|
* @param {Options} [options] - function options
|
||
|
* @param {number} [options.alpha=0.05] - significance level
|
||
|
* @param {Array} [options.groups] - array of group indicators
|
||
|
* @throws {Error} must provide at least two array-like arguments if `groups` is not set
|
||
|
* @throws {TypeError} must provide array-like arguments
|
||
|
* @throws {TypeError} options has to be simple object
|
||
|
* @throws {TypeError} must provide valid options
|
||
|
* @throws {RangeError} alpha option has to be a number in the interval `[0,1]`
|
||
|
* @returns {Object} test results
|
||
|
*
|
||
|
* @example
|
||
|
* // Data from Hollander & Wolfe (1973), p. 116:
|
||
|
* var x = [ 2.9, 3.0, 2.5, 2.6, 3.2 ];
|
||
|
* var y = [ 3.8, 2.7, 4.0, 2.4 ];
|
||
|
* var z = [ 2.8, 3.4, 3.7, 2.2, 2.0 ];
|
||
|
*
|
||
|
* var out = kruskal( x, y, z );
|
||
|
* // returns {...}
|
||
|
*/
|
||
|
function kruskal() {
|
||
|
var groupsIndicators;
|
||
|
var groupRankSums;
|
||
|
var tieSumTerm;
|
||
|
var ngroups;
|
||
|
var options;
|
||
|
var levels;
|
||
|
var alpha;
|
||
|
var param;
|
||
|
var ranks;
|
||
|
var vals;
|
||
|
var opts;
|
||
|
var pval;
|
||
|
var stat;
|
||
|
var ties;
|
||
|
var arg;
|
||
|
var err;
|
||
|
var key;
|
||
|
var out;
|
||
|
var i;
|
||
|
var j;
|
||
|
var n;
|
||
|
var N;
|
||
|
var x;
|
||
|
var v;
|
||
|
|
||
|
ngroups = arguments.length;
|
||
|
opts = {};
|
||
|
if ( isPlainObject( arguments[ ngroups - 1 ] ) ) {
|
||
|
options = arguments[ ngroups - 1 ];
|
||
|
ngroups -= 1;
|
||
|
err = validate( opts, options );
|
||
|
if ( err ) {
|
||
|
throw err;
|
||
|
}
|
||
|
}
|
||
|
groupRankSums = {};
|
||
|
n = {};
|
||
|
if ( opts.groups ) {
|
||
|
x = arguments[ 0 ];
|
||
|
if ( x.length !== opts.groups.length ) {
|
||
|
throw new RangeError( 'invalid arguments. First argument and `opts.groups` must be arrays of the same length.' );
|
||
|
}
|
||
|
n = countBy( opts.groups, identity );
|
||
|
levels = objectKeys( n );
|
||
|
ngroups = levels.length;
|
||
|
for ( i = 0; i < ngroups; i++ ) {
|
||
|
key = levels[ i ];
|
||
|
groupRankSums[ key ] = 0;
|
||
|
}
|
||
|
if ( ngroups < 2 ) {
|
||
|
throw new Error( 'invalid number of groups. `groups` array must contain at least two unique elements. Value: `' + levels + '`.' );
|
||
|
}
|
||
|
groupsIndicators = opts.groups;
|
||
|
} else {
|
||
|
x = [];
|
||
|
groupsIndicators = [];
|
||
|
if ( ngroups < 2 ) {
|
||
|
throw new Error( 'invalid number of input arguments. Must provide at least two array-like arguments. Value: `' + arg + '`.' );
|
||
|
}
|
||
|
for ( i = 0; i < ngroups; i++ ) {
|
||
|
arg = arguments[ i ];
|
||
|
if ( !isCollection( arg ) ) {
|
||
|
throw new TypeError( 'invalid argument. Must provide array-like arguments. Value: `' + arg + '`.' );
|
||
|
}
|
||
|
if ( arg.length === 0 ) {
|
||
|
throw new Error( 'invalid argument. Supplied arrays cannot be empty. Value: `' + arg + '`.' );
|
||
|
} else {
|
||
|
n[ i ] = arg.length;
|
||
|
}
|
||
|
groupRankSums[ i ] = 0;
|
||
|
for ( j = 0; j < n[ i ]; j++ ) {
|
||
|
groupsIndicators.push( i );
|
||
|
x.push( arg[ j ] );
|
||
|
}
|
||
|
}
|
||
|
levels = incrspace( 0, ngroups, 1 );
|
||
|
}
|
||
|
if ( opts.alpha === void 0 ) {
|
||
|
alpha = 0.05;
|
||
|
} else {
|
||
|
alpha = opts.alpha;
|
||
|
}
|
||
|
if ( alpha < 0.0 || alpha > 1.0 ) {
|
||
|
throw new RangeError( 'invalid option. `alpha` must be a number in the range 0 to 1. Value: `' + alpha + '`.' );
|
||
|
}
|
||
|
|
||
|
N = x.length;
|
||
|
ranks = rank( x );
|
||
|
|
||
|
// Calculate # ties for each value & rank sums per group:
|
||
|
ties = {};
|
||
|
for ( i = 0; i < N; i++ ) {
|
||
|
groupRankSums[ groupsIndicators[ i ] ] += ranks[ i ];
|
||
|
if ( x[ i ] in ties ) {
|
||
|
ties[ x[ i ] ] += 1;
|
||
|
} else {
|
||
|
ties[ x[ i ] ] = 1;
|
||
|
}
|
||
|
}
|
||
|
|
||
|
// Calculate test statistic using short-cut formula:
|
||
|
stat = 0.0;
|
||
|
for ( i = 0; i < ngroups; i++ ) {
|
||
|
key = levels[ i ];
|
||
|
stat += pow( groupRankSums[ key ], 2.0 ) / n[ key ];
|
||
|
}
|
||
|
stat = ( ( 12.0 / ( N * (N+1) ) ) * stat ) - ( 3.0 * (N+1) );
|
||
|
|
||
|
// Correction for ties:
|
||
|
tieSumTerm = 0;
|
||
|
vals = objectKeys( ties );
|
||
|
for ( i = 0; i < vals.length; i++ ) {
|
||
|
v = ties[ vals[ i ] ];
|
||
|
tieSumTerm += pow( v, 3.0 ) - v;
|
||
|
}
|
||
|
|
||
|
stat /= 1.0 - ( ( tieSumTerm ) / ( pow( N, 3 ) - N ) );
|
||
|
param = ngroups - 1;
|
||
|
pval = 1.0 - chisqCDF( stat, param );
|
||
|
|
||
|
out = {};
|
||
|
setReadOnly( out, 'rejected', pval <= alpha );
|
||
|
setReadOnly( out, 'alpha', alpha );
|
||
|
setReadOnly( out, 'df', param );
|
||
|
setReadOnly( out, 'pValue', pval );
|
||
|
setReadOnly( out, 'statistic', stat );
|
||
|
setReadOnly( out, 'method', 'Kruskal-Wallis Test' );
|
||
|
setReadOnly( out, 'print', print );
|
||
|
return out;
|
||
|
}
|
||
|
|
||
|
|
||
|
// EXPORTS //
|
||
|
|
||
|
module.exports = kruskal;
|