@stdlib/stats-ranks
v0.2.2
Published
Compute ranks for values of an array-like object.
Downloads
1,497
Readme
ranks
Compute ranks for values of an array-like object.
Installation
npm install @stdlib/stats-ranks
Usage
var ranks = require( '@stdlib/stats-ranks' );
ranks( arr[, opts] )
Returns the sample ranks of the elements in arr
, which can be either an array
or typed array
.
var arr = [ 1.1, 2.0, 3.5, 0.0, 2.4 ];
var out = ranks( arr );
// returns [ 2, 3, 5, 1, 4 ]
// Ties are averaged:
arr = [ 2, 2, 1, 4, 3 ];
out = ranks( arr );
// returns [ 2.5, 2.5, 1, 5, 4 ];
// Missing values are placed last:
arr = [ null, 2, 2, 1, 4, 3, NaN, NaN ];
out = ranks( arr );
// returns [ 6, 2.5, 2.5, 1, 5, 4, 7 ,8 ]
The function accepts the following options:
- method:
string
indicating how ties are handled. Can be one of the following values:'average'
,'min'
,'max'
,'ordinal'
and'dense'
. Default:'average'
. - missing:
string
specifying how missing values are handled. Must be either'last'
,'first'
or'remove'
. Default:'last'
. - encoding:
array
holding all values which will be regarded as missing values. Default:[ NaN, null]
.
When all elements of the array
are different, the ranks are uniquely determined. When there are equal elements (called ties), the method
option determines how they are handled. The default, 'average'
, replace the ranks of the ties by their mean. Other possible options are 'min'
and 'max'
, which replace the ranks of the ties by their minimum and maximum, respectively. 'dense'
works like 'min'
, with the difference that the next highest element after a tie is assigned the next smallest integer. Finally, ordinal
gives each element in arr
a distinct rank, according to the position they appear in.
var data = [ 2, 2, 1, 4, 3 ];
// Max method:
var out = ranks( data, {
'method': 'max'
});
// returns [ 3, 3, 1, 5, 4 ]
// Min method:
out = ranks( data, {
'method': 'min'
});
// returns [ 2, 2, 1, 5, 4 ]
// Ordinal method
out = ranks( data, {
'method': 'ordinal'
});
// returns [ 2, 3, 1, 5, 4 ]
// Dense method:
out = [ 2, 2, 1, 4, 3 ];
out = ranks( data, {
'method': 'dense'
});
// returns [ 2, 2, 1, 4, 3 ]
The missing
option is used to specify how to handle missing data. By default, NaN
or null
are treated as missing values. 'last'
specifies that missing values are placed last, 'first'
that the are assigned the lowest ranks and 'remove'
means that they are removed from the array before the ranks are calculated.
var data = [ NaN, 2, 2, 1, 4, 3, null, null ];
var out = ranks( data, {
'missing': 'first'
});
// returns [ 1, 5.5, 5.5, 4, 8, 7, 2, 3 ]
out = ranks( data, {
'missing': 'last'
});
// returns [ 6, 2.5, 2.5, 1, 5, 4, 7 ,8 ]
out = ranks( data, {
'missing': 'remove'
});
// returns [ 2.5, 2.5, 1, 5, 4 ]
Custom encoding for missing values is supported via the encoding
option, which allows to supply the function with an array
of values which should be treated as missing.
var Int32Array = require( '@stdlib/array-int32' );
var data = new Int32Array( [ 2, 1, -999, 3, 4 ] );
var out = ranks( data, {
'encoding': [ -999 ]
});
// returns [ 2, 1, 5, 3, 4 ]
Examples
var Int32Array = require( '@stdlib/array-int32' );
var round = require( '@stdlib/math-base-special-round' );
var randu = require( '@stdlib/random-base-randu' );
var ranks = require( '@stdlib/stats-ranks' );
var data;
var out;
var i;
// Plain arrays...
data = new Array( 10 );
for ( i = 0; i < data.length; i++ ) {
data[ i ] = round( randu()*10.0 );
}
out = ranks( data );
// returns <array>
// Typed arrays...
data = new Int32Array( 10 );
for ( i = 0; i < data.length; i++ ) {
data[ i ] = randu() * 10.0;
}
out = ranks( data );
// returns <array>
Notice
This package is part of stdlib, a standard library for JavaScript and Node.js, with an emphasis on numerical and scientific computing. The library provides a collection of robust, high performance libraries for mathematics, statistics, streams, utilities, and more.
For more information on the project, filing bug reports and feature requests, and guidance on how to develop stdlib, see the main project repository.
Community
License
See LICENSE.
Copyright
Copyright © 2016-2024. The Stdlib Authors.