Random

Constructor for creating ndarrays filled with pseudorandom values drawn from a unary PRNG.

Installation

npm install @stdlib/random-tools-unary

Usage

var Random = require( '@stdlib/random-tools-unary' );

Random( prng, idtypes, odtypes, policies[, options] )

Returns an interface for creating ndarrays filled with pseudorandom values drawn from a unary PRNG.

var dtypes = require( '@stdlib/ndarray-dtypes' );
var exponential = require( '@stdlib/random-base-exponential' );

var idt = dtypes( 'real_and_generic' );
var odt = dtypes( 'real_floating_point_and_generic' );

var policies = {
    'output': 'real_floating_point_and_generic'
};
var options = {
    'order': 'row-major'
};

var rand = new Random( exponential, idt, odt, policies, options );

The constructor has the following parameters:

• prng: unary pseudorandom value generator.

• idtypes: list of supported input data types.

• odtypes: list of supported output data types.

• policies: interface policies. Must have the following properties:

  • output: output data type policy.

• options: function options (optional).

The constructor supports the following options:

• order: default memory layout.

Random.prototype.generate( shape, param1[, options] )

Returns an ndarray filled with pseudorandom values drawn from a unary PRNG.

var dtypes = require( '@stdlib/ndarray-dtypes' );
var exponential = require( '@stdlib/random-base-exponential' );

var idt = dtypes( 'real_and_generic' );
var odt = dtypes( 'real_floating_point_and_generic' );

var policies = {
    'output': 'real_floating_point_and_generic'
};
var options = {
    'order': 'row-major'
};

var rand = new Random( exponential, idt, odt, policies, options );

var v = rand.generate( [ 2, 2 ], 2.0 );
// returns <ndarray>

The method has the following parameters:

• shape: output ndarray shape.
• param1: PRNG parameter. May be either a scalar or an ndarray. If an ndarray, must be broadcast compatible with the specified output ndarray shape.
• options: function options (optional).

The method accepts the following options:

• dtype: output ndarray data type. Setting this option overrides the output data type policy.
• order: memory layout. Setting this option overrides the default memory layout.
• mode: specifies how to handle indices which exceed ndarray dimensions.
• submode: specifies how to handle subscripts which exceed ndarray dimensions on a per dimension basis.
• readonly: boolean indicating whether an ndarray should be read-only.

By default, the method returns an ndarray having a data type determined by the output data type policy. To override the default behavior, set the dtype option.

var dtypes = require( '@stdlib/ndarray-dtypes' );
var getDType = require( '@stdlib/ndarray-dtype' );
var exponential = require( '@stdlib/random-base-exponential' );

var idt = dtypes( 'real_and_generic' );
var odt = dtypes( 'real_floating_point_and_generic' );

var policies = {
    'output': 'real_floating_point_and_generic'
};
var options = {
    'order': 'row-major'
};

var rand = new Random( exponential, idt, odt, policies, options );

var v = rand.generate( [ 2, 2 ], 2.0, {
    'dtype': 'generic'
});
// returns <ndarray>

var dt = String( getDType( v ) );
// returns 'generic'

Random.prototype.assign( param1, out )

Fills an ndarray with pseudorandom values drawn from a unary PRNG.

var dtypes = require( '@stdlib/ndarray-dtypes' );
var ndzeros = require( '@stdlib/ndarray-zeros' );
var exponential = require( '@stdlib/random-base-exponential' );

var idt = dtypes( 'real_and_generic' );
var odt = dtypes( 'real_floating_point_and_generic' );

var policies = {
    'output': 'real_floating_point_and_generic'
};
var options = {
    'order': 'row-major'
};

var rand = new Random( exponential, idt, odt, policies, options );

var out = ndzeros( [ 2, 2 ] );
var v = rand.assign( 2.0, out );
// returns <ndarray>

var bool = ( v === out );
// returns true

The method has the following parameters:

• param1: PRNG parameter. May be either a scalar or an ndarray. If an ndarray, must be broadcast compatible with the output ndarray.
• out: output ndarray.

Notes

• The output data type policy only applies to the generate method. For the assign method, the output ndarray is allowed to have any supported output data type.

Examples

var exponential = require( '@stdlib/random-base-exponential' );
var dtypes = require( '@stdlib/ndarray-dtypes' );
var ndarray = require( '@stdlib/ndarray-ctor' );
var ndarray2array = require( '@stdlib/ndarray-to-array' );
var Random = require( '@stdlib/random-tools-unary' );

// Create a new PRNG instance...
var idt = dtypes( 'real_and_generic' );
var odt = dtypes( 'real_floating_point_and_generic' );
var policies = {
    'output': 'real_floating_point_and_generic'
};
var random = new Random( exponential, idt, odt, policies );

// Generate a 3x3 matrix of pseudorandom numbers:
var x = random.generate( [ 3, 3 ], 2.0 );
console.log( ndarray2array( x ) );

// Generate another matrix with a specified data type:
x = random.generate( [ 3, 3 ], 2.0, {
    'dtype': 'float32'
});
console.log( ndarray2array( x ) );

// Define an array of distribution parameters:
var param = new ndarray( 'generic', [ 2.0, 20.0, 200.0 ], [ 3, 1 ], [ 1, 1 ], 0, 'row-major' );

// Broadcast the parameters to generate another 3x3 matrix of pseudorandom numbers:
x = random.generate( [ 3, 3 ], param );
console.log( ndarray2array( x ) );

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 © 2016-2026. The Stdlib Authors.