@stdlib/ndarray-slice-assign NPM

sliceAssign

Assign element values from a broadcasted input ndarray to corresponding elements in an output ndarray view.

Installation

npm install @stdlib/ndarray-slice-assign

Usage

var sliceAssign = require( '@stdlib/ndarray-slice-assign' );

sliceAssign( x, y, ...s[, options] )

Assigns element values from a broadcasted input ndarray to corresponding elements in an output ndarray view.

var Slice = require( '@stdlib/slice-ctor' );
var MultiSlice = require( '@stdlib/slice-multi' );
var ndarray = require( '@stdlib/ndarray-ctor' );
var ndzeros = require( '@stdlib/ndarray-zeros' );
var ndarray2array = require( '@stdlib/ndarray-to-array' );

// Define an input array:
var buffer = [ 1.0, 2.0, 3.0, 4.0, 5.0, 6.0 ];
var shape = [ 3, 2 ];
var strides = [ 2, 1 ];
var offset = 0;

var x = ndarray( 'generic', buffer, shape, strides, offset, 'row-major' );
// returns <ndarray>

var sh = x.shape;
// returns [ 3, 2 ]

var arr = ndarray2array( x );
// returns [ [ 1.0, 2.0 ], [ 3.0, 4.0 ], [ 5.0, 6.0 ] ]

// Define an output array:
var y = ndzeros( [ 2, 3, 2 ], {
    'dtype': x.dtype
});

// Create a slice:
var s0 = null;
var s1 = new Slice( null, null, -1 );
var s2 = new Slice( null, null, -1 );
var s = new MultiSlice( s0, s1, s2 );
// returns <MultiSlice>

// Perform assignment:
var out = sliceAssign( x, y, s );
// returns <ndarray>

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

arr = ndarray2array( y );
// returns [ [ [ 6.0, 5.0 ], [ 4.0, 3.0 ], [ 2.0, 1.0 ] ], [ [ 6.0, 5.0 ], [ 4.0, 3.0 ], [ 2.0, 1.0 ] ] ]

The function accepts the following arguments:

x: input ndarray.
y: output ndarray.
s: a MultiSlice instance, an array of slice arguments, or slice arguments as separate arguments.
options: function options.

The function supports three (mutually exclusive) means for providing slice arguments:

providing a single MultiSlice instance.
providing a single array of slice arguments.
providing slice arguments as separate arguments.

The following example demonstrates each invocation style achieving equivalent results.

var Slice = require( '@stdlib/slice-ctor' );
var MultiSlice = require( '@stdlib/slice-multi' );
var scalar2ndarray = require( '@stdlib/ndarray-from-scalar' );
var ndzeros = require( '@stdlib/ndarray-zeros' );
var ndarray2array = require( '@stdlib/ndarray-to-array' );

// 1. Using a MultiSlice:
var x = scalar2ndarray( 10.0 );
var y = ndzeros( [ 2, 3 ] );

var s0 = 0;
var s1 = new Slice( 1, null, 1 );
var s = new MultiSlice( s0, s1 );
// returns <MultiSlice>

var out = sliceAssign( x, y, s );
// returns <ndarray>

var arr = ndarray2array( out );
// returns [ [ 0.0, 10.0, 10.0 ], [ 0.0, 0.0, 0.0 ] ]

// 2. Using an array of slice arguments:
x = scalar2ndarray( 10.0 );
y = ndzeros( [ 2, 3 ] );

out = sliceAssign( x, y, [ s0, s1 ] );
// returns <ndarray>

arr = ndarray2array( out );
// returns [ [ 0.0, 10.0, 10.0 ], [ 0.0, 0.0, 0.0 ] ]

// 3. Providing separate arguments:
x = scalar2ndarray( 10.0 );
y = ndzeros( [ 2, 3 ] );

out = sliceAssign( x, y, s0, s1 );
// returns <ndarray>

arr = ndarray2array( out );
// returns [ [ 0.0, 10.0, 10.0 ], [ 0.0, 0.0, 0.0 ] ]

The function supports the following options:

strict: boolean indicating whether to enforce strict bounds checking.

By default, the function throws an error when provided a slice which exceeds array bounds. To ignore slice indices exceeding array bounds, set the strict option to false.

var Slice = require( '@stdlib/slice-ctor' );
var MultiSlice = require( '@stdlib/slice-multi' );
var scalar2ndarray = require( '@stdlib/ndarray-from-scalar' );
var ndzeros = require( '@stdlib/ndarray-zeros' );
var ndarray2array = require( '@stdlib/ndarray-to-array' );

// Define an input array:
var x = scalar2ndarray( 10.0 );

// Define an output array:
var y = ndzeros( [ 3, 2 ], {
    'dtype': x.dtype
});

// Create a slice:
var s0 = new Slice( 1, null, 1 );
var s1 = new Slice( 10, 20, 1 );
var s = new MultiSlice( s0, s1 );
// returns <MultiSlice>

// Perform assignment:
var out = sliceAssign( x, y, s, {
    'strict': false
});
// returns <ndarray>

var arr = ndarray2array( y );
// returns [ [ 0.0, 0.0 ], [ 0.0, 0.0 ], [ 0.0, 0.0 ] ]

Notes

An output ndarray must be writable. If provided a read-only ndarray, the function throws an error.
A slice argument must be either a Slice, an integer, null, or undefined.
The number of slice dimensions must match the number of output array dimensions. Hence, if y is a zero-dimensional ndarray, then, if s is a MultiSlice, s should be empty, and, if s is an array, s should not contain any slice arguments. Similarly, if y is a one-dimensional ndarray, then, if s is a MultiSlice, s should have one slice dimension, and, if s is an array, s should contain a single slice argument. And so on and so forth.
The input ndarray must be broadcast compatible with the output ndarray view.
The input ndarray must have a data type which can be safely cast to the output ndarray data type. Floating-point data types (both real and complex) are allowed to downcast to a lower precision data type of the same kind (e.g., element values from a 'float64' input ndarray can be assigned to corresponding elements in a 'float32' output ndarray).

Examples

var E = require( '@stdlib/slice-multi' );
var scalar2ndarray = require( '@stdlib/ndarray-from-scalar' );
var ndarray2array = require( '@stdlib/ndarray-to-array' );
var ndzeros = require( '@stdlib/ndarray-zeros' );
var slice = require( '@stdlib/ndarray-slice' );
var sliceAssign = require( '@stdlib/ndarray-slice-assign' );

// Alias `null` to allow for more compact indexing expressions:
var _ = null;

// Create an output ndarray:
var y = ndzeros( [ 3, 3, 3 ] );

// Update each matrix...
var s1 = E( 0, _, _ );
sliceAssign( scalar2ndarray( 100 ), y, s1 );

var a1 = ndarray2array( slice( y, s1 ) );
// returns [ [ 100, 100, 100 ], [ 100, 100, 100 ], [ 100, 100, 100 ] ]

var s2 = E( 1, _, _ );
sliceAssign( scalar2ndarray( 200 ), y, s2 );

var a2 = ndarray2array( slice( y, s2 ) );
// returns [ [ 200, 200, 200 ], [ 200, 200, 200 ], [ 200, 200, 200 ] ]

var s3 = E( 2, _, _ );
sliceAssign( scalar2ndarray( 300 ), y, s3 );

var a3 = ndarray2array( slice( y, s3 ) );
// returns [ [ 300, 300, 300 ], [ 300, 300, 300 ], [ 300, 300, 300 ] ]

// Update the second rows in each matrix:
var s4 = E( _, 1, _ );
sliceAssign( scalar2ndarray( 400 ), y, s4 );

var a4 = ndarray2array( slice( y, s4 ) );
// returns [ [ 400, 400, 400 ], [ 400, 400, 400 ], [ 400, 400, 400 ] ]

// Update the second columns in each matrix:
var s5 = E( _, _, 1 );
sliceAssign( scalar2ndarray( 500 ), y, s5 );

var a5 = ndarray2array( slice( y, s5 ) );
// returns [ [ 500, 500, 500 ], [ 500, 500, 500 ], [ 500, 500, 500 ] ]

// Return the contents of the entire ndarray:
var a6 = ndarray2array( y );
/* returns
  [
    [
      [ 100, 500, 100 ],
      [ 400, 500, 400 ],
      [ 100, 500, 100 ]
    ],
    [
      [ 200, 500, 200 ],
      [ 400, 500, 400 ],
      [ 200, 500, 200 ]
    ],
    [
      [ 300, 500, 300 ],
      [ 400, 500, 400 ],
      [ 300, 500, 300 ]
    ]
  ]
*/

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.

@stdlib/ndarray-slice-assign v0.2.1