@stdlib/blas-base-wasm-daxpy

v0.1.0

Published

8 days ago

Multiply a vector `x` by a constant and add the result to `y`.

daxpy

Multiply a vector x by a constant alpha and add the result to y.

Installation

npm install @stdlib/blas-base-wasm-daxpy

Usage

var daxpy = require( '@stdlib/blas-base-wasm-daxpy' );

daxpy.main( N, alpha, x, strideX, y, strideY )

Multiplies a vector x by a constant alpha and adds the result to y.

var Float64Array = require( '@stdlib/array-float64' );

var x = new Float64Array( [ 1.0, 2.0, 3.0, 4.0, 5.0 ] );
var y = new Float64Array( [ 1.0, 1.0, 1.0, 1.0, 1.0 ] );

daxpy.main( x.length, 5.0, x, 1, y, 1 );
// y => <Float64Array>[ 6.0, 11.0, 16.0, 21.0, 26.0 ]

The function has the following parameters:

N: number of indexed elements.
alpha: scalar constant.
x: input Float64Array.
strideX: index increment for x.
y: input Float64Array.
strideY: index increment for y.

The N and stride parameters determine which elements in the strided arrays are accessed at runtime. For example, to multiply every other value in x by alpha and add the result to the first N elements of y in reverse order,

var Float64Array = require( '@stdlib/array-float64' );

var x = new Float64Array( [ 1.0, 2.0, 3.0, 4.0, 5.0, 6.0 ] );
var y = new Float64Array( [ 1.0, 1.0, 1.0, 1.0, 1.0, 1.0 ] );

var alpha = 5.0;

daxpy.main( 3, alpha, x, 2, y, -1 );
// y => <Float64Array>[ 26.0, 16.0, 6.0, 1.0, 1.0, 1.0 ]

Note that indexing is relative to the first index. To introduce an offset, use typed array views.

var Float64Array = require( '@stdlib/array-float64' );

// Initial arrays...
var x0 = new Float64Array( [ 1.0, 2.0, 3.0, 4.0, 5.0, 6.0 ] );
var y0 = new Float64Array( [ 7.0, 8.0, 9.0, 10.0, 11.0, 12.0 ] );

// Create offset views...
var x1 = new Float64Array( x0.buffer, x0.BYTES_PER_ELEMENT*1 ); // start at 2nd element
var y1 = new Float64Array( y0.buffer, y0.BYTES_PER_ELEMENT*3 ); // start at 4th element

daxpy.main( 3, 5.0, x1, -2, y1, 1 );
// y0 => <Float64Array>[ 7.0, 8.0, 9.0, 40.0, 31.0, 22.0 ]

daxpy.ndarray( N, alpha, x, strideX, offsetX, y, strideY, offsetY )

Multiplies a vector x by a constant alpha and adds the result to y using alternative indexing semantics.

var Float64Array = require( '@stdlib/array-float64' );

var x = new Float64Array( [ 1.0, 2.0, 3.0, 4.0, 5.0 ] );
var y = new Float64Array( [ 1.0, 1.0, 1.0, 1.0, 1.0 ] );
var alpha = 5.0;

daxpy.ndarray( x.length, alpha, x, 1, 0, y, 1, 0 );
// y => <Float64Array>[ 6.0, 11.0, 16.0, 21.0, 26.0 ]

The function has the following additional parameters:

offsetX: starting index for x.
offsetY: starting index for y.

While typed array views mandate a view offset based on the underlying buffer, the offset parameters support indexing semantics based on starting indices. For example, to multiply every other value in x by a constant alpha starting from the second value and add to the last N elements in y where x[i] -> y[n], x[i+2] -> y[n-1],...,

var Float64Array = require( '@stdlib/array-float64' );

var x = new Float64Array( [ 1.0, 2.0, 3.0, 4.0, 5.0, 6.0 ] );
var y = new Float64Array( [ 7.0, 8.0, 9.0, 10.0, 11.0, 12.0 ] );

var alpha = 5.0;

daxpy.ndarray( 3, alpha, x, 2, 1, y, -1, y.length-1 );
// y => <Float64Array>[ 7.0, 8.0, 9.0, 40.0, 31.0, 22.0 ]

Module

daxpy.Module( memory )

Returns a new WebAssembly module wrapper instance which uses the provided WebAssembly memory instance as its underlying memory.

var Memory = require( '@stdlib/wasm-memory' );

// Create a new memory instance with an initial size of 10 pages (640KiB) and a maximum size of 100 pages (6.4MiB):
var mem = new Memory({
    'initial': 10,
    'maximum': 100
});

// Create a BLAS routine:
var mod = new daxpy.Module( mem );
// returns <Module>

// Initialize the routine:
mod.initializeSync();

daxpy.Module.prototype.main( N, α, xp, sx, yp, sy )

Multiplies a vector x by a constant and adds the result to y.

var Memory = require( '@stdlib/wasm-memory' );
var oneTo = require( '@stdlib/array-one-to' );
var ones = require( '@stdlib/array-ones' );
var zeros = require( '@stdlib/array-zeros' );
var bytesPerElement = require( '@stdlib/ndarray-base-bytes-per-element' );

// Create a new memory instance with an initial size of 10 pages (640KiB) and a maximum size of 100 pages (6.4MiB):
var mem = new Memory({
    'initial': 10,
    'maximum': 100
});

// Create a BLAS routine:
var mod = new daxpy.Module( mem );
// returns <Module>

// Initialize the routine:
mod.initializeSync();

// Define a vector data type:
var dtype = 'float64';

// Specify a vector length:
var N = 5;

// Define pointers (i.e., byte offsets) for storing two vectors:
var xptr = 0;
var yptr = N * bytesPerElement( dtype );

// Write vector values to module memory:
mod.write( xptr, oneTo( N, dtype ) );
mod.write( yptr, ones( N, dtype ) );

// Perform computation:
mod.main( N, 5.0, xptr, 1, yptr, 1 );

// Read out the results:
var view = zeros( N, dtype );
mod.read( yptr, view );

console.log( view );
// => <Float64Array>[ 6.0, 11.0, 16.0, 21.0, 26.0 ]

The function has the following parameters:

N: number of indexed elements.
α: scalar constant.
xp: input Float64Array pointer (i.e., byte offset).
sx: index increment for x.
yp: input Float64Array pointer (i.e., byte offset).
sy: index increment for y.

daxpy.Module.prototype.ndarray( N, α, xp, sx, ox, yp, sy, oy )

Multiplies a vector x by a constant and adds the result to y using alternative indexing semantics.

var Memory = require( '@stdlib/wasm-memory' );
var oneTo = require( '@stdlib/array-one-to' );
var ones = require( '@stdlib/array-ones' );
var zeros = require( '@stdlib/array-zeros' );
var bytesPerElement = require( '@stdlib/ndarray-base-bytes-per-element' );

// Create a new memory instance with an initial size of 10 pages (640KiB) and a maximum size of 100 pages (6.4MiB):
var mem = new Memory({
    'initial': 10,
    'maximum': 100
});

// Create a BLAS routine:
var mod = new daxpy.Module( mem );
// returns <Module>

// Initialize the routine:
mod.initializeSync();

// Define a vector data type:
var dtype = 'float64';

// Specify a vector length:
var N = 5;

// Define pointers (i.e., byte offsets) for storing two vectors:
var xptr = 0;
var yptr = N * bytesPerElement( dtype );

// Write vector values to module memory:
mod.write( xptr, oneTo( N, dtype ) );
mod.write( yptr, ones( N, dtype ) );

// Perform computation:
mod.ndarray( N, 5.0, xptr, 1, 0, yptr, 1, 0 );

// Read out the results:
var view = zeros( N, dtype );
mod.read( yptr, view );

console.log( view );
// => <Float64Array>[ 6.0, 11.0, 16.0, 21.0, 26.0 ]

The function has the following additional parameters:

ox: starting index for x.
oy: starting index for y.

Notes

If N <= 0 or alpha == 0, y is left unchanged.
This package implements routines using WebAssembly. When provided arrays which are not allocated on a daxpy module memory instance, data must be explicitly copied to module memory prior to computation. Data movement may entail a performance cost, and, thus, if you are using arrays external to module memory, you should prefer using @stdlib/blas-base/daxpy. However, if working with arrays which are allocated and explicitly managed on module memory, you can achieve better performance when compared to the pure JavaScript implementations found in @stdlib/blas/base/daxpy. Beware that such performance gains may come at the cost of additional complexity when having to perform manual memory management. Choosing between implementations depends heavily on the particular needs and constraints of your application, with no one choice universally better than the other.
daxpy() corresponds to the BLAS level 1 function daxpy.

Examples

var discreteUniform = require( '@stdlib/random-array-discrete-uniform' );
var daxpy = require( '@stdlib/blas-base-wasm-daxpy' );

var opts = {
    'dtype': 'float64'
};
var x = discreteUniform( 10, 0, 100, opts );
console.log( x );

var y = discreteUniform( x.length, 0, 10, opts );
console.log( y );

daxpy.ndarray( x.length, 5.0, x, 1, 0, y, -1, y.length-1 );
console.log( y );

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.