<!--

@license Apache-2.0

Copyright (c) 2025 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.

-->

# Uint32Vector

> Create an unsigned 32-bit integer vector (i.e., a one-dimensional [ndarray][@stdlib/ndarray/ctor]).

<!-- Section to include introductory text. Make sure to keep an empty line after the intro `section` element and another before the `/section` close. -->

<section class="intro">

</section>

<!-- /.intro -->

<!-- Package usage documentation. -->

<section class="usage">

## Usage

```javascript
var Uint32Vector = require( '@stdlib/ndarray/vector/uint32' );
```

#### Uint32Vector( \[options] )

Returns a one-dimensional unsigned 32-bit integer [ndarray][@stdlib/ndarray/ctor].

```javascript
var numel = require( '@stdlib/ndarray/numel' );

var arr = new Uint32Vector();
// returns <ndarray>

var len = numel( arr );
// returns 0
```

The function accepts the following options:

-   **order**: specifies whether an [ndarray][@stdlib/ndarray/ctor] is `'row-major'` (C-style) or `'column-major'` (Fortran-style). Default: `'row-major'`.
-   **mode**: specifies how to handle indices which exceed array dimensions (see [`ndarray`][@stdlib/ndarray/ctor]). Default: `'throw'`.
-   **readonly**: boolean indicating whether an array should be **read-only**. Default: `false`.

#### Uint32Vector( length\[, options] )

Returns a one-dimensional unsigned 32-bit integer [ndarray][@stdlib/ndarray/ctor] having a specified `length`.

```javascript
var numel = require( '@stdlib/ndarray/numel' );

var arr = new Uint32Vector( 5 );
// returns <ndarray>

var len1 = numel( arr );
// returns 5
```

#### Uint32Vector( obj\[, options] )

Creates a one-dimensional unsigned 32-bit integer [ndarray][@stdlib/ndarray/ctor] from an array-like object or iterable.

```javascript
var numel = require( '@stdlib/ndarray/numel' );

var arr = new Uint32Vector( [ 1, 2, 3 ] );
// returns <ndarray>

var len1 = numel( arr );
// returns 3
```

#### Uint32Vector( buffer\[, byteOffset\[, length]]\[, options] )

Returns a one-dimensional unsigned 32-bit integer [ndarray][@stdlib/ndarray/ctor] view of an [`ArrayBuffer`][@stdlib/array/buffer].

```javascript
var ArrayBuffer = require( '@stdlib/array/buffer' );
var numel = require( '@stdlib/ndarray/numel' );

var buf = new ArrayBuffer( 32 );

var arr1 = new Uint32Vector( buf );
// returns <ndarray>

var len1 = numel( arr1 );
// returns 8

var arr2 = new Uint32Vector( buf, 16 );
// returns <ndarray>

var len2 = numel( arr2 );
// returns 4

var arr3 = new Uint32Vector( buf, 16, 1 );
// returns <ndarray>

var len3 = numel( arr3 );
// returns 1
```

</section>

<!-- /.usage -->

<!-- Package usage notes. Make sure to keep an empty line after the `section` element and another before the `/section` close. -->

<section class="notes">

</section>

<!-- /.notes -->

<!-- Package usage examples. -->

<section class="examples">

## Examples

<!-- eslint no-undef: "error" -->

```javascript
var discreteUniform = require( '@stdlib/random/array/discrete-uniform' );
var sum = require( '@stdlib/blas/ext/sum' );
var map = require( '@stdlib/ndarray/map' );
var Uint32Vector = require( '@stdlib/ndarray/vector/uint32' );

// Create a vector containing random values:
var x = new Uint32Vector( discreteUniform( 10, 0, 100 ) );

// Compute the sum:
var v = sum( x );
console.log( v.get() );

// Define a function which applies a threshold to individual values:
function threshold( v ) {
    return ( v > 10 ) ? v : 0;
}

// Apply threshold:
var y = map( x, threshold );

// Recompute the sum:
v = sum( y );
console.log( v.get() );
```

</section>

<!-- /.examples -->

<!-- Section to include cited references. If references are included, add a horizontal rule *before* the section. Make sure to keep an empty line after the `section` element and another before the `/section` close. -->

<section class="references">

</section>

<!-- /.references -->

<!-- Section for related `stdlib` packages. Do not manually edit this section, as it is automatically populated. -->

<section class="related">

* * *

## See Also

-   <span class="package-name">[`@stdlib/ndarray/vector/float64`][@stdlib/ndarray/vector/float64]</span><span class="delimiter">: </span><span class="description">create a double-precision floating-point vector (i.e., a one-dimensional ndarray).</span>
-   <span class="package-name">[`@stdlib/ndarray/vector/float32`][@stdlib/ndarray/vector/float32]</span><span class="delimiter">: </span><span class="description">create a single-precision floating-point vector (i.e., a one-dimensional ndarray).</span>
-   <span class="package-name">[`@stdlib/ndarray/vector/int16`][@stdlib/ndarray/vector/int16]</span><span class="delimiter">: </span><span class="description">create a signed 16-bit integer vector (i.e., a one-dimensional ndarray).</span>
-   <span class="package-name">[`@stdlib/ndarray/vector/int32`][@stdlib/ndarray/vector/int32]</span><span class="delimiter">: </span><span class="description">create a signed 32-bit integer vector (i.e., a one-dimensional ndarray).</span>
-   <span class="package-name">[`@stdlib/ndarray/vector/int8`][@stdlib/ndarray/vector/int8]</span><span class="delimiter">: </span><span class="description">create a signed 8-bit integer vector (i.e., a one-dimensional ndarray).</span>
-   <span class="package-name">[`@stdlib/ndarray/vector/uint16`][@stdlib/ndarray/vector/uint16]</span><span class="delimiter">: </span><span class="description">create an unsigned 16-bit integer vector (i.e., a one-dimensional ndarray).</span>
-   <span class="package-name">[`@stdlib/ndarray/vector/uint8`][@stdlib/ndarray/vector/uint8]</span><span class="delimiter">: </span><span class="description">create an unsigned 8-bit integer vector (i.e., a one-dimensional ndarray).</span>
-   <span class="package-name">[`@stdlib/ndarray/vector/uint8c`][@stdlib/ndarray/vector/uint8c]</span><span class="delimiter">: </span><span class="description">create a clamped unsigned 8-bit integer vector (i.e., a one-dimensional ndarray).</span>
-   <span class="package-name">[`@stdlib/ndarray/vector/complex128`][@stdlib/ndarray/vector/complex128]</span><span class="delimiter">: </span><span class="description">create a double-precision complex floating-point vector (i.e., a one-dimensional ndarray).</span>
-   <span class="package-name">[`@stdlib/ndarray/vector/complex64`][@stdlib/ndarray/vector/complex64]</span><span class="delimiter">: </span><span class="description">create a single-precision complex floating-point vector (i.e., a one-dimensional ndarray).</span>
-   <span class="package-name">[`@stdlib/ndarray/vector/bool`][@stdlib/ndarray/vector/bool]</span><span class="delimiter">: </span><span class="description">create a boolean vector (i.e., a one-dimensional ndarray).</span>
-   <span class="package-name">[`@stdlib/ndarray/vector/ctor`][@stdlib/ndarray/vector/ctor]</span><span class="delimiter">: </span><span class="description">create a vector (i.e., a one-dimensional ndarray).</span>
-   <span class="package-name">[`@stdlib/ndarray/ctor`][@stdlib/ndarray/ctor]</span><span class="delimiter">: </span><span class="description">multidimensional array constructor.</span>

</section>

<!-- /.related -->

<!-- Section for all links. Make sure to keep an empty line after the `section` element and another before the `/section` close. -->

<section class="links">

[@stdlib/array/buffer]: https://www.npmjs.com/package/@stdlib/array-buffer

[@stdlib/ndarray/ctor]: https://github.com/stdlib-js/ndarray/tree/main/ctor

<!-- <related-links> -->

[@stdlib/ndarray/vector/float64]: https://github.com/stdlib-js/ndarray/tree/main/vector/float64

[@stdlib/ndarray/vector/float32]: https://github.com/stdlib-js/ndarray/tree/main/vector/float32

[@stdlib/ndarray/vector/int16]: https://github.com/stdlib-js/ndarray/tree/main/vector/int16

[@stdlib/ndarray/vector/int32]: https://github.com/stdlib-js/ndarray/tree/main/vector/int32

[@stdlib/ndarray/vector/int8]: https://github.com/stdlib-js/ndarray/tree/main/vector/int8

[@stdlib/ndarray/vector/uint16]: https://github.com/stdlib-js/ndarray/tree/main/vector/uint16

[@stdlib/ndarray/vector/uint8]: https://github.com/stdlib-js/ndarray/tree/main/vector/uint8

[@stdlib/ndarray/vector/uint8c]: https://github.com/stdlib-js/ndarray/tree/main/vector/uint8c

[@stdlib/ndarray/vector/complex128]: https://github.com/stdlib-js/ndarray/tree/main/vector/complex128

[@stdlib/ndarray/vector/complex64]: https://github.com/stdlib-js/ndarray/tree/main/vector/complex64

[@stdlib/ndarray/vector/bool]: https://github.com/stdlib-js/ndarray/tree/main/vector/bool

[@stdlib/ndarray/vector/ctor]: https://github.com/stdlib-js/ndarray/tree/main/vector/ctor

<!-- </related-links> -->

</section>

<!-- /.links -->
