# map
> Apply a callback function to elements in an input [ndarray][@stdlib/ndarray/ctor] and assign results to elements in a new output [ndarray][@stdlib/ndarray/ctor].
## Usage
```javascript
var map = require( '@stdlib/ndarray/map' );
```
#### map( x\[, options], fcn\[, thisArg] )
Applies a callback function to elements in an input [ndarray][@stdlib/ndarray/ctor] and assigns results to elements in a new output [ndarray][@stdlib/ndarray/ctor].
```javascript
var Float64Array = require( '@stdlib/array/float64' );
var ndarray = require( '@stdlib/ndarray/ctor' );
function scale( z ) {
return z * 10.0;
}
var buffer = new Float64Array( [ 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0, 12.0 ] );
var shape = [ 2, 3 ];
var strides = [ 6, 1 ];
var offset = 1;
var x = ndarray( 'float64', buffer, shape, strides, offset, 'row-major' );
// returns
var y = map( x, scale );
// returns [ [ 20.0, 30.0, 40.0 ], [ 80.0, 90.0, 100.0 ] ]
```
The function accepts the following arguments:
- **x**: input [ndarray][@stdlib/ndarray/ctor].
- **options**: function options _(optional)_.
- **fcn**: callback to apply.
- **thisArg**: callback execution context _(optional)_.
The function accepts the following options:
- **dtype**: output ndarray [data type][@stdlib/ndarray/dtypes]. If not specified, the output ndarray [data type][@stdlib/ndarray/dtypes] is inferred from the input [ndarray][@stdlib/ndarray/ctor].
By default, the output ndarray [data type][@stdlib/ndarray/dtypes] is inferred from the input [ndarray][@stdlib/ndarray/ctor]. To return an ndarray with a different [data type][@stdlib/ndarray/dtypes], specify the `dtype` option.
```javascript
var Float64Array = require( '@stdlib/array/float64' );
var ndarray = require( '@stdlib/ndarray/ctor' );
var dtype = require( '@stdlib/ndarray/dtype' );
function scale( z ) {
return z * 10.0;
}
var buffer = new Float64Array( [ 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0, 12.0 ] );
var shape = [ 2, 3 ];
var strides = [ 6, 1 ];
var offset = 1;
var x = ndarray( 'float64', buffer, shape, strides, offset, 'row-major' );
// returns
var opts = {
'dtype': 'float32'
};
var y = map( x, opts, scale );
// returns [ [ 20.0, 30.0, 40.0 ], [ 80.0, 90.0, 100.0 ] ]
var dt = String( dtype( y ) );
// returns 'float32'
```
To set the callback function execution context, provide a `thisArg`.
```javascript
var Float64Array = require( '@stdlib/array/float64' );
var ndarray = require( '@stdlib/ndarray/ctor' );
function scale( z ) {
this.count += 1;
return z * 10.0;
}
var buffer = new Float64Array( [ 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0, 12.0 ] );
var shape = [ 2, 3 ];
var strides = [ 6, 1 ];
var offset = 1;
var x = ndarray( 'float64', buffer, shape, strides, offset, 'row-major' );
// returns
var ctx = {
'count': 0
};
var y = map( x, scale, ctx );
// returns [ [ 20.0, 30.0, 40.0 ], [ 80.0, 90.0, 100.0 ] ]
var count = ctx.count;
// returns 6
```
The callback function is provided the following arguments:
- **value**: current array element.
- **indices**: current array element indices.
- **arr**: the input [ndarray][@stdlib/ndarray/ctor].
## Notes
- The function does **not** perform explicit casting (e.g., from a real-valued floating-point number to a complex floating-point number). Any such casting should be performed by a provided callback function.
```javascript
var Float64Array = require( '@stdlib/array/float64' );
var ndarray = require( '@stdlib/ndarray/ctor' );
var Complex128 = require( '@stdlib/complex/float64/ctor' );
function toComplex( z ) {
return new Complex128( z, 0.0 );
}
var buffer = new Float64Array( [ 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0, 11.0, 12.0 ] );
var shape = [ 2, 3 ];
var strides = [ 6, 1 ];
var offset = 1;
var x = ndarray( 'float64', buffer, shape, strides, offset, 'row-major' );
// returns
var opts = {
'dtype': 'complex128'
};
var y = map( x, opts, toComplex );
// returns
```
- The function **always** returns an [ndarray][@stdlib/ndarray/ctor] having the same shape and [order][@stdlib/ndarray/orders] as the input [ndarray][@stdlib/ndarray/ctor].
- For very high-dimensional ndarrays which are non-contiguous, one should consider copying the underlying data to contiguous memory before applying a callback function in order to achieve better performance.
## Examples
```javascript
var discreteUniform = require( '@stdlib/random/array/discrete-uniform' );
var abs = require( '@stdlib/math/base/special/abs' );
var ndarray2array = require( '@stdlib/ndarray/to-array' );
var naryFunction = require( '@stdlib/utils/nary-function' );
var ndarray = require( '@stdlib/ndarray/ctor' );
var map = require( '@stdlib/ndarray/map' );
var buffer = discreteUniform( 10, -100, 100, {
'dtype': 'generic'
});
var shape = [ 5, 2 ];
var strides = [ 2, 1 ];
var offset = 0;
var x = ndarray( 'generic', buffer, shape, strides, offset, 'row-major' );
console.log( ndarray2array( x ) );
var y = map( x, naryFunction( abs, 1 ) );
console.log( ndarray2array( y ) );
```
[@stdlib/ndarray/ctor]: https://github.com/stdlib-js/ndarray/tree/main/ctor
[@stdlib/ndarray/dtypes]: https://github.com/stdlib-js/ndarray/tree/main/dtypes
[@stdlib/ndarray/orders]: https://github.com/stdlib-js/ndarray/tree/main/orders
[@stdlib/ndarray/filter]: https://github.com/stdlib-js/ndarray/tree/main/filter
[@stdlib/ndarray/filter-map]: https://github.com/stdlib-js/ndarray/tree/main/filter-map
[@stdlib/ndarray/for-each]: https://github.com/stdlib-js/ndarray/tree/main/for-each
[@stdlib/ndarray/slice]: https://github.com/stdlib-js/ndarray/tree/main/slice