## Usage ```javascript var itermsumabs = require( '@stdlib/stats/iter/msumabs' ); ``` #### itermsumabs( iterator, W ) Returns an [iterator][mdn-iterator-protocol] which iteratively computes a moving sum of absolute values. The `W` parameter defines the number of iterated values over which to compute the moving sum. ```javascript var array2iterator = require( '@stdlib/array/to-iterator' ); var arr = array2iterator( [ 2.0, 1.0, 3.0, -7.0, -5.0 ] ); var it = itermsumabs( arr, 3 ); // Fill the window... var s = it.next().value; // [2.0] // returns 2.0 s = it.next().value; // [2.0, 1.0] // returns 3.0 s = it.next().value; // [2.0, 1.0, 3.0] // returns 6.0 // Window begins sliding... s = it.next().value; // [1.0, 3.0, -7.0] // returns 11.0 s = it.next().value; // [3.0, -7.0, -5.0] // returns 15.0 ```

## Notes - If an iterated value is non-numeric (including `NaN`), the function returns `NaN` for **at least** `W-1` future invocations. If non-numeric iterated values are possible, you are advised to provide an [`iterator`][mdn-iterator-protocol] which type checks and handles non-numeric values accordingly. - As `W` values are needed to fill the window buffer, the first `W-1` returned values are calculated from smaller sample sizes. Until the window is full, each returned value is calculated from all previously iterated values.

## Examples ```javascript var runif = require( '@stdlib/random/iter/uniform' ); var itermsumabs = require( '@stdlib/stats/iter/msumabs' ); // Create an iterator for generating uniformly distributed pseudorandom numbers: var rand = runif( -10.0, 10.0, { 'seed': 1234, 'iter': 100 }); // Create an iterator for iteratively computing a moving sum of absolute values: var it = itermsumabs( rand, 3 ); // Perform manual iteration... var v; while ( true ) { v = it.next(); if ( typeof v.value === 'number' ) { console.log( 'sumabs: %d', v.value ); } if ( v.done ) { break; } } ```