# incrnansumabs
> Compute a sum of absolute values incrementally, ignoring `NaN` values.
The sum of absolute values is defined as
## Usage
```javascript
var incrnansumabs = require( '@stdlib/stats/incr/nansumabs' );
```
#### incrnansumabs()
Returns an accumulator `function` which incrementally computes a sum of absolute values, ignoring `NaN` values.
```javascript
var accumulator = incrnansumabs();
```
#### accumulator( \[x] )
If provided an input value `x`, the accumulator function returns an updated sum. If not provided an input value `x`, the accumulator function returns the current sum.
```javascript
var accumulator = incrnansumabs();
var sum = accumulator( 2.0 );
// returns 2.0
sum = accumulator( -1.0 );
// returns 3.0
sum = accumulator( NaN );
// returns 3.0
sum = accumulator( -3.0 );
// returns 6.0
sum = accumulator();
// returns 6.0
```
## Notes
- Input values are **not** type checked. If non-numeric inputs are possible, you are advised to type check and handle accordingly **before** passing the value to the accumulator function.
- For long running accumulations or accumulations of large numbers, care should be taken to prevent overflow.
## Examples
```javascript
var randu = require( '@stdlib/random/base/randu' );
var incrnansumabs = require( '@stdlib/stats/incr/nansumabs' );
var accumulator;
var v;
var i;
// Initialize an accumulator:
accumulator = incrnansumabs();
// For each simulated datum, update the sum...
for ( i = 0; i < 100; i++ ) {
if ( randu() < 0.2 ) {
v = NaN;
} else {
v = ( randu()*100.0 ) - 50.0;
}
accumulator( v );
}
console.log( accumulator() );
```
