Skip to content

feat(stats): add package stats/incr/nanmmda #6455

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 5 commits into
base: develop
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
173 changes: 173 additions & 0 deletions lib/node_modules/@stdlib/stats/incr/nanmmda/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,173 @@
<!--

@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.

-->

# incrnanmmda

> Compute a moving [mean directional accuracy][mean-directional-accuracy] (MDA) incrementally, ignoring `NaN` values.

<section class="intro">

For a window of size `W`, the [mean directional accuracy][mean-directional-accuracy] is defined as

<!-- <equation class="equation" label="eq:mean_directional_accuracy" align="center" raw="\operatorname{MDA} = \begin{cases} 1 & \textrm{if}\ W = 1 \\ \frac{1}{W} \sum_{i=1}^{W} \delta_{\operatorname{sgn}(\Delta f_{i,i-1}),\ \operatorname{sgn}(\Delta a_{i,i-1})} & \textrm{if}\ W > 1 \end{cases}" alt="Equation for the mean directional accuracy."> -->

```math
\mathop{\mathrm{MDA}} = \begin{cases} 1 & \textrm{if}\ W = 1 \\ \frac{1}{W} \sum_{i=1}^{W} \delta_{\mathop{\mathrm{sgn}}(\Delta f_{i,i-1}),\ \mathop{\mathrm{sgn}}(\Delta a_{i,i-1})} & \textrm{if}\ W > 1 \end{cases}
```

<!-- <div class="equation" align="center" data-raw-text="\operatorname{MDA} = \begin{cases} 1 & \textrm{if}\ W = 1 \\\frac{1}{W} \sum_{i=1}^{W} \delta_{\operatorname{sgn}(\Delta f_{i,i-1}),\ \operatorname{sgn}(\Delta a_{i,i-1})} & \textrm{if}\ W > 1 \end{cases}" data-equation="eq:mean_directional_accuracy">
<img src="https://cdn.jsdelivr.net/gh/stdlib-js/stdlib@99730afbace8256ce53cfbc0714c7f3cac92466a/lib/node_modules/@stdlib/stats/incr/mmda/docs/img/equation_mean_directional_accuracy.svg" alt="Equation for the mean directional accuracy.">
<br>
</div> -->

<!-- </equation> -->

where `f_i` is the forecast value, `a_i` is the actual value, `sgn(x)` is the [signum][@stdlib/math/base/special/signum] function, and `δ` is the [Kronecker delta][@stdlib/math/base/special/kronecker-delta].

</section>

<!-- /.intro -->

<section class="usage">

## Usage

```javascript
var incrnanmmda = require( '@stdlib/stats/incr/nanmmda' );
```

#### incrnanmmda( window )

Returns an accumulator `function` which incrementally computes a moving [mean directional accuracy][mean-directional-accuracy], ignoring `NaN` values. The `window` parameter defines the number of values over which to compute the moving [mean directional accuracy][mean-directional-accuracy].

```javascript
var accumulator = incrnanmmda( 3 );
```

#### accumulator( \[f, a] )

If provided input values `f` and `a`, the accumulator function returns an updated [mean directional accuracy][mean-directional-accuracy]. If not provided input values `f` and `a`, the accumulator function returns the current [mean directional accuracy][mean-directional-accuracy].

```javascript
var accumulator = incrnanmmda( 3 );

var m = accumulator();
// returns null

m = accumulator( NaN, 4.0 );
// returns null

// Fill the window...
m = accumulator( 2.0, 3.0 ); // [(+,+)]
// returns 1.0

m = accumulator( 1.0, 4.0 ); // [(+,+), (-,+)]
// returns 0.5

m = accumulator( 3.0, NaN ); // [(+,+), (-,+)]
// returns 0.5

m = accumulator( 3.0, 9.0 ); // [(+,+), (-,+), (+,+)]
// returns ~0.67

// Window begins sliding...
m = accumulator( 7.0, 3.0 ); // [(-,+), (+,+), (+,-)]
// returns ~0.33

m = accumulator( NaN, NaN ); // [(-,+), (+,+), (+,-)]
// returns -0.33

m = accumulator( 5.0, 3.0 ); // [(+,+), (+,-), (-,0)]
// returns ~0.33

m = accumulator();
// returns ~0.33
```

</section>

<!-- /.usage -->

<section class="notes">

## Notes

- Input values are **not** type checked. If provided `NaN` or a value which, when used in computations, results in `NaN`, the accumulated value is **not** changed and the current input values get **ignored**. If non-numeric inputs are possible, you are advised to type check and handle accordingly **before** passing the value to the accumulator function.
- As `W` (f,a) pairs 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 provided values.

</section>

<!-- /.notes -->

<section class="examples">

## Examples

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

```javascript
var bernoulli = require( '@stdlib/random/base/bernoulli' );
var randu = require( '@stdlib/random/base/randu' );
var incrnanmmda = require( '@stdlib/stats/incr/nanmmda' );

var accumulator;
var v1;
var v2;
var i;

// Initialize an accumulator:
accumulator = incrnanmmda( 5 );

// For each simulated datum, update the moving mean directional accuracy...
for ( i = 0; i < 100; i++ ) {
v1 = ( bernoulli( 0.1 ) ) ? NaN : ( randu()*100.0 ) - 50.0;
v2 = ( bernoulli( 0.1 ) ) ? NaN : ( randu()*100.0 ) - 50.0;
accumulator( v1, v2 );
}
console.log( accumulator() );
```

</section>

<!-- /.examples -->

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

<section class="related">


</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">

[mean-directional-accuracy]: https://en.wikipedia.org/wiki/Mean_Directional_Accuracy_%28MDA%29

[@stdlib/math/base/special/signum]: https://github.com/stdlib-js/stdlib/tree/develop/lib/node_modules/%40stdlib/math/base/special/signum

[@stdlib/math/base/special/kronecker-delta]: https://github.com/stdlib-js/stdlib/tree/develop/lib/node_modules/%40stdlib/math/base/special/kronecker-delta


</section>

<!-- /.links -->
69 changes: 69 additions & 0 deletions lib/node_modules/@stdlib/stats/incr/nanmmda/benchmark/benchmark.js
Original file line number Diff line number Diff line change
@@ -0,0 +1,69 @@
/**
* @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.
*/

'use strict';

// MODULES //

var bench = require( '@stdlib/bench' );
var randu = require( '@stdlib/random/base/randu' );
var pkg = require( './../package.json' ).name;
var incrnanmmda = require( './../lib' );


// MAIN //

bench( pkg, function benchmark( b ) {
var f;
var i;
b.tic();
for ( i = 0; i < b.iterations; i++ ) {
f = incrnanmmda( ( i%5 ) + 1 );
if ( typeof f !== 'function' ) {
b.fail( 'should return a function' );
}
}
b.toc();
if ( typeof f !== 'function' ) {
b.fail( 'should return a function' );
}
b.pass( 'benchmark finished' );
b.end();
});

bench( pkg+'::accumulator', function benchmark( b ) {
var acc;
var v;
var i;

acc = incrnanmmda( 5 );

b.tic();
for ( i = 0; i < b.iterations; i++ ) {
v = acc( randu()-0.5, randu()-0.5 );
if ( v !== v ) {
b.fail( 'should not return NaN' );
}
}
b.toc();
if ( v !== v ) {
b.fail( 'should not return NaN' );
}
b.pass( 'benchmark finished' );
b.end();
});
Loading