|
| 1 | +<!-- |
| 2 | +
|
| 3 | +@license Apache-2.0 |
| 4 | +
|
| 5 | +Copyright (c) 2018 The Stdlib Authors. |
| 6 | +
|
| 7 | +Licensed under the Apache License, Version 2.0 (the "License"); |
| 8 | +you may not use this file except in compliance with the License. |
| 9 | +You may obtain a copy of the License at |
| 10 | +
|
| 11 | + http://www.apache.org/licenses/LICENSE-2.0 |
| 12 | +
|
| 13 | +Unless required by applicable law or agreed to in writing, software |
| 14 | +distributed under the License is distributed on an "AS IS" BASIS, |
| 15 | +WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 16 | +See the License for the specific language governing permissions and |
| 17 | +limitations under the License. |
| 18 | +
|
| 19 | +--> |
| 20 | + |
| 21 | +# incrmvmr |
| 22 | + |
| 23 | +> Compute a moving [variance-to-mean ratio][variance-to-mean-ratio] (VMR) incrementally. |
| 24 | +
|
| 25 | +<section class="intro"> |
| 26 | + |
| 27 | +For a window of size `W`, the [unbiased sample variance][sample-variance] is defined as |
| 28 | + |
| 29 | +<!-- <equation class="equation" label="eq:unbiased_sample_variance" align="center" raw="s^2 = \frac{1}{W-1} \sum_{i=0}^{W-1} ( x_i - \bar{x} )^2" alt="Equation for the unbiased sample variance."> --> |
| 30 | + |
| 31 | +<div class="equation" align="center" data-raw-text="s^2 = \frac{1}{W-1} \sum_{i=0}^{W-1} ( x_i - \bar{x} )^2" data-equation="eq:unbiased_sample_variance"> |
| 32 | + <img src="" alt="Equation for the unbiased sample variance."> |
| 33 | + <br> |
| 34 | +</div> |
| 35 | + |
| 36 | +<!-- </equation> --> |
| 37 | + |
| 38 | +and the [arithmetic mean][arithmetic-mean] is defined as |
| 39 | + |
| 40 | +<!-- <equation class="equation" label="eq:arithmetic_mean" align="center" raw="\bar{x} = \frac{1}{n} \sum_{i=0}^{W-1} x_i" alt="Equation for the arithmetic mean."> --> |
| 41 | + |
| 42 | +<div class="equation" align="center" data-raw-text="\bar{x} = \frac{1}{n} \sum_{i=0}^{W-1} x_i" data-equation="eq:arithmetic_mean"> |
| 43 | + <img src="" alt="Equation for the arithmetic mean."> |
| 44 | + <br> |
| 45 | +</div> |
| 46 | + |
| 47 | +<!-- </equation> --> |
| 48 | + |
| 49 | +The [variance-to-mean ratio][variance-to-mean-ratio] (VMR) is thus defined as |
| 50 | + |
| 51 | +<!-- <equation class="equation" label="eq:variance_to_mean_ratio" align="center" raw="F = \frac{s^2}{\bar{x}}" alt="Equation for the variance-to-mean ratio (VMR)."> --> |
| 52 | + |
| 53 | +<div class="equation" align="center" data-raw-text="F = \frac{s^2}{\bar{x}}" data-equation="eq:variance_to_mean_ratio"> |
| 54 | + <img src="" alt="Equation for the variance-to-mean ratio (VMR)."> |
| 55 | + <br> |
| 56 | +</div> |
| 57 | + |
| 58 | +<!-- </equation> --> |
| 59 | + |
| 60 | +</section> |
| 61 | + |
| 62 | +<!-- /.intro --> |
| 63 | + |
| 64 | +<section class="usage"> |
| 65 | + |
| 66 | +## Usage |
| 67 | + |
| 68 | +```javascript |
| 69 | +var incrmvmr = require( '@stdlib/stats/incr/mvmr' ); |
| 70 | +``` |
| 71 | + |
| 72 | +#### incrmvmr( window\[, mean] ) |
| 73 | + |
| 74 | +Returns an accumulator `function` which incrementally computes a moving [variance-to-mean ratio][variance-to-mean-ratio]. The `window` parameter defines the number of values over which to compute the moving [variance-to-mean ratio][variance-to-mean-ratio]. |
| 75 | + |
| 76 | +```javascript |
| 77 | +var accumulator = incrmvmr( 3 ); |
| 78 | +``` |
| 79 | + |
| 80 | +If the mean is already known, provide a `mean` argument. |
| 81 | + |
| 82 | +```javascript |
| 83 | +var accumulator = incrmvmr( 3, 5.0 ); |
| 84 | +``` |
| 85 | + |
| 86 | +#### accumulator( \[x] ) |
| 87 | + |
| 88 | +If provided an input value `x`, the accumulator function returns an updated accumulated value. If not provided an input value `x`, the accumulator function returns the current accumulated value. |
| 89 | + |
| 90 | +```javascript |
| 91 | +var accumulator = incrmvmr( 3 ); |
| 92 | + |
| 93 | +var F = accumulator(); |
| 94 | +// returns null |
| 95 | + |
| 96 | +// Fill the window... |
| 97 | +F = accumulator( 2.0 ); // [2.0] |
| 98 | +// returns 0.0 |
| 99 | + |
| 100 | +F = accumulator( 1.0 ); // [2.0, 1.0] |
| 101 | +// returns ~0.33 |
| 102 | + |
| 103 | +F = accumulator( 3.0 ); // [2.0, 1.0, 3.0] |
| 104 | +// returns 0.5 |
| 105 | + |
| 106 | +// Window begins sliding... |
| 107 | +F = accumulator( -7.0 ); // [1.0, 3.0, -7.0] |
| 108 | +// returns -28.0 |
| 109 | + |
| 110 | +F = accumulator( -5.0 ); // [3.0, -7.0, -5.0] |
| 111 | +// returns ~-9.33 |
| 112 | + |
| 113 | +F = accumulator(); |
| 114 | +// returns ~-9.33 |
| 115 | +``` |
| 116 | + |
| 117 | +</section> |
| 118 | + |
| 119 | +<!-- /.usage --> |
| 120 | + |
| 121 | +<section class="notes"> |
| 122 | + |
| 123 | +## Notes |
| 124 | + |
| 125 | +- Input values are **not** type checked. If provided `NaN` or a value which, when used in computations, results in `NaN`, the accumulated value is `NaN` for **at least** `W-1` future invocations. If non-numeric inputs are possible, you are advised to type check and handle accordingly **before** passing the value to the accumulator function. |
| 126 | + |
| 127 | +- 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 provided values. |
| 128 | + |
| 129 | +- The following table summarizes how to interpret the [variance-to-mean ratio][variance-to-mean-ratio]: |
| 130 | + |
| 131 | + | VMR | Description | Example Distribution | |
| 132 | + | :---------------: | :-------------: | :--------------------------: | |
| 133 | + | 0 | not dispersed | constant | |
| 134 | + | 0 < VMR < 1 | under-dispersed | binomial | |
| 135 | + | 1 | -- | Poisson | |
| 136 | + | >1 | over-dispersed | geometric, negative-binomial | |
| 137 | + |
| 138 | + Accordingly, one can use the [variance-to-mean ratio][variance-to-mean-ratio] to assess whether observed data can be modeled as a Poisson process. When observed data is "under-dispersed", observed data may be more regular than as would be the case for a Poisson process. When observed data is "over-dispersed", observed data may contain clusters (i.e., clumped, concentrated data). |
| 139 | + |
| 140 | +- The [variance-to-mean ratio][variance-to-mean-ratio] is also known as the **index of dispersion**, **dispersion index**, **coefficient of dispersion**, **relative variance**, and the [**Fano factor**][fano-factor]. |
| 141 | + |
| 142 | +</section> |
| 143 | + |
| 144 | +<!-- /.notes --> |
| 145 | + |
| 146 | +<section class="examples"> |
| 147 | + |
| 148 | +## Examples |
| 149 | + |
| 150 | +<!-- eslint no-undef: "error" --> |
| 151 | + |
| 152 | +```javascript |
| 153 | +var randu = require( '@stdlib/random/base/randu' ); |
| 154 | +var incrmvmr = require( '@stdlib/stats/incr/mvmr' ); |
| 155 | + |
| 156 | +var accumulator; |
| 157 | +var v; |
| 158 | +var i; |
| 159 | + |
| 160 | +// Initialize an accumulator: |
| 161 | +accumulator = incrmvmr( 5 ); |
| 162 | + |
| 163 | +// For each simulated datum, update the moving variance-to-mean ratio... |
| 164 | +for ( i = 0; i < 100; i++ ) { |
| 165 | + v = randu() * 100.0; |
| 166 | + accumulator( v ); |
| 167 | +} |
| 168 | +console.log( accumulator() ); |
| 169 | +``` |
| 170 | + |
| 171 | +</section> |
| 172 | + |
| 173 | +<!-- /.examples --> |
| 174 | + |
| 175 | +<section class="links"> |
| 176 | + |
| 177 | +[variance-to-mean-ratio]: https://en.wikipedia.org/wiki/Index_of_dispersion |
| 178 | + |
| 179 | +[arithmetic-mean]: https://en.wikipedia.org/wiki/Arithmetic_mean |
| 180 | + |
| 181 | +[sample-variance]: https://en.wikipedia.org/wiki/Variance |
| 182 | + |
| 183 | +[fano-factor]: https://en.wikipedia.org/wiki/Fano_factor |
| 184 | + |
| 185 | +</section> |
| 186 | + |
| 187 | +<!-- /.links --> |
0 commit comments