# incrkurtosis > Compute a [corrected sample excess kurtosis][sample-excess-kurtosis] incrementally.
The [kurtosis][sample-excess-kurtosis] for a random variable `X` is defined as
Equation for the kurtosis.
Using a univariate normal distribution as the standard of comparison, the [excess kurtosis][sample-excess-kurtosis] is the kurtosis minus `3`. For a sample of `n` values, the [sample excess kurtosis][sample-excess-kurtosis] is
Equation for the sample excess kurtosis.
where `m_4` is the sample fourth central moment and `m_2` is the sample second central moment. The previous equation is, however, a biased estimator of the population excess kurtosis. An alternative estimator which is unbiased under normality is
Equation for the corrected sample excess kurtosis.
## Usage ```javascript var incrkurtosis = require( '@stdlib/stats/incr/kurtosis' ); ``` #### incrkurtosis() Returns an accumulator `function` which incrementally computes a [corrected sample excess kurtosis][sample-excess-kurtosis]. ```javascript var accumulator = incrkurtosis(); ``` #### accumulator( \[x] ) If provided an input value `x`, the accumulator function returns an updated [corrected sample excess kurtosis][sample-excess-kurtosis]. If not provided an input value `x`, the accumulator function returns the current [corrected sample excess kurtosis][sample-excess-kurtosis]. ```javascript var accumulator = incrkurtosis(); var kurtosis = accumulator( 2.0 ); // returns null kurtosis = accumulator( 2.0 ); // returns null kurtosis = accumulator( -4.0 ); // returns null kurtosis = accumulator( -4.0 ); // returns -6.0 ```
## 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 `NaN` for **all** 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.
## Examples ```javascript var randu = require( '@stdlib/random/base/randu' ); var incrkurtosis = require( '@stdlib/stats/incr/kurtosis' ); var accumulator; var v; var i; // Initialize an accumulator: accumulator = incrkurtosis(); // For each simulated datum, update the corrected sample excess kurtosis... for ( i = 0; i < 100; i++ ) { v = randu() * 100.0; accumulator( v ); } console.log( accumulator() ); ```
* * *
## References - Joanes, D. N., and C. A. Gill. 1998. "Comparing measures of sample skewness and kurtosis." _Journal of the Royal Statistical Society: Series D (The Statistician)_ 47 (1). Blackwell Publishers Ltd: 183–89. doi:[10.1111/1467-9884.00122][@joanes:1998].