Description Usage Arguments Details Value Author(s) References Examples
Mardia's estimate of multivariate kurtosis is given by
b_{2, k} = \frac{1}{n} ∑_{i = 1}^{n} ≤ft[ ≤ft( \mathbf{X}_{i} - \mathbf{\bar{X}} \right)^{T} \boldsymbol{\hat{Σ}}^{-1} ≤ft( \mathbf{X}_{i} - \mathbf{\bar{X}} \right) \right]^{2}
where
\mathbf{X} is the n \times k sample data
\mathbf{\bar{X}} represent sample means
\mathbf{X}_{i} - \mathbf{\bar{X}} represent deviations from the mean
\boldsymbol{\hat{Σ}} is the estimated variance-covariance matrix of \mathbf{X} using sample data
1 | mardiakurt(X)
|
X |
Matrix or data frame. |
If the null hypothesis that \mathbf{X} comes from a multivariate normal distribution is true, b_{2, k} follows a normal distribution with a mean of k ≤ft( k + 2 \right) and a variance of \frac{8 k ≤ft( k + 2 \right)}{n}. Consequently,
\frac{b_{2, k} - k ≤ft( k + 2 \right)}{√{\frac{8 k ≤ft( k + 2 \right)}{n}}}
asymptotically follows a standard normal distribution \mathcal{N} ≤ft( 0, 1 \right) .
Returns a vector with the following elements
Estimate of multivariate kurtosis ≤ft( b_{2, k} \right) .
z-statistic ≤ft( \frac{b_{2, k} - k ≤ft( k + 2 \right)}{√{\frac{8 k ≤ft( k + 2 \right)}{n}}} \right) .
p-value associated with the z-statistic.
Ivan Jacob Agaloos Pesigan
Mardia, K. V. (1970). Measures of multivariate skewness and kurtosis with applications. Biometrika, 57(3), 519-530. doi:10.2307/2334770.
Mardia, K. V. (1974). Applications of Some Measures of Multivariate Skewness and Kurtosis in Testing Normality and Robustness Studies. Sankhyā: The Indian Journal of Statistics, Series B (1960-2002), 36(2), 115-128.
1 2 3 4 5 6 7 8 9 |
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.