| mixture_kurtosis {ICSKAT} | R Documentation |
mixture_kurtosis.R
Description
Calculate the kurtosis of Qrho when performing SKATO with bootstrapped moments. This function is included to allow for the potential to match the SKAT package, however we generally don't call it because we can just bootstrap the kurtosis of Qrho directly if we are already doing bootstrap, thus avoiding this calculation. Also it's only used in calculating the qmin values, not in the final p-value calculation, which uses a kappa expression that is only the first two terms of Qrho.
Usage
mixture_kurtosis(tempDF1, tempDF2, v1, a1, a2)
Arguments
tempDF1 |
Generally the bootstrapped kurtosis of the mixture of chi-squares in kappa. |
tempDF2 |
Generally 1 because it's for the chi-square1 RV in kappa. |
v1 |
Generally the variance of the mixture of chi-squares plus the variance of zeta in kappa. |
a1 |
Generally the 1-rho in front of the first part of the kappa term. |
a2 |
Generally the tau(rho) term in front of the chi-square1 RV in kappa. |
Value
Kurtosis (excess kurtosis to be more precise), use df = 12 / kurtosis.