wu.params: Non-central chi-square approximation to mixture of 1-DF...

In baolinwu/KATSP: Sample size and power calculation for variant-set association test with SKAT

Description

Given the mixture of 1-DF chi-square distribution, ∑_iλ_iχ_1^2(δ_i), the non-central chi-square distribution approximation is based on matching higher moments. See the ref.

Usage

wu.params(lambda, delta, N = 12)

Arguments

lambda	mixing coefficients for chi-square mixtures
delta	vector of non-centrality parameters for chi-square mixtures
N	match (N-1,N)-th moments

Value

l

estimated DF

d

estimated non-centrality parameter

muQ,sigmaQ

first two moments based on the chi-square mixture distribution

muX,sigmaX

first two moments based on the estimated non-central chi-square distribution

References

Wu,B. and Pankow,J.S. (2016) On sample size and power calculation for variant set-based association tests. /AHG/, 80(2), 136-143.

Wu,B. and Pankow,J.S. (2017) On computing the tail probability of non-negative definite quadratic forms in central normal variables. tech rep.

baolinwu/KATSP documentation built on May 29, 2019, 12:04 p.m.