R: Feasible GLS...

FGLS {MixABEL}R Documentation

Feasible GLS...

Description

Feasible GLS

Usage

FGLS(Y, X, W, test="wald", whichtest=c(FALSE, rep(TRUE, dim(X)[2] - 1)))

Arguments

Y dependent variable
X design matrix (including intercept, if necessary)
test test to be applied, one of 'wald', 'score' or 'robust'
whichtest which independent variables to be tested (set to 'TRUE')
W for GLS, inverse variance-covariance matrix, as such returned by GenABEL's polygenic(...)$InvSigma, or NULL for LS

Details

Feasible Generalised Least Squares

Value

List with elements 'beta' – estimates fo the regression coefficients; 'V' – variance covariance matrix for parameters estimates; 'T2' – test statistics (distributed as Chi-squared under the null) for the testing of whichtest parameters; 'df' – the number of degrees of freedom of the T2 test; 'tested' – which parameters were tested with T2; 'meanColX' – mean value of variable in columns of X; 'n' – length of Y (== height of X)

Author(s)

Yurii Aulchenko


[Package MixABEL version 0.0-9 Index]