tableEg: The result for the RSR method with respect to
the extended Fisher's G test
Description
This data set is used to build up the response surface regressions
in order to obtain the p-values for the extended Fisher's G test.
The simulation set-up of the response surface regression method is
i.i.d. statistics of length 10^6 with 500 replications and the
interpolated series sizes are c(8:19,seq(20,50,2),seq(55,100,5)).
Format
A list object with the following components: model_matrix
: A matrix with the estimated coefficents
for the Response Surface Regressions (one row per quantile). nc
: a vectpr of the lengths of the simulated series. pc
: A vector containing all probabilities at which
the quantiles are estimated. r
: the rate in the basis function g(n)=(1/n)^r. q
: the number of parameters (except the intercept) used in
the response surface regression, the intercept term is removed if negative.References
MacKinnon, James (2001) :
Computing numerical distribution functions in
econometrics, Queen's Economics Department Working Paper, No. 1037.