d_effchoiceupdate: Fast multinomial logistic d-efficiency update function.

Description Usage Arguments Value Examples

Description

Fast d-efficiency update function for a single point change to a multinomial logistic model. Supply a subset of the existing design corresponding to a single question from the model matrix and the information matrix for the remaining design for all other questions.

Usage

1
d_effchoiceupdate(CurrentMatrix, paramestimates, info_mat)

Arguments

CurrentMatrix

Subset of model matrix with attribute factors coded using contr.sum coding. Matrix must be standardized before passing to function if standardization is desired. Opt out choices, if included, should be coded as a row of all zeroes for every parameter. Subset should be 1 row per alternative and subset should only include all alternatives for one question only.

paramestimates

Vector of estimates for each effect (column) of model matrix sized corresponding to standardized model matrix. If this is not supplied, all effect estimates will be set to 0.

info_mat

The information matrix for all other questions in the model matrix that were not passed as the CurrentMatrix argument.

Value

D-efficiency of the supplied model.

Examples

1
2
3
##---- Should be DIRECTLY executable !! ----
##-- ==>  Define data, use random,
##--	or do  help(data=index)  for the standard data sets.

taalbrecht/MultiEqOptimizer documentation built on May 31, 2019, 12:51 a.m.