Calculate D-efficiency of an experimental desgin in R -
i have experimental design. want calculate d-efficiency. thought r package algdesign help. found function optfederov generates design , - if user wants - returns d efficiency. however, don't want use optfederov generate design - have design! tried eval.design(~.,mydesign). metrics gives me are: determinant, a, diagonality, , gmean.variances. maybe there way determinant or d-efficiency (i not mathematician, not sure). or maybe other way calculate d-efficiency "manually" say?
thanks lot hint!
i working on similar project. found out formula deff = (|x'x|^(1/p))/nd in link. x model matrix, p number of betas in linear model , nd number of runs experiment has. make code , trick.
det(t(x)%*%x)^(1/beta)/(numruns)
i tested results using jmp project believe correct formula
Comments
Post a Comment