I'm doing an experimental study of 5 factors with a different amount of levels in each. More specifically: 9, 13, 12, 6 & 15.
Using the AlgDesign library in R, i ran the following syntax to generate a D-efficient design:
levels.design <- c(9,13,12,6,15) #5 factors for the data #Full factorial: f.design <- gen.factorial(levels.design) #Fractional factorial design: start <- Sys.time() fract.design <- optFederov(data=f.design, nTrials=sum(levels.design), approximate=TRUE, criterion="D", nRepeats=100, eval=TRUE)
This gave me the following measures:
My question is: How do I implement the table/design above? And how do I interpret the result on the D-measure?
Every level is a specific setting/alternative in a an algorithm. But how do I interpret which of the levels to be run, based on the table?