getEffFactor {infoDecompuTE}

Construct the Matrix from Information Decomposition and Compute the Efficiency Factors of Treatment effects

Description

Perform the information decomposition for either the block or treatment effects within a single stratum and Compute the Efficiency Factors for every treatment effect within a single stratum.

Usage

getEffFactor(z, T, N, Rep, trt.Sca)

Arguments

z

a matrix containing the orthogonal projector of a stratum generated by makeOrthProjectors.

T

a list of contrast matrices generated by makeContrMat.

N

a matrix containing the design matrix generated by makeOverDesMat.

Rep

a matrix containing the treatment replication number and is generated by getTrtRep.

trt.Sca

a numeric vector for computing a coefficients of the fixed effect parameter in EMS and is generated by getTrtRep.

Details

The main purpose of this function is to construct a list of resultant matrices associated with each source of variation after the information decomposition and to compute the canonical or average efficiency factors for each treatment effects in each stratum of ANOVA table.

The canonical efficiency factors are generated when the user input the treatment contrasts, otherwise the average efficiency factors, which is the harmonic mean of the canonical efficiency factors, are generated.

Value

A list of matrices and numeric vectors containing the efficiency factors of every treatment effect.

Author(s)

Kevin Chang

Examples

design1 <- local({ 
  Ani = as.factor(LETTERS[c(1,2,3,4,
                            5,6,7,8)])
  Trt = as.factor(letters[c(1,1,1,1,
                            2,2,2,2)])
  data.frame(Ani, Trt, stringsAsFactors = TRUE )
})

blk.str = "Ani"
    
rT = terms(as.formula(paste("~", blk.str, sep = "")), keep.order = TRUE) 
blkTerm = attr(rT,"term.labels")
     
Z = makeBlkDesMat(design1, blkTerm)

trt.str = "Trt"              
fT <- terms(as.formula(paste("~", trt.str, sep = "")), keep.order = TRUE)  #fixed terms

trtTerm <- attr(fT, "term.labels")
effectsMatrix <- attr(fT, "factor")        

T <- makeContrMat(design1, trtTerm, effectsMatrix, contr.vec = NA)

N =  makeOverDesMat(design1, trtTerm)

Replist = getTrtRep(design1, trtTerm)   
 
Rep <- Replist$Rep
trt.Sca <- Replist$Sca
    
effFactors = lapply(makeOrthProjectors(Z), function(z) 
      getEffFactor(z, T, N, Rep, trt.Sca))