MEET (version 5.1.1)

JacksonParameters: JacksonParameters: To calculates the parameters needed to transform a Q-residual to a confidence interval

Description

Given PCA model, this function calculates the parameters to perform the Jackson statistics of the Q-residuals.

Usage

JacksonParameters(nPCs, TFBS)

Arguments

nPCs

number of principal components of the model

TFBS

numerical TFBS matrix

Value

output:list including the parameters h0 x1, x2 and x3 needed to calculate the Q-residuals statistics

See Also

kfold.PCA, Predict.PCA, QtoJackson

Examples

require("MEET")
data(TranscriptionFactor)
data(iicc)
nPCs<-1
TFBS<-iicc$Transcriptionfactor
Prob<-iicc$background
missing<-50
NumericalMatrix<-numericalDNA(Prob)
suma<-apply(TFBS,2,function(y){sum(y=="-")})
threshold<-floor(nrow(TFBS)*missing/100)
TFBS<-TFBS[, suma<=threshold]
ncolTFBS<-ncol(TFBS)
TFBSnum<-apply(TFBS,1,function(x){as.vector(t(NumericalMatrix[x,]))})
TFBSnum<-t(TFBSnum)
model<-pca(TFBSnum, nPcs=nPCs, method="svd", center=TRUE)
JacksonPars<-JacksonParameters(nPCs,TFBSnum)