PLSbiplot1 (version 0.1)

SPLS.biplot_Bmat: A zoomed-in display of the coefficient points in the Partial Least Squares (PLS) biplot for Sparse Partial Least Squares (SPLS)

Description

Takes in a set of predictor variables and a set of response variables and produces a zoomed-in display of the coefficient points in the PLS biplot for SPLS.

Usage

SPLS.biplot_Bmat(X, Y, algorithm = NULL, eps, lambdaY = NULL, lambdaX = NULL, ax.tickvec.B = NULL, ...)

Arguments

X

A (NxP) predictor matrix

Y

A (NxM) response matrix

algorithm

The SPLS algorithm

eps

Cut off value for convergence step

lambdaY

A value for the penalty parameters for the soft-thresholding penalization function for Y-weights

lambdaX

A value for the penalty parameters for the soft-thresholding penalization function for X-weights

ax.tickvec.B

(purple) tick marker length for the Y-variable axes in the biplot

...

Other arguments. Currently ignored

Value

A zoomed-in display of the coefficient points in the PLS biplot of a SPLS-GLM of D=[X Y] with some parameters

Examples

if(require(robustbase))
data(toxicity, package="robustbase")
Y1 = as.matrix(cbind(toxicity$toxicity))
dimnames(Y1) = list(paste(1:nrow(Y1)), "toxicity")
X1 = as.matrix(cbind(toxicity[,2:10]))
rownames(X1) = paste(1:nrow(X1))
#choosing a value for the penalty parameters lambdaY and lambdaX for this data
main2 = opt.penalty.values(X=scale(X1), Y=scale(Y1), A=2, algorithm=mod.SPLS, eps=1e-5,
from.value.X=0, to.value.X=10, from.value.Y=0, to.value.Y=0, lambdaY.len=1, lambdaX.len=100)
min.RMSEP.value = main2$min.RMSEP.value
lambdaY.to.use = main2$lambdaY.to.use
lambdaX.to.use = main2$lambdaX.to.use
list(lambdaY.to.use=lambdaY.to.use, lambdaX.to.use=lambdaX.to.use, min.RMSEP.value=min.RMSEP.value)
#SPLS analysis
main3 = mod.SPLS(X=scale(X1), Y=scale(Y1), A=2,
lambdaY=lambdaY.to.use, lambdaX=lambdaX.to.use,
eps=1e-5)
X.to.use = main3$X.select
Y.to.use = main3$Y.select
X.new = as.matrix(X1[,X.to.use])
Y.new = as.matrix(Y1[,Y.to.use])
colnames(Y.new) = colnames(Y1)
SPLS.biplot_Bmat(X.new, Y.new, algorithm=mod.SPLS,
lambdaY=lambdaY.to.use, lambdaX=lambdaX.to.use,
eps=1e-5, ax.tickvec.B=rep(5,ncol(Y.new)))

#ash data
if(require(chemometrics))
data(ash, package="chemometrics")
X1 = as.matrix(ash[,10:17], ncol=8)
Y1 = as.matrix(ash$SOT)
colnames(Y1) = paste("SOT")
#choosing a value for the penalty parameters lambdaY and lambdaX for this data
main2 = opt.penalty.values(X=scale(X1), Y=scale(Y1), A=2, algorithm=mod.SPLS, eps=1e-5,
from.value.X=0, to.value.X=500, from.value.Y=0, to.value.Y=0, lambdaY.len=1, lambdaX.len=100)
min.RMSEP.value = main2$min.RMSEP.value
lambdaY.to.use = main2$lambdaY.to.use
lambdaX.to.use = main2$lambdaX.to.use
list(lambdaY.to.use=lambdaY.to.use, lambdaX.to.use=lambdaX.to.use, min.RMSEP.value=min.RMSEP.value)
#SPLS analysis
main3 = mod.SPLS(X=scale(X1), Y=scale(Y1), A=2,
lambdaY=lambdaY.to.use, lambdaX=lambdaX.to.use,
eps=1e-5)
X.to.use = main3$X.select
Y.to.use = main3$Y.select
X.new = as.matrix(X1[,X.to.use])
colnames(X.new)  #P=6
colnames(X1)  #P=8
Y.new = as.matrix(Y1[,Y.to.use])
colnames(Y.new) = colnames(Y1)
colnames(Y.new)
SPLS.biplot_Bmat(X.new, Y.new, algorithm=mod.SPLS,
lambdaY=lambdaY.to.use, lambdaX=lambdaX.to.use,
eps=1e-5, ax.tickvec.B=5)