[R] pls - No variable selection

Barry King barry.king at qlx.com
Wed May 27 15:03:27 CEST 2015


I am attempting to use mixOmics' pls() on a problem with 16 rows
and 27 columns.  When I run pls I get the message "No variable selection."
I have included a small problem below that produces this message.
What am I doing wrong?  Any help is appreciated.


#===== Begin small reproducible sample program =====
library(mixOmics)

X <- matrix(c(2766, 2610, 3306, 3630,
              1492, 1419, 1369, 1158,
              2450, 2379, 2400, 2055,
              2751, 2883, 3492, 3570,
              2652, 2691, 3225, 3285,
              3993, 4722, 6147, 6720,
              4032, 4350, 5430, 5763,
              4530, 5190, 6910, 7580,
              4077, 4410, 5460, 5857,
              3450, 3432, 3969, 4020,
              4989, 5301, 6807, 7425,
              5340, 5790, 7590, 8390,
              3162, 3477, 4365, 4650,
              4380, 4695, 6018, 6510,
              4587, 4200, 5040, 5289,
              4017, 4725, 6090, 6570),
            ncol=4)
colnames(X) <- c("v1","v2","v3","v4")

Y <- matrix(c( 3.0110,
               0.0000,
               0.0000,
               1.4820,
               1.1160,
               3.3970,
               2.4280,
               4.0240,
               2.2750,
               0.9588,
               3.1900,
               4.1320,
               2.1600,
               3.0940,
               1.6040,
               3.1620),
            ncol=1)
colnames(Y) <- c("y")

mpls <- pls(X, Y,  mode="regression", ncomp=2)
mpls

#===== End small reproducible sample program, begin console output =====

Call:
 pls(X = X, Y = Y, mode = "regression")

 PLS with a 'regression' mode with 2 PLS components.
 You entered data X of dimensions: 16 4
 You entered data Y of dimensions: 16 1

 No variable selection.

 Available components:
 --------------------
 loading vectors: see object$loadings
 variates: see object$variates
 variable names: see object$names

#===== End console output =====

Barry King
Associate Professor of Information Technology
Butler University

	[[alternative HTML version deleted]]



More information about the R-help mailing list