Hi Farnoosh,
Try this:
set.seed(151)
mat1<-matrix(sample(1:400,100,replace=TRUE),ncol=20)
set.seed(15)
colnames(mat1)<-paste(sample(c("proc","Lab","other"),20,replace=TRUE),sample(1:45,20,replace=FALSE),sep="")
mat1[,grepl("proc|Lab",colnames(mat1))]
#???? Lab37 proc35 Lab3 Lab24 proc6 proc36 Lab9 Lab17 Lab32 proc34
#[1,]?? 192??? 246? 241?? 126?? 374??? 122? 281?? 116?? 215????? 4
#[2,]?? 347??? 182? 237?? 393?? 300???? 49? 125??? 90?? 245??? 148
#[3,]?? 312??? 329? 144?? 285?? 224??? 110? 378?? 329?? 275???? 79
#[4,]?? 216??? 139? 300?? 230?? 249??? 206? 231?? 357?? 188???? 87
#[5,]?? 177??? 383? 102?? 211??? 26??? 283?? 20?? 351?? 392??? 286
#if you want columns not "Lab" or "proc"
mat1[,!grepl("proc|Lab",colnames(mat1))]
A.K.
________________________________
From: farnoosh sheikhi <farnoosh_81 at yahoo.com>
To: "smartpink111 at yahoo.com" <smartpink111 at yahoo.com>
Cc: "r-help at R-project.org" <r-help at R-project.org>
Sent: Wednesday, December 12, 2012 3:39 PM
Subject: Subset of Data
Hi Arun,
I have a question about choosing a subset of data.
I have a matrix of 5000 in 3500.
I want to choose specific variables such as proc1 to proc1000 and Lab1 to
Lab1600 and put it into a new matrix to run some correlation analysis.
Since I have 3500?variables, I don't know from what column my lab or
procedures start.
I really appreciate your help.
Thanks.
Best,Farnoosh Sheikhi?