Skip to content

Using string vectors as for loop arguments

2 messages · Kevin Middleton, Bert Gunter

Kevin Middleton
# 
Is there a mechanism to interate through a vector of strings? Say I  
have a data frame of 50 variables (VAR1 to VAR50), each with 100  
measurements along with some coding factors (EXP and DOSE). I want to  
calculate the mean of a subset of each of VAR1 to VAR 50 (selecting  
by EXP and DOSE). Rather than just copy/pasting the same code 50  
times, I thought of using a for loop to go through a vector of VAR1  
to VAR50 (VAR.LIST below). Each iteration would be the variable name  
for which the mean would be calculated.

Trying this, I get an error that the argument is not numeric or  
logical. So I assume that for loops with vectors of strings are  
disallowed. What I am wondering is whether there a way to mimic this  
sort of procedure?

The code looked something like this:

VAR.LIST<-paste(c("VAR"), 1:50, sep="")

for (i in VAR.LIST){
	mean(i[EXP==1 & DOSE==1], na.rm=T)
	}



Thanks
Kevin Middleton
Bert Gunter
# 
?apply

as in:

answer<-apply(yourframe,2,function(x)x[EXP==1 & DOSE==1])

Note that there are slicker ways to do this call.
Note also for that for the particular case of column means, you have a
built-in much faster alternative:

answer<-colMeans(yourframe[EXP==1 & DOSE==1,],na.rm=TRUE)

Also note that the == construction may be problematic if EXP or DOSE are not
factors, say.

Finally, you need to start reading the R docs to learn about this yourself.
"An Introduction to R" is a good place to start.

Oh... and your attempt below will certainly not work. After you read the
basic docs, you'll see why not.


-- Bert Gunter
Genentech Non-Clinical Statistics
South San Francisco, CA