R handling column names in a for loop

Question

I have a data frame that's ~ 50,000 X 200. The column names are 4 different types with numbers on the end ranging from 1-50 (store1, price1, time1, rate1, store2, price2, time2, rate2,...,store50, price50, time50, rate50). I'm trying to create dummy variables depending on the values of each column but am having trouble getting R to handle the column names inside a loop.

store1    price1       time1      rate1     store2     price2    time2     rate2 ....
   A        55.55      08:09      1.44        B         44.44     11:09     1.46
   C        55.55      08:09      1.44        G         44.44     11:09     1.46
   X        55.55      08:09      1.44        E         44.44     11:09     1.46
   D        55.55      08:09      1.44        S         44.44     11:09     1.46

Here's what I have tried so far with no luck.

xform_data <- function(x) { 
       for(i in 1:50){
       storeX <-  (paste("store",i,sep="")) 
       storeX2  <- ifelse(storeX  == "A", 1, 2)
       x <- cbind(x, storeX2  )
       }
    x    
}

Any suggestions?

Answer 1

The following compares the name instead of the comparing the value:

ifelse(storeX  == "A", ...

Try:

ifelse(x[,storeX]  == "A", ...

Also, all the new columns will be called storeX2 . You might prefer to rename them:

x <- cbind(x, storeX2)
colnames(x)[length(colnames(x))] <- storeX2

(I am sure there exist more elegant ways to do it.)

Answer 2

@aix gave the proper way to do this with a loop, however you may find it quicker or easier to use some other tools, depending on what you want your final result to be. Functions like sapply and lapply can be used to process every column of a data frame (or subset of a data frame) the same way. The model.matrix function will convert variables into dummy variables (0's and 1's) in one step. Other tools that may help include factors, switch , and match .