'for' loop on columns with multiple numbers in name, R

Question

My problem is very simple but I cannot get it to work (despite searching)... I'm a newbie to R.

I have a dataframe with light measurements. Measurements below the canopy were taken along 4 transects, on 13 positions within each transect. Measurements above the canopy were taken once per transect.

df$below_position_1_transect_1
df$below_position_2_transect_1
# Et cetera with position 1:13 and transect 1:4

df$above_transect_1
df$above_transect_2
# Et cetera with transect 1:4

I want to perform a simple function on those columns and put the results in new columns which are correctly named. There should be a total of 13*4 new columns.

lightavailable <- function(below,above) {below / above * 100}

for (i in c(1:13)) {
  for (j in c(1:4)) {
    df[,i] <- lightavailable(
      below  = df[,paste0("below_position_",i,"_transect_",j)],
      above  = df[,paste0("above_transect_",j)]
      )
    colnames(df)[i] <- paste0("transmitted_transect_",j,"_position_",i)
  }
}

As a result, 13 new columns were added but only for transect 4, meaning that the results from transect 1 up to 3 were overwritten. I'm sure it has to do with writing everything to df[,i], but I'm not sure how to fix it.

How should the code be edited such that new columns are created for all combinations of transects and positions?

Thank you!

Can you share some sample data and show us that the intended result should look like. Have a look at this post for help: https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example — Phil, Jun 15 '17 at 10:39
Thank you Phil, I will make sure do so next time. The problem is already solved now. — Eva, Jun 15 '17 at 12:23

score 1 · Accepted Answer · answered Jun 15 '17 at 11:18

Rather than trying to figure out what integer index your new column will have, just name the column using the df[["string"]] syntax.

for (i in c(1:13)) {
  for (j in c(1:4)) {
        #---- new variable name -----------------------#
    df[[paste0("transmitted_transect_",j,"_position_",i)]] <- 
      lightavailable(
        below  = df[,paste0("below_position_",i,"_transect_",j)],
        above  = df[,paste0("above_transect_",j)]
      )
  }
}

Yes! Thank you very much Benjamin, that works perfectly! – Eva Jun 15 '17 at 12:20 — Eva, Jun 15 '17 at 12:20

'for' loop on columns with multiple numbers in name, R

1 Answers1