Question

I’d like to apply a function to each row of a data frame, as below. I know how to use apply in the case where the data frame contains only numbers, but what if the rows contain, say, booleans / logicals, strings and integers? Example:

df <- data.frame(x=1:10,
                 y=c(TRUE, FALSE),
                 z=letters[1:10],
                 stringsAsFactors=FALSE)

RowFunction <- function(row) {
  if (row$y) return(row$x)
  return (row$z)
}

sapply(1:dim(df)[1], function(i) { RowFunction(df[i, ]) })

Is there a better way to do this? My first thought was to use apply(df, 1, RowFunction) after adding row <- as.list(row) to the beginning of RowFunction, but this doesn’t work because apply coerces df into an array, which can’t handle rows containing different data types.

Just for my R knowledge, I’d like to know if there is a cleaner way to do this than sapply(1:dim(df)[1], ... ). Any ideas?

Thanks in advance!

Answer 1

In this case, you can simply use ifelse:

sapply(1:dim(df)[1], function(i) { RowFunction(df[i, ]) })
 [1] "1" "b" "3" "d" "5" "f" "7" "h" "9" "j"

with(df, ifelse(y, x, z))
 [1] "1" "b" "3" "d" "5" "f" "7" "h" "9" "j"

---

For convenience and readability I also used with – this allows you to refer to a column just by name, without using the $ operator.