-1

I have to change the range 1-4 to 0-3 for five columns. (replace 1 with 0, replace 2 with 1,...). I already replaced those values for one column below. I still have to do the same for 9 other columns (AM01_02 - AM01_10). How could I do this in a simpler way without using that much space? Thanks a lot

#change range 1-4 to 0-3 
#replace 1 with 0
ba_data$AM01_01[ba_data$AM01_01  == 1] <- 0
#replace 2 with 1
ba_data$AM01_01[ba_data$AM01_01  == 2] <- 1
#replace 3 with 2
ba_data$AM01_01[ba_data$AM01_01  == 3] <- 2
#replace 4 with 3
ba_data$AM01_01[ba_data$AM01_01  == 4] <- 3
Anna
  • 1
  • 2
    Does this answer your question? [R - Subtract the same value from multiple columns](https://stackoverflow.com/questions/65182275/r-subtract-the-same-value-from-multiple-columns) or https://stackoverflow.com/questions/70794961/wanting-to-subtract-1-from-several-specific-columns-in-a-dataframe – harre Jul 23 '22 at 13:09

1 Answers1

2

As an alternative to Zheyuan Li's simpler method that works really well with math operations, this is a little more generalized for operations that might be a little more complex:

cols_to_change <- c(2,4,5,8,10)
# or
cols_to_change <- c("AM01_01", "AM01_02", ...)

myfun <- function(z) z-1
ba_data[cols_to_change] <- lapply(ba_data[cols_to_change], myfun)

Walk-through:

  • lapply(L, F) iterates the function F over each "element" in L (a list). In R, a data.frame is mostly just a list where each element (column) is the same length.
  • Because lapply(..) returns a list, and the columns you're working on are likely a subset of the entire frame, we need to assign it back to the respective columns; ergo ba_data[cols_to_change] <-

The reason this is more general and can be useful: if your operation is more of a "lookup" than a "subtract one", you can change myfun to be more specific. For instance, if in all of these columns you need to replace 1 with 21, 2 with 97, and 3 with -1, and leave all other values intact, then you might write the function as:

myfun <- function(z, lookup) {
  for (nm in names(lookup)) {
    z <- ifelse(as.character(z) == nm, lookup[[nm]], z)
  }
  z
}
ba_data[cols_to_change] <- 
  lapply(ba_data[cols_to_change],
         function(x) myfun(x, c("1"=21, "2"=97, "3"=-1)))

If you were to use a lookup like this, realize that I named them as strings regardless of what class the original data is, because "names" of things in R should not start with (or be entirely) numbers.

r2evans
  • 141,215
  • 6
  • 77
  • 149