R - change values in multiple selected rows based on multiple conditional column-CodePudding

In SAS data step it is easy and intuitive to change value of selected row and column based on conditions. For example:

data want;
  set have;
  if col1 = "A" and col2 = "B" then do;
    col3 = "C";
    col4 = catx(",", Col1, Col2);
  end;
run;

I want to do the same thing in R. I have the following:

want <- have %>%
  mutate(
    col3 = ifelse(col1 == "A" & col2 == "B", "C", col3),
    col4 = ifelse(col1 == "A" & col2 == "B", paste(col1, col2, sep = ","))
  )

If I have more than 1 column need to change, I have to repeat the ifelse() for each column I want to change.

Is there a easier way to do it?

Thanks

=====================================

I tried this:

for (row in 1:nrow(have)) { 
    if (have$col1 == "A" & have$col2 == "B") { 
        have$col3 = 'C' 
        have$col4 = paste(have$col1, have$col2, sep=",") 
    } 
}

However, mydataset is too big to loop through.

CodePudding user response：

One approach would be to save the condition in a column and use it for different columns.

library(dplyr)

want <- have %>%
  mutate(
    condition = col1 == "A" & col2 == "B",
    col3 = ifelse(condition, "C", col3),
    col4 = ifelse(condition, paste(col1, col2, sep = ","), col4)
  )

CodePudding user response：

Here is an alternative way without loop:

have <- have %>% mutate(cond = col1 == "A" & col2 == "B", id = 1:nrow(have))
met  <- have %>% filter(cond == TRUE) %>% 
        mutate(col3 = "C", col4 = paste(cyl, hp, sep = ","))
have[met$id,] <- met