Row-wise sum of logical values in a dataframe

user_13046806 0
Viewed 487

I have a data frame like this (with more vars)

tb = data.frame(ID=c("a","b","c"),V1=c(TRUE,FALSE,TRUE),
                V2=c(FALSE,FALSE,TRUE),V3=c(TRUE,TRUE,FALSE) )
tb 

  ID    V1    V2    V3
1  a  TRUE FALSE  TRUE
2  b FALSE FALSE  TRUE
3  c  TRUE  TRUE FALSE

I need to add a fourth variable with the sum of true values by row like this but conserving all the other vars

tb %>%
 select(V1:V3) %>%
 mutate(out = rowSums(.))

     V1    V2    V3 out
1  TRUE FALSE  TRUE   2
2 FALSE FALSE  TRUE   1
3  TRUE  TRUE FALSE   2
dplyr r
2 Answers

The dplyr solution

In a single call, you can use the selection helper where inside acrossto feed only the columns that meet a condition (is.logical) to rowSums.

tb %>% mutate(sum = rowSums(across(where(is.logical))))

  ID    V1    V2    V3 sum
1  a  TRUE FALSE  TRUE   2
2  b FALSE FALSE  TRUE   1
3  c  TRUE  TRUE FALSE   2

You can also select the columns by name inside across, or with name patterns with starts_with or matches:

tb %>% mutate(sum = rowSums(across(V1:V3)))
#OR
tb %>% mutate(sum = rowSums(across(starts_with("V"))))
#OR
tb %>% mutate(sum = rowSums(across(matches("V\\d"))))
user_139723330

Maybe not the dplyr way but you could use select() within mutate() like so:

tb %>% mutate(out = rowSums(tb %>% select(V1:V3)))


  ID    V1    V2    V3 out
1  a  TRUE FALSE  TRUE   2
2  b FALSE FALSE  TRUE   1
3  c  TRUE  TRUE FALSE   2
user_45522950
Related