dplyr - Is there a way to turn a specific amount of columns to rows in R?

Question

A	B	C	n	n%
FALSE	FALSE	FALSE	randomnum1	0.86
FALSE	FALSE	TRUE	randomnum2	0.6
FALSE	TRUE	FALSE	randomnum3	0.3
TRUE	FALSE	FALSE	randomnum4	0.84

question from:https://stackoverflow.com/questions/65902741/is-there-a-way-to-turn-a-specific-amount-of-columns-to-rows-in-r

Answer 1

This might be not the most efficient way, but it does the job.

Note: your "n%" column does not have a proper R column name so I chose "n_pct" instead.

df <- data.frame(A = c(F, F, F, T),
                 B = c(F, F, T, F),
                 C = c(F, T, F, F),
                 n = c(1, 2, 3, 4),
                 n_pct = c(0.86, 0.6, 0.3, 0.84))

library(tidyverse)
df %>%
  pivot_longer(cols = c(A, B, C)) %>%
  group_by(n) %>%
  mutate(sum_value = sum(value),
         name      = if_else(sum_value == 0, "---", name),
         helper    = 1:n()) %>%
  ungroup() %>%
  filter(value == TRUE | (sum_value == 0 & helper == 1)) %>%
  select(name, n, n_pct)

which gives:

# A tibble: 4 x 3
  name      n n_pct
  <chr> <dbl> <dbl>
1 ---       1  0.86
2 C         2  0.6 
3 B         3  0.3 
4 A         4  0.84