Expanding values from multiple columns into binary values

I have columns with either numeric or NA values that I want to reorder so that the new columns are named with the distinct numeric values, and the values are binary 0 or 1 values.

The actual dataset contains many other variables, so the code would need to refine the function only to the 'stress' variables.

Example dataset:

df <- data.frame(
  stress1 = c('A', 'A', 'B', 'A'),
  stress2 = c(NA, 'B', 'C', 'B'),
  stress3 = c(NA, NA, NA, 'C'),
  stress4 = c(NA, NA, NA, 'D')
)

Desired outcome:

desiredoutcome <- data.frame (
  A = c(1, 1, 0, 1),
  B = c(0, 1, 1, 1),
  C = c(0, 0, 1, 1),
  D = c(0, 0, 0, 1)
)

In tidyr + dplyr.

library(tidyr)
library(dplyr)

df %>% 
  mutate(id = row_number()) %>% 
  pivot_longer(-id, values_drop_na = TRUE) %>% 
  pivot_wider(names_from = "value", values_from = "name", 
              values_fill = 0, values_fn = length)

#   id A B C D
# 1  1 1 0 0 0
# 2  2 1 1 0 0
# 3  3 0 1 1 0
# 4  4 1 1 1 1

Or in base R:

df$ID <- seq_along(df)
table(cbind(df['ID'], unlist(df[1:4]))) |>
  as.data.frame.matrix()

#   A B C D
# 1 1 0 0 0
# 2 1 1 0 0
# 3 0 1 1 0
# 4 1 1 1 1

Base R way, generalised to hopefully work for real data with any number of different strings in:

# updated sample data frame - NA's should not be quoted, removed trailing comma
df <- data.frame(
  stress1 = c('A', 'A', 'B', 'A'),
  stress2 = c(NA, 'B', 'C', 'B'),
  stress3 = c(NA, NA, NA, 'C'),
  stress4 = c(NA, NA, NA, 'D')
)


desiredout <- data.frame (
  A = c(1, 1, 0, 1),
  B = c(0, 1, 1, 1),
  C = c(0, 0, 1, 1),
  D = c(0, 0, 0, 1)
)


out = data.frame( # for data frame output
  lapply( # iterative
    unique(
      unlist(df[!is.na(df)]) # all unique, non-NA values in df 
      ),
    \(x) {
      rowSums(df == x, na.rm = TRUE)
    }
    ))
            
names(out) <- unique(unlist(df[!is.na(df)]))

check:

> all.equal(desiredout, out)
[1] TRUE

and:

> out
  A B C D
1 1 0 0 0
2 1 1 0 0
3 0 1 1 0
4 1 1 1 1

If you find you have figures other than 1 or 0 in your final column then you may have rows with more than one instance of a given string - if so, come back and we can edit appropriately.

A purrr solution with pmap_dfr + table:

library(purrr)

pmap_dfr(df, ~ unclass(table(c(...)))) %>%
  replace(is.na(.), 0)

# # A tibble: 4 × 4
#       A     B     C     D
#   <int> <int> <int> <int>
# 1     1     0     0     0
# 2     1     1     0     0
# 3     0     1     1     0
# 4     1     1     1     1