CODEWITHSUNDEEP
rdplyr
NBK
NBK

Reputation: 905

r - pivot wider multiple option questionnaire

I have a dataframe where rows are persons, and columns are how they responded in multiple option questions:

df <- data.frame(person = c("A", "B", "C"), question_1 = c(1, 3, 2), question_2 = c(1, 1, 2))
> df
  person question_1 question_2
1      A          1          1
2      B          3          1
3      C          2          2

I need to pivot_wider this table in the following way: each value of the original table becomes a column, and the values of the new table are whether the option was chosen (0 = not chosen, 1 = chosen). The final output should look this way:

person q1_1 q1_2 q1_3 q2_1 q2_2
1      A    1    0    0    1    0
2      B    0    0    1    1    0
3      C    0    1    0    0    1
>

I've tried with pivot_wider(df, names_from = c(question_1, question_2)), but it doesn't work. Any suggestions?

Upvotes: 2

Views: 158

Answers (2)

Onyambu
Onyambu

Reputation: 79208

In base R you could do:

with(stack(df[-1]), table(cbind(df[1], q = sprintf("q%s_%d", gsub("\\D", "", ind), values))))

     q
person q1_1 q1_2 q1_3 q2_1 q2_2
     A    1    0    0    1    0
     B    0    0    1    1    0
     C    0    1    0    0    1

Upvotes: 2

akrun
akrun

Reputation: 887048

We could reshape to 'long' format before converting to 'wide

library(dplyr)
library(tidyr)
library(stringr)
df %>% 
  pivot_longer(cols = -person) %>%
  mutate(name = str_c('q', str_extract(name, "\\d"), "_", value), value = 1) %>% 
  pivot_wider(names_from = name, values_from = value, values_fill = 0) %>%
  select(gtools::mixedorder(names(.)))

-output

# A tibble: 3 x 6
#  person  q1_1  q1_2  q1_3  q2_1  q2_2
#  <chr>  <dbl> <dbl> <dbl> <dbl> <dbl>
#1 A          1     0     0     1     0
#2 B          0     0     1     1     0
#3 C          0     1     0     0     1

Upvotes: 2

Related Questions