How to merge 3 variables in one and then transform count to cases

[phsousa123]

Hi everyone, so I have this csv from a few studies I'm analyzing, and I don't know how to merge the last 3 columns into one (a single variable with 3 levels) and then transform each number in a case (I managed to do this by creating another data frame with the function "countsToCases", but it only works for two variables, and I would like to add the other ones as well. Is it possible?

For example: a single case of [rocha2013, sudeste, Sao Paulo, negation].

[scottyd22]

I think your desired output can be achieved with pivot_longer() and uncount().

library(tidyverse)

# one row of your data
df = data.frame(
  pesquisa = 'Nunes',
  regiao = 'Sudeste',
  estado = 'Rio de Janeiro',
  preverbal = 616,
  duplanegacao = 214,
  posverbal = 13
)

out = df %>%
  # reshape counts into one column
  pivot_longer(cols = c('preverbal', 'duplanegacao', 'posverbal'),
               names_to = 'type') %>%
  # expand the counts into rows for individual cases
  group_by(pesquisa, regiao, estado, type) %>%
  uncount(value) %>%
  ungroup()

out
#> # A tibble: 843 × 4
#>    pesquisa regiao  estado         type     
#>    <chr>    <chr>   <chr>          <chr>    
#>  1 Nunes    Sudeste Rio de Janeiro preverbal
#>  2 Nunes    Sudeste Rio de Janeiro preverbal
#>  3 Nunes    Sudeste Rio de Janeiro preverbal
#>  4 Nunes    Sudeste Rio de Janeiro preverbal
#>  5 Nunes    Sudeste Rio de Janeiro preverbal
#>  6 Nunes    Sudeste Rio de Janeiro preverbal
#>  7 Nunes    Sudeste Rio de Janeiro preverbal
#>  8 Nunes    Sudeste Rio de Janeiro preverbal
#>  9 Nunes    Sudeste Rio de Janeiro preverbal
#> 10 Nunes    Sudeste Rio de Janeiro preverbal
#> # … with 833 more rows

Created on 2023-01-27 with reprex v2.0.2.9000

[phsousa123]

thank you so so so much