For example

Will transform into this

Another example

will turn into this

So, if we say that n="sum of all numbers in the original data.frame", i.e., the number of all individuals, the final output will be a data.frame with 6 columns and n rows.

I want to do this in R but i don't have any idea how. Once I have this, what i want to do is apply a generalized linear model with family binomial and link = probit.

Now, this page can explain some of what I tried to do:

https://www.datanalytics.com/libro_r/la-funcion-melt-y-datos-en-formato-largo.html

Solution

Okay... I have an answer, but... i was wondering if there exists any generalization. Here it goes:

library(readxl)
library(dplyr)

# Información original ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
byssinosis <- read_xls(path = "byssinosis.xls",range = "B4:K27",col_names = F)
names(byssinosis) <- c("Employment","Smoking","Sex","Race",
                       "W1y","W1n","W2y","W2n","W3y","W3n")
# View(byssinosis)

# Procesando la información a individuos ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# Primero pasamos las columnas a una sola.
datos <- reshape2::melt(byssinosis)
# Separamos estas columnas en las dos características deseadas.
datos <- datos %>%
  mutate(Workplace = ifelse(variable %in% c("W1y", "W1n"),1,
                            ifelse(variable %in% c("W2y", "W2n"),2,3)),
         Byssinosis = ifelse(variable %in% c("W1y", "W2y", "W3y"),"yes","no"))
# Repetimos con base en value.
individuos=rep(seq_len(nrow(datos)),datos$value)
datos <- datos[individuos,]
# Nos quedamos solo las columnas deseadas
datos <- datos %>% select(-c(variable,value))
# View(datos)

# Comprobación ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
tabla <-
  table(datos) %>%
  as.data.frame() %>%  
  arrange(Employment, desc(Smoking), desc(Sex), desc(Race), Workplace, desc(Byssinosis))
# View(tabla)