Search code examples
rgtsummary

In R how can I make a contingency table with five binary factors with gtsummary (or any other r package)

I am trying to replicate a table:

enter image description here which shows counts of people who are grouped on five binary variables. I can load the data and get the counts to display with lousy labels like this:

Drugs2 <- read.table("http://users.stat.ufl.edu/~aa/cat/data/Substance2.dat",
                        header = TRUE, stringsAsFactors = TRUE)
library(dplyr)

`Table 7.5` <-  bind_cols(`Alcohol Use` = c("Yes", "", "No", ""),
  `Cigarette Use` = rep(c("Yes", "No"),2),
  matrix(Drugs2$count, ncol = 8,byrow = T,
         dimnames = list(NULL,
         c("Y_F_W", "N_F_W", "Y_M_W", "N_M_W", "Y_F_O", "N_F_O", "Y_M_O", "N_M_O")))) 

knitr::kable(`Table 7.5`)

I see that the gtsummary package can add one level of column labels with the modify_spanning_header() function but I don't see how to nest spanning headers. Does anybody know how I can add nested headers with gtsummary?

Can anybody suggest how to produce a publication quality version of this table using gtsummary or some other package?

Solution

  • You're correct that gtsummary cannot add multiple spanning headers (it's written that way to ensure we support output to multiple formats).

    I would try flextable to print this table. They have functions for adding multiple headers row, and the ability to merge cells (both vertically and horizontally). https://davidgohel.github.io/flextable/

    enter image description here