Creating {gt} frequency tables with {ivo.table}

ivo.table version 0.7 is now available on CRAN. It adds support for creating gt tables, meaning that you can now create great-looking frequency tables and crosstables in the gt format with a single line of code.

Creating gt crosstables

In a previous post, I described how to create and style flextable frequency tables using the ivo_table function. In this post, I’ll describe the new ivo_table_gt function.

To create a three-way contingency table for the penguins data, we select the variables we wish to tabulate and pipe them to ivo_table_gt as follows:

library(dplyr)
library(ivo.table)

penguins |> select(species, sex, island) |> ivo_table_gt()

	species
	Adelie	Chinstrap	Gentoo
female
Biscoe	22	0	58
Dream	27	34	0
Torgersen	24	0	0
male
Biscoe	22	0	61
Dream	28	34	0
Torgersen	23	0	0
(Missing)
Biscoe	0	0	5
Dream	1	0	0
Torgersen	5	0	0

We can easily add row and column sums:

penguins |>
  select(species, sex, island) |>
  ivo_table_gt(sums = c("cols", "rows"))

	species			Total
	Adelie	Chinstrap	Gentoo
female
Biscoe	22	0	58	80
Dream	27	34	0	61
Torgersen	24	0	0	24
Total	73	34	58	165
male
Biscoe	22	0	61	83
Dream	28	34	0	62
Torgersen	23	0	0	23
Total	73	34	61	168
(Missing)
Biscoe	0	0	5	5
Dream	1	0	0	1
Torgersen	5	0	0	5
Total	6	0	5	11

Because ivo_table_gt creates a gt object, we can change the style of the table using standard gt functions like tab_style:

library(gt)
penguins |>
  select(species, sex, island) |>
  ivo_table_gt(color = "red",
               font_name = "Courier") |> 
  tab_style(style = list(cell_text(weight = "bold")),
                locations = cells_stub())

	species
	Adelie	Chinstrap	Gentoo
female
Biscoe	22	0	58
Dream	27	34	0
Torgersen	24	0	0
male
Biscoe	22	0	61
Dream	28	34	0
Torgersen	23	0	0
(Missing)
Biscoe	0	0	5
Dream	1	0	0
Torgersen	5	0	0

The caption´ argument lets us add a caption.tab_style` can also be used to highlight cells of particular interest:

penguins |>
  select(species, sex, island) |>
  ivo_table_gt(caption = "A table with penguins in it") |> 
    tab_style(style = list(cell_fill(color = "darkgreen")),
              locations = cells_body(columns = 5, rows = 4))

A table with penguins in it
	species
	Adelie	Chinstrap	Gentoo
female
Biscoe	22	0	58
Dream	27	34	0
Torgersen	24	0	0
male
Biscoe	22	0	61
Dream	28	34	0
Torgersen	23	0	0
(Missing)
Biscoe	0	0	5
Dream	1	0	0
Torgersen	5	0	0

The best part

What I think makes using gt tables instead of flextable tables an appealing option is that we can paste them with ggplot objects using the patchwork package, like so:

library(ggplot2)
library(patchwork)

penguins_plot <- ggplot(penguins, aes(body_mass, fill = species)) +
                    geom_density(alpha = 0.7) +
                    theme(legend.position = "bottom")

penguins_table <- penguins |>
                    select(species, sex, island) |>
                    ivo_table_gt()

penguins_plot + penguins_table

This should be really useful for presentations and other situations where we don’t need the table to be in a text format.

ivo_table_gt was developed by Stefan Furne. Thanks Stefan!