Formatting Tables in R

Sometime data viz people dump on tables a little bit (I’ve done it), as not being able to very efficiently communicate information. But this is probably because they are not doing it right.

Data visualization uses all these super-speed channels: position, color, size. In fact, visuals can also be quite strong in well formatted tables; they also use encoding and meaning for visual cues. With the help of table formatting packages, we can do better at making more attractive tables that aid efficient understanding and use! Nice!

Okay, but there are a lot of table formatting tools out there – just in R. How do I know which one to use?

This is the outline of an resource page on table formatting in R. Some parts are more filled in, but a lot is missing.

I’m using flipbooks. Because the formatting tools are so visual – it is a fun to watch the outputs evolve and it is easy to identify parts of the pipeline that are relevant to what you want to get done. I think they can help show how the different API’s work in a detailed and accessible way. Lot’s of info, but presented in stacks of slides, it doesn’t feel overwhelming – I hope!

I’m using a new flipbook function: embed_flipbook() and there are some wrinkles to work out. You may even see some error messages related to this! Work in progress. Embedded flipbooks benefit a lot from xaringanExtra, and flipbooks benefit from {xaringan} itself!

Tables have been an inspiration for moving the flipbook project forward too, to thinking more about flipbooks in documentation and specifically motivating “rotate” and “replace” modes for such work. Thanks especially {gt}!

Motivation

Let’s start with a motivating example. Of my tables, my favorite is probably this named colors table. It is nice and compact and the you can copy and paste the names of the colors. I made it with {gt}. Then we’ll give you a birds eye view on some table formatting packages in R.

Cross-Package Comparisons

output formats

This comparison table summarizes a lot of information about table packages: what outputs the package supports. Even if you like the way a package’s table looks in html, if you can’t output to the format you need, you probably don’t want to go too far down that road.

It is put together by Daniel Sjoberg and use with permission! He built the table with {gt}, and we reproduce much of the build here.

Print Engine	Function	HTML	PDF	RTF	Word
gt	as_gt()
kable	as_kable()
flextable	as_flex_table()
kableExtra	as_kable_extra()
huxtable	as_hux_table()
tibble	as_tibble()

---

Default table appearance

Error : Problem with `filter()` input `..1`.
x subscript out of bounds
ℹ Input `..1` is `line %in% which_highlight_frame`.

list column

Error : Problem with `filter()` input `..1`.
x subscript out of bounds
ℹ Input `..1` is `line %in% which_highlight_frame`.

Longer builds

This section just compares to builds from two separate packages, but would ideally compare the full set of packages. The results are similar, but APIs are a bit different.

{flextable}

Year		The Country Name	lifeExp	Population	gdpPercap
2002	Asia	Afghanistan	42.129	25268405	726.7341
	Europe	Albania	75.651	3508512	4604.2117
	Africa	Algeria	70.994	31287142	5288.0404
		Angola	41.003	10866106	2773.2873
	Americas	Argentina	74.340	38331121	8797.6407
	Oceania	Australia	80.370	19546792	30687.7547

{huxtable}

year	continent	country	lifeExp	pop	gdpPercap
2002	Asia	Afghanistan	42.1	25268405	727
	Europe	Albania	75.7	3508512	4.6e+03
	Africa	Algeria	71	31287142	5.29e+03
		Angola	41	10866106	2.77e+03
	Americas	Argentina	74.3	38331121	8.8e+03
	Oceania	Australia	80.4	19546792	3.07e+04

Links to step-by-step tutorials mostly the authors’ examples, animated

• {gt}

• {flextable}

• {huxtable}

• {kableExtra}

---

Future comparisons of model output formatting?

• gtsummary

• stargazer

• modelsummary