Making Pretty Tables in RMarkdown

Why?

Default tables are not that pretty when compiled:

tipjoke <- read.csv('https://uoepsy.github.io/data/TipJoke.csv')
tiptable <- table(tipjoke$Card, tipjoke$Tip)
tiptable

      
        0  1
  Ad   60 14
  Joke 42 30
  None 49 16

The output of table() (above) can sometimes look neat enough, but data in other formats is hard to make look nice (it will just get printed as the object it is, e.g. a tibble below)

library(tidyverse)
pass_scores <- read.csv("https://edin.ac/2wJgYwL") 
pass_scores %>% group_by(school) %>%
  summarise(
    meanPASS = mean(PASS),
    sdPASS = sd(PASS)
  )

# A tibble: 7 × 3
  school      meanPASS sdPASS
  <chr>          <dbl>  <dbl>
1 ECA             32.3   8.02
2 Economics       34     1.41
3 GeoSciences     31    NA   
4 LAW             30.7   1.75
5 PPLS            28.8   1.30
6 SLLC            28    NA   
7 SSPS            31     0   

Functions for pretty tables

kable()

knitr’s kable() is a fairly simple way to make tables and data look a little nicer. Simply pass something to kable() and will usually spit out something nicer. You can find a guide here: https://bookdown.org/yihui/rmarkdown-cookbook/kable.html
We can also add captions to our table!

library(knitr)
kable(tiptable, caption = "Oh what a lovely table")

Table 1: Oh what a lovely table

	0	1
Ad	60	14
Joke	42	30
None	49	16

The kable() function also has various useful bits of functionality such as rounding all numeric values to the same number of digits, or setting a value for which all NAs will be displayed.

options(knitr.kable.NA = "")

pass_scores %>% group_by(school) %>%
  summarise(
    meanPASS = mean(PASS),
    sdPASS = sd(PASS)
  ) %>% kable(., digits = 1)

school	meanPASS	sdPASS
ECA	32.3	8.0
Economics	34.0	1.4
GeoSciences	31.0
LAW	30.7	1.8
PPLS	28.8	1.3
SLLC	28.0
SSPS	31.0	0.0

gt()

The gt package has a lot of useful functionality, and works well with grouped data. The best resource for this is at https://gt.rstudio.com. gt() also allows us to specify captions.

library(tidyverse)
library(gt)
tipjoke %>% 
  count(Card, Tip) %>%
  gt(caption = "Another lovely table!")

Table 2: Another lovely table!

Card	Tip	n
Ad	0	60
Ad	1	14
Joke	0	42
Joke	1	30
None	0	49
None	1	16

You can shape your data in different ways before passing it to gt().

tipjoke %>% 
  group_by(Card) %>%
  summarise(
    TipYes = sum(Tip==1),
    TipNo = sum(Tip==0)
  ) %>%
  gt()

Card	TipYes	TipNo
Ad	14	60
Joke	30	42
None	16	49

pass_scores %>% group_by(school) %>%
  summarise(
    meanPASS = mean(PASS),
    sdPASS = sd(PASS)
  ) %>% gt()

school	meanPASS	sdPASS
ECA	32.33333	8.020806
Economics	34.00000	1.414214
GeoSciences	31.00000	NA
LAW	30.66667	1.751190
PPLS	28.80000	1.303840
SLLC	28.00000	NA
SSPS	31.00000	0.000000

pander()

Pander is nice and simple, and works on the idea of converting R output into neat markdown syntax, which then get compiled into nice looking tables.
You can read up on some of the functionality here: https://cran.r-project.org/web/packages/pander/vignettes/pander.html

library(pander)
pander(tiptable)

	0	1
Ad	60	14
Joke	42	30
None	49	16

pass_scores %>% group_by(school) %>%
  summarise(
    meanPASS = mean(PASS),
    sdPASS = sd(PASS)
  ) %>% pander()

school	meanPASS	sdPASS
ECA	32.33	8.021
Economics	34	1.414
GeoSciences	31	NA
LAW	30.67	1.751
PPLS	28.8	1.304
SLLC	28	NA
SSPS	31	0