Easy plots with ggplot2

Using curly curly brackets in functions (tidyeval)

Background

I was watching Julia Silge’s youtube video on tidymodels, and saw that she had a very neat way of creating visualizations for exploratory data analysis.

A function is created to reduce the copying and pasting of code, and you can use curly curly brackets to refer to variables directly.

Let’s try it out!

Packages

library(tidyverse)

Data

mtcars_df <- mtcars %>% 
  as.data.frame()

An easy plot

ggplot(aes(mpg, cyl), data = mtcars_df) +
  geom_point() +
  labs(title = "Plot of cyl against mpg") +
  theme_classic() +
  theme(axis.text = element_text(size = 14),
        axis.title = element_text(face = "bold", size = 16),
        title = element_text(face = "bold", size = 20))

Writing a function

# function
 
# var_x : x variable
# var_y : y variable
# title : title of chart, in string

plot_mtcars <- function(var_x, var_y, title) {

  ggplot(aes({{var_x}}, {{var_y}}), data = mtcars_df) +
  geom_point() +
  labs(title = title) +
  theme_classic() +
  theme(axis.text = element_text(size = 14),
        axis.title = element_text(face = "bold", size = 16),
        title = element_text(face = "bold", size = 20))
 
}

Using the function

The customizations are saved in the function, simplifying the code and reducing the need for copying and pasting.

plot_mtcars(mpg, vs, "mpg vs vs")

Sleep dataset

glimpse(sleep)

Rows: 20
Columns: 3
$ extra <dbl> 0.7, -1.6, -0.2, -1.2, -0.1, 3.4, 3.7, 0.8, 0.0, 2.0, …
$ group <fct> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, …
$ ID    <fct> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 1, 2, 3, 4, 5, 6, 7, 8,…

Simple plot

sleep %>% 
  group_by(group) %>% 
  summarise(mean_extra_hrs = mean(extra), na.rm = T) %>% 
  ggplot(aes(group, mean_extra_hrs)) +
  geom_boxplot(aes(fill = group)) +
  geom_text(aes(label = mean_extra_hrs), vjust = -1) +
  theme_classic() +  
  labs(title = "Plot of mean_extra_hrs against group") +
  theme(axis.text = element_text(size = 14),
        axis.title = element_text(face = "bold", size = 16),
        title = element_text(face = "bold", size = 18))

Function

plot_boxplot_group_comparison <- function(data_df, var_x, var_y) {
  
  {{data_df}} %>% 
  group_by({{var_x}}) %>% 
  summarise(mean = mean({{var_y}}), na.rm = T) %>% 
  ggplot(aes({{var_x}}, mean)) +
  geom_boxplot(aes(fill = {{var_x}})) +
      geom_text(aes(label = mean), vjust = -1) +
    labs(title = 
           paste("Plot of ", as_label( enquo(var_y)), "against ", as_label(enquo(var_x))),
         y = paste("mean_", as_label(enquo(var_y)))) +
  theme_classic() +  
  theme(axis.text = element_text(size = 14),
        axis.title = element_text(face = "bold", size = 16),
        title = element_text(face = "bold", size = 20))
  
}


# using the function  
plot_boxplot_group_comparison(sleep, group, extra)

Using it on another dataset

# iris dataset

plot_boxplot_group_comparison(iris, Species, Petal.Width)

Thoughts

I really like this! Coding can be miminalist and functional at the same time.

References:

• https://sharla.party/post/tidyeval/
• https://www.youtube.com/channel/UCTTBgWyJl2HrrhQOOc710kA