Descriptive Statistics in R

Introduction to plotting in the tidyverse: ggplot2

Michael Luu, MPH | Marcio Diniz, PhD
September 22, 2022

ggplot2

Objective

We won’t be able to cover how to create all the various types of plots
The goal is to provide you with the understanding of the basic components and tools on HOW these plots are created using ggplot2
All plots in ggplot2 are built upon the same fundamental principals and concepts

Plots we WILL go over

Dotplot

Histogram

Boxplot

Barplot

Components of a Plot

We will be using the emergency dataset as an example

# A tibble: 149 × 23
      id   age gender    hr   sbp   dbp ascites glasgow   cpr encephal…¹ creat…²
   <dbl> <dbl> <fct>  <dbl> <dbl> <dbl>   <dbl>   <dbl> <dbl>      <dbl>   <dbl>
 1     1    49 Female   109   114    76       1      15  NA            0    1.17
 2     2    61 Female    78   130    80       0      14  77.5          1    4.4 
 3     3    65 Female    63   138    71       1      14  11.9          1    1.69
 4     4    54 Female   100    92    80       1      15 184            0    0.62
 5     5    66 Female    86   150    70       0      NA  NA            1    0.91
 6     6    54 Female    86    60    40       1      14 200            1    1.09
 7     7    67 Male      NA    NA    NA       0      13  23.7          1    0.58
 8     8    51 Female    68   120    70       1      15  40.1          0    2.35
 9     9    54 Female   112    81    58       1      15 192.           1    2.68
10    10    53 Female    90   103    60       1      15  66.9          0    0.41
# … with 139 more rows, 12 more variables: sodium <dbl>, inr <dbl>,
#   leukocytes <dbl>, neutrophil <dbl>, lymphocytes <dbl>, albumin <dbl>,
#   bilirubin <dbl>, child <dbl>, meld <dbl>, los <dbl>, death <dbl>,
#   infection <dbl>, and abbreviated variable names ¹encephalopathy,
#   ²creatinine

Components of a Plot

The data for the plot comes from the emergency dataset
The plot contains a x and y coordinate system
- A categorical X describing gender (Male, Female)
- A numeric Y describing the age
The plot contains a color
- The color differs by gender
The plot is being depicted using a dotplot (geometries)

The Grammar of Graphics

The grammar of graphics was originally proposed by Leland Wilkinson
A variation of this framework was further proposed by Hadley Wickham which is used in ggplot2
This framework is a tool that allows us to concisely describe the components of a plot or graphic

The Grammar of Graphics

ggplot(data = df)

The Grammar of Graphics

ggplot(data = df, 
       mapping = aes(x = gender, y = age, fill = gender))

The Grammar of Graphics

ggplot(data = df, 
       mapping = aes(x = gender, y = age, fill = gender)) +
  geom_dotplot(binaxis = 'y', 
               binwidth = 1, 
               stackdir = 'center')

The Grammar of Graphics

ggplot(data = df, 
       mapping = aes(x = gender, y = age, fill = gender)) +
  geom_dotplot(binaxis = 'y', 
               binwidth = 1, 
               stackdir = 'center') +
  labs(x = NULL, 
       y = 'Age')

The Grammar of Graphics

ggplot(data = df,
       mapping = aes(x = gender, y = age, fill = gender)) +
  geom_dotplot(binaxis = 'y',
               binwidth = 1,
               stackdir = 'center') +
  labs(x = NULL, 
       y = 'Age') +
  theme_light(base_size = 20) + 
  theme(legend.position = 'none',
        axis.title = element_text(face = 'bold'))

Another Example

The Grammar of Graphics

ggplot(data = df)

The Grammar of Graphics

ggplot(data = df, 
       mapping = aes(x = gender, y = age, fill = gender))

The Grammar of Graphics

ggplot(data = df,
       mapping = aes(x = gender, y = age, fill = gender)) +
  geom_boxplot(width = .25)

The Grammar of Graphics

ggplot(data = df,
       mapping = aes(x = gender, y = age, fill = gender)) +
  geom_boxplot(width = .25) +
  labs(x = "Gender", y = "Heart Rate")

The Grammar of Graphics

ggplot(data = df,
       mapping = aes(x = gender, y = age, fill = gender)) +
  geom_boxplot(width = .25) +
  labs(x = "Gender", y = "Heart Rate") +
  theme_light(base_size = 15) +
  theme(legend.position = "none",
        axis.title = element_text(face = "bold"))

Resources

Website

Cheatsheet

Book