Getting to know ggplot2

What is ggplot2?

ggplot2 is a way to declaratively create graphics, meaning that you control all aspects of how you want your plot to look. ggplot2 takes in data frames, you provide the aesthetics and the details of what you’d like to see on your plot.

Structure of a ggplot

• Data

  • What information do we want to show from the data collected? Can we determine if there is a connection between two variables?

    • Anu example: relationship between light recipe and plant weight.
• Geometry
  • What is the best way to communicate what the data is showing us?
    • Scatter plot? Bar plot? Line plot? Other?
• Aesthetics

  • How do we want the plot to look?

    • Theme, colors, shapes, error bars.

Build a simple plot - scatter plot

# Install tidyverse package, or alternatively just install ggplot2
# install.packages("tidyverse")
# Install palmerpenguins to get access to penguins dataset
# install.packages("palmerpenguins")
# install.packages("shinylive")

# Load library
library(ggplot2) 
library(palmerpenguins)
# Load palmer penguins data
# data(package = 'palmerpenguins')

# Read in your df
df <- penguins

# Aesthetic mappings describe how variables in the data are mapped to visual properties (aesthetics) of geoms. Aesthetic mappings can be set in ggplot() and in individual layers.

# Basic plot
p_simple <- ggplot(df, aes(x = body_mass_g, y = bill_length_mm, color = species)) +
  # Include a "+" symbol at the end of each line when adding additional components
  geom_point()
p_simple

How do we build on this?

# Additional features to make the plot look prettier
# Change the shapes of the dots of each species
p <- ggplot(df, aes(x = body_mass_g, y = bill_length_mm, color = species, shape = species)) + 
  geom_point() +
  # Add trend line, with method = linear regression (lm = linear model)
  geom_smooth(method = lm, se=FALSE) +
  # Add a title
  ggtitle(paste0("Body Mass (g) vs. Bill Length (mm)")) +
  # Increase the number of x-axis ticks
  scale_x_continuous(breaks = round(seq(min(df$body_mass_g, na.rm = TRUE), max(df$body_mass_g, na.rm = TRUE), by = 500),1)) +
  # Increase the number of y-axis ticks
  scale_y_continuous(breaks = round(seq(min(df$bill_length_mm, na.rm = TRUE), max(df$bill_length_mm, na.rm = TRUE), by = 5),1)) +
  # Makes background white, looks a little cleaner
  theme_bw()

p

Other ways to visualize this data

# Facet data by species
facet_plot <- p + 
  # Facet data based on penguin species
  # Scales options are: "fixed", "free", "free_x", "free_y" - fixed = default
  facet_wrap(facets = df$species, scales = "fixed") +
  # Format the text at the bottom of the plots so that it's at a 45 degree angle
  theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust = 1))
facet_plot

Other plot types

Bar plot

# Want to plot average size of each variety of penguin in a specific year - bar plot
library(Rmisc) 
library(magrittr)
library(ggplot2)
# Need to summarize data
summ_data <- summarySE(data = df, measurevar = "body_mass_g", groupvars = c("species", "year", "sex"))

# Filter data to only include data from 2007 for female penguins
filtered_data <- summ_data %>% 
  dplyr::filter(year == "2007" & sex == "female")

ggplot(filtered_data, aes(x = species, y= body_mass_g, fill = species)) +
  geom_bar(stat="identity", color="black", position=position_dodge(), size = 0.5, width = 0.5) +
  # Add error bars
  geom_errorbar(aes(ymin=body_mass_g - se, ymax=body_mass_g + se), width=.1) +
  theme_bw()

# geom_bar() notes
  # Stat = identity tells R to use the provided y values directly, instead of doing other statistics
  # position = position_dodge() adjusts the placement of the bars on a bar plot
  # Specify size and width of the bars in the bar plot

# geom_errorbar() notes
  # calculate ymin and ymax
  # position_dodge() - same as geom_bar
  # set width of error bars

ggplot2 in R Shiny

#| standalone: true
#| viewerHeight: 600
library(shiny)
library(bslib)
library(palmerpenguins)
library(ggplot2)


# Define UI for app that draws a histogram ----
ui <- page_sidebar(
  sidebar = sidebar(open = "open",
    #selectInput(inputId = "select_year", label = "Select Year", choices = c(choices_year)),
  selectInput(inputId = "select_year", label = "Select Year", choices = c(2007, 2008, 2009)),
    selectInput(inputId = "sex", label = "Select Sex", choices = c("female", "male"))
  ),
  plotOutput("plot")
)

server <- function(input, output, session) {
  reactive_data <- reactive({
    df <- palmerpenguins::penguins
    # Summarize data
    sum_data <- Rmisc::summarySE(df, measurevar = "body_mass_g", groupvars = c("species", "year", "sex"), na.rm = TRUE)
    # sum_data$sex <- as.character(sum_data$sex)
    # Filter data
    dplyr::filter(sum_data, year == input$select_year & sex == input$sex)
    
    # ggplot(new_data, aes(x = species, y= body_mass_g, fill = species)) +
    # geom_bar(stat="identity", color="black", position=position_dodge(), size = 0.5, width = 0.5) +
    # # Add error bars
    # geom_errorbar(aes(ymin=body_mass_g - se, ymax=body_mass_g + se), width=.1) +
    # theme_bw()
  })
  
  output$plot <- renderPlot({
  ggplot(reactive_data(), aes(x = species, y= body_mass_g, fill = species)) +
    geom_bar(stat="identity", color="black", position=position_dodge(), size = 0.5, linewidth = 0.5) +
  #   # Add error bars
    geom_errorbar(aes(ymin=body_mass_g - se, ymax=body_mass_g + se), width=.1) +
    theme_bw()
    # reactive_data()
  })
}

# Create Shiny app ----
shinyApp(ui = ui, server = server)

ggplotly

• Gives the ability to turn a ggplot object to a plotly object

• Allows for easy viewing of data points in a congested plot

library(plotly)
ggplotly(facet_plot)

Resources

• ggplot2
• ggplotly