• 2. ggplot summary

Links to: Summary, Chatbot Tutor, Practice Questions, Glossary, R functions and R packages, and Additional resources + BONUS CONTENT: ggplotly.

Alternative formats: 🎥 Watch  ·  🎧 Listen

Chapter Summary

Solid data visualization starts with clear biological questions and curiosity about their potential answers. Before beginning coding, consider what you want to learn from your data and how best to visualize that information. After sketching out soe ideas, you are ready to use ggplot to bring these ideas to paper (or maybe your computer screen). ggplot is built around the idea that plots are constructed by layering components: you begin by mapping variables to aesthetic properties like position, color, or size, and then choose how to display those mapped variables using geometric elements like histograms, barplots, or points. With this approach, you can iteratively build visualizations that reflect your questions and highlight meaningful patterns.

Chatbot tutor

Please interact with this custom chatbot (ChatGPT link here, Gemini link here) I have made to help you with this chapter. I suggest interacting with at least ten back-and-forths to ramp up and then stopping when you feel like you got what you needed from it.

Practice Questions

Try the questions below. The R environment below will allow you to work without changing tabs.

Q1) How would you modify the code above to make a scatterplot to evaluate the relationship between petal area and anther stigma distance? 
ggplot(ril_data, aes(x = petal_area_mm , 
                     y = asd_mm))+
  geom_point()

Q2) After fixing the code above, add a best fit trendline. What pattern do you see? 
ggplot(ril_data, aes(x = petal_area_mm , y = asd_mm))+
  geom_point()+
  geom_smooth(method = "lm")
`geom_smooth()` using formula = 'y ~ x'

Q3) Now add petal color to the mix, what do you see (choose the best answer)?

The slopes may be a bit different, but it is more obvious that the red line is higher than the turquoise.

ggplot(ril_data, aes(x = petal_area_mm , 
                     y = asd_mm, 
                     color = petal_color))+
  geom_point()+
  geom_smooth(method = "lm")
`geom_smooth()` using formula = 'y ~ x'

Q4 Make a histogram of asd_mm with a binwidth of 0.25. What is the most common anther stigma distance (asd_mm) with this binning? 
ggplot(ril_data, aes(x = asd_mm))+
  geom_histogram(binwidth = .25, color = "white")

Q5 Consider the code below, what color are the bars? 
ggplot(ril_data, aes(x = asd_mm, color = "green"))+
    geom_histogram()

As you can see the bar’s color is the standard dark grey. The lines around the bars are reddish. This is because we mapped the variable “green” onto color. Instead keep it out of an aes() call. Also be sure to note the difference between color and fill arguments.

ggplot(ril_data, aes(x = asd_mm))+
    geom_histogram(color = "green")
`stat_bin()` using `bins = 30`. Pick better value `binwidth`.

ggplot(ril_data, aes(x = asd_mm))+
    geom_histogram(fill = "green")
`stat_bin()` using `bins = 30`. Pick better value `binwidth`.

Q6 Which code will make a histogram with green bars? 
#A 
ggplot(ril_data, aes(x = asd_mm, fill = "green"))+
    geom_histogram()

#B
ggplot(ril_data, aes(x = asd_mm))+
    geom_histogram(color = "green")

#C
ggplot(ril_data, aes(x = asd_mm))+
    geom_histogram(fill = "green")

Glossary of Terms

  • Aesthetic Mappings (aes()): Defines how variables map onto plot elements (e.g., x/y position, color, size).
  • Geometric Objects (geom_*()): Defines how data is represented (e.g., geom_point() for scatterplots, geom_bar() for bar plots).
  • Faceting (facet_wrap() and facet_grid()): Divides plots into multiple panels based on categorical variables.
  • Overplotting: When multiple data points overlap, making them hard to see. Can be addressed with alpha, geom_jitter(), or geom_violin().
  • Small Multiples: A visualization technique where multiple small plots show variations across categories (e.g., faceting by location).

Key R functions

R Packages Introduced

  • ggplot2: The core package for data visualization in the tidyverse.

  • plotly: Turns a ggplot into an interactive, hoverable/zoomable plot via ggplotly().

Additional resources

R Recipes:

Other web resources:

Videos:

Bonus content: Interactive plots with ggplotly

Data often contain strange outliers, ambiguous patterns, or otherwise interesting individual points. When I run into these issues during exploratory data analysis I often want to know more about individual data points. To do so, I make interactive graphs with the ggplotly() function in the plotly package.

The example below shows how to do this. Note that you can make up random aesthetics that you never use and they show up when you hover over points – this helps with understanding outliers. You can also zoom in!

library(plotly)
big_plot <- ril_data |>
  filter(!is.na(petal_color))|>
  ggplot(aes(x = petal_area_mm, 
             y = prop_hybrid,
             ril = ril,
             mean_visits =  mean_visits))+ 
  geom_point(size = 3, alpha = .4)+
  facet_grid(petal_color ~ location, labeller = "label_both")+
  geom_smooth(method = "lm", se = FALSE)

ggplotly(big_plot)