Graphing with ggplot2

Grayson White

Math 241
Week 2 | Spring 2026

Announcements

Office Hours Schedule is now posted on the course website.
- We will try to stick to this schedule, but if one of us has to reschedule, we will send a Slack message and update the schedule.
P-Set 0 due at 9am on WEDNESDAY (all other p-sets will be due on Thursdays).

Week 2 Goals

Mon Lecture

Basics of ggplot2
Explore several geoms.
And a little data wrangling with dplyr as needed!

Wed Lecture

GitHub workflow overview
Graphing context!
- Labels
- Highlighting
- Useful text
Look at more geoms.
Explore further customizations.
- Color
- Themes
Learn how to ask coding questions well.

Recall: The Grammar of Graphics

data: dataset that contains the data
geom: geometric shape that the data are mapped to
- point, line, bar, text, …
aesthetic: visual properties of the geom
- x position, y position, color, fill, shape
coord: coordinate system
- Cartesian, polar, geographic
scale: controls how data are mapped to the visual values of the aesthetic
- EX: particular colors, linear
guide: legend to help user convert visual display back to the data

ggplot2 example code

ggplot(data = ---, mapping = aes(---)) +
  geom_---(---) + 
  coord_---() + 
  scale_---_---() +
  ---

Example: Over the course of a year, how does the daily number of births vary?

What patterns do you see?

Example

# Load library that has dataset of interest
library(mosaicData)

# Grab data
data(Births2015)

# Load tidyverse (which contains ggplot2)
library(tidyverse)

Example

# Example code
ggplot(data = ---, mapping = aes(---)) +
  geom_---(---) + 
  coord_---() + 
  scale_---_---() +
  ---

# Create plot
ggplot(data = Births2015, 
       mapping = aes(x = date, y = births)) + 
  geom_point()

Example: Cycles related to day of the week?

# Look at structure of data with dplyr function
glimpse(Births2015)

Rows: 365
Columns: 8
$ date         <date> 2015-01-01, 2015-01-02, 2015-01-03, 2015-01-04, 2015-01-…
$ births       <dbl> 8068, 10850, 8328, 7065, 11892, 12425, 12141, 12094, 1186…
$ wday         <ord> Thu, Fri, Sat, Sun, Mon, Tue, Wed, Thu, Fri, Sat, Sun, Mo…
$ year         <dbl> 2015, 2015, 2015, 2015, 2015, 2015, 2015, 2015, 2015, 201…
$ month        <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
$ day_of_year  <int> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17…
$ day_of_month <dbl> 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17…
$ day_of_week  <dbl> 5, 6, 7, 1, 2, 3, 4, 5, 6, 7, 1, 2, 3, 4, 5, 6, 7, 1, 2, …

Example: Cycles related to day of the week?

# Create plot
ggplot(data = Births2015, 
       mapping = aes(x = date, y = births, 
                     color = day_of_week)) + 
  geom_point()

Additional aesthetic:
- Why is this not quite what we want? What do we want?
What happened to the aspect ratio when we added the legend?

Example: Cycles related to day of the week?

# Create plot
ggplot(data = Births2015, 
       mapping = aes(x = date, y = births, 
                     color = wday)) + 
  geom_point() + 
  theme(legend.position = "bottom")

Positioning the legend on the bottom of the graph helped give us a nicer aspect ratio.
What if we want to see the direction that the number of births take over time for each day of the week?
- New visual cue/geom?

Example

# Create plot
ggplot(data = Births2015, 
       mapping = aes(x = date, y = births, 
                     color = wday)) + 
  geom_line() + 
  theme(legend.position = "bottom")

What if we want visual cues for both position and direction?

Example

# Create plot
ggplot(data = Births2015, 
       mapping = aes(x = date, y = births, 
                     color = wday)) + 
  geom_line() +
  geom_point() +
  theme(legend.position = "bottom")

Coordinate System Layer

library(lubridate)

ggplot(data = Births2015, 
       mapping = aes(x = date, y = births, 
                     color = wday)) + 
  geom_line() + 
  geom_point() +
  theme(legend.position = "bottom") +
  coord_cartesian(xlim = 
                    as_date(c("2015-01-01",
                              "2015-01-31")))

How did this new layer change our plot?
What if we want all the points to be colored “midnightblue”?

Setting instead of Mapping

ggplot(data = Births2015, 
       mapping = aes(x = date, y = births, 
                     color = "midnightblue")) + 
  geom_point()

Setting instead of Mapping

ggplot(data = Births2015, 
       mapping = aes(x = date, y = births)) + 
  geom_point(color = "midnightblue")

If you want to set an aesthetic to a specific value (instead of mapping the aesthetic to a variable), do so in the geom_--() function.

Layer order (sometimes) matters

ggplot(data = Births2015, 
       mapping = aes(x = date, y = births,
                     color = wday)) + 
  geom_point(color = "#ff006e") +
  geom_line() + 
  theme(legend.position = "bottom")

Layer order (sometimes) matters

ggplot(data = Births2015, 
       mapping = aes(x = date, y = births,
                     color = wday)) + 
  geom_line(color = "#ff006e") + 
  geom_point() + 
  theme(legend.position = "bottom")

Layer order (sometimes) matters

ggplot(data = Births2015, 
       mapping = aes(x = date, y = births,
                     color = wday)) + 
  geom_line() + 
  geom_point(color = "#ff006e") + 
  theme(legend.position = "bottom")

Inheriting aesthetics discussion.

Let’s explore other geoms

Many are listed on the first page of the ggplot2 cheatsheet.

Can also ask R:

apropos("geom_")

 [1] "geom_abline"            "geom_area"              "geom_bar"              
 [4] "geom_bin_2d"            "geom_bin2d"             "geom_blank"            
 [7] "geom_boxplot"           "geom_col"               "geom_contour"          
[10] "geom_contour_filled"    "geom_count"             "geom_crossbar"         
[13] "geom_curve"             "geom_density"           "geom_density_2d"       
[16] "geom_density_2d_filled" "geom_density2d"         "geom_density2d_filled" 
[19] "geom_dotplot"           "geom_errorbar"          "geom_errorbarh"        
[22] "geom_freqpoly"          "geom_function"          "geom_hex"              
[25] "geom_histogram"         "geom_hline"             "geom_jitter"           
[28] "geom_label"             "geom_line"              "geom_linerange"        
[31] "geom_map"               "geom_path"              "geom_point"            
[34] "geom_pointrange"        "geom_polygon"           "geom_qq"               
[37] "geom_qq_line"           "geom_quantile"          "geom_raster"           
[40] "geom_rect"              "geom_ribbon"            "geom_rug"              
[43] "geom_segment"           "geom_sf"                "geom_sf_label"         
[46] "geom_sf_text"           "geom_smooth"            "geom_spoke"            
[49] "geom_step"              "geom_text"              "geom_tile"             
[52] "geom_violin"            "geom_vline"             "get_geom_defaults"     
[55] "reset_geom_defaults"    "update_geom_defaults"

Adding Curve(s)

ggplot(data = Births2015, 
       mapping = aes(x = date, y = births,
                     color = wday)) + 
  geom_point() + 
  geom_smooth(method = "lm", se = FALSE) + 
  theme(legend.position = "bottom")

Does a multiple linear regression line(s) capture the trend?

Adding Curve(s)

ggplot(data = Births2015, 
       mapping = aes(x = date, y = births,
                     color = wday)) + 
  geom_point() + 
  geom_smooth(se = FALSE) + 
  theme(legend.position = "bottom")

The default LOESS smoother usually does a reasonable job.

Adding Curve(s)

ggplot(data = Births2015, 
       mapping = aes(x = date, y = births,
                     color = wday)) + 
  geom_smooth(color = "black", se = FALSE) +
  geom_point() + 
  theme(legend.position = "bottom")

What happened?
Inheriting aesthetics discussion.

New Example: Movies and the Bechdel Test

Need a new dataset with more categorical variables

The Alison Bechdel Test: A movie passes the test if:
- There are at least two named women in the picture
- They have a conversation with each other at some point
- That conversation isn’t about a male character

Movies from 1970 - 2013

movies <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2021/2021-03-09/movies.csv') %>%
  filter(rated %in% c("R", "PG-13", "PG", "G"))

New Example: Movies and the Bechdel Test

glimpse(movies)

Rows: 1,549
Columns: 34
$ year          <dbl> 2013, 2013, 2013, 2013, 2013, 2013, 2013, 2013, 2013, 20…
$ imdb          <chr> "tt2024544", "tt1272878", "tt0453562", "tt1335975", "tt1…
$ title         <chr> "12 Years a Slave", "2 Guns", "42", "47 Ronin", "A Good …
$ test          <chr> "notalk-disagree", "notalk", "men", "men", "notalk", "ok…
$ clean_test    <chr> "notalk", "notalk", "men", "men", "notalk", "ok", "ok", …
$ binary        <chr> "FAIL", "FAIL", "FAIL", "FAIL", "FAIL", "PASS", "PASS", …
$ budget        <dbl> 2.00e+07, 6.10e+07, 4.00e+07, 2.25e+08, 9.20e+07, 1.20e+…
$ domgross      <chr> "53107035", "75612460", "95020213", "38362475", "6734919…
$ intgross      <chr> "158607035", "132493015", "95020213", "145803842", "3042…
$ code          <chr> "2013FAIL", "2013FAIL", "2013FAIL", "2013FAIL", "2013FAI…
$ budget_2013   <dbl> 2.00e+07, 6.10e+07, 4.00e+07, 2.25e+08, 9.20e+07, 1.20e+…
$ domgross_2013 <chr> "53107035", "75612460", "95020213", "38362475", "6734919…
$ intgross_2013 <chr> "158607035", "132493015", "95020213", "145803842", "3042…
$ period_code   <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,…
$ decade_code   <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,…
$ imdb_id       <chr> "2024544", "1272878", "0453562", "1335975", "1606378", "…
$ plot          <chr> "In the antebellum United States, Solomon Northup, a fre…
$ rated         <chr> "R", "R", "PG-13", "PG-13", "R", "R", "PG-13", "PG-13", …
$ response      <lgl> TRUE, TRUE, TRUE, TRUE, TRUE, TRUE, TRUE, TRUE, TRUE, TR…
$ language      <chr> "English", "English, Spanish", "English", "English, Japa…
$ country       <chr> "USA, UK", "USA", "USA", "USA", "USA", "UK", "USA", "USA…
$ writer        <chr> "John Ridley (screenplay), Solomon Northup (based on \"T…
$ metascore     <dbl> 97, 55, 62, 29, 28, 55, 48, 33, 90, 58, 52, 78, 83, 53, …
$ imdb_rating   <dbl> 8.3, 6.8, 7.6, 6.6, 5.4, 7.8, 5.7, 5.0, 7.5, 7.4, 6.2, 7…
$ director      <chr> "Steve McQueen", "Baltasar Kormákur", "Brian Helgeland",…
$ released      <chr> "08 Nov 2013", "02 Aug 2013", "12 Apr 2013", "25 Dec 201…
$ actors        <chr> "Chiwetel Ejiofor, Dwight Henry, Dickie Gravois, Bryan B…
$ genre         <chr> "Biography, Drama, History", "Action, Comedy, Crime", "B…
$ awards        <chr> "Won 3 Oscars. Another 131 wins & 137 nominations.", "1 …
$ runtime       <chr> "134 min", "109 min", "128 min", "118 min", "98 min", "1…
$ type          <chr> "movie", "movie", "movie", "movie", "movie", "movie", "m…
$ poster        <chr> "http://ia.media-imdb.com/images/M/MV5BMjExMTEzODkyN15BM…
$ imdb_votes    <dbl> 143446, 87301, 43608, 25735, 123837, 85871, 18973, 10826…
$ error         <lgl> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, …

What are useful `geom`s for describing amounts/frequencies?

Amounts: `geom_bar`

ggplot(data = movies,
       mapping = aes(x = binary)) + 
  geom_bar()

Verbalize the mapping of data to geom_bar().
- How is this mapping different from geom_point()?

Another option: `geom_col`

# First wrangle with dplyr
movies_ag <- count(movies, binary) 
movies_ag

# A tibble: 2 × 2
  binary     n
  <chr>  <int>
1 FAIL     863
2 PASS     686

ggplot(data = movies_ag, 
       mapping = aes(x = binary, y = n)) + 
  geom_col()

`geom_point` again

ggplot(data = movies_ag, 
       mapping = aes(x = binary, y = n)) + 
  geom_point(size = 4)

If you are worried about the data-ink ratio…

`geom_point` + `geom_segment`

ggplot(data = movies_ag, 
       mapping = aes(x = binary, y = n)) + 
  geom_segment(mapping = aes(xend = binary), 
               yend = 0) + 
  geom_point(size = 10, color = "orange") +
  ylim(c(0, 875))

Lollipop chart: compromise?

Two categorical variables: `geom_bar`

ggplot(data = movies,
       mapping = aes(x = rated,
                     fill = binary)) + 
  geom_bar()

Describe the mapping.

Two categorical variables: `geom_bar`

ggplot(data = movies,
       mapping = aes(x = rated,
                     fill = binary)) + 
  geom_bar(position = "fill")

Describe the mapping.

Two categorical variables: `geom_bar`

ggplot(data = movies,
       mapping = aes(x = rated,
                     fill = binary)) + 
  geom_bar(position = "dodge")

Describe the mapping.

Two categorical variables: `geom_tile`

movies_ag <- count(movies, rated, binary)

ggplot(data = movies_ag,
       mapping = aes(x = rated,
                     y = binary,
                     fill = n)) + 
  geom_tile()

Describe the mapping.

Two categorical variables: `geom_tile`

movies_ag <- count(movies, rated, binary)

ggplot(data = movies_ag,
       mapping = aes(x = rated,
                     y = binary,
                     fill = n)) + 
  geom_tile() +
  scale_fill_distiller(palette = 4)

Change the fill scale!

Can display more than frequencies!

movies_ag <- group_by(movies,
                      rated, binary) %>%
  summarize(mean_budget = mean(budget))

ggplot(data = movies_ag,
       mapping = aes(x = rated,
                     y = binary,
                     fill = mean_budget)) + 
  geom_tile() +
  scale_fill_viridis_c(direction = -1) +
  theme(legend.position = "bottom")

Can display more than frequencies!

options(scipen = 999) # turn off scientific notation

movies_ag <- group_by(movies,
                      rated, binary) %>%
  summarize(mean_budget = mean(budget))

ggplot(data = movies_ag,
       mapping = aes(x = rated,
                     y = binary,
                     fill = mean_budget)) + 
  geom_tile() +
  scale_fill_viridis_c(direction = -1,
                       guide = guide_colorbar(angle = 90)) +
  theme(legend.position = "bottom")

What are useful `geom`s (graphs) for visualizing distributions?

Distributions: `geom_histogram`

ggplot(movies, aes(x = budget)) +
  geom_histogram()

Describe the mapping.

Distributions: `geom_histogram`

ggplot(movies, aes(x = budget)) +
  geom_histogram(bins = 50, 
                 color = "white",
                 fill = "darkcyan")

Can modify the mapping via the binwidth or bins arguments

Distributions: `geom_histogram`

ggplot(movies, aes(x = budget, 
                   fill = binary)) +
  geom_histogram(bins = 50, 
                 color = "white")

What is problematic about this graph?

Distributions: `geom_histogram`

ggplot(movies, aes(x = budget, 
                   fill = binary)) +
  geom_histogram(bins = 50,
                 alpha = 0.4,
                 position = "identity")

Still problematic.

One option: Faceting

ggplot(movies, aes(x = budget, 
                   fill = binary)) +
  geom_histogram(bins = 50) +
  facet_wrap(~binary) +
  guides(fill = "none")

One option: Faceting

ggplot(movies, aes(x = budget, 
                   fill = binary)) +
  geom_histogram() +
  facet_grid(rated ~ binary, scales = "free_y") +
  guides(fill = "none")

Another option: `geom_density`

ggplot(movies, aes(x = budget, 
                   fill = binary)) +
  geom_density(alpha = 0.4)  +
  theme(legend.position = "bottom")

Distributions: `geom_boxplot`

ggplot(movies, aes(x = binary, 
                   y = budget)) +
  geom_boxplot()

Distributions: `geom_boxplot`

ggplot(movies, aes(x = binary, 
                   y = budget)) +
  geom_boxplot(varwidth = TRUE,
               notch = TRUE)

What does varwidth do?
Why might we add notch = TRUE?

Distributions: `geom_boxplot`

ggplot(movies, aes(x = binary, 
                   y = budget,
                   fill = rated)) +
  geom_boxplot()

Distributions: `geom_violin`

ggplot(movies, aes(x = binary, 
                   y = budget)) +
  geom_violin()

Utility of the violin over the box?

Distributions: `geom_violin`

ggplot(movies, aes(x = binary, 
                   y = budget)) +
  geom_violin() + 
  geom_jitter(alpha = .1,
              width = .1,
              color = "darkcyan")

Reminders

Office Hours Schedule
P-Set 1 released at 9am on Thursday.
- Will discuss how to access the p-sets through GitHub on Wednesday.

Announcements

Week 2 Goals

Recall: The Grammar of Graphics

ggplot2 example code

Example: Over the course of a year, how does the daily number of births vary?

Example

Example

Example: Cycles related to day of the week?

Example: Cycles related to day of the week?

Example: Cycles related to day of the week?

Example

Example

Coordinate System Layer

Setting instead of Mapping

Setting instead of Mapping

Layer order (sometimes) matters

Layer order (sometimes) matters

Layer order (sometimes) matters

Let’s explore other geoms

Adding Curve(s)

Adding Curve(s)

Adding Curve(s)

New Example: Movies and the Bechdel Test

New Example: Movies and the Bechdel Test

What are useful geoms for describing amounts/frequencies?

Amounts: geom_bar

Another option: geom_col

geom_point again

geom_point + geom_segment

Two categorical variables: geom_bar

Two categorical variables: geom_bar

Two categorical variables: geom_bar

Two categorical variables: geom_tile

Two categorical variables: geom_tile

Can display more than frequencies!

Can display more than frequencies!

What are useful geoms (graphs) for visualizing distributions?

Distributions: geom_histogram

Distributions: geom_histogram

Distributions: geom_histogram

Distributions: geom_histogram

One option: Faceting

One option: Faceting

Another option: geom_density

Distributions: geom_boxplot

Distributions: geom_boxplot

Distributions: geom_boxplot

Distributions: geom_violin

Distributions: geom_violin

Reminders

What are useful `geom`s for describing amounts/frequencies?

Amounts: `geom_bar`

Another option: `geom_col`

`geom_point` again

`geom_point` + `geom_segment`

Two categorical variables: `geom_bar`

Two categorical variables: `geom_bar`

Two categorical variables: `geom_bar`

Two categorical variables: `geom_tile`

Two categorical variables: `geom_tile`

What are useful `geom`s (graphs) for visualizing distributions?

Distributions: `geom_histogram`

Distributions: `geom_histogram`

Distributions: `geom_histogram`

Distributions: `geom_histogram`

Another option: `geom_density`

Distributions: `geom_boxplot`

Distributions: `geom_boxplot`

Distributions: `geom_boxplot`

Distributions: `geom_violin`

Distributions: `geom_violin`