Connecting artpack Assets to ggplot2 Geoms

Introduction

This vignette goes over different artpack asset creations and their recommended ggplot2 geoms that can be used when creating art in R with artpack and ggplot2. To start, let’s look at a handy chart that can be used as a reference for artpack users.

artpack Assets and Their Recommended ggplot2 geom

artpack Asset Function	Recommended ggplot2 Geoms	grouping variable needed?	Miscellaneous Info
circle_data()	- geom_polygon - geom_path()	Maybe	- set group_var to TRUE if more than one circle is being plotted - Use coord_equal() for best results
grid_maker()	- geom_polygon - geom_path()	Yes	Note that grid_maker() was originally created just for geom_polygon(), but it CAN be used with geom_path(). geom_path() does not have a fill argument. If you wish to create different sizes or colors for the grid squares with geom_path() you’ll need to create unique values for each grouping. You can learn more about this in the later section Making Sense of Grouping in ggplot2
packer()	- geom_polygon - geom_path()	Yes	Note that the data produced by packer() can be fed into geom_polygon() and geom_path() regardless of the circle_type that is set. When circle_type = "swirl", geom_path() is suggested, but geom_polygon() can be used with varying results.
square_data()	- geom_polygon - geom_path()	Maybe	N/A
wave_data()	- geom_polygon - geom_path()	Maybe	- set group_var to TRUE if more than one wave is being plotted - Use coord_equal() for best results

Feeding in the Created artpack Data to ggplot Plots

The focus of artpack is to provide you with crafted data frames that can be easily used within ggplot2 plots. Note that there are different ways to place data into a ggplot:

Via aes()

With I()

# install.packages("artpack")
library(ggplot2)
library(artpack)
set.seed(0515)

df_packed_circles <-
   packer(
   n = 25, big_r = 7, med_r = 3, small_r = 1,
   min_x = 0, max_x = 100, min_y = 0, max_y = 100,
   color_pal = art_pals("rainbow", 15),
   circle_type = "swirl"
 )
#> ⠙ Sampling for big-sized circles started
#> ✔ Sampling for big-sized circles started [92ms]
#> 
#> ℹ Big-sized circles complete!
#> ✔ Big-sized circles complete! [14ms]
#> 
#> ⠙ Sampling for medium-sized circles started
#> ✔ Sampling for medium-sized circles started [137ms]
#> 
#> ℹ Medium-sized circles complete!
#> ✔ Medium-sized circles complete! [13ms]
#> 
#> ⠙ Sampling for small-sized circles started
#> ✔ Sampling for small-sized circles started [84ms]
#> 
#> ℹ Small-sized circles complete!
#> ✔ Small-sized circles complete! [20ms]
#> 
 
 df_packed_circles |>
   ggplot(aes(x, y, group = group, color = I(color))) + # I (base R) takes the color values as-is and passes it through
   theme_void() +
   theme(plot.background = element_rect(fill = "#333333")) +
   geom_path() +
   coord_equal()

With scale_{xx}-type functions

df_packed_circles |>
  ggplot(aes(x, y, group = group, color = color)) +
  theme_void() +
  theme(plot.background = element_rect(fill = "#333333")) +
  scale_color_identity(guide = "none") + # Takes color values as-is and removes plot legend
  geom_path() +
  coord_equal()

Via geom layers

df_packed_circles |>
  ggplot(aes(x, y, group = group)) +
  theme_void() +
  theme(plot.background = element_rect(fill = "#333333")) +
  geom_path(aes(x = x + 1, y = y - .7), color = "#000000") + # Changes made directly in the layer
  geom_path(color = df_packed_circles$color) + # Column from df placed directly in the layer
  coord_equal()

Making Sense of Grouping In ggplot2

One last important thing to remember about using ggplot2 for art, is grouping. Getting comfortable with ggplot grouping is an essential skill for creating generative art. When creating art on a ggplot, you can think of it as drawing on a piece of paper in real life. Think of “grouping” as your way of telling ggplot2 to “pick the pencil up” off the paper”. These grouping variables help ggplot to understand when to “stop drawing” a line for a shape. Let’s look at a quick example using the grid_maker() function:

df_grid <-
  grid_maker(
    xlim = c(0,10),
    ylim = c(0,10),
    size = 10,
    fill_pal = art_pals("sunnyside", 5),
    color_pal = sapply(art_pals("sunnyside", 5), \(x) set_brightness(x, 0.40))
  )

df_grid |>
  ggplot(aes(x,y, group = group)) +
  geom_polygon(
    fill = df_grid$fill,
    color = df_grid$color
  ) +
  theme_void() +
  coord_equal()

With the grouping function, the image looks as expected with nicely colored squares in a grid. If we remove the group argument from aes() in the ggplot() function:

df_grid |>
  ggplot(aes(x,y)) +
  geom_polygon(
    fill = df_grid$fill,
    color = df_grid$color
  ) +
  theme_void() +
  coord_equal()

It turns into a hot mess! This is because without a grouping variable, ggplot doesn’t understand where the data for a square stops and ends. Using groups allows more control over creative decisions and can simplify your workflow by allowing you to create and use multiple variables from a data frame within one geom layer. Remember that artpack has additional group-related functions to help you wrangle your group variables, like the group_sample() function which allows you to sample a data frame by groups (instead of just rows):

set.seed(01234)

df_grid |>
  group_sample(group = group, prop = .70) |>
  ggplot(aes(x,y, group = group, fill = I(fill), color = I(color))) +
  geom_polygon() +
  theme_void() +
  coord_equal()

Wrapping Up

This vignette demonstrated how artpack’s asset creation functions can be integrated within ggplot2’s geom layers.

Remember that:

• Each artpack asset function works best with specific ggplot2 geoms (see reference table)
• Colors can be applied using I(), scale_*_identity(), or directly in geom layers
• Proper grouping is essential for multi-shape plots
• Functions like group_sample() provide additional control over your artistic output

See Also:

• Thinking Outside The Grid - A “bare bones” intro to Rtistry concepts in R using ggplot.
• It’s all about Perspective: “Making A Case for Generative Art”