Graphic Design with ggplot2

Concepts of the {ggplot2} Package Pt. 2:
Facets, Scales, and Coordinate Systems

Cédric Scherer // rstudio::conf // July 2022

Setup

library(tidyverse)

bikes <- readr::read_csv(
  here::here("data", "london-bikes-custom.csv"),
  col_types = "Dcfffilllddddc"
)

bikes$season <- forcats::fct_inorder(bikes$season)

theme_set(theme_light(base_size = 14, base_family = "Roboto Condensed"))

theme_update(
  panel.grid.minor = element_blank(),
  plot.title = element_text(face = "bold"),
  legend.position = "top",
  plot.title.position = "plot"
)

invisible(Sys.setlocale("LC_TIME", "C"))

Facets

Facets

= split variables to multiple panels

Facets are also known as:

• small multiples
• trellis graphs
• lattice plots
• conditioning

Setup

g <-
  ggplot(
    bikes,
    aes(x = temp_feel, y = count,
        color = season)
  ) +
  geom_point(
    alpha = .3,
    guide = "none"
  )

g

Wrapped Facets

g +
  facet_wrap(
    vars(day_night)
  )

Wrapped Facets

g +
  facet_wrap(
    ~ day_night
  )

Facet Multiple Variables

g +
  facet_wrap(
    ~ is_workday + day_night
  )

Facet Options: Cols + Rows

g +
  facet_wrap(
    ~ day_night,
    ncol = 1
  )

Facet Options: Free Scaling

g +
  facet_wrap(
    ~ day_night,
    ncol = 1,
    scales = "free"
  )

Facet Options: Free Scaling

g +
  facet_wrap(
    ~ day_night,
    ncol = 1,
    scales = "free_y"
  )

Facet Options: Switch Labels

g +
  facet_wrap(
    ~ day_night,
    ncol = 1,
    switch = "x"
  )

Gridded Facets

g +
  facet_grid(
    rows = vars(day_night),
    cols = vars(is_workday)
  )

Gridded Facets

g +
  facet_grid(
    day_night ~ is_workday
  )

Facet Multiple Variables

g +
  facet_grid(
    day_night ~ is_workday + season
  )

Facet Options: Free Scaling

g +
  facet_grid(
    day_night ~ is_workday,
    scales = "free"
  )

Facet Options: Switch Labels

g +
  facet_grid(
    day_night ~ is_workday,
    scales = "free",
    switch = "y"
  )

Facet Options: Proportional Spacing

g +
  facet_grid(
    day_night ~ is_workday,
    scales = "free",
    space = "free"
  )

Facet Options: Proportional Spacing

g +
  facet_grid(
    day_night ~ is_workday,
    scales = "free_y",
    space = "free_y"
  )

Your Turn!

Create the following facet from the diamonds data.

Your Turn!

Bonus: Create this bloody-dark version.

Diamonds Facet

ggplot(
    diamonds,
    aes(x = carat, y = price)
  ) +
  geom_point(
    alpha = .3
  ) +
  geom_smooth(
    method = "lm",
    se = FALSE,
    color = "dodgerblue"
  )

Diamonds Facet

ggplot(
    diamonds,
    aes(x = carat, y = price)
  ) +
  geom_point(
    alpha = .3
  ) +
  geom_smooth(
    method = "lm",
    se = FALSE,
    color = "dodgerblue"
  ) +
  facet_grid(
    cut ~ clarity,
    space = "free_x",
    scales = "free_x"
  )

Diamonds Facet (Dark Theme Bonus)

ggplot(
    diamonds,
    aes(x = carat, y = price)
  ) +
  geom_point(
    alpha = .3,
    color = "white"
  ) +
  geom_smooth(
    method = "lm",
    se = FALSE,
    color = "dodgerblue"
  ) +
  facet_grid(
    cut ~ clarity,
    space = "free_x",
    scales = "free_x"
  ) +
  theme_dark(
    base_size = 14,
    base_family = "Roboto Condensed"
  )

Scales

Scales

= translate between variable ranges and property ranges

• feels-like temperature  ⇄  x
• reported bike shares  ⇄  y
• season  ⇄  color
• year  ⇄  shape
• …

Scales

The scale_*() components control the properties of all the
aesthetic dimensions mapped to the data.

Consequently, there are scale_*() functions for all aesthetics such as:

• positions via scale_x_*() and scale_y_*()

• colors via scale_color_*() and scale_fill_*()

• sizes via scale_size_*() and scale_radius_*()

• shapes via scale_shape_*() and scale_linetype_*()

• transparency via scale_alpha_*()

Scales

The scale_*() components control the properties of all the
aesthetic dimensions mapped to the data.

The extensions (*) can be filled by e.g.:

• continuous(), discrete(), reverse(), log10(), sqrt(), date() for positions

• continuous(), discrete(), manual(), gradient(), gradient2(), brewer() for colors

• continuous(), discrete(), manual(), ordinal(), area(), date() for sizes

• continuous(), discrete(), manual(), ordinal() for shapes

• continuous(), discrete(), manual(), ordinal(), date() for transparency

Illustration by Allison Horst

Continuous vs. Discrete in {ggplot2}

Continuous:
quantitative or numerical data

• height
• weight
• age
• counts

Discrete:
qualitative or categorical data

• species
• sex
• study sites
• age group

Continuous vs. Discrete in {ggplot2}

Continuous:
quantitative or numerical data

• height (continuous)
• weight (continuous)
• age (continuous or discrete)
• counts (discrete)

Discrete:
qualitative or categorical data

• species (nominal)
• sex (nominal)
• study site (nominal or ordinal)
• age group (ordinal)

Aesthetics + Scales

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point()

Aesthetics + Scales

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_x_date() +
  scale_y_continuous() +
  scale_color_discrete()

Scales

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_x_continuous() +
  scale_y_continuous() +
  scale_color_discrete()

Scales

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_x_continuous() +
  scale_y_log10() +
  scale_color_discrete()

Scales

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_x_continuous() +
  scale_y_log10() +
  scale_color_viridis_d()

`scale_x|y_continuous`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +  
  geom_point() +
  scale_y_continuous(
    trans = "log10"
  )

`scale_x|y_continuous`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_y_continuous(
    name = "Reported bike shares"
  ) 

`scale_x|y_continuous`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_y_continuous(
    name = "Reported bike shares",
    breaks = seq(0, 60000, by = 15000)
  ) 

`scale_x|y_continuous`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_y_continuous(
    name = "Reported bike shares",
    breaks = 0:4*15000
  )

`scale_x|y_continuous`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_y_continuous(
    name = "Reported bike shares",
    breaks = c(0, 2:12*2500, 40000, 50000)
  ) 

`scale_x|y_continuous`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_y_continuous(
    name = "Reported bike shares in thousands",
    breaks = 0:4*15000,
    labels = 0:4*15
  ) 

`scale_x|y_continuous`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_y_continuous(
    name = "Reported bike shares in thousands",
    breaks = 0:4*15000,
    labels = paste(0:4*15000, "bikes")
  ) 

`scale_x|y_continuous`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_y_continuous(
    name = "Reported bike shares",
    breaks = 0:4*15000,
    limits = c(NA, 60000)
  ) 

`scale_x|y_continuous`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_y_continuous(
    name = "Reported bike shares",
    breaks = 0:4*15000,
    expand = c(0, 0)
  ) 

`scale_x|y_continuous`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_y_continuous(
    name = "Reported bike shares",
    breaks = -1:5*15000,
    expand = c(.5, .5) ## c(add, mult)
  ) 

`scale_x|y_continuous`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_y_continuous(
    name = "Reported bike shares",
    breaks = -1:5*15000,
    expand = expansion(add = 2000)
  ) 

`scale_x|y_continuous`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_y_continuous(
    name = "Reported bike shares",
    breaks = 0:4*15000,
    guide = "none"
  )

`scale_x|y_date`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_x_date(
    name = NULL,
    date_breaks = "4 months"
  )

`scale_x|y_date`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_x_date(
    name = NULL,
    date_breaks = "20 weeks"
  )

`scale_x|y_date` with `strftime()`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_x_date(
    name = NULL,
    date_breaks = "6 months",
    date_labels = "%Y/%m/%d"
  )

`scale_x|y_date` with `strftime()`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_x_date(
    name = NULL,
    date_breaks = "6 months",
    date_labels = "%b '%y"
  )

`scale_x|y_discrete`

ggplot(
    bikes,
    aes(x = season, y = count)
  ) +
  geom_boxplot() +
  scale_x_discrete(
    name = "Period",
    labels = c("Dec-Feb", "Mar-May", "Jun-Aug", "Sep-Nov")
  )

`scale_x|y_discrete`

ggplot(
    bikes,
    aes(x = season, y = count)
  ) +
  geom_boxplot() +
  scale_x_discrete(
    name = "Season",
    expand = c(.5, 0) ## add, mult
  )

Discrete or Continuous?

ggplot(
    bikes,
    aes(x = as.numeric(season), y = count)
  ) +
  geom_boxplot(
    aes(group = season)
  )

Discrete or Continuous?

ggplot(
    bikes,
    aes(x = as.numeric(season),
        y = count)
  ) +
  geom_boxplot(
    aes(group = season)
  ) +
  scale_x_continuous(
    name = "Season",
    breaks = 1:4,
    labels = levels(bikes$season)
  )

Discrete or Continuous?

ggplot(
    bikes,
    aes(x = as.numeric(season) + 
            as.numeric(season) / 8,
        y = count)
  ) +
  geom_boxplot(
    aes(group = season)
  ) +
  scale_x_continuous(
    name = "Season",
    breaks = 1:4,
    labels = levels(bikes$season)
  )

`scale_color|fill_discrete`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_color_discrete(
    name = "Season:",
    type = c("#69b0d4", "#00CB79", "#F7B01B", "#a78f5f")
  )

Inspect Assigned Colors

g <- ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_color_discrete(
    name = "Season:",
    type = c("#3ca7d9", "#1ec99b", "#F7B01B", "#bb7e8f")
  )

gb <- ggplot_build(g)

gb$data[[1]][c(1:5, 200:205, 400:405), 1:5]

     colour     x     y PANEL group
1   #3ca7d9 16439  6830     1     1
2   #3ca7d9 16439  2404     1     1
3   #3ca7d9 16440 14763     1     1
4   #3ca7d9 16440  5609     1     1
5   #3ca7d9 16441 14501     1     1
200 #1ec99b 16538  8830     1     2
201 #1ec99b 16539 24019     1     2
202 #1ec99b 16539 10500     1     2
203 #1ec99b 16540 25640     1     2
204 #1ec99b 16540 11830     1     2
205 #1ec99b 16541 22216     1     2
400 #F7B01B 16638 12079     1     3
401 #F7B01B 16639 26646     1     3
402 #F7B01B 16639 12446     1     3
403 #F7B01B 16640 11312     1     3
404 #F7B01B 16640  4722     1     3
405 #F7B01B 16641 22748     1     3

`scale_color|fill_discrete`

my_colors <- c(
  `winter` = "#3c89d9",
  `spring` = "#1ec99b",
  `summer` = "#F7B01B",
  `autumn` = "#a26e7c"
)

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_color_discrete(
    name = "Season:",
    type = my_colors
  )

`scale_color|fill_discrete`

my_colors_alphabetical <- c(
  `autumn` = "#a26e7c",
  `spring` = "#1ec99b",
  `summer` = "#F7B01B",
  `winter` = "#3c89d9"
)

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_color_discrete(
    name = "Season:",
    type = my_colors_alphabetical
  )

`scale_color|fill_discrete`

library(RColorBrewer)

ggplot(
    bikes,
    aes(x = date, y = count,
        color = season)
  ) +
  geom_point() +
  scale_color_discrete(
    name = "Season:",
    type = brewer.pal(
      n = 4, name = "Dark2"
    )
  )

`scale_color|fill_manual`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = weather_type)
  ) +
  geom_point() +
  scale_color_manual(
    name = "Season:",
    values = brewer.pal(n = 6, name = "Pastel1"),
    na.value = "black"
  )

`scale_color|fill_carto_d`

ggplot(
    bikes,
    aes(x = date, y = count,
        color = weather_type)
  ) +
  geom_point() +
  rcartocolor::scale_color_carto_d(
    name = "Season:",
    palette = "Pastel",
    na.value = "black"
  )

Your Turn!

Modify our diamonds facet like this:

Diamonds Facet

facet <-
  ggplot(
    diamonds,
    aes(x = carat, y = price)
  ) +
  geom_point(
    alpha = .3
  ) +
  geom_smooth(
    aes(color = cut),
    method = "lm",
    se = FALSE
  ) +
  facet_grid(
    cut ~ clarity,
    space = "free_x",
    scales = "free_x"
  )

facet

Diamonds Facet

facet +
  scale_x_continuous(
    breaks = 0:5
  ) +
  scale_y_continuous(
    limits = c(0, 30000),
    breaks = 0:3*10000,
    labels = c("$0", "$10,000", "$20,000", "$30,000")
  )

Diamonds Facet

facet +
  scale_x_continuous(
    breaks = 0:5
  ) +
  scale_y_continuous(
    limits = c(0, 30000),
    breaks = 0:3*10000,
    labels = paste0(
      "$", format(
        0:3*10000, 
        big.mark = ",", 
        trim = TRUE
      )
    )
  )

Diamonds Facet

facet +
  scale_x_continuous(
    breaks = 0:5
  ) +
  scale_y_continuous(
    limits = c(0, 30000),
    breaks = 0:3*10000,
    labels = function(y) paste0(
      "$", format(
        y, big.mark = ",",
        trim = TRUE
      )
    )
  )

Diamonds Facet

facet +
  scale_x_continuous(
    breaks = 0:5
  ) +
  scale_y_continuous(
    limits = c(0, 30000),
    breaks = 0:3*10000,
    labels = scales::dollar_format()
  )

Diamonds Facet

facet +
  scale_x_continuous(
    breaks = 0:5
  ) +
  scale_y_continuous(
    limits = c(0, 30000),
    breaks = 0:3*10000,
    labels = scales::dollar_format()
  ) +
  scale_color_brewer(
    palette = "Set2",
    guide = "none"
  )

Diamonds Facet

facet +
  scale_x_continuous(
    breaks = 0:5
  ) +
  scale_y_continuous(
    limits = c(0, 30000),
    breaks = 0:3*10000,
    labels = scales::dollar_format()
  ) +
  scale_color_brewer(
    palette = "Set2"
  ) +
  theme(
    legend.position = "none"
  )

Coordinate Systems

Coordinate Systems

= interpret the position aesthetics

• linear coordinate systems: preserve the geometrical shapes
  • coord_cartesian()
  • coord_fixed()
  • coord_flip()
• non-linear coordinate systems: likely change the geometrical shapes
  • coord_polar()
  • coord_map() and coord_sf()
  • coord_trans()

Cartesian Coordinate System

ggplot(
    bikes,
    aes(x = season, y = count)
  ) +
  geom_boxplot() +
  coord_cartesian()

Cartesian Coordinate System

ggplot(
    bikes,
    aes(x = season, y = count)
  ) +
  geom_boxplot() +
  coord_cartesian(
    ylim = c(NA, 15000)
  )

Changing Limits

ggplot(
    bikes,
    aes(x = season, y = count)
  ) +
  geom_boxplot() +
  coord_cartesian(
    ylim = c(NA, 15000)
  )

ggplot(
    bikes,
    aes(x = season, y = count)
  ) +
  geom_boxplot() +
  scale_y_continuous(
    limits = c(NA, 15000)
  )

Clipping

ggplot(
    bikes,
    aes(x = season, y = count)
  ) +
  geom_boxplot() +
  coord_cartesian(
    ylim = c(NA, 15000),
    clip = "off"
  )

Clipping

ggplot(
    filter(bikes, is_holiday == TRUE),
    aes(x = temp_feel, y = count)
  ) +
  geom_point() +
  geom_text(
    aes(label = season),
    nudge_x = .3,
    hjust = 0
  ) +
  coord_cartesian(
    clip = "off"
  )

… or better use {ggrepel}

ggplot(
    filter(bikes, is_holiday == TRUE),
    aes(x = temp_feel, y = count)
  ) +
  geom_point() +
  ggrepel::geom_text_repel(
    aes(label = season),
    nudge_x = .3,
    hjust = 0
  ) +
  coord_cartesian(
    clip = "off"
  )

Remove All Padding

ggplot(
    bikes,
    aes(x = temp_feel, y = count)
  ) +
  geom_point() +
  coord_cartesian(
    expand = FALSE,
    clip = "off"
  )

Fixed Coordinate System

ggplot(
    bikes,
    aes(x = temp_feel, y = temp)
  ) +
  geom_point() +
  coord_fixed()

ggplot(
    bikes,
    aes(x = temp_feel, y = temp)
  ) +
  geom_point() +
  coord_fixed(ratio = 4)

Flipped Coordinate System

ggplot(
    bikes,
    aes(x = weather_type)
  ) +
  geom_bar() +
  coord_cartesian()

ggplot(
    bikes,
    aes(x = weather_type)
  ) +
  geom_bar() +
  coord_flip()

Flipped Coordinate System

ggplot(
    bikes,
    aes(y = weather_type)
  ) +
  geom_bar() +
  coord_cartesian()

ggplot(
    bikes,
    aes(x = weather_type)
  ) +
  geom_bar() +
  coord_flip()

Reminder: Sort Your Bars!

ggplot(
    filter(bikes, !is.na(weather_type)),
    aes(y = fct_infreq(weather_type))
  ) +
  geom_bar()

Reminder: Sort Your Bars!

ggplot(
    filter(bikes, !is.na(weather_type)),
    aes(y = fct_rev(
      fct_infreq(weather_type)
    ))
  ) +
  geom_bar()

Circular Corrdinate System

ggplot(
    filter(bikes, !is.na(weather_type)),
    aes(x = weather_type,
        fill = weather_type)
  ) +
  geom_bar() +
  coord_polar()

ggplot(
    filter(bikes, !is.na(weather_type)),
    aes(x = weather_type,
        fill = weather_type)
  ) +
  geom_bar() +
  coord_cartesian()

Circular Cordinate System

ggplot(
    filter(bikes, !is.na(weather_type)),
    aes(x = fct_infreq(weather_type),
        fill = weather_type)
  ) +
  geom_bar(width = 1) +
  coord_polar()

ggplot(
    filter(bikes, !is.na(weather_type)),
    aes(x = fct_infreq(weather_type),
        fill = weather_type)
  ) +
  geom_bar(width = 1) +
  coord_cartesian()

Circular Corrdinate System

ggplot(
    filter(bikes, !is.na(weather_type)),
    aes(x = fct_infreq(weather_type),
        fill = weather_type)
  ) +
  geom_bar() +
  coord_polar(theta = "x")

ggplot(
    filter(bikes, !is.na(weather_type)),
    aes(x = fct_infreq(weather_type),
        fill = weather_type)
  ) +
  geom_bar() +
  coord_polar(theta = "y")

Circular Corrdinate System

ggplot(
    filter(bikes, !is.na(weather_type)),
    aes(x = 1, fill = weather_type)
  ) +
  geom_bar(position = "stack") +
  coord_polar(theta = "y") 

ggplot(
    filter(bikes, !is.na(weather_type)),
    aes(x = 1, fill = weather_type)
  ) +
  geom_bar(position = "stack") +
  coord_cartesian() 

Circular Corrdinate System

ggplot(
    filter(bikes, !is.na(weather_type)),
    aes(x = 1,
        fill = fct_rev(fct_infreq(weather_type)))
  ) +
  geom_bar(position = "stack") +
  coord_polar(theta = "y") +
  scale_fill_discrete(name = NULL)

ggplot(
    filter(bikes, !is.na(weather_type)),
    aes(x = 1,
        fill = fct_rev(fct_infreq(weather_type)))
  ) +
  geom_bar(position = "stack") +
  coord_cartesian() +
  scale_fill_discrete(name = NULL)

Transform a Coordinate System

ggplot(
    bikes,
    aes(x = temp, y = count)
  ) +
  geom_point() +
  coord_trans(y = "log10")

Transform a Coordinate System

ggplot(
    bikes,
    aes(x = temp, y = count,
        group = day_night)
  ) +
  geom_point() +
  geom_smooth(method = "lm") +
  coord_trans(y = "log10")

ggplot(
    bikes,
    aes(x = temp, y = count,
        group = day_night)
  ) +
  geom_point() +
  geom_smooth(method = "lm") +
  scale_y_log10()

Recap

• facet_*() functions allow to create small multiples
• scale_*() functions translate between aesthetics and properties
• use *_continuous() for numerical data
• … and *_discrete() for categorical data
• use scale_color|fill_*() to customize data-related colors
• coord_*() functions interpret the positional aesthetics
  • e.g. create polar plots or work with spatial projections
• be careful when adjusting axis limits:
  • scale_*_continuous(limits) removes the data outside the range
  • coord_*(*lim) keeps the data and zooms into the range

Exercises

Exercise 1

• Have a look at the following visualization of the cumulative time that cosmo- and astronauts have spent in outer space. The data also contains information on the year of their first and last travel, respectively.
• Together with your group, discuss which layers and modifications are needed to create such a chart with {ggplot2}.
• Note down the aesthetics, geometries, and scales used for each element of this graphic.
• What is the coordinate system? Have any adjustments been made?
• Which theme was used and how was it modified?

Exercise 2

• Create a circular lollipop plot of reported bike shares per season.
• The data is not in the right format as there are no totals.
  How can you solve it?
• Remove all legend elements (with a single line of code).
• How can you add the labels next to the starting point of each lollipop?
• How could you add a baseline?

Appendix

Spatial Coordinate (Reference) Systems

countries <- rnaturalearth::ne_countries(
  returnclass = "sf"
)

ggplot() +
  geom_sf(
    data = countries,
    color = "#79dfbd",
    fill = "#28a87d",
    size = .3
  )

Spatial Coordinate (Reference) Systems

ggplot() +
  geom_sf(
    data = countries,
    color = "#79dfbd",
    fill = "#28a87d",
    size = .3
  ) +
  coord_sf(
    crs = "+proj=moll"
  )

Spatial Coordinate (Reference) Systems

ggplot() +
  geom_sf(
    data = countries,
    color = "#79dfbd",
    fill = "#28a87d",
    size = .3
  ) +
  coord_sf(
    crs = "+proj=bonne +lat_1=10"
  )

Spatial Coordinate (Reference) Systems

oceans <- rnaturalearth::ne_download(
  category = "physical", type = "ocean", returnclass = "sf"
)

OGR data source with driver: ESRI Shapefile 
Source: "C:\Users\DataVizard\AppData\Local\Temp\RtmpQHth9z", layer: "ne_110m_ocean"
with 2 features
It has 3 fields

Spatial Coordinate (Reference) Systems

ggplot() +
  geom_sf(
    data = oceans,
    fill = "#d8f1f6"
  ) +
  geom_sf(
    data = countries,
    color = "#79dfbd",
    fill = "#28a87d",
    size = .3
  ) +
  coord_sf(
    crs = "+proj=bonne +lat_1=10"
  ) +
  theme_void()

Mapping of Visual Properties

ggplot() +
  geom_sf(
    data = oceans,
    fill = "#d8f1f6",
    color = "white"
  ) +
  geom_sf(
    data = countries,
    aes(fill = economy),
    color = "white",
    size = .3
  ) +
  coord_sf(
    crs = "+proj=bonne +lat_1=10"
  ) +
  scale_fill_viridis_d(option = "magma") +
  theme_void() +
  theme(legend.position = "top")

Better Borders

borders <- rmapshaper::ms_innerlines(countries)

ggplot() +
  geom_sf(
    data = oceans,
    fill = "#d8f1f6",
    color = "white"
  ) +
  geom_sf(
    data = countries,
    aes(fill = economy),
    color = "transparent"
  ) +
  geom_sf(
    data = borders,
    fill = "transparent",
    color = "white",
    size = .3
  ) +
  coord_sf(
    crs = "+proj=bonne +lat_1=10"
  ) +
  scale_fill_viridis_d(option = "magma") +
  theme_void() +
  theme(legend.position = "top")