2 Visualization

We start with visualization because, well, you can see the results.

2.1 Reading

To design good visuals, you need both whys and hows. You may have come here for the hows, but both are important. Our tools are changing more rapidly than ever, so if we want knowledge that lasts, we really need to know the why.

2.1.1 Why

Read Look at Data from Healy “Data Visualization”.

The text is wordy but well organized, so your speed reading skills should work well. Look at the examples: can you explain to someone else what those examples show?

2.1.2 How

Read Data Visualization from ModernDive.

Try to actually answer the “Learning Check” questions for yourself. Yes this takes longer than just skimming right past them. But they may show up on a quiz…

2.2 References

2.2.1 Visualization Design

2.2.2 Implementation

2.3 Tweaks

2.3.1 Reordering bars in a bar plot

Use fct_reorder on the categorical variable.

starwars %>% 
  drop_na(height) %>% 
  ggplot(aes(x = height, y = species)) +
  geom_boxplot()

starwars %>% 
  drop_na(height) %>% 
  ggplot(aes(x = height, y = fct_reorder(species, height))) +
  geom_boxplot()

starwars %>% 
  drop_na(height) %>% 
  ggplot(aes(x = height, y = fct_reorder(species, height, .fun = max))) +
  geom_boxplot()

For more info, see the forcats vignette.

2.3.2 Tweaking scales

A common request: scientific notation vs not. A few options:

  1. Use different units. e.g., millions of people.
gapminder::gapminder %>% 
  filter(country == "United States") %>% 
  ggplot(aes(x = year, y = pop / 1e6)) +
  geom_line() +
  labs(y = "Population (millions)")

  1. Use scale_y_continuous with labels = scales::comma.
gapminder::gapminder %>% 
  filter(country == "United States") %>% 
  ggplot(aes(x = year, y = pop)) +
  geom_line() +
  scale_y_continuous(labels = scales::comma) + 
  labs(y = "Population")

  1. Use scales::label_number for even more control (see the help page).
gapminder::gapminder %>% 
  filter(country == "United States") %>% 
  ggplot(aes(x = year, y = pop)) +
  geom_line() +
  scale_y_continuous(labels = scales::label_number(scale = 1e-6, suffix = "M")) + 
  labs(y = "Population")

2.3.3 Direct Labels

When you have many lines, colors don’t work well for labels. Instead, use two tricks:

  1. Create a data frame with just the rightmost point of each line:
gapminder_filtered <-
  gapminder::gapminder %>% 
  group_by(country) %>% 
  filter(max(pop) > 100000000)

last_pop <- gapminder_filtered %>% 
  group_by(country) %>% 
  slice_tail(n = 1)
  1. Use text geoms to label those points:
gapminder_filtered %>% 
  ggplot(aes(x = year, y = pop, color = country)) +
  geom_line() +
  geom_text(
    data = last_pop, aes(label = country), # use different data
    color = "black", hjust = "left" # text starts at "x" and faces right
  ) +
  scale_x_continuous(expand = expansion(mult = c(0, .2))) + # make some room
  scale_y_log10() +
  theme(legend.position = "none") # turn off legend since it's redundant

  1. Use ggrepel::geom_text_repel to keep them from running into each other:
gapminder_filtered %>% 
  ggplot(aes(x = year, y = pop, color = country)) +
  geom_line() +
  ggrepel::geom_text_repel(
    data = last_pop, aes(label = country),
    color = "black", hjust = "left",
    direction = "y", # only move up or down, never left/right
    segment.alpha = .1, # lighten the connecting lines
    nudge_x = 3,
    seed = 0 # make this plot reproducible.
  ) +
  scale_x_continuous(expand = expansion(mult = c(0, .3))) +
  scale_y_log10() +
  theme(legend.position = "none")

2.3.4 Legends and Labels

If you need multiple rows for your legend, you probably have too many different values. But you can grit your teeth and do it…

starwars %>% skimr::skim()
Table 2.1: Data summary
Name Piped data
Number of rows 87
Number of columns 14
_______________________
Column type frequency:
character 8
list 3
numeric 3
________________________
Group variables None

Variable type: character

skim_variable n_missing complete_rate min max empty n_unique whitespace
name 0 1.00 3 21 0 87 0
hair_color 5 0.94 4 13 0 12 0
skin_color 0 1.00 3 19 0 31 0
eye_color 0 1.00 3 13 0 15 0
sex 4 0.95 4 14 0 4 0
gender 4 0.95 8 9 0 2 0
homeworld 10 0.89 4 14 0 48 0
species 4 0.95 3 14 0 37 0

Variable type: list

skim_variable n_missing complete_rate n_unique min_length max_length
films 0 1 24 1 7
vehicles 0 1 11 0 2
starships 0 1 17 0 5

Variable type: numeric

skim_variable n_missing complete_rate mean sd p0 p25 p50 p75 p100 hist
height 6 0.93 174.36 34.77 66 167.0 180 191.0 264 ▁▁▇▅▁
mass 28 0.68 97.31 169.46 15 55.6 79 84.5 1358 ▇▁▁▁▁
birth_year 44 0.49 87.57 154.69 8 35.0 52 72.0 896 ▇▁▁▁▁ 
starwars %>% ggplot(aes(x = height, y = mass, color = species)) +
  geom_point() +
  theme(
    legend.position = "bottom", 
    legend.key.size = unit(0.3, "cm")
#    legend.box.margin = margin(t = 0, r = 0, b = 0, l = 0, unit = "pt")
    ) +
  guides(fill = guide_legend(nrow = 2, byrow = TRUE))
## Warning: Removed 28 rows containing missing values (geom_point).

2.4 Mapping

2.4.1 Plotly

This document shows examples of two simple mapping tasks using Plotly. More details are available in the plotly-r book.

We’ll be using the tidyverse and the plotly package.

library(tidyverse)
library(plotly)

2.4.1.1 Markers

When you just want to mark something on a map, you can give lat/long coordinates to add_markers.

For example, let’s use a dataset of US cities:

maps::us.cities %>% head()
##         name country.etc    pop   lat    long capital
## 1 Abilene TX          TX 113888 32.45  -99.74       0
## 2   Akron OH          OH 206634 41.08  -81.52       0
## 3 Alameda CA          CA  70069 37.77 -122.26       0
## 4  Albany GA          GA  75510 31.58  -84.18       0
## 5  Albany NY          NY  93576 42.67  -73.80       2
## 6  Albany OR          OR  45535 44.62 -123.09       0

Here’s how to draw it on a map.

maps::us.cities %>% 
  # Fix the column names.
  rename(state = country.etc) %>%
  # Keep only larger cities.
  filter(pop > 100000) %>% 
  # Construct the "geo" projection.
  plot_geo() %>%
  # Add state markers
  add_markers(
    # Set marker position.
    x = ~long, 
    y = ~lat, 
    # Set other aesthetics (here, redundantly encode population)
    size = ~pop, 
    color = ~pop,
    # Customize the label.
    text = ~ glue::glue("{name}, population {scales::comma(pop)}"),  
    hoverinfo = "text"
  ) %>% 
  layout(
    # Zoom into just USA.
    geo = list(
      scope = 'usa'
    )
  )
## Warning: `line.width` does not currently support multiple values.

2.4.1.2 Choropleths

Plotly has builtin support for countries and US states. Any other granularity requires manually working with GeoJSON files; see the documentation.

Let’s make a world population map. First, let’s construct a dataset of the most recent data that Gapminder has for each country:

library(gapminder)
latest_country_data <- gapminder::gapminder_unfiltered %>%
  arrange(year) %>% 
  group_by(country) %>%
  slice_tail(n = 1) %>% 
  left_join(gapminder::country_codes, by = "country")

Now we add a “choropleth” trace. Note that this has the typical problem of choropleth maps and densities; see Fundamentals of Data Visualization for some discussion of this.

latest_country_data %>% 
  plot_geo() %>% 
  add_trace(
    type = "choropleth",
    # Specify that the "country" column contains the country names.
    locations = ~country,
    locationmode = "country names",
    # Use fill to show population. (I don't know why it's called 'z' and not 'fill'.)
    z = ~pop
  )