Installation

• On SCOTS, install in the same way you install other R packages

• Easiest way is to use pkginstaller add-in

• More guidance available on eRDM

• Use remotes::install_github() if not working on SCOTS

Get some data

# Load packages
library(ggplot2)
library(dplyr)
library(gapminder)

# Create data set
life_exp <-
  gapminder |>
  select(year, country, lifeExp) |>
  filter(country %in% c("United Kingdom", "China"))

head(life_exp)

# A tibble: 6 × 3
   year country lifeExp
  <int> <fct>     <dbl>
1  1952 China      44  
2  1957 China      50.5
3  1962 China      44.5
4  1967 China      58.4
5  1972 China      63.1
6  1977 China      64.0

ggplot2 defaults

# Create line chart
ggplot(life_exp) +
  geom_line(aes(x = year, y = lifeExp, colour = country))

use_sgplot()

library(sgplot)

use_sgplot()

ℹ Default ggplot2 theme set to `theme_sg`.

ℹ Default colours set.

ℹ Default geom aesthetics set.

use_sgplot()

# Create line chart
life_exp_chart <- 
  ggplot(life_exp) +
  geom_line(aes(x = year, y = lifeExp, colour = country))

life_exp_chart

Other improvements

# Existing line chart
life_exp_chart <- 
  life_exp_chart +
  
  # Customise x and y axes
  scale_y_continuous(limits = c(0, 82),
                     breaks = seq(0, 80, 20),
                     expand = c(0, 0)) +
  scale_x_continuous(breaks = seq(1952, 2007, 5)) +
  
  # Add titles and labels
  labs(
    x = "Year",
    y = NULL,
    title = "Living Longer",
    subtitle = "Life Expectancy in the United Kingdom and China 1952-2007",
    caption = "Source: Gapminder",
    colour = NULL
  )

life_exp_chart

Labels instead of legend

# Calculate label positions using data
life_exp_labs <- life_exp |>
  group_by(country) |>
  filter(year == max(year)) |>
  ungroup()

life_exp_chart <- life_exp_chart +

  # Remove legend
  guides(colour = "none") +
  
  # Adjust x axis to give extra room for labels  
  scale_x_continuous(limits = c(1952, 2017),
                     breaks = seq(1952, 2017, 5)) +
  
  # Add labels
  geom_label(data = life_exp_labs,
             aes(x = year, y = lifeExp, label = country, colour = country),
             hjust = 0,
             vjust = 0.5,
             nudge_x = 0.5,
             label.size = NA)

life_exp_chart  

Apply to individual charts

ggplot(life_exp) +
  geom_line(aes(x = year, y = lifeExp, colour = country)) +
  theme_sg() +
  scale_colour_discrete_sg("main2")

Apply to individual charts

ggplot(life_exp) +
  geom_line(aes(x = year, y = lifeExp, colour = country)) +
  theme_sg(legend = "top") +
  scale_colour_discrete_sg("main2")

Package website

https://datasciencescotland.github.io/sgplot

• Cookbook; lots of example code

• Reference; help files for sgplot functions

• Further guidance on accessibility, colours and saving charts

• Guidance on how to ask for help or contribute

Support and feedback

• How to get help

  • R Users Yammer Community

  • Open an issue on GitHub if you’ve found a bug

  • Contact the maintainers

• Please let us know if you’re using sgplot - we would love to see examples of publications using sgplot

Future developments

• Support for interactive chart packages

• Add Scottish Government colour palettes when available

• Use Roboto Google font as per Scottish Government design system

• Anything that would make your life easier!

  • If your team has its own branding style and/or colour palettes, these can be added to sgplot (as long as they’re accessible!)

Get involved

• Open to anybody

  • Make a suggestion

  • Fix a typo

  • Write a new function

  • Become a package maintainer

• Contributing guidance

Analysis Function guidance

• Using sgplot does not guarantee accessibility

• sgplot should be used in conjunction with Analysis Function data visualisation guidance

  • Their e-learning is a good place to start

  • Review the charts checklist before sharing or publishing your chart

Important to remember

• sgplot package website has details of other things to consider for accessibility

  • Save charts in SVG image format

  • Include titles, subtitles and captions in the main body of the page (e.g. in html or Word document) rather than embedding within the chart image

  • Provide alternative text

Telling stories with charts

• Our best practice guidance discusses story-driven publications – charts are very much a part of this

• Charts should have a clear key message you are trying to communicate

• This key message should be interesting or notable in some way

• Analysis Function guidance: If you can’t write down the message your chart is giving, you should think again about the chart you have chosen

Chart titles

• Most effective way to communicate your message is in the title

• Analysis Function Guidance recommends two titles for charts

  • A main title that describes the main message of the chart

  • A statistical subtitle which gives details about the data, geography, time period etc.

• These should be included in the main text, above the image

• Currently updating our standard Word Template for Statistical Publications to incorporate this

Data visualisation is actually about text

• John Burn-Murdoch, Financial Times (YouTube)

  • Charts can be extremely effective at communicating information

  • However, to create effective charts you need to understand how people consume charts

  • People consume charts passively, you need to make your message clear and easy to digest

  • Evidence suggests users tend to focus on the title first

FT

BBC

Economist

Chartr

Guardian

ONS

ONS - Census 2021

Links

• sgplot website

  • Cookbook

  • How to contribute

  • How to get help

• GitHub repository

• Analysis Function Data Visualisation Guidance

• Scottish Government Design System charts guidance

Contact us

• Lee Bunce, Office of the Chief Statistician (OCS)

  lee.bunce@gov.scot

• Alice Byers, Data Innovation Team

  alice.hannah@gov.scot