mojchart is an R package to help with formatting charts in ggplot2. It includes a theme function based on Government Analysis Function guidance and colour palettes based on both MoJ corporate branding (MoJ internal link) and Government Analysis Function guidance on colours for data visualisation.

You can install mojchart with:

install.packages("devtools")
devtools::install_github("moj-analytical-services/mojchart")

The package provides the following custom ggplot2 functions:

• theme_gss() - a theme based on Government Analysis Function guidance;
• scale_colour_moj() and scale_fill_moj() for applying mojchart colour palettes.

To access the package colour palettes without using the scale functions:

• moj_palette() returns a colour palette as a vector of colour hex codes.

To obtain the hex codes for individual colours:

• moj_colours() provides MoJ corporate colours;
• govanal_colours() provides colours from the Government Analysis Function colour palettes
• palette_colours() provides the colours used in the mojchart colour palettes.

To display colour palettes:

• display_palette() displays a vector of colour hex codes;
• display_scheme() displays all the colour palettes within an mojchart colour scheme.

The function help pages (accessed using ? or help() in RStudio) contain full details and examples.

The main principle behind the theme function theme_gss() is to simplify the chart and remove unecessary formatting. The theme allows for a main title plus a subtitle for statistical details. The data source can be shown in the caption below the chart.

Axis labels should be horizontal according to the Analysis Function guidance To that end, theme_gss() suppresses the ggplot2 y axis label; a label can instead be included within the ggplot2 subtitle if required. This is demonstrated by example 2 below.

theme_gss() includes parameters to show or hide the x axis label and x axis tick marks. Setting the parameter flipped = TRUE reverses the x and y axes for use with coord_flip().

Note that the name theme_gss() is because the Analysis Function Guidance was previously provided by the Government Statistical Service (GSS).

The package provides colour palettes for discrete data, grouped into several colour schemes. Each colour scheme consists of six colour palettes, containing from one to six colours. Note that the Analysis Function guidance recommends aiming for a maximum of four colours on a chart to avoid the chart becomming too cluttered.

The Analysis Function colour palettes are taken from the Analysis Function guidance. There is one colour scheme for bar charts and one for line charts.

For MoJ corporate colours, there are two types of colour scheme: vibrant and muted. The vibrant schemes use MoJ corporate colours directly. The muted schemes are based on MoJ colours but with reduced saturation. Reduced saturation was previously recommended in Government Statistical Service guidance for shading areas, such as for bar charts, and might still be appropriate in some settings.

For sequential colour palettes or continuous data, you could look at the inbuilt ggplot2 options, or try the colorspace package.

scheme_names() returns the colour scheme names.

library(mojchart)
scheme_names()
#> [1] "muted1"        "muted2"        "muted3"        "vibrant1"     
#> [5] "vibrant2"      "govanal_bars"  "govanal_lines"

You can use scale_colour_moj() or scale_fill_moj() to apply a colour palette to a ggplot2 chart, passing the number of colours and the name of a colour scheme as arguments. There is an optional order parameter to change the colour order.

To obtain a colour palette as a vector, use moj_palette(), again specifying the number of colours and the colour scheme. The display_palette() function displays the colours.

pal <- moj_palette(n = 5, scheme = "muted3")
pal
#> mojdarkblue    midblue1       teal2       pink1     yellow1 
#>   "#003057"   "#3975ac"   "#59a6a4"   "#f3bee0"   "#feebb3"
display_palette(pal)

The display_scheme() function shows all the colour palettes within a colour scheme, as shown for every colour scheme here.

moj_colours() provides the hex codes for all MoJ corporate colours. govanal_colours provides the Government Analysis Function colours. palette_colours() provides all colours used in the mojchart colour palettes. Running one of these functions with no arguments returns a vector containing all of the available colours. Passing the names of individual colours returns just those.

moj_colours("mojblue", "mojgreen")
#>   mojblue  mojgreen 
#> "#1d609d" "#30aa51"

See the Analysis Function guidance for a full discussion of accessibility considerations regarding the use of colour in charts.

The mojchart colour palettes aim to be accessible to those with the most common forms of colour blindness, but accessibility does decrease the more colours you use.

Below are two resources to help assess colour blindness accessibility.

• To simulate how an image file could appear to individuals with forms of colour blindness:
  https://www.color-blindness.com/coblis-color-blindness-simulator/
• To simulate the appearance of a colour palette of hex codes:
  https://davidmathlogic.com/colorblind/

Note that for line charts, the Analysis Function guidance recommends labelling lines directly rather than using a separate legend. The directlabels package provides one way to implement this in R.

The examples below show the use of theme_gss(), scale_colour_moj() and scale_fill_moj(), together with various ggplot2 functions, to create fully formatted charts. The data frames for the charts are produced by two mojchart functions, familystarts_reduced() and bars().

• In this example note that the line thickness has been increased in geom_line() so that the colours stand out more, making the categories easier to identify.

library(mojchart)
library(ggplot2)
library(xts)
library(scales)
ggplot(familystarts_reduced(6), aes(x = year_qtr, y = count, colour = case_type)) +
  geom_line(size = 1.5) +
  labs(title = "Number of cases started in family courts",
       subtitle = "England and Wales, Q1 2011 - Q2 2020",
       caption = "Source: Family court statistics quarterly, April to June 2020 (table 1)") +
  scale_y_continuous(expand = expansion(mult = c(0, 0.05)), label = comma) +
  scale_x_yearqtr(labels = date_format("%Y-Q%q")) +
  expand_limits(y = 0) +
  theme_gss(xticks = TRUE) +
  scale_colour_moj(6, scheme = "vibrant1", order = c(1, 3, 2, 6, 4, 5))

ggplot(bars(5), aes(x = col1, y = col3, fill = col2)) +
  geom_col(position = position_dodge2(padding = 0), width = 0.8) +
  labs(title = "Example grouped bar chart",
       subtitle = "Additional details here\n\nAxis label",
       caption = "Source") +
  scale_y_continuous(expand = expansion(mult = c(0, 0.05))) +
  expand_limits(y = 10) +
  theme_gss() +
  scale_fill_moj(5, scheme = "muted2")

ggplot(bars(3), aes(x = col1, y = col3, fill = col2)) +
  geom_col(position = "dodge", width = 0.7) +
  coord_flip() +
  labs(title = "Example horizontal bar chart",
       subtitle = "Additional details here\n",
       caption = "Source",
       y = "Axis label") +
  scale_y_continuous(expand = expansion(mult = c(0, 0.05))) +
  expand_limits(y = 10) +
  guides(fill = guide_legend(reverse = TRUE)) +
  theme_gss(flipped = TRUE, xlabel = TRUE) +
  scale_fill_moj(3, scheme = "govanal_bars")