waratah

Fonts

The NSW Government typeface is Public Sans. Recommendations:

  • Install the font on your computer. This package will attempt to register an embedded copy of the font so that it can be used even without installing the font first.
  • If possible you should use a graphics device that supports modern features such as ragg::agg_png(). These devices should be used by default by RStudio if {ragg} is installed.
  • If using HTML output (such as for interactive plots of tables), ensure that the document loads Public Sans from Google Fonts (see below).
Show font import code for HTML documents.

Either add this to <head>:

<link href="https://fonts.googleapis.com/css2?family=Public+Sans:ital,wght@0,400;0,700;1,400&display=swap" rel="stylesheet">

or add this to your CSS:

@import url('https://fonts.googleapis.com/css2?family=Public+Sans:ital,wght@0,400;0,700;1,400&display=swap');

Theme for ggplot2

theme_waratah() sets default fonts, styles, and colour palettes for plots. You can set the theme globally with ggplot2::set_theme() or add it to an individual plot. If using any global options like options(waratah.colour_theme = "aboriginal"), make sure to do this before setting the theme.

See vignette("cookbook") for some examples of using the theme.

Colours

The NSW palettes are designed around two grids consisting of colour columns and tonal rows. Guidelines for charts suggest techniques for ensuring good contrast so that your charts are accessible.

Individual colours are available by name, e.g. nsw_colours$blue_01.

Grid-based method

To access colours using the grid system, use pal_nsw(). We can visualise the colours with a utility from the scales package:

pal_nsw(hue = c("blue", "red"), tone = 1:2) |> scales::show_col()

This gave us the first two tones of blue followed by the first two tones of red. Importantly, pal_nsw() always provides a discrete colour palette.

To use that with ggplot as a discrete colour scale, you can use:

scale_colour_discrete(palette = pal_nsw(hue = c("blue", "red"), tone = 1:2))

or scale_fill_discrete() to set the fill scale. In case you need a continuous scale, ggplot will automatically interpolate the colours:

scale_colour_continuous(palette = pal_nsw(hue = "blue"))

Flexible palettes

Instead of working directly with the grid, you can instead use pal_waratah(). Specify the pattern of your data and you will get a reasonable discrete or continous scale. Note that this makes it harder to follow the chat recommendations mentioned above, but it can be more convenient for scientific data visualisation.

For qualitative data, a mix of tonal rows 1 and 2 are chosen to avoid overly-similar colours:

pal_waratah("qual") |> scales::show_col()

There are also palettes for sequential and diverging continuous data that allow a choice of the base hue:

pal_waratah("div", hue = "green") |> scales::show_col(labels = FALSE)

Because pal_waratah() considers colour similarity, you can request that it also take into account colour vision disorders:

pal_waratah("div", hue = "green", cvd = TRUE) |> scales::show_col(labels = FALSE)

To make that choice globally, set:

options(waratah.cvd = TRUE)

Colour themes

Colour palettes in waratah allow a variant to be specified, including the default variant = "base" and the Aboriginal palette variant = "aboriginal". See pal_nsw() for details.