A (very) brief introduction to wfaplot

Installing wfaplot

To get started, install wfaplot from the github repository using install_github(). This function is part of the remotes package which can be installed from CRAN using the more familiar install_packages() command.

During the installation process, we may be prompted to update the package dependencies. If so, choose which packages to update or hit Enter to skip this step.

If the installation fails, you can also try to manually download the package files from GitHub and move these to the local directory where your other R packages are stored:

https://github.com/ciantudur/wfaplot

install.packages("remotes", repos = "http://cran.us.r-project.org")

remotes::install_github("ciantudur/wfaplot", dependencies = TRUE)

Once wfaplot is installed, don’t forget to load the package using library(wfaplot). You will need to do this every time you use R, but you will only be required to install the package once.

library(wfaplot)

---

Using colour with wfaplot

Now that wfaplot is installed and loaded into the environment, let’s begin by creating a basic scatter plot using the Iris dataset which is bundled with base R.

If you have never used ggplot2 before, you will need to install the package using install_packages("ggplot2") before running the next chunk of script.

library(ggplot2)

iris_plot <- ggplot(data = iris,
               aes(x = Sepal.Length,
                   y = Sepal.Width,
                   colour = Species))+
  geom_point()

plot(iris_plot)

We can apply an WFA colour palette to the chart with just one additional line of code.

iris_plot_wfa <- iris_plot +
  scale_colour_wfa()

plot(iris_plot_wfa)

By default, scale_color_wfa() uses the main WFA colour palette (red, dark grey, light grey), but we can also specify different palettes.

iris_plot_wfa <- iris_plot +
  scale_colour_wfa(palette = "fuchsia")

plot(iris_plot_wfa)

There are several colour palettes to choose from. Don’t worry if there aren’t enough unique colours to assign to each legend item, just choose any palette and wfaplot will automatically interpolate additional colours if required.

We can easily reverse the order of the colour palette by setting the reverse argument to TRUE.

iris_plot_wfa <- iris_plot +
  scale_colour_wfa(palette = "fuchsia",
                   reverse = TRUE)

plot(iris_plot_wfa)

There might be times when we need more granular control over the colours used. Let’s bring up a list of all WFA-defined colours using wfa_cols().

wfa_cols()

##        red     orange     yellow       lime      green       teal       blue 
##  "#CE0538"  "#ED8035"  "#F9D362"  "#C0DE72"  "#84E296"  "#14D9CC"  "#1AC8ED" 
##       pink     purple     indigo  dark grey light grey 
##  "#D14099"  "#9D7CE1"  "#41A9FB"  "#373737"  "#CECECE"

Now, let’s create a manual colour scale with Red, Purple, and Orange.

iris_plot_wfa <- iris_plot +
  scale_colour_manual(values = c(
    wfa_cols("red"),
    wfa_cols("purple"),
    wfa_cols("orange")))

plot(iris_plot_wfa)

Excellent! Now suppose rather than a categorical variable, we wanted to shade the points using a continuous value. This can easily be accomplished by setting the discrete argument to FALSE.

iris_plot_continuous <- ggplot(data = iris,
               aes(x = Sepal.Length,
                   y = Sepal.Width,
                   colour = Petal.Length))+
  geom_point()+
  scale_colour_wfa(palette = "grey",
                   discrete = FALSE)
  

plot(iris_plot_continuous)

The same principles apply when changing the colour of fill objects (e.g. bar charts / choropleth maps). Just note that we would use scale_fill_wfa() rather than scale_color_wfa() in these instances.

iris_bar_wfa <- ggplot(data = iris,
               aes(x = Sepal.Width,
                   y = Species,
                   fill = Species))+
  geom_bar(stat = "identity")+
  scale_fill_wfa(palette = "colourful")
  

plot(iris_bar_wfa)

---

Themeing charts with wfaplot

Themeing an WFA chart couldn’t be simpler. Just add theme_wfa() to the ggplot object. This function takes no additional arguments.

Just bear in mind that you will need the Fira Sans and Source Sans Pro fonts installed on your operating system to apply the theme properly. Both fonts are available to download for free from Google Fonts:

• Fira Sans: https://fonts.google.com/specimen/Fira+Sans

• Source Sans Pro: https://fonts.google.com/specimen/Source+Sans+Pro

iris_bar_themed <- iris_plot_wfa +
  theme_wfa()

plot(iris_bar_themed)

There might still be times when we want to tweak some of the theme elements. If so, just make sure that any additional theme customisations come after the theme_wfa() function.

iris_bar_themed <- iris_plot_wfa +
  theme_wfa() +
  theme(legend.title = element_text(family = "fira-sans",
                                    colour = wfa_cols("dark grey")))

plot(iris_bar_themed)

---

Saving charts with wfaplot

Once we’re happy with the chart, we’ll want to export it as an image using the wfa_save() function.

But before that, we must familiarise ourselves with two categories of graphic files:

• Pixel graphics – the most common image file extensions (png and jpeg) fall into this category. They can be opened using most operating system, the file size is relatively small, but they will start to look blurry once we zoom in enough. We can increase (or decrease) the file resolution by changing the resolution argument in wfa_save() e.g. wfa_save(export_res = 600) . The default is 400dpi (which will be more than sufficient for most scenarios).

• Vector graphics – these files retain their sharpness no matter how much we zoom in, but they take up much more disk space than their pixel counterparts. Common file extensions include svg and pdf. We typically use svg files for WFA reports. There is no need to specify a resolution for these type of files.

Let’s assume we want to share a chart on Twitter.

First, make sure that wfalogo.png is placed in the root directory. You can download the logo here:

https://raw.githubusercontent.com/ciantudur/wfaplot/main/wfalogo.png

Let’s add some titles then save the chart as a png with logo. We’ll use the default dimension and resolution options.

iris_bar_export <- iris_bar_themed +
  labs(title = "Short title goes here",
       subtitle = "Longer subtitle goes here",
       caption = "The caption / attribution goes here.")

wfa_save(iris_bar_export, "iris.png", include_logo = TRUE)

And voila… the chart is ready to be shared on Twitter.

Suppose we wanted to include the chart in an WFA report instead. Let’s save it as a vectorized svg file. In this case, we won’t need to include the logo and titles.

wfa_save(iris_bar_export, "iris2.svg", include_logo = FALSE)

Great, although it looks a bit cramped.

The default dimensions are a good place to start, but we can always override these settings.

As a general rule, leave the width at its default setting of 6in, and adjust the height from the default of 3 (e.g. if you want a taller chart, try specifying a height of 4 or 4.5).

This will take a bit of trial-and-error.

wfa_save(iris_bar_export, "iris_stretched.svg", include_logo = FALSE,
         export_height = 4, export_width = 6)

That’s better!

You can always find more information about this or any of the other wfaplot functions by typing help(wfa_save) in the console.

Since wfaplot is built on ggplot, we can take advantage of near-infinite data visualisation options. Once you’re ready to get stuck in, this website is a good place to start…

https://r-graphics.org/