Uploaded July 2026 | Updated November 2026, 10 hours ago
How to build a data visualization in R from raw data to a finished chart: read the dataset first, set up a project with the packages you need, join and filter your data down to what the chart is actually about, then build the plot up in layers with ggplot2. This session adds a sunburst with coord_radial, a custom palette with rcartocolor, coloured subtitle text with ggtext, interactivity with ggiraph, and automated captions. Live, with Nicola Rennie.
What you will learn
- What to look at in a dataset before you decide what chart to make
- How to set up a repeatable project template so each week starts faster
- How to join two datasets and filter down to the subset the chart is about
- How to build a sunburst chart with coord_radial
- How to apply a custom colour palette with rcartocolor
- How to colour individual words in a subtitle with ggtext
- How to make a static ggplot interactive with ggiraph
- How to automate the caption and source credit on every chart
- How to save a plot without altering the underlying ggplot object
Chapters
0:00 What a TidyTuesday workflow looks like start to finish
1:18 Reading the dataset before you plot anything
3:41 Project setup: template files and the packages you need
7:25 Do showtext and systemfonts do different things?
9:04 Loading and joining the two datasets
11:45 Filtering down to one city
17:25 Reusing chart code from a previous week
23:56 Building a sunburst chart with coord_radial
25:47 Why mutate rather than summarise?
31:09 Custom colour palettes with rcartocolor
33:57 Coloured subtitle text with ggtext
40:38 Making the chart interactive with ggiraph
43:32 Adding automated captions and source credits
46:26 Saving a plot without breaking the ggplot object
49:55 Where to follow Nicola's work
Learn more
ggplot2: ggplot2.tidyverse.org
TidyTuesday: github.com/rfordatascience/tidytuesday
Positron: positron.posit.co
The Data Science Lab is a live weekly call. Register at https://pos.it/dslab
How to build a data visualization in R from raw data to a finished chart: read the dataset first, set up a project with the packages you need, join and filter your data down to what the chart is actually about, then build the plot up in layers with ggplot2. This session adds a sunburst with coord_radial, a custom palette with rcartocolor, coloured subtitle text with ggtext, interactivity with ggiraph, and automated captions. Live, with Nicola Rennie.
What you will learn
- What to look at in a dataset before you decide what chart to make
- How to set up a repeatable project template so each week starts faster
- How to join two datasets and filter down to the subset the chart is about
- How to build a sunburst chart with coord_radial
- How to apply a custom colour palette with rcartocolor
- How to colour individual words in a subtitle with ggtext
- How to make a static ggplot interactive with ggiraph
- How to automate the caption and source credit on every chart
- How to save a plot without altering the underlying ggplot object
Chapters
0:00 What a TidyTuesday workflow looks like start to finish
1:18 Reading the dataset before you plot anything
3:41 Project setup: template files and the packages you need
7:25 Do showtext and systemfonts do different things?
9:04 Loading and joining the two datasets
11:45 Filtering down to one city
17:25 Reusing chart code from a previous week
23:56 Building a sunburst chart with coord_radial
25:47 Why mutate rather than summarise?
31:09 Custom colour palettes with rcartocolor
33:57 Coloured subtitle text with ggtext
40:38 Making the chart interactive with ggiraph
43:32 Adding automated captions and source credits
46:26 Saving a plot without breaking the ggplot object
49:55 Where to follow Nicola's work
Learn more
ggplot2: ggplot2.tidyverse.org
TidyTuesday: github.com/rfordatascience/tidytuesday
Positron: positron.posit.co
The Data Science Lab is a live weekly call. Register at https://pos.it/dslab










