“So much more than pretty graphs”: Recent adventures in automating dataviz solutions

The Data Lab Community - Women in Data and AI Meetup | I talked about dataviz design systems, parameterised plots and WebR, before enjoying a talk by Martina Pugliese, and joining her for a fireside chat with Ally Orr, to discuss how we ended up here and what we’re excited about.

I used to say that I come in “towards the end of the pipeline”, to help organisations create compelling publication-ready data visualisations, based on the outcomes of all the work that has gone before. Dataviz design systems and parameterised plots are among my favourite things to build to streamline organisations’ data-to-viz processes. And the more of them I’ve built, the more I’ve come to realise their value much earlier in that data pipeline.

After looking at a few examples of what I mean by these, and seeing them in action, I’ll bring it all together with the example of an app I built over the last few days using WebR, in response to a bar graph I spotted in the wild. The app invites users to create a set of graphs based on the information they enter into the app, to compare it to the original, and to reflect on the consequence of different choices of visualisations.

We’ll explore how the “parameterisation” works, the traps I fell into, and how I coded my way out of them. We’ll also explore one thing the original bar graph did get right.

Stick around for the fireside chat (my favourite part of this event!) in which we unpacked what we do, how we combat impostor syndrome, what gets us excited about our work, and share tips for folks starting our on their data science journeys.

Recording

Slides with copy-pastable code

BarsBeRight

If you want to give #barsBeRight a go, you can do so on your phone! The R code just runs locally in your browser, with no need to install anything. Enjoy!