Introducing Deckard for large scale data visualisation
Deckard is our first open source contribution, combining deck.gl with R. We saw the great work done at Uber with deck.gl and some impressive visualisations produced by the team over at Small Multiples (detailed here). We quickly realised it was a great tool and something we wanted to use for our own geospatial analysis (see this post for an overview) and large scale visualisations. The main reasons we were looking at Deck.gl was the large amount of data we required be visualised, the types of interactions that were possible and the ability to extend the library by adding in our own layers. It’s also been great for visualising modelling output and deeper insights from analysis, not just raw geospatial data.
To fully utilise deck.gl we wanted to integrate it with our existing workflow which is predominately in R, where we leverage some of the amazing spatial packages like sf, sp and leaflet. So we created a package that allows you to visualise data in deck.gl directly from R, either in the Viewer, Markdown Docs or using Shiny.
We can visualise our data using the existing layers provided in Deck.gl such as; line, hexagon, icon, geojson, screen grid, arc, scatter plot, point cloud, text layer and custom layers. Although the visualisations are geospatial centric they can be extended to other visualisation types, something we will be exploring in future iterations.
So give it a try over here and let us know what you think. Also, if you’re looking for a more complete solution that takes care of data ingestion, data manipulation, modelling and visualisation of geospatial data feel free to get in touch.
Links:
Where did the name come from? Deckard stands for “Deck.gl and R, Deckard” (recursive acronyms). Happy coincidence that it’s the main character from a great sci-fi film 😉