Florence is a data visualization framework, based on the grammar of graphics, built on top of Svelte's template syntax. We designed Florence to address some challenges that we experienced in cartography education and practice. It is built on top of existing open web standards that are already in intensive use for online mapmaking today, but provides a framework that is firmly based on cartographic and visualization theory. Specifically, we adopt concepts from Bertin’s Semiology of Graphics and Wilkinson’s Grammar of Graphics to create a language with a limited number of core concepts and verbs that are combined in a declarative style of “writing” visualizations.
For a more detailed discussion of the framework and its design goals, please have a look at corresponding in article in Cartographic Perspectives.
Poorthuis, A., van der Zee, L., Guo, G., Keong, J. H., & Dy, B. (2020). Florence: a Web-based Grammar of Graphics for Making Maps and Learning Cartography. Cartographic Perspectives, (96). https://doi.org/10.14714/CP96.1645.
Inspiration & Foundations
Obviously, Florence is inspired by, and builds directly on top of, amazing work by a wide range of people.
- The constellation of libraries around D3 has become a staple in any data visualization toolbox. We make heavy use of
d3-scaleand try to ensure compatibility with other d3 libraries.
- Vega, created by the UW Interactive Data Lab, is an amazing example of the power of building visualization software on top of the grammar of graphics. The design goals and emphasis of Florence are slightly different from Vega but we have learned much from both Vega's approach and implementation.
Florence is named after two trailblazing 19th century data visualization professionals that used visualization to great effect.
- Florence Nightingale, who used data visualization to increase the adoption of sanitary standards in hospitals.
- Florence Kelley, who fought for social welfare in late 19th century America and used data visualization and mapping to show socio-economic disparities on a micro-scale in Chicago.
Both women were not data visualization 'professionals' by training but were able to adopt existing visualization practices and find novel, new approaches using relatively simple tools (pen, paper etc). Inspired by their example, we hope that our data visualization framework can help contribute to a creative and inclusive data visualization practice as well.
Florence is built by Ate Poorthuis (KU Leuven), Luuc van der Zee (KU Leuven), Grace Guo (Georgia Tech), Jo Hsi Keong (SUTD), and Bianchi Dy (SUTD). Development started in 2019 at the Singapore University of Technology and Design and is now being continued at KU Leuven.