Lyra 2: Designing Interactive Visualizations by Demonstration

Abstract

Fig. 1. An example interactive visualization designed in Lyra 2, a visualization design environment. Users can (a) brush in the scatterplot to re-aggregate the histogram, and (b) click histogram bars to filter for corresponding points in the scatterplot. This visualization was designed by demonstration -users did not have to write any textual code.

Integrated Visualization Editing via Parameterized Declarative Templates

“…Many VA tools are backed by JSON-based declarative grammars (such as Lyra [74,102] or Voyager [97,98]), yet do not allow users to modify the underlying specifcations. Restricting modifcations to what can be created within such tools misses a signifcant opportunity, as there is a wealth of online knowledge and examples that ought to be utilized [9].…”

“…Data manipulation tasks, for instance, might be better facilitated by the incorporation of ideas from datafow programming and spreadsheets, as well as programming-by-example techniques to help with data wrangling [17,36,38]. To more fully support presentation tasks, it would be fruitful to extend our combination of modalities to include visual builders [21]-which ofer a variety of direct manipulation features [71,79] for creating chartsand visualization by demonstration [72,73,91,102].…”

User-Driven Support for Rapid Visualization Prototyping in D3

“…There is a plethora of visualization authoring tools available to users for producing complex visualization designs, including specification based toolkits (e.g [10,28,48,50]), automated tools (e.g [12,52,59], PowerBI), or graphic driven tools (e.g [34]). Specification languages are expressive, but require considerable programming skills for users to fully utilize them [13,27,53].…”

“…Templates have also been combined with Code synthesis algorithms to create tools that automatically infer a user's intent based on their direct manipulation input and generates the corresponding code output [15,31,54,59]. Tools like Wrex [15], Falx [54], and Mage [31] provide readable code in response to users' interactions with GUI widgets while Lyra 2 [59] only focuses on using direct manipulation to make static charts interactive without exposing the underlying code to users. Our work extends this line of research of applying templates and code synthesis techniques for implementing visualizations.…”

Exploring D3 Implementation Challenges on Stack Overflow

“…D3 has made incredible contributions to the visualization community and the world; and like Mike Bostock, we simply seek to answer the question: what more can we do to provide these users with the best visualization experience possible? Many systems (e.g., [35,37,63]), recommendation engines (e.g., [18,26,41]), and languages (e.g., [48,59]) have been developed to ease the implementation burden of visualization users. Though powerful, they require the user to change or learn entirely new workflows, which may not match the user's implementation goals.…”