Links
Contribute
We are happy to have you on board.
ProVega and Pro-Ex are open source. The public Pro-Ex instance is open and writable, so your exemplar can sit next to ours in the gallery.
- 1
Add your progressive visualization to the gallery
Upload it from the Pro-Ex Gallery View.
- 2
Open an issue
Bugs, missing properties, ideas for the grammar.
- 3
Send a pull request
How it works
One block on top of Vega-Lite
All progressive constructs live under a single top-level property. Marks, encodings and native transforms are untouched and refine each partial result. ProVega supports data chunking, process chunking and mixed chunking, and accepts complete or streamed input (WebSocket, Socket.IO).
Vega-Lite
{ "data": { "url": "fars.csv" },
"mark": "circle",
"encoding": { "longitude": …, "latitude": … },+ ProVega "provega": { "progression": { "chunking": { "type": "data", "reading": { "method": "sequential", "chunk_size": 2000, "frequency": 200 } }, "control": { … }, "monitoring": { "change": { "mark": true } } } } }
Use case 1: 5.6 × 10⁵ fatal US road crashes (FARS 2001–2009), one update every 250 ms.
Come to the talk
Tuesday, Nov 10
11:12–11:24 ET
America Center, Boston · VIS Full Papers
Session “Please don’t just stare at the picture”, 10:00–11:30
Paper
ProVega: A Grammar to Ease the Prototyping, Creation, and Reproducibility of Progressive Data Analysis and Visualization Solutions
Matteo Filosa¹*, Graziano Blasilli¹*, Emilio Martino¹, Marco Angelini²
¹ Sapienza University of Rome · ² Link Campus University · * joint first authors
