vega / roadmap Goto Github PK

The Vega and Vega-Lite Roadmap

roadmap's Introduction

Vega: A Visualization Grammar

Vega is a visualization grammar, a declarative format for creating, saving, and sharing interactive visualization designs. With Vega you can describe data visualizations in a JSON format, and generate interactive views using either HTML5 Canvas or SVG.

For documentation, tutorials, and examples, see the Vega website. For a description of changes between Vega 2 and later versions, please refer to the Vega Porting Guide.

Build Instructions

For a basic setup allowing you to build Vega and run examples:

Clone https://github.com/vega/vega.
Run yarn to install dependencies for all packages. If you don't have yarn installed, see https://yarnpkg.com/en/docs/install. We use Yarn workspaces to manage multiple packages within this monorepo.
Once installation is complete, run yarn test to run test cases, or run yarn build to build output files for all packages.
After running either yarn test or yarn build, run yarn serve to launch a local web server — your default browser will open and you can browse to the "test" folder to view test specifications.

This repository includes the Vega website and documentation in the docs folder. To launch the website locally, first run bundle install in the docs folder to install the necessary Jekyll libraries. Afterwards, use yarn docs to build the documentation and launch a local webserver. After launching, you can open http://127.0.0.1:4000/vega/ to see the website.

Internet Explorer Support

For backwards compatibility, Vega includes a babel-ified IE-compatible version of the code in the packages/vega/build-es5 directory. Older browser would also require several polyfill libraries:

<script src="https://cdnjs.cloudflare.com/ajax/libs/babel-polyfill/7.4.4/polyfill.min.js"></script>
<script src="https://cdn.jsdelivr.net/npm/[email protected]/runtime.min.js"></script>
<script src="https://cdn.jsdelivr.net/npm/[email protected]/dist/fetch.umd.min.js"></script>

Contributions, Development, and Support

Interested in contributing to Vega? Please see our contribution and development guidelines, subject to our code of conduct.

Looking for support, or interested in sharing examples and tips? Post to the Vega discussion forum or join the Vega slack organization! We also have examples available as Observable notebooks.

If you're curious about system performance, see some in-browser benchmarks. Read about future plans in our roadmap.

roadmap's People

Contributors

Stargazers

Watchers

roadmap's Issues

Revamping Selections

Vega-Lite’s award-winning selection abstraction enables high-level interactive charts. We are extending the selection abstraction to support the parameterization of any encoding in Vega-Lite. For example, you will be able to dynamically change or compute the color of a bar (similar to signals in Vega).

Exporting Selections

Vega-Lite selections should be accessible through an API and we hope to expose this API in Python as well. With this API users can create an interactive visualization, select some data in the chart, and finally put only the selected subset of the data into a new variable and run some computation on it. See vega/vega-lite#1830 for details.

Regression Modeling

Data analysts often calculate statistical regressions and then visualize them on top of the raw data. We plan to add basic modeling and regression to Vega (and later Vega-Lite) so that they become easier to use, can be optimized by our runtime system, and be part of interactive visualizations.

Regular Vega Developers Call?

Thanks, @domoritz for setting up this repository! It's really helpful to see some of these roadmap items. I had some questions about some of these and I was thinking if it might make sense to set up a regular Vega developers call? A space where folks could show what they are working on and have higher bandwidth discussions about the project direction?

I wasn't sure where to start this discussion, so feel free to redirect me if there is a better place.

Support for Mobile

Vega-Lite’s implementation of selections currently only listens to desktop interactions. In the future, Vega-Lite’s selections should also be triggered by touch events. Since mobile displays are much smaller, some of the interactions and defaults have to be adjusted. See vega/vega-lite#4661 for details.

Improved React Support

The React Vega bindings need to be refactored to be more beginner-friendly and performant.

Interactions with Maps

Vega-Lite has support for geographic shapes and projections but only some of the interactive primitives support interactions with maps.

Serialization and Transfer

The data that is being visualized with Vega and Vega-Lite needs to be loaded into a browser. Right now, we serialize this data into text formats (e.g., JSON/CSV) which incurs a large overhead. Thus, for larger datasets, data transfer is often the main bottleneck in Vega. We plan to integrate more efficient binary formats (such as Apache Arrow) into the Vega runtime, to improve transfer in JupyterLab and other data science environments. See vega/vega#1300 for details.

Dataflow Analysis and Optimization

Vega-Lite’s internal dataflow is similar to a query execution plan in a database. We are applying simple optimizations to avoid redundant or unnecessary computation. We plan to extend the set of optimizations we apply.

Text Wrapping and Multiline Text

Vega supports different rendering backends including SVG and Canvas. Unfortunately, the Canvas API does not support text wrapping and so Vega needs to implement logic to support text wrapping. Text wrapping would also enable multiline text, but it’s not necessary. Vega could support multiline text with manually inserted line breaks. See vega/vega#488 for details.

Animations

Animations help viewers to keep track of elements in a visualization. Native support in Vega is an often requested feature. See vega/vega#641 for details.

Issuing Queries to Various Backends

Vega runs completely in the browser but much of the computation could be run on a scalable backend. We are working on developing infrastructure to automatically separate client-side and backend processing in Vega or Vega-Lite. This will particularly benefit visualizations that use aggregation where the size of the raw data is much larger than the number rendered marks.

Wikipedia integration

Wikipedia supports Vega v2 charts. The Wikimedia Foundation would like to update to Vega 4 and also support Vega-Lite. See https://phabricator.wikimedia.org/T165118 for details.

Extend the Core Language

Vega-Lite already supports an expressive range of visualizations. Nonetheless, we hope to expand the core building blocks of Vega-Lite with arc marks, jittering, support for Vega timers, and graph and tree layouts. We are also interested in support for density visualizations.

Labeling

We are working on labeling support for Vega and Vega-Lite charts such as scatter plots, bar charts, and line charts.