From the implementation standpoint, it tries to keep a fair decoupling between all steps in the pipeline. Data, scale, layout, attribute, component, renderer, interaction. It starts with the basic charts: simple/stacked/percent bar charts and simple/multiple/stacked/area charts. But the goal is to validate if this datavis pipeline architecture makes it easy to expand indefinitely.
The charts function takes a config object. The only required one is the container, which can be a CSS selector or a node
The list of configuration attributes exposed is available in core.js and are pretty self-explanatory:
If you are not sure what a config attribute is doing, just try it here.
You can find some demos in the examples folder as well as a minimal test suite. Enjoy!