We propose a constraint driven model of visualizing multidimensional, multivariate, and labeled data to drive a functional architecture of visualization library components that developers can use to build structure preserving linked, interactive, and animated visualizations. <-tweet length - 15 chars left >
data point, such as a row in a table or a node in a network,
some examples of structure preserving: translation, rotation, etc... observation <-section>
structure: continuity + scale preserving, data types -> data types, set of all objects + maps to preserve properties, maps are pushed through
is about the API ->
Notes: probably need to define structure
associated repos
notes, API sketches, drafts, etc. leading towards dissertation proposal: