Datashader is a graphics pipeline system for creating meaningful

representations of large datasets quickly and flexibly. Datashader

breaks the creation of images into a series of explicit steps that

allow computations to be done on intermediate representations.

This approach allows accurate and effective visualizations to be

produced automatically without trial-and-error parameter tuning,

and also makes it simple for data scientists to focus on particular

data and relationships of interest in a principled way.