Visualization is oftentimes the best way to explore raw data. But as data grows to include millions and billions of points, traditional visualization techniques break down. Whether you're loading the data into limited memory, or separating the signal from the noise when thousands of data points occupy each pixel, as data gets big, visualization gets challenging.
In this talk, Peter will describe an approach called "datashading" that deconstructs the classical infovis pipeline to place statistical processing at the heart of the visualization task. The result is a scalable, interactive system that is easy to use and produces perceptually accurate renderings of extremely large datasets. He will show the open-source Datashader library, which implements these ideas, and makes them available within Jupyter notebooks and Bokeh data applications.