Streaming Cross-Sectional Data Visualization with Perspective + Spark

Many data visualization libraries are built with static data in mind. However, analyzing data streams in real-time is crucial for many industries. In this talk, we will look at Perspective. Perspective is an open source library built for high-performance streaming data visualization that lets users quickly and easily dissect their data. It is written in C++ and Rust and bound to WebAssembly in the browser and Python on the backend, allowing it greater performance than similar tools. This talk is built around a tangible demonstration of a real-world application. We'll start with raw data in a Jupyter Notebook. With Spark Streaming, we'll parse, reformat, and join data streams with reference information and feed it into Perspective - all within a notebook. We'll then breakout of the notebook into a full-featured streaming data web application, demonstrating how quickly and easily we can move from prototype to production with Perspective and Spark Streaming.