Description
bqplot is a Python plotting library based on d3.js that offers its functionality directly in the Jupyter Notebook, including selections, interactions, and arbitrary css customizations. bqplot is built to generate professional looking, highly interactive plots and dashboards with minimal code. This enables the rapid generation of full fledged web applications directly in the Notebook with just a few lines of Python code. We will then go over bqplot's library of statistical and machine learning focused higher level abstractions and tools which are built to aid a Data Scientist or a Machine Learning researcher to interactively analyze the data, and arrive at useful insights. The seamless interactivity enables the creation of custom visualizations that allow researchers to understand their models and data in a much richer fashion. The talk will also cover bqplot's seamless integration with the numerous other libraries in the Jupyter widgets ecosystem such as ipyleaflet or ipyvolume. We will then explain how to link these widgets together using callback functions and share a few examples of advanced applications from various disciplines built using other open source projects. We will wrap up by previewing the latest features we are adding to bqplot.Presenter(s): Speaker: Sylvain Corlay, QuantStack