The recent success of D3.js in visualization tasks in data science is due to its paradigm that leads to graphical computations happening in the browser, not on the server side. Visualizations can be delivered faster (no latency due to data going to a server and graphics getting back to the web browser) and in an interactive and personalized way.
Bokeh, from the Japanese "blurred" (a kind of photographic rendering that emphasizes the subject of the photo while blurring the background), is a Python package and part of the pydata stack that allows using web browser for presentations, mimicking the graphic style and interactivity of D3.js. It is a package that strives to make otherwise difficult representations an easy task, such as building dashboards, creating interactive plots, and representing large datasets.
We are just going to introduce it briefly, focusing on how to render matplotlib-based plots to be posted on the Web and leaving exploration of its...