Summary

• Working in a web browser avoid dependencies, wider audience
• Python and JavaScript complement each other.
• Avoid JavaScript's dodgy parts.

Reality

• You need to know JavaScript for the best web visualisations.
• Trick is to use Python to clean and deliver data to JavaScript's tools (D3, Raphael, Leaflet, Crossfilter).

JavaScript is rough and terrible defaults

• Default global variables.
• No namespacing.
• Religious wars surrounding imports, confused.
• Crockford's "JavaScript: The Good Parts"
• Use a linter, e.g. JSLint, or preferably JSHint.
• JavaScript frameworks are unlike buses - there are always three frameworks coming at once.
• People's default answer to "how do I get this to work in X?" is "but Y is far better, use that".

..and the good

• 100% installation base.
• Orders of magnitude faster than Python, arms race to make it faster.
• Visualization libraries written by real programmers.

d3

• Learning curve is worth it

Crossfilter

• Large multivariate datasets, pairs with d3 and visualises large complex data sets well

Server-side JavaScript

• Compared to Python there is no compelling use-case.
• Maybe Node.js for massively concurrently traffic.
• But JavaScript has no standard imports, missing NumPy and SciPy.

Python server to JavaScript client

• JSON, as:
  1. static file, sure works but not dynamic.
  2. API call, most common. RESTful recommended.
  3. Web sockets, realtime but still esoteric.

 What Python server?

• SimpleHTTPServer for static, starts off fine.
• Flask, lightweight when something very simple is all you need.

 Websockets use-case

• Move from Python -> QT, OpenGL GUI, to
• Python -> Flask -> WebGL / Angular / Bootstrap
• Wider audience, no dependencies

 Project to visualise Met Office weather data

• Very, very dirty data. Arbitrarily change delimiters and null values over time.
• Python makes short work of this.
• Geo-JSON: scipy, geojson.
• scipy.interpolate.interp2d to put coordinates onto 2D grid for front-end.
• key thing about D3, as opposed to Bokeh or Vega, is to be able to interact with the visualisation.