by Giancarlo Ruffo, Università degli Studi di Torino
Disclaimer: This is a work in progress. I made a lot of changes to this repo during 2020, and I plan to continue updating this tutorial for a while.
Access is public to my students and to everyone else.
There are thousands of useful resources that you can easily access on the Web to learn how to plot your data, and how to use programming languages, libraries, packages and other tools to master the art of visualising information. With this set of ipython notebooks I want to provide the students of my "Complex Networks Analysis and Visualisation" course at the University of Turin with many practical examples to learn basic data and information visualization principles.
The main learning objective of the dataviz part of the course is to understand the most important design principles behind a good data visualization. Hence our students are invited to understand how to exploit the viewer's perception biases and characteristics to improve their (interactive and non interactive) graphics, how to measure and to assess the quality of a design, and how to create complex data visualization platforms and dashboards. All these design and evaluation principles must be applied, and so more practical skills are considered part of the learning objectives.
In order to better understand the differences between basic and mainly static charts, we decided to adopt matplotlib, one of the most popular python libraries that have been specifically tailored to create static, animated, and interactive visualizations in Python. Matplotlib project is open source, and this is one of the reasons behind the selection of the related library for this series of notebooks accompanying the course. The other reason behind the selectin of this library is its wide adoption in the python programmers community, and the huge availability of resources, sample codes, suggestions, tutorials and handbooks you can find on the Web and in bookstores.
Under this perspective, this series of nootebooks is yet another tutorial on matplotlib. This is true, of course, but also partially unfair. Our goal is to present the most important design principles to make sense of data visually. matplotlib will be used mainly as a platform to give practical examples of how to apply our guidelines, and also how our design principles can be misinterpreted and misused. However, we need to learn matplotlib a little bit before we can make some good example. Naturally, this notebooks series is public and will be mantained on github. Feel free to create new threads of these notebooks to improve, edit, and update the code. Even if you are not attending my complex network class!
To start, let's introduce basic plotting tools and methods. We will make use of many numerical examples, so do not be surprised if we will import other python libraries such as numpy, scipy, pandas, and so on. Otherwise, if you need to refresh your skills on python and pandas, you can also use the related series on this repository or... search the Web for your favorite tutorial or handbook!