I will teach how to organize, transform, analyse and visualize data, as well as how to effectively communicate the outcomes of the workflow, with a strong focus on network data.

The course will be multi-task (learn, make, use, watch, glance, read, dig, listen; see more below) and multi-teacher (I will be assisted by other real and virtual teachers). Some basics in programming, linear algebra and matrix theory, and statistics are desirable.

Play

1. Getting started
2. Data science in a nutshell
3. Essential linear algebra
4. Essential graph theory
5. Network science
6. Communication

You will go through different tasks: learn, make, use, watch, glance, read, dig, listen. A legend is below:

• learn: I teach, you listen (and hopefully learn).
• make: I give you an assignment, you make it during the class. We discuss the solutions during the next class.
• use: you use a software: download, install and run it for the first time. I give you a brief practical introduction to it.
• watch: We watch a video together. By and large, the video acts as a teaser, introducing the next topic in an informal and attractive way.
• glance: You give a brief and fast look at something, generally an informative website. I steer you towards the most important sections.
• read: You read a story, typically at home. We discuss it together during the following class.
• dig: You read a theoretical deepening of the current topic, normally at home. We talk about it during one of the next classes.
• listen: The class is given by an invited speaker, an expert in the field.

Books

• N10 Networks. Mark Newman. Oxford University Press, 2010.
• EK10 Networks, crowds and markets. David Easley and Jon Kleinberg. Cambridge University Press, 2010.
• WG17 R for Data Science. Hadley Wickham and Garrett Grolemund. O’Reilly. 2017.
• T11 R Cookbook. Paul Teetor. O’Reilly Media. 2011.

Data challenges

Data challenges have 3 components:

• Input, which consists of:
1. a dataset of raw data. No data model is assumed. The data should be open so it can be freely distributed.
2. a set of data questions and challenges, formulated in natural language, whose answers might be (but not necessarily are) hidden behind the raw data. Questions should be sufficiently general and compelling to tease the attention and curiosity of scholars.
• Analysis notebook: a stream of analyses and visualizations aimed at approaching the given data questions and challenges. Ideally, the notebook is written in some popular, free language (like R or Python) and it is self-containing so that it can be easily distributed, executed and modified by other scholars. Issues like readability, conciseness, elegance, efficiency of the notebook are relevant, although not crucial.
• Output: these are the suggested answers to the given data questions and challenges. Answers might be partial (not definitive). The same question can be answered with different notebooks. A (modest) degree of subjectivity in the interpretation of the data answers is expected.

The following are examples of data challenges you are invited to try:

1. Are female dolphins more social than male dolphins? (markup, markdown)
2. Which are the most powerful countries in the European natural gas market? (markup, markdown)
3. Detect the most dangerous terrorists involved in Madrid train bombing attack of 2011 (markup, markdown)
4. Discover the most interdisciplinary and autarchic disciplines in science (markup, markdown)
5. Detect communities in a Katate club friendship network (markup, markdown)