Just a day before the qualification round of 2016's hashcode, some thoughts about this programming competition.

In this post, we will use an open dataset from the Vélib (Paris shared bicycle system) provider and do data visualization using pandas and bokeh.

In plasma physics, charged particles are subject to magnetic and electric fields. In this notebook, we investigate their trajectories in special cases of these fields.

In this notebook, we look at matter from two point of views: a local description and a global point of view. It turns out that the average vision is different from the local vision. Like in plasma physics.

Another look at the Kaggle What's Cooking dataset, from the visualization side. This time, we'll use Bokeh for interactive visualization of QQ plots.

In this post, we set out to reproduce a plot by Randy Olson, using Wikipedia data about the Roman emperors before the fall of the empire. We then propose a new visualization of the reign durations that shows stability and turmoil in the Roman empire from a different perspective.

This post is a little tutorial about the Sankey diagram capabilities of matplotlib. We'll start by drawing simple Sankey boxes, connecting them and as a final project we will build a graphics showing how hard MOOCs are.

Dans ce billet, nous calculons l'âge du sénateur français moyen à partir des données disponibles sur le site du sénat. Contrairement au billet de la semaine dernière sur les députés, nous allons directement récupérer ces informations à l'aide de requêtes web. Nous utilisons ensuite des données de l'INSEE pour évaluer la représentation des différentes classes d'âge françaises par les députés.

In this post, I reflect upon the MOOCs that I have taken in 2015. Lessons learned, what it means for me.

In this post, we go through the basics of training a logistic regression on the Kaggle What's Cooking data. We also do some in-depth analysis on recipes and their ingredients.