Graph clustering and visualization

I've been working on graph clustering for visualization since a few years. Following the publication of my latest publication on the subject, I've released the corresponding code.

Self-Organising Map

I'm using a lot R for implementing and testing new research idea, as a data analysis tool and as a teaching support. My modest contribution to this software is a quite complete implementation of Kohonen's Self-Organising Map in the yasomi package.

Software developed by colleagues


CloudDALVQ is a Cloud (Azure) implementation of the stochastic vector quantization algorithm (a.k.a., on-line k-means). It has been developed by my (former) PhD student, Matthieu Durut with his colleague Benoît Patra from Lokad. See our paper for details.

Older projects

I've been involved in two other open source projects in the past:

  1. I coded the first version of the multidimensional optimisation code of the GNU Scientific Library.
  2. I worked on the sound engine and on the pathfinder of the Stratagus open source real time strategy game engine inspired from Warcraft II (we used to call it freecraft, but Blizzard requested a name change…).