HomologyLive


What the software HomologyLive does: Statistical Physics simulations involve a large number of particles, whose geometrical distance to each other is calculated precisely and efficiently. This means that we have a natural distance function on the set of particles, and hence on its Vietoris-Rips complex, we can interpret naturally the 1- and 2-dimensional homology generators as loops and bubbles in the particle configuration. With HomologyLive, Alexander D. Rahm has made an implementation of the Vietoris-Rips complex adapted to this setting: Truncated at dimension 3, to stay within the physical dimensions of the ambient space, and with boundaries recorded as sparse matrices, in order to allow for huge numbers of particles in the simulations while keeping the linear algebra feasible on the machine. While a dense matrix computation with the full Vietoris-Rips complex - as implemented in SAGE - takes a big multiple of the machine resources of the simulations, HomologyLive in fact only takes a little fraction of them.

The current stable release (version 1.1) of HomologyLive is downloadable from github free of charge, covered by a GNU General Public License , University of French Polynesia, 2019.



Back to the homepage of Alexander D. Rahm