Package volesti provides several geometric random walks for high dimensional sampling from convex polytopes. The current implementations can be used for any logĀ­concave distribution restricted to a convex polytope but they do not support parallel execution and sparse linear algebra computations. The aim of this project is to provide parallel implementations for all the existing random walks in volesti and efficient sparse linear algebra computations when the matrix of a polytope is sparse. I believe that this project would be of special interest for high dimensional geometric and statistical computing, e.g. in Markov Chain Monte Carlo multivariate integration, convex optimization and Bayesian inference.



Konstantinos Pallikaris


  • Apostolos Chalkis
  • Vissarion Fisikopoulos
  • Marios Papachristou
  • Elias