Codes & Experiments

Code & Material

• Small explanatory note about Quadratic OT manifold learning - QuadOT 

• Implementation of SGD for Wasserstein distance (Genevay et al., NIPS, 2016).  Code  for R with parts written in C for speedup. Available on my GitHub page.

• I made Mérigot 's code usable in R by writing several wrappers.  This code computes optimal transportation plans in the semi-discrete setting for the squared Euclidean metric in 2D (see Kitagawa et al., 2016,).  Available upon request.

• Visualisation of grids for OT-based multivariate ranks - Download 

• Visualisation of "Depth" for OT-based multivariate ranks for various different reference grids - Download 

• Code for the paper "On the finite-sample performance of measure transportation-based multivariate rank tests" -Download 

• Check my GitHub for available code.

• Check C. Weitkamp's GitHub page for code related to statistical shape comparison using Gromov—Wasserstein lower bounds. 

Experiments 

Second order stochastic gradient descent for large scale regularized optimal transport - Numerical tests and comparison with Genevay et al. -  Code available upon request

I remain available to discuss the codes and provide information.