Skip to content

Conversation

rflamary
Copy link
Collaborator

The PR add the following:

  • LP solver for for Wassersein barycenters (do not scale at all but still nice for comparison).
  • Can choose the solver between scipy.optimize and cvxopt (the new interior point solver of scipy as pretty good)
  • Documentation and test for the solver, update of the README.
  • Example of comparison between LP and regularized barycenter :

image

@rflamary rflamary requested a review from ncourty May 14, 2018 10:02
ot/bregman.py Outdated
@@ -844,6 +844,8 @@ def barycenter(A, M, reg, weights=None, numItermax=1000,
loss matrix for OT
reg : float
Regularization term >0
weights : np.ndarray (n,)
Weights of each histogram i_i on the simplex
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

each histogram i_i -> each histogram bin i

ot/lp/cvx.py Outdated
reg : float
Regularization term >0
weights : np.ndarray (n,)
Weights of each histogram i_i on the simplex
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

same here

@rflamary rflamary merged commit 90efa5a into master May 29, 2018
@rflamary rflamary deleted the bary branch May 29, 2018 14:16
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants