Rémi Flamary
Rémi Flamary
Hello, do you still have the bug?
Hello I'm closing this Issue since we have been working on the new API that adersses most of the comments. I will open a new issue with a list of...
can you try the impoart from another folder? the cpp file is not compiled locally but if you are in thje POT repo it is trying to import the local...
in anay case you should install with `pîp install pot` or diecrtly using the git url for master version
ok i'm closing this then, since pot need ompilation you should not import from the source unless you install it with `develop`or compile with `-inplace`
Closing this issue because we implemented the solver with option `method='sinkhorn_log'` that is stable numerically. See : https://pythonot.github.io/all.html#ot.sinkhorn
Closing this one as a duplicate of #532 that is more detailed
yes acually I now provide both losses in the new ot.solve API and i plan tu provise also the proper divergence for empirical distributions when impelmenting the new api (with...
I think its OK now (doc for fyunction is done exemples and example update will be done later when doing the full API V2 release). @agramfort care for a quick...
I agree but it is not ready yet because we need to implement also GW, OT on sample and OT on grid which is a lot of work in addition...