POT
POT copied to clipboard
POT : Python Optimal Transport
## Describe the bug Running `ot.solve` with tensors on the CPU allows memory on the GPU (thisis documented in `get_backend_list`) but also seems to use the GPU, as the Watts...
## Describe the bug We needed to force an older version of jax to prevent a bug when importing pymanopt because of a very recent deprecation of the configuration setup...
## Describe the bug It appears UnbalancedSinkhornTransport fails on .transform() calls. It seems to fail on the `nx.array_equal(self.xs_, Xs)` call. It seems the transporter is failing to set nx or...
## TODO - [ ] make proj_SDP batch-friendy and backend-compatible - [ ] finish flow example - [ ] re-implement pdfs - [ ] documentation - [ ] faster bures...
## Describe the bug in `ot.gaussian` the empirical means and computed with nx.mean it shoudl use weights ### To Reproduce Steps to reproduce the behavior: 1. ... 2. #### Screenshots...
## Types of changes This PR introduces an additional check for `NaN`s in the loss matrix of the emd computation. If `NaN`s are detected we raise an error in order...
> I would like to have the skip so that the tests do not fail if setuptools is not installed (for new developers this could be frustrating). If you do...
## Describe the bug As far as I understand it, I should be able to use this library paired with jax via your backend switching (depending on the input types)?...
## Describe the bug Hello, First of all, thank you for sharing the POT library as open source. When I try to import the POT library in an Apple Silicon...