Closed marcocuturi closed 2 days ago
Check out this pull request on
See visual diffs & provide feedback on Jupyter Notebooks.
Powered by ReviewNB
Attention: Patch coverage is 85.71429%
with 1 line
in your changes missing coverage. Please review.
Project coverage is 88.17%. Comparing base (
828acca
) to head (b7f8624
). Report is 1 commits behind head on main.
Files with missing lines | Patch % | Lines |
---|---|---|
src/ott/tools/plot.py | 75.00% | 0 Missing and 1 partial :warning: |
This PR introduces a few enhancements:
introduces a
sinkdiv
convenience wrapper to compute the Sinkhorn divergence between two point clouds. Using thesinkhorn_divergence
method was a bit clumsy in that case, since it required passing first thePointCloud
geometry type, clearly overkill for most usages.introduces the possiblity to store the
reg_ot_cost
between the second point cloud (in a sinkhorn divergence) and itself and use it as an offset. This is useful when one wants to track the true Sinkhorn divergence between two point clouds but do not want to recompute that term at every iteration, say, of a Gradient descent. that quantity is calledoffset_static_b
corrects a bug in the
plot
class, wherebytitles
were ignored when theshow_lines
flag wasFalse
expands
glossary.rst