-
-
Notifications
You must be signed in to change notification settings - Fork 2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add JAX implementation fol MatrixIsPositiveDefinite
Op
#6853
Conversation
Codecov Report
Additional details and impacted files@@ Coverage Diff @@
## main #6853 +/- ##
==========================================
- Coverage 92.03% 90.62% -1.42%
==========================================
Files 96 96
Lines 16398 16404 +6
==========================================
- Hits 15092 14866 -226
- Misses 1306 1538 +232
|
We have to check that HSGP failing test |
The tests pass locally 🤔 |
Some random (unrelated) tests failed 🤷 |
I am not sure is random, doesn't it use the Op we modified? |
Well, before the last commit where I simply changed the test parameterization all the tests passed. Can you maybe rerun the failed jobs? I can double check from my side. |
I don't know if the test is deterministic, so passing once isn't a conclusive thing. I can have a look on Monday |
There was a bug in the MatrixNormal test, it was supposed to try a different draw in the k2_samp test when it failed (max 10 times), but it was reusing the same draws everytime. |
b3da211
to
ed83025
Compare
I fixed the test and cleaned the PR git history so we can merge without squash. Let me know if it looks correct (and if you need chages be carefull with fetching the remote branch) |
Test behavior was accidentally changed in 9dad9c2
ed83025
to
39136a5
Compare
MatrixIsPositiveDefinite
Op
MatrixIsPositiveDefinite
Op
MatrixIsPositiveDefinite
Op
Great, thank you! Looks great! |
@@ -178,7 +178,7 @@ def test_prior(self, model, cov_func, X1, parameterization): | |||
gp = pm.gp.Latent(cov_func=cov_func) | |||
f2 = gp.prior("f2", X=X1) | |||
|
|||
idata = pm.sample_prior_predictive(samples=1000) | |||
idata = pm.sample_prior_predictive(samples=1000, random_seed=rng) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
🙏
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
One day we will get this 5 line PR merged. Just you wait :D
Closes: #6849
Following https://pytensor.readthedocs.io/en/latest/extending/creating_a_numba_jax_op.html
📚 Documentation preview 📚: https://pymc--6853.org.readthedocs.build/en/6853/