Fix JAX warnings in tests #306
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
* fix unitary case * fix pre commit
* Add wrapper for running blackjax pathfinder. * Only look for tests in test subdir. Co-authored-by: Ricardo Vieira <[email protected]>
…ymc-devs#78) This reverts commit 75aa37b.
* remove bart * black
* added model_builder * added explanation * added tests * formatting * change in save and load method * updated save and load methods * fixed more errors * fixed path variable in save and load * Documentation * Documentation * fixed formatting and tests * fixed docstring * fixed minor issues * fixed minor issues * Update pymc_experimental/model_builder.py Co-authored-by: Michael Osthege <[email protected]> * Update pymc_experimental/model_builder.py Co-authored-by: Michael Osthege <[email protected]> * fixed spelling errors * Update pymc_experimental/model_builder.py Co-authored-by: Thomas Wiecki <[email protected]> * Update pymc_experimental/model_builder.py * Update pymc_experimental/model_builder.py * added build method again * Update pymc_experimental/tests/test_model_builder.py * removed unecessary imports * changed arviz to az * fixed codes to pass build tests * linespace -> linspace * updated test_model_builder.py * updated model_builder.py * fixed overloading of test_fit() * fixed indentation in docstring * added some better examples * fixed test.yml * indetation * Apply suggestions from code review * Update pymc_experimental/model_builder.py * Update pymc_experimental/model_builder.py Co-authored-by: Michael Osthege <[email protected]> Co-authored-by: Thomas Wiecki <[email protected]>
Co-authored-by: Ricardo Vieira <[email protected]>
* Initial commit of the linear conjugate gradients * Updated __init__ so that we can import linear_cg * Since we are invoking the function as a class attribute, removing the self * Fixed the import * Cleaned up some unused code * Fixed the name of the function * If vector is not equal to itself, it might have NaN * removed temporary variable * Moved configuration from settings class to function arguments * Ensure output has same datatype regardless of value of n_tridiag * if n_tridiag = 0 then tridiagonal matrices are not computed instead an identity matrix is returned which doesn't affect downstream computation
* added predict_posterior * changed tests * Update pymc_experimental/model_builder.py Co-authored-by: Thomas Wiecki <[email protected]>
) * Initial commit of the linear conjugate gradients * Updated __init__ so that we can import linear_cg * Initial commit of pivoted cholesky * Fixed the name of pivoted cholesky function * Since we are invoking the function as a class attribute, removing the self * Removing linear conjugate gradients from this branch * Also removing linear cg from __init__ * Fixed the import * Adding dependencies for pivoted_cholesky, they are also needed for tests * Added correct package for pytorch * Added try...catch block to ensure users who don't have these packages installed, and don't want to use pivoted Cholesky, can use pymc * Added import checks in the test file as well * removed unused commits * removing pytorch and gpytorch from requirements. * pre-commit wouldn't let me commit print statements * removing the test for now * Raising an ImportError instead of printing * Addressing stylistic comment * formatting fixes * formatting fixes * pre-commit modifications
* Make xhistogram dependency optional * Update setup.py * Update setup.py Co-authored-by: Ricardo Vieira <[email protected]> * Update setup.py * Update setup.py * pretty extras require * Update setup.py Co-authored-by: Ricardo Vieira <[email protected]> * fix black Co-authored-by: Maxim Kochurov <[email protected]>
…variable Warning is now only issued for IntervalTransforms
* added predict_posterior * changed tests * Update pymc_experimental/model_builder.py * updated load method * updated load method * added required changes * added required changes * fixed laod function * fixed laod functio * Update pymc_experimental/model_builder.py * Update pymc_experimental/model_builder.py * Apply suggestions from code review * Restructure tests. Fix load(). * Bump isort. * Skip saving on windows. * Different approach. * Different approach. * Different approach. * Different approach. --------- Co-authored-by: Thomas Wiecki <[email protected]>
…s#110) Co-authored-by: Michal Raczycki <[email protected]>
* changed id of model_builder to property, added type hinting * added id property test * autolinting * added test for model_builder predict function * extended test_model_builder.py with test_extract_samples * last test for model_builder * addressing requested changes, created support function from code repetitions --------- Co-authored-by: Michal Raczycki <[email protected]>
* add more typehints * Update pymc_experimental/tests/distributions/test_multivariate.py Co-authored-by: Ricardo Vieira <[email protected]> * Update pymc_experimental/tests/distributions/test_multivariate.py Co-authored-by: Ricardo Vieira <[email protected]> * change a letter * change a letter * wrap results into a named tuple --------- Co-authored-by: Ricardo Vieira <[email protected]>
Expand test coverage fix bugs in `structural.py`
* Cycle initial state has shape (2,) and two named dimensions Test initial state against statsmodels * Add tests of coords and dims for structural models * Add tests of params_info for structural models
…el` (pymc-devs#285) Adding recover_marginals utility function
…variances (pymc-devs#296) * Model variables named "sigma" no longer expect variances * Rename variables "sigma" to "sigma2" in `create_structural_model_and_equivalent_statsmodel` test function
updates: - [github.com/pre-commit/pre-commit-hooks: v4.3.0 → v4.5.0](pre-commit/pre-commit-hooks@v4.3.0...v4.5.0) - [github.com/PyCQA/isort: 5.12.0 → 5.13.2](PyCQA/isort@5.12.0...5.13.2) - [github.com/asottile/pyupgrade: v2.37.1 → v3.15.0](asottile/pyupgrade@v2.37.1...v3.15.0) - [github.com/psf/black: 22.6.0 → 23.12.1](psf/black@22.6.0...23.12.1) - [github.com/PyCQA/pylint: v2.14.5 → v3.0.3](pylint-dev/pylint@v2.14.5...v3.0.3) - [github.com/MarcoGorelli/madforhooks: 0.3.0 → 0.4.1](MarcoGorelli/madforhooks@0.3.0...0.4.1)
for more information, see https://pre-commit.ci
updates: - [github.com/psf/black: 23.12.1 → 24.1.1](psf/black@23.12.1...24.1.1)
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
|
Sorry I posted cringe, fixing the git history on this branch. |
|
I think you still need to do a force push to fix the git history
…On Sun, 11 Feb 2024, 14:02 Jesse Grabowski, ***@***.***> wrote:
Sorry I posted cringe, fixing the git history on this branch.
—
Reply to this email directly, view it on GitHub
<#306 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAACCUKZRT5BHLH3NYWOEF3YTC6P5AVCNFSM6AAAAABDDQ2X76VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMZXG4ZTGMRRGQ>
.
You are receiving this because you are subscribed to this thread.Message
ID: ***@***.***>
|
jessegrabowski
force-pushed
the
fix-tests
branch
2 times, most recently
from
02b5665 to
6d6ee91
Compare
jessegrabowski
force-pushed
the
fix-tests
branch
from
6d6ee91 to
0623e58
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
JAX warning mentioned #305 seems to be caused by using multiple cores when sampling with
pm.samplein tests. This PR setscores=1in all the tests to avoid it.I also fixed a future warning emitted by
arvizrelated to the use of thecoordsproperty to get the size of dimension -- thesizesproperty should be used instead.Finally (and perhaps controversially?), I took the opportunity to move the test folder up one level, out of the project module. This matches the directory structure of
pymcandpytensor, and makes more sense to me.