Bringing main changes to JAX branch - #309
Merged
Merged
Conversation
I'm not sure what the bad class name complaint is about. Checking if name too long.
Consistently using booleans
oldest tox task passing locally.
Update oldest supported versions
For Python. Identical to MATLAB version, except for the abs err tolerance for one special case test that is significantly larger in Python. All tests passing locally.
Synchronize `247NoKwargs` Branch
The changes were made in both the Python and MATLAB codes. The following
discussion is w.r.t. the Python code and hopefully helps explain the changes.
After designing the new test for this hfun, I suspected that the sign in front
of sumH in line 59 was incorrect. However, the tests were passing with small
errors. To determine if the test is capable of detecting that sign error, I
reran the tests but with the sign changed. There was effectively no difference
in the results despite the fact that the Hres and sumH variables contain
non-trivial numbers. This can be understood by seeing that
sumH * sum_{i=1}^m (Cres[i] - m_sumF) = sumH * 0.
Therefore, I removed the sumH term, which allowed for some simplifications,
especially in line 63.
Both Python and MATLAB tests were passing locally with no need to adjust error
tolerances in the test.
Clean parameterized hfun interface
This establishes the notion/design of parameterized hfuns for MSP. This also puts MSP on the same footing as POUNDERS then. The same will be done for the MATLAB version of this hfun.
Tests passing on GCE.
Note that mustBeColumn was introduced at 2024b. It was removed as this is too new of a feature for our MATLAB support.
This includes making a decision about path management for the path. Since we know where the product_of_cells file is relative to the test's file, we can use the scheme used elsewhere.
I tried and failed locally to get a good name. Ignore class naming rules just as we do for function and variable names.
This included backporting to the Python version improvements that were made in the MATLAB version. Testing of both versions passing on GCE.
The test script presently does *not* consider a skipped test in its count of total tests run.
Test MATLAB v1 against MATLAB v1
I was able to test a v2 against a v1 successfully.
Establish MATLAB v2 benchmark format & test
Was able to create v3 results and successfully compare against v2.
Establish MATLAB v3 Benchmark Format
Using the current test suites, comparing Python v3 against MATLAB v3 won't work since the two suites are run with different nf_max, which the test scripts interprets as incompatible problem definitions.
This was done based on manually acquiring Python and MATLAB results with nf_max=500 and fixing up the Python code temporarily to align those two results such that the test script compares matched results. The logging is now more verbose and it's easier to judge if a difference is significant or not.
…bad. So log differences to help determine if it is bad.
This should get actions passing.
Fixing a bug with success
Synchronize feature branch
Finalize benchmark test script
Acknowledging FM
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
No description provided.