Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 76.36%. Comparing base (a1be331) to head (9b3ebb7).
Report is 1 commits behind head on main.

@@            Coverage Diff             @@
##             main    #1465       /-   ##
==========================================
- Coverage   76.36%   76.36%   -0.01%     
==========================================
  Files         394      394              
  Lines       65130    65130              
  Branches    14429    14429              
==========================================
- Hits        49737    49734       -3     
- Misses      12831    12835        4     
  Partials     2562     2561       -1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

@lbianchi-lbl @ksbeattie this is ready to review and merge. I've updated the requirement to the 6.8.0 Pyomo release which we cut on Tuesday.

It looks like a few of the integration checks (namely those where IDAES is installed in "user mode", i.e. not in editable mode) are failing, seemingly because of Pint:

I wonder if this is caused by the fact that we're pinning Pint to a previous version since #1438 (IIRC because of an incompatibility with the way dimensionless and/or unitless quantities are rendered).

I'm not entirely sure why there would be a difference with respect to the editable/developer mode, so I'll have to look more closely into this.

@lbianchi-lbl looks like more recent versions of Pint started supporting Python 3.9 again (see: hgrecco/pint#1974) so we can try unpinning Pint (or requiring >=0.24.1) and fixing the tests with the updated rendering

@lbianchi-lbl looks like more recent versions of Pint started supporting Python 3.9 again (see: hgrecco/pint#1974) so we can try unpinning Pint (or requiring >=0.24.1) and fixing the tests with the updated rendering

Thanks for looking into it @blnicho. I think your proposal sounds good, and IIRC, the decision to set a maximum version was mostly due to version incompatibilities than the need to update the tests (although that will still need to happen, not only in this repository, but also in our dependencies). I'll try updating to pint >= 0.24.1 as you suggest and see how it goes; hopefully, only those tests involving the outputs we expect to be affected will fail.

There is an issue with the same Pint version constraint being specified in idaes-ui, which is causing the test failures. I'm not sure there are other good solutions other than having to create a patch release for idaes-ui.

In the meantime, I'll try to remove ui from the list of optional dependencies install idaes-ui from a PR branch with the changes (IDAES/idaes-ui#52) and see how that goes so that we can first address the Pint-specific issues here.

I'm currently trying to understand why installing in dev mode causes NumPy v1 to be installed:

while, in the user-mode installation, NumPy 2.1.0 gets installed instead:

I imagine this might be caused by one of the dev dependencies having a numpy <2 constraint that causes this. I'm trying to reproduce this locally to try to figure out which one is doing that.

It looks like the (and as far as I can tell, the only) package distribution requiring numpy < 2 is Tensorflow:

"pythondistribution[tensorflow]": {
    "version": "2.17.0",
    "requires": [
        "absl-py >=1.0.0",
        "astunparse >=1.6.0",
        "flatbuffers >=24.3.25",
        "gast !=0.5.0,!=0.5.1,!=0.5.2,>=0.2.1",
        "google-pasta >=0.1.1",
        "h5py >=3.10.0",
        "libclang >=13.0.0",
        "ml-dtypes <0.5.0,>=0.3.1",
        "opt-einsum >=2.3.2",
        "packaging",
        "protobuf !=4.21.0,!=4.21.1,!=4.21.2,!=4.21.3,!=4.21.4,!=4.21.5,<5.0.0dev,>=3.20.3",
        "requests <3,>=2.21.0",
        "setuptools",
        "six >=1.12.0",
        "termcolor >=1.1.0",
        "typing-extensions >=3.6.6",
        "wrapt >=1.11.0",
        "grpcio <2.0,>=1.24.3",
        "tensorboard <2.18,>=2.17",
        "keras >=3.2.0",
        "tensorflow-intel ==2.17.0 ; platform_system == \"Windows\"",
        "tensorflow-io-gcs-filesystem >=0.23.1 ; python_version < \"3.12\"",
        "numpy <2.0.0,>=1.23.5 ; python_version <= \"3.11\"",
        "numpy <2.0.0,>=1.26.0 ; python_version >= \"3.12\"",
        "nvidia-cublas-cu12 ==12.3.4.1 ; extra == 'and-cuda'",
        "nvidia-cuda-cupti-cu12 ==12.3.101 ; extra == 'and-cuda'",
        "nvidia-cuda-nvcc-cu12 ==12.3.107 ; extra == 'and-cuda'",
        "nvidia-cuda-nvrtc-cu12 ==12.3.107 ; extra == 'and-cuda'",
        "nvidia-cuda-runtime-cu12 ==12.3.101 ; extra == 'and-cuda'",
        "nvidia-cudnn-cu12 ==8.9.7.29 ; extra == 'and-cuda'",
        "nvidia-cufft-cu12 ==11.0.12.1 ; extra == 'and-cuda'",
        "nvidia-curand-cu12 ==10.3.4.107 ; extra == 'and-cuda'",
        "nvidia-cusolver-cu12 ==11.5.4.101 ; extra == 'and-cuda'",
        "nvidia-cusparse-cu12 ==12.2.0.103 ; extra == 'and-cuda'",
        "nvidia-nccl-cu12 ==2.19.3 ; extra == 'and-cuda'",
        "nvidia-nvjitlink-cu12 ==12.3.101 ; extra == 'and-cuda'"
    ]
}

(this was generated for all dependencies of NumPy, i.e. all packages containing a numpy requirement, using a modified version of https://github.com/watertap-org/getinfo; full output here: dependenciesof-numpy.json)

As far as I can tell, the impact for us is:

• We'll have to continue testing and working with NumPy v1 for developer installs, unless we split the current dev installation into two: one with Tensorflow (restricting us to NumPy v1) and one without (where NumPy v2 will be installed)
• Even if we continue to use a single dev installation with Tensorflow, NumPy v2 will still get tested as part of the user-mode installation in the integration checks (which are currently failing)
• Once we fix those failures, we should be able to test (and therefore support, to the extent that uses of the NumPy API in IDAES are exercised by our test suite) compatibility for both NumPy major versions (which is good!)

The NumPy failures mostly look straightforward to fix, so I'll try to do that now.