Create fake hourly profiles for github tests; fix logs

[patrickbrown4]

Summary

This PR addresses #22 and #19:

• A function is added to recf.py to create fake CF profiles with the new GSw_FakeData switch is active, allowing us to avoid ~12 GB of data download and caching for the runner tests
• The logs had inadvertently been reformatted by bokeh; now they're back to normal

Technical details

Implementation notes

Fake profiles

• A new make_fake_profiles() function is added to recf.py. To make the profiles somewhat realistic in terms of min/mean/max (but not temporal shape), we randomly sample from the two-state [0,1] array, where probability(1) = CF [fraction].
  • These profiles should NOT be used for any kind of analysis; they are just for testing when you want to avoid downloading the actual data via reeds/remote.py.
  • Some example histograms are below in the validation section.
• In the current implementation, we only use it for the site-level upv, wind-ons, and wind-ofs profiles.
  • Those are by far the biggest files we download from Zenodo, so avoiding their download speeds up the test and reduces the peak storage needs of the runner
  • It's probably good to keep automatically testing the other Zenodo downloads anyway
    • Though we might want to extend the new functionality to distPV, since that file is still kind of large at ~440 MB
• In runbatch.py, we check to make sure all the cases in a batch use the same GSw_FakeData setting (since we download the files shared by all runs)

Logs

The verbose logging came from this line: https://github.nrel.gov/ReEDS/ReEDS-2.0/pull/1927/files#diff-a48c953794de025c1a499dbd77153a1da54ebb6bcd68499d4509b32bc4b8477fR21. Because a bokehpivot import was added to reeds.output_calc, and because bokeh uses its own debug-level logging, all the python scripts started using debug-level logging. This PR reverts that change, so now the logs are back to normal.

Additional changes

• Fixed the git metadata at the top of meta.csv (it wasn't working since we don't have any tags on this repo yet)

Switches added/removed/changed

• GSw_FakeData (default 0): Use fake capacity factor profiles (for TESTING ONLY when you do not want to download the real CF data)

Issues resolved

• Clean up logs #19
• Programmatically create hourly profiles for github tests #22

Validation, testing, and comparison report(s)

Zero change for the Pacific test case: results-v20260416_mainM0_Pacific,v20260416_fakeM0_Pacific.pptx

The github_Pacific case finished for me locally.

Here are histograms of the CF for a couple random samples:

Checklist for author

Details to double-check

• Included comparison reports for appropriate test cases
• Documentation updated if necessary
• Code formatting standardized
• Reusable functions used where possible instead of copy/pasted code

General information to guide review

• Zero impact on results of default case
• No large data file(s) added/modified
• No substantive impact on runtime for full-US reference case
• No substantive impact on folder size for full-US reference case
• No change to process flow (runbatch.py, d_solve_iterate.py)
• No change to code organization
• No change to package requirements (environment.yml or Project.toml)

Did you use LLM tools (chatbot or copilot) in the preparation of this PR? If so, describe how

No

[ahamilton5]

Verbose logs fix works great. Thanks, Patrick.

[kodiobika]

Thanks for doing this! LGTM