Clean LROM Demonstration

This folder contains a cleaned-up notebook series for the learned reduced-operator model (LROM) experiments.

The notebooks are meant to be portable: they use the public ROSE package from PyPI, installed as:

pip install nuclear-rose

The package imports as rose.

Setup

Use a fresh environment rather than a pre-existing Anaconda/base environment. The notebooks rely on the public PyPI package nuclear-rose; older locally installed rose packages can shadow it and cause import errors.

python -m venv .venv
source .venv/bin/activate  # Windows PowerShell: .venv\Scripts\Activate.ps1
python -m pip install --upgrade pip
pip install -r requirements.txt
python -m ipykernel install --user --name clean-lrom-demo --display-name "clean-lrom-demo"

Then open the notebooks with the clean-lrom-demo kernel. In VS Code, use the kernel picker in the upper-right of the notebook. If you work from WSL, create and select the kernel inside WSL as well.

If you edit scripts/create_clean_demo_notebooks.py, regenerate the notebooks with:

python scripts/create_clean_demo_notebooks.py

Notebooks

The notebooks are generated from scripts/create_clean_demo_notebooks.py and scripts/notebook04_source.py. Edit those source files and regenerate the notebooks rather than hand-editing notebook JSON.

1. notebooks/01_fom_and_rose_basics.ipynb

   • FOM wavefunctions, phase shifts, cross sections, and a standard ROSE ROM benchmark.

2. notebooks/02_lrom_method_walkthrough.ipynb

   • Residual-fit LROM training, central-reference coordinates, raw-parameter predictors, potential-collocation predictors, and wavefunction diagnostics.

3. notebooks/03_cross_section_cat_comparison.ipynb

   • Cross-section CAT comparison between ROSE and the packed predictor LROM.

4. notebooks/04_energy_isotope_wavefunction_extrapolation.ipynb

   • Koning-Delaroche calcium isotope/energy extrapolation with wavefunction and cross-section diagnostics, predictor-point rainbows, and K/n sweeps.

The helper code lives in lrom_demo/. It is intentionally local to this demo so the notebooks can be handed to collaborators as one folder.

Cached Outputs

The notebooks are saved with rendered figures. Some expensive datasets are also cached under outputs/:

• outputs/cached_cat/: CAT-plot and representative cross-section data for notebook 03.
• outputs/cached_ae/: isotope/energy wavefunction and cross-section sweeps for notebook 04.
• outputs/notebook_outputs/: PNG exports of the notebook figures.

These caches are not required for understanding the notebooks, but they make rerunning the heavier cells much faster. If repository size is a concern, keep the executed notebooks and selected PNGs, and share the larger outputs/cached_ae files through a release artifact or separate data link.

Troubleshooting

If the first cell reports an error such as cannot import name CustomBasis from rose.basis, the active Python environment is probably importing an older rose package. Create a fresh environment with the commands above and confirm:

python -c "import rose; print(rose.__file__)"

The printed path should point inside the fresh environment's site-packages.