Scripts and data to reproduce main figures in the "Estimating coil features from an equilibrium" paper

This repository bundles scripts and datasets needed to reproduce the main figures in the paper "Estimating coil features from an equilibrium" by E. Rodriguez and W. Sengupta. The main figures are:

• coil_set_k1.png (Figure 1)
• regcoil_phi_comparison_simple_true.png (Figure 3)
• excursion_distance_scan.pdf (Figure 4)

All paths below are relative to the repository root.

Included files

Root files:

• README.md
• LICENSE
• requirements.txt
• .gitattributes
• .gitignore

Script files:

• scripts/streamlined_regcoil_test.py
• scripts/downsample_vmec.py
• scripts/regcoil_distance_scan.py
• scripts/regcoil_utils.py
• scripts/utils_surface_current.py
• scripts/make_coil_set_k1.py
• scripts/make_regcoil_phi_comparison_simple_true.py
• scripts/make_excursion_distance_scan.py

Figure outputs:

• figures/coil_set_k1.png
• figures/regcoil_phi_comparison_simple_true.png
• data/precise_QH/excursion_distance_scan.pdf

Data folders:

• data/precise_QH/
• data/precise_QA/
• data/W7X/

Included case families

• precise_QH
  • data/precise_QH/regcoil_scan_results_precise_QH_s_target_0.12
  • data/precise_QH/regcoil_scan_results_precise_QH_s_target_0.30
  • data/precise_QH/regcoil_scan_results_precise_QH_s_target_1.00
• precise_QA
  • data/precise_QA/regcoil_scan_results_preciseQA
  • data/precise_QA/regcoil_scan_results_preciseQA_s0p12
  • data/precise_QA/regcoil_scan_results_preciseQA_s0p72
• W7X
  • data/W7X/regcoil_scan_results_d23p4_tm_s_target_0.12
  • data/W7X/regcoil_scan_results_d23p4_tm_s_target_0.30
  • data/W7X/regcoil_scan_results_d23p4_tm_s_target_1.00

Included scripts

• scripts/streamlined_regcoil_test.py
  • Main script to make the analysis of REGCOIL on surfaces at different distances.
  • By default, does not rerun VMEC/REGCOIL (uses bundled data).
• scripts/downsample_vmec.py
  • Helper for building downsampled VMEC inputs to consider configs at different radii.
• scripts/regcoil_distance_scan.py
• scripts/regcoil_utils.py
• scripts/utils_surface_current.py
• scripts/make_coil_set_k1.py
  • Recreates figures/coil_set_k1.png.
• scripts/make_regcoil_phi_comparison_simple_true.py
  • Recreates figures/regcoil_phi_comparison_simple_true.png.
• scripts/make_excursion_distance_scan.py
  • Recreates data/precise_QH/excursion_distance_scan.pdf from bundled coil_excursion_results.pkl files.

The utility scripts were slimmed for publication by removing non-essential top-level test/demo entry points while preserving the workflow and plotting paths used in this repository.

Environment

Install Python dependencies:

pip install -r requirements.txt

This repository contains large binary data files. Install Git LFS before cloning/pushing:

git lfs install

External tools for full end-to-end reruns:

• VMEC (xvmec2000) for --run-vmec
• REGCOIL executable (regcoil) for --run-regcoil

If regcoil is not on your PATH, set:

export REGCOIL_EXECUTABLE=/absolute/path/to/regcoil

Quick start (bundled-data mode)

From this folder:

python scripts/streamlined_regcoil_test.py
python scripts/make_coil_set_k1.py
python scripts/make_regcoil_phi_comparison_simple_true.py
python scripts/make_excursion_distance_scan.py

Expected outputs:

• figures/coil_set_k1.png
• figures/regcoil_phi_comparison_simple_true.png
• data/precise_QH/excursion_distance_scan.pdf

How excursion_distance_scan.pdf is made:

• scripts/make_excursion_distance_scan.py calls make_excursion_plot(...) in scripts/regcoil_distance_scan.py.
• It reads precomputed coil_excursion_results.pkl files from:
  • data/precise_QH/regcoil_scan_results_precise_QH_s_target_1.00/
  • data/precise_QH/regcoil_scan_results_precise_QH_s_target_0.30/
  • data/precise_QH/regcoil_scan_results_precise_QH_s_target_0.12/
• It reads VMEC reference excursions from:
  • data/precise_QH/coil_excursion_results_vmec.pkl

If you want to recompute those coil_excursion_results.pkl files (instead of using bundled ones), run compute_coil_excursion_distance(load_or_compute="compute", folder=...) for each run folder before generating the PDF.

Optional full rerun

You may need additional VMEC wout inputs and external executables for a full rerun.

python scripts/streamlined_regcoil_test.py \
  --run-vmec --run-regcoil --run-regcoil-analysis \
  --vmec-executable /absolute/path/to/xvmec2000

Case-specific example for Precise QA data layout:

python scripts/streamlined_regcoil_test.py \
  --wout-name preciseQA \
  --target-dir ./data/precise_QA \
  --s-target 0.12 0.30 1.00 \
  --vmec-file ./data/precise_QA/regcoil_scan_results_preciseQA_s0p72/wout_downsampled_preciseQAs_target_0.72.nc

Notes

The bundle in this repository was done by compiling and downsizing scripts and data originally used. GitHub Copilot was used (with guidance) to achieve this.