Skip to content

AGENTS.md

Latest commit

 

History

History
33 lines (27 loc) · 2.88 KB

AGENTS.md

File metadata and controls

33 lines (27 loc) · 2.88 KB

Repository Guidelines

Project Structure & Module Organization

  • Root contains Julia scripts for ACOPF dataset generation: acopf_gen_parallel.jl (parallel scenario batches), acopf_gen.jl (single-run/debug), acopf_model.jl (JuMP/PowerModels model builder with slack logic), perturbations.jl (load perturbations), and hdf5_writer.jl (chunked output + verification helpers).
  • Dependencies are declared in Project.toml; artifacts such as PGLib OPF cases are managed via Artifacts.toml and downloaded on demand.
  • Outputs are written under results/<instance> by default; generated HDF5 chunks follow chunk_####.h5 naming.

Build, Test, and Development Commands

  • Install deps: julia --project -e 'using Pkg; Pkg.instantiate()'.
  • Parallel dataset generation (typical workflow): julia --project acopf_gen_parallel.jl --instance=pglib_opf_case118_ieee --nprocs=8 --n_scenarios=50 --chunk_size=5 --output_dir=results/case118.
  • Single run for debugging: julia --project acopf_gen.jl --solver=madnlp --instance=pglib_opf_case24_ieee_rts --write_output=true (writes test_scenario_001.h5).
  • Inspect an HDF5 chunk: julia --project -e 'using HDF5; h5open("results/case118/chunk_0001.h5") do f; println(keys(f)); end'.

Coding Style & Naming Conventions

  • Julia style: 4-space indentation, snake_case for variables/functions, CONSTANT_CASE for constants.
  • Prefer type-stable functions; avoid global mutation except for configuration constants.
  • Keep CLI flags lowercase with hyphens; mirror existing ArgParse naming.
  • Log messages should stay concise and progress-oriented (see ProgressMeter usage in acopf_gen_parallel.jl).

Testing Guidelines

  • No formal test suite yet; validate changes by running a small scenario set (e.g., --n_scenarios=2 --chunk_size=1) and confirming HDF5 outputs are created.
  • When modifying model or perturbation logic, compare solver status/objective before and after changes on a fixed instance to catch regressions.
  • Record solver options used (Ipopt vs MadNLP, HSL/MA27 availability) because numerical differences matter.

Commit & Pull Request Guidelines

  • Follow the existing short, imperative commit style (refactored to make it callable with args, added ArgParse). Keep subjects under ~60 chars.
  • PRs should include: a short summary of the change, sample commands used for verification, any new flags or defaults, and notes on output directories/files touched.
  • Add logs or screenshots only when they clarify solver behavior or performance impacts; otherwise keep PRs text-focused.

Security & Configuration Notes

  • PGLib artifacts download automatically; avoid committing downloaded data or generated HDF5 outputs.
  • HSL/MA27 has licensing constraints—ensure you are authorized to use the binaries you point to; do not commit private license files.
  • For GPU paths (MadNLPGPU), clearly document CUDA version and driver assumptions in your PR description.