Script Opinions

Script Opinions

Scripts must be stateless, headless, reproducible, observable. This extends the coding-style skill for standalone Python scripts.

Scripts should be:

• General: reusable across datasets and parameters. No hardcoded paths, magic numbers, or assumptions.
• Atomic: Do one thing well — when scope grows, note it and ask whether to extend or split.
• Opinionated: Typed ingestion good; fallbacks and mode switches bad.

Non-negotiables

These exist to make scripts debuggable on a cluster you don't control — they are infrastructure, not style.

1. Top-to-bottom execution. No hidden state, no sys.path hacks. Anchor imports to __file__ or pip-install the repo.
2. argparse for all config. No Hydra, no env-var configs. (If the repo already uses a framework, match it and note the deviation.)
3. stdlib logging only. logger.info/debug not print. (Same: match existing frameworks if present.)
4. Headless. No plt.show(), display(), input(). Use headless backends.
5. Run directory per invocation with config snapshot, logs/, artifacts/.
6. Fail fast. Validate paths and assumptions before heavy work.
7. Atomic writes. Write to <path>.tmp, then os.replace(tmp, path). For checkpoints and anything preemption could corrupt.
8. Strict imports. Loud ImportError at top level unless truly optional.
9. No env vars for config between script and submitter. Explicit args only.

Standard skeleton

The skeleton is a floor, not a ceiling — instrument, annotate, and structure the core logic however best serves the script. Use @dataclass by default; pydantic.BaseModel only for runtime validation.

• @dataclass ExperimentConfig — typed, flat, hierarchical field names.
• parse_args() -> ExperimentConfig — argparse. Pull defaults from dataclass.
• setup_dir_save(config) -> str — creates dir_save/{logs,artifacts,checkpoints}.
• setup_logging(dir_save, level) — dual handlers: FileHandler(logs/run.log) + StreamHandler(stdout). Both get the full log — no post-hoc copying from SLURM stdout.
• set_determinism(seed, deterministic_torch=True), save_config(dir_save, config) (writes .yaml or .json).
• run_experiment(config, logger) — the scientific meat. Readable, flat. The design here is yours — bring judgment about what to measure, log, and save.
• main() wires: parse → dir → log → config → determinism → run.

Run directory layout

<dir_save>/
├── config.yaml        # parsed args snapshot
├── logs/run.log       # mirrored to stdout
├── artifacts/         # arrays, metrics, figures
└── checkpoints/       # optional; resumable state

CLI arg names

--dir_save, --name_run, --path_config, --path_data_*, --seed, --deterministic, --save-plots, --checkpoint-every-seconds, --resume.

Long-running jobs

• Preemption safety: --checkpoint-every-{seconds,steps} + --resume. Atomic writes. Validate checkpoint before new work.
• Large I/O: try to implement lazy loading, streaming, or chunking.
• Multiprocess: each worker opens its own file handle. Never share HDF5/zarr handles across processes.
• Analysis plots: save underlying arrays alongside figures for local regeneration.