[Code scan] Missing NVE systems crash stability scoring instead of receiving the failure penalty

[njzjz]

This issue was found by a Codex global repository scan of tracked non-test files at commit 8c93925cb10b401b2b83c738bd9263fd74474468.

Relevant code

LAMBench/lambench/metrics/vishelper/metrics_calculations.py

Lines 166 to 189 in 8c93925

	def calculate_stability_results(self) -> dict[str, float]:
	"""This calculates the stability score for a given LAM."""
	stability_results = self.fetcher.fetch_stability_results()
	# filter out models with missing stability results
	stability_results = {
	model: metrics for model, metrics in stability_results.items()
	}
	stability_results = pd.DataFrame.from_dict(stability_results, orient="index")
	stability_results = stability_results.map(
	lambda cell: self._calculate_instability_error(cell)
	)
	# average over all systems
	stability_results = stability_results.mean(axis=1)
	return stability_results.to_dict()
	def _calculate_instability_error(self, cell: dict, lambda_0: float = 5e-4) -> float:
	"""
	Private method applied to each cell of the stability_results DataFrame.
	Calculates the instability error for a given LAM on a given NVE traj.
	Penalty is applied if the simulation is not complete, by returning a large value.
	"""
	if cell is None or cell.get("steps", None) is None:

Impact

DataFrame.from_dict(..., orient="index") fills missing per-system NVE entries with NaN. _calculate_instability_error() handles None, but then calls cell.get(...) on the cell value. For NaN, that raises AttributeError instead of returning the documented failure penalty.

A model with one missing NVE system can crash stability scoring rather than receiving the intended penalty.

Suggested fix

Treat non-dict, null, and NaN cells as failed simulations and return 5. Add a regression test where one model is missing a system present for another model.