mcframework.validation.ConvergenceReport

class mcframework.validation.ConvergenceReport

Bases: object

Outcome of a convergence check against a simulation’s analytic reference.

Attributes:

status{“pass”, “fail”, “no-oracle”}

"no-oracle" if the simulation does not declare an analytic_reference(); otherwise "pass"/"fail" from within_tol.

nint

Number of simulation draws used.

oraclefloat or None

The known reference value (None when no oracle exists).

estimatefloat or None

The Monte Carlo estimate of the mean.

abs_errorfloat or None

|estimate - oracle|.

rel_errorfloat or None

abs_error / |oracle| (None when oracle is 0 or missing).

sefloat or None

Standard error of the mean.

ci_low, ci_highfloat or None

Confidence-interval endpoints for the mean.

within_cibool

Whether oracle lies within [ci_low, ci_high].

within_tolbool

Whether abs_error <= sigma_tol * se.

sigma_tolfloat

Standard-error multiplier used for within_tol.

confidencefloat

Confidence level used for the interval.

reference_sourcestr

Citation for the oracle, copied from the simulation.

reference_kindstr

Kind of reference: "closed-form", "benchmark", "limit", or "".

Methods

classmethod __new__(*args, **kwargs)

__init__(status: str, n: int, oracle: float | None = None, estimate: float | None = None, abs_error: float | None = None, rel_error: float | None = None, se: float | None = None, ci_low: float | None = None, ci_high: float | None = None, within_ci: bool = False, within_tol: bool = False, sigma_tol: float = 5.0, confidence: float = 0.99, reference_source: str = '', reference_kind: str = '') → None