Uncertainty-modelling value

Use uncertainty_modelling_value to compare a declared deterministic point-estimate solution with an explicitly enumerated nonanticipative policy class.

import json
from pathlib import Path
from voiage import uncertainty_modelling_value

request = json.loads(Path("request.json").read_text())
result = uncertainty_modelling_value(request).to_contract_dict()
print(result["decomposition"])

The v1 result exposes the expected-value problem and selected deterministic solution, EEV, recourse/stochastic-program value, wait-and-see value, VSS/EVIU, EVPI, complete ties, feasibility and exact-enumeration assurance. It supports both minimization and maximization. One decision per shared history encodes nonanticipativity; every candidate policy supplies one outcome per finite state.

Run the same contract from the CLI:

voiage --format json calculate-uncertainty-modelling-value request.json

This is an uncertainty-representation diagnostic, not an acquisition model. No samples or signals are purchased or observed. DVSS/VMS, approximate solvers, external solver claims and non-expected-value risk criteria are deferred. Python execution is experimental; Rust, R and Julia are not implemented and Mojo remains external.