certfcm: code, data and results for "Certified scenario analysis of fuzzy cognitive maps under interval weight uncertainty"

Python implementation (certfcm) of certified scenario analysis for sigmoid and hyperbolic-tangent fuzzy cognitive maps with clamped concepts under interval uncertainty in the weights and steepness: verified interval arithmetic, a box-wise weighted-contraction well-posedness certificate, coupled slope and adjoint enclosures of scenario effects, and a verified branch-and-bound procedure. The archive contains the package, the synthetic-instance generator with all random seeds, the experiment, analysis and figure/table scripts, the raw results (JSON/JSONL) and soundness-validation logs, and unchanged copies of the openly licensed retail FCM (Petukhova & Fachada 2022, doi:10.5281/zenodo.6046893) and Foodlevers FCM (Ciolfi 2024, doi:10.5281/zenodo.14516877) data. See README.md for reproduction commands.