fptp-seat-audit: a numerical audit of factor-based FPTP seat distributions (Great Britain 2024 case study)

Code, data and stored outputs for the computation note "A numerical audit of factor-based FPTP seat distributions: a Great Britain 2024 case study" (draft v0.2.2, not peer reviewed). Given common factors, each party's seat distribution is computed with the Poisson-binomial recursion and a convolution across regions; the factor integral uses scrambled Sobol points with regional Gauss-Hermite rules. The note reports how the usual scramble error bar missed a regional-rule bias (Green) and an under-resolved rare outer contribution (Other), and how each was located by a different computation path. The model is a fixed computational case, not a forecasting product.

Licences: code MIT; the note and the stored outputs CC BY 4.0; data under their sources' terms (Open Parliament Licence v3.0; BES 2024 constituency results file CC BY 4.0), see DATA_LICENSES.md. Contains Parliamentary information licensed under the Open Parliament Licence v3.0.

AI use: Developed with extensive assistance from Claude Opus 5.5 (Anthropic) and reviewed with GPT 6.1 Sol (OpenAI); the author directed and reviewed the work and is responsible for the content.