Skip to content

Changelog

Unreleased

  • Publish 19 Chinese documentation pages with localized navigation and search, shared math/link handling, and bilingual maintenance instructions.
  • Correct translated physics terminology, source links, and the homepage image; retain the paper equations and copyable agent requests.

0.1.2 — 2026-09-15

  • Use GitHub-supported LaTeX for traces, real parts, IPR, and standard errors; verify every equation on GitHub as well as the documentation website.

0.1.1 — 2026-09-15

  • State the paper Hamiltonian, triangular lattice, supported parameters, and code-to-paper conventions in both READMEs and the algorithm guide.
  • Document all active physical outputs, operator definitions, normalization, pair phase, and measurement availability with GitHub-rendered LaTeX.
  • Add a model-development guide and agent skill for coordinated lattice, hopping, interaction, solver, ED, and observable extensions.
  • Make README agent requests individually copyable.
  • Remove archived-data checksum manifests and byte-level validation gates; retain numerical data checks and calculation continuation consistency.

0.1.0 — 2026-09-15

Initial public release of BAFQMC under the MIT license.

  • Number-conserving and full Nambu bosonic auxiliary-field Monte Carlo solvers, with portable MPI Fortran and BLAS/LAPACK support.
  • Exact-diagonalization references and complete production workflows for the main and supplemental benchmark figures: 22 parameter points.
  • Original inputs, seeds, compact means and standard errors, block means, reference values, and reusable plotting code. Raw measurements are generated by the reproduction command.
  • Custom-campaign interfaces, physics and development guides, numerical tests, and continuous integration.
  • Independent finite-Fock ED, analytic Gaussian, fixed interacting-HS-field, and thermodynamic-identity tests, with real MPI execution in CI.
  • Agent instructions, research task skills, an environment doctor, and a development container for GNU Fortran, MPI, and Python ED.