q2m3 v0.1.1 — Full one-electron embedding API

Highlights

This patch release exposes the full one-electron embedding API as a stable public surface and adds faster, parallelised test infrastructure.

Public API additions

• q2m3.interfaces.build_fixed_mo_embedding_integrals — fixed-MO one-electron embedding with active-space diagnostics, exported at the top level.
• q2m3.core.estimate_resources(..., embedding_mode="full_oneelectron") — EFTQC resource estimation now distinguishes none / diagonal / full_oneelectron embedding modes and surfaces scalar diagnostics for fixed-MO one-electron embedding.
• FixedMOEmbeddingResult, FixedMOEmbeddingDiagnostics — typed result containers exported from q2m3.interfaces.

Tutorials and docs

• New end-to-end tutorial: examples/full_oneelectron_embedding.py — H2 with diagonal vs full-one-electron resource rows side-by-side.
• New documentation section: doc/source/tutorials/full-oneelectron-embedding.md.
• Architecture and core-concepts docs updated to mark the fixed-MO boundary and dynamic-coefficient limitation.

Test infrastructure

• pytest-xdist parallelism enabled (-n auto by default).
• Deterministic QPE / PySCF doubles in tests/solvation/conftest.py for stable parallel runs.
• CPU-pin JAX during tests to avoid GPU contention.

Compatibility

• Fully backward compatible with v0.1.0 public API.
• No breaking changes in existing entry points (PySCFPennyLaneConverter, QPEEngine, QuantumQMMM, solvation orchestrator).

Installation

pip install git+https://github.com/yjmaxpayne/q2m3@v0.1.1
# or via uv inside this repo
uv sync --all-extras

Citation

This release is archived on Zenodo. Use the version-specific DOI shown in the README badge for reproducibility, or the BibTeX entry in README.md.

Full Changelog

See CHANGELOG.md for the auto-generated commit list.

Diff: v0.1.0...v0.1.1