qec-canvas v0.1.0 is the first complete public vertical slice: define and inspect a small stabilizer code, run a reproducible phenomenological memory experiment, decode it in batch or sliding windows, and explain the result on a shared timeline.
Highlights
- Universal immutable code algebra and validation.
- Odd-distance rotated planar code generation with a distance-three golden fixture.
- Explicit extraction schedules, experiment plans, error locations, and sampled fault records.
- Seeded Pauli and measurement noise.
- Exact-within-model Python reference syndrome simulation.
- Transparent exhaustive batch decoding with ambiguity reporting.
- Stateful sliding-window commits and software Pauli-frame tracking.
- Running, explicitly labelled logical-observable flip parity at every decoder commit.
- Deterministic code, extraction, trace, and decoder visualisations.
- A copyable twelve-round tutorial with labelled data and measurement faults.
- A controlled beyond-distance example with an explicit logical failure.
- Independent Stim conformance for Pauli algebra, stabilizer structure, logical commutation, and distance-three single-error syndromes.
- Checked-in code-layout, extraction-circuit, and experiment-timeline figures.
- Python 3.11–3.14 support, strict typing, and 100% statement/branch coverage.
Install
Download the attached wheel and install it with Python 3.11 or newer:
python -m pip install qec_canvas-0.1.0-py3-none-any.whlThe release is intentionally not on PyPI.
Scope
The Python backend is exact within a phenomenological phase-free Pauli model. It does not simulate quantum-state amplitudes, extraction-gate noise, coherent channels, leakage, or continuous time. The exhaustive decoder is a small-instance correctness oracle, not a threshold-scale performance decoder.