Add Pauli input initialization

[masa10-f]

#229 has been merged into master; this PR now targets master directly.

Summary

• Add per-input positive Pauli eigenstate initialization (X+, Y+, Z+) through GraphState, Pattern, and qompile().
• Compile input initialization to Stim reset instructions (RX, RY, R) and mirror the same states in PatternSimulator.
• Import leading Stim reset instructions (R, RX, RY) as the corresponding input initialization states; reject mid-circuit resets explicitly.
• Sample all pattern measurements from exact Born probabilities by default; retain the legacy non-output 50/50 assumption with calc_prob=False.
• Bump exported .ptn files to format version 2 with .input_basis, while keeping version 1 files readable with X+ input defaults.
• Document initial reset import, simulator sampling modes, and the .ptn version 2 format.
• Add regression coverage for graph copying/composition, initialization validation, .ptn roundtrip, simulator behavior, and Stim import/export.

Rebase notes

• Merged the latest master after Add Stim circuit importer #229 landed, resolving the changelog overlap while preserving both entries.
• Kept current dependency versions and the Add Stim circuit importer #229 Stim importer fixes.
• Dropped the duplicate QEC sparse-matrix typing commit already present through Add Stim MPP stabilizer import #217.

Validation

• uv run --no-sync pytest -q -s -m "not pyzx" — 622 passed, 13 deselected.
• uv run --no-sync pytest -q -s -m pyzx — 13 passed, 622 deselected.
• uv run --no-sync ruff check . and uv run --no-sync ruff format --check .
• uv run --no-sync mypy graphqomb and uv run --no-sync pyright graphqomb
• uv run --no-sync sphinx-build -W --keep-going -b html docs/source <output>
• GitHub Actions matrix: triggered for the latest commit.

[codecov]

Codecov Report

❌ Patch coverage is 88.67925% with 18 lines in your changes missing coverage. Please review.
✅ Project coverage is 86.19%. Comparing base (133621a) to head (0291fd2).
⚠️ Report is 2 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #230      +/-   ##
==========================================
+ Coverage   86.09%   86.19%   +0.09%     
==========================================
  Files          29       29              
  Lines        4451     4577     +126     
  Branches      783      802      +19     
==========================================
+ Hits         3832     3945     +113     
- Misses        438      448      +10     
- Partials      181      184       +3     

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 5989bd7c53

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Sample Pauli-initialized non-output measurements

When simulating a pattern where an input initialized with a new non-X basis is later measured as a non-output in the same basis (for example, a one-edge input/output graph with init_axis=Axis.Z and an MZ on the input), the state is now deterministic but apply_cmd(M) still samples all non-output measurements as 50/50. Half of those runs project onto an impossible branch and normalize a zero vector, producing NaNs or wrong downstream outputs; the simulator needs to sample these measurements from the current state when the Pauli input initialization can bias them.

Useful? React with 👍 / 👎.