release: 0.2.0

[chaoming0625]

Prepares the 0.2.0 release: bumps brainevent/_version.py to 0.2.0 and turns the existing [Unreleased] section of changelog.md into a full [0.2.0] entry covering everything landed since v0.1.2.

Why a minor bump

Breaking changes since v0.1.2:

• Bare todense() / tocsr() / tocsc() / tocoo() on the JITC families now raise NotImplementedError; callers go through the new mat.mv / mat.mm materialization views, because the mv (32-lane) and mm (AW-T4) light kernels draw different matrices.
• jits / jitn / jitu require the new matrix_mode keyword.
• CSR / CSC now validate their structure at construction, and index_dtype=int64 is rejected.
• Re-registering an FFI target with changed content raises KernelRegistrationError.
• 154 explicit CUDA entry points and the brainevent.pararnn subpackage are removed.

Plus a behavioural change worth flagging loudly: the JITC numba kernels were rebuilt on the CUDA light-RNG walk, so CPU and GPU finally draw bit-identical matrices — but CPU results recorded against 0.1.x will not reproduce.

Scope covered

PR
#179	jit normal/uniform/scalar CPU/GPU parity, f64 indptr in CSR, Dense, hybrid scheduler config
#188	CUDA source scan: shared-memory OOB, INT32_MAX index wrap, restored CSRMV warp tier
#187	operator-registration hardening (19 audit findings)
#185	unused CUDA operators, orphan headers, dead dispatch paths
#184 / #186	test co-location + the AGENTS.md rule
#183	package-wide dead-code removal and duplicate consolidation
#182	CONTRIBUTING/SECURITY/CoC rewrite, jax 0.11.x validated
#181	CLAUDE.md → AGENTS.md
#177	remove the pararnn module
#175 / #180	mypy 2.3.0, init pytest fix

Verification

• python -c "import brainevent" reports 0.2.0; __all__ stable at 165 (4 additions since 0.1.2: Dense, HybridConfig, get_hybrid_config, init_csr_config)
• pytest brainevent/_deprecation_test.py — 55 passed

🤖 Generated with Claude Code

[sourcery-ai]

Reviewer's Guide

Prepares the 0.2.0 release by bumping the library version and adding a detailed 0.2.0 section to the changelog that documents breaking changes, new features (Dense representation, int64 indptr support, tunable CSR hybrid scheduler, cublas backend, JITC materialization views, numba CPU kernels, light-RNG helpers, matrix_mode), structural validations, and removal of the pararnn subpackage.

Flow diagram for CSR hybrid config resolution order

flowchart TD
    Start([init_csr_config / get_hybrid_config invoked])
    Env{BRAINEVENT_CSR_HYBRID_CONFIG set?}
    Cache{csr_hybrid_config.json has device_kind entry?}
    Default[Use defaults baked into .cu sources]
    UseEnv[Use config from BRAINEVENT_CSR_HYBRID_CONFIG]
    UseCache[Load per-GPU HybridConfig from cache]
    Benchmark[Benchmark candidate configs with production CSR kernel]
    Save[Persist winning HybridConfig to csr_hybrid_config.json]
    Output([HybridConfig and task capacity returned])

    Start --> Env
    Env -- Yes --> UseEnv
    Env -- No --> Cache
    Cache -- Yes --> UseCache
    Cache -- No --> Default
    UseEnv --> Output
    UseCache --> Output
    Default --> Benchmark
    Benchmark --> Save
    Save --> Output

Loading

File-Level Changes

Change	Details	Files
Add detailed 0.2.0 release notes describing breaking changes, new features, and internal updates.	Document JITC CPU/GPU parity and resulting non-reproducibility for 0.1.x CPU results. Document new mat.mv / mat.mm materialization views and required matrix_mode argument on JITC APIs. Describe added Dense representation, int64 indptr support for CSR/CSC, tunable CSR hybrid scheduler configuration, and cublas dense backend. Describe structural validation for CSR/CSC, index coercions, CSR task capacity single-sourcing, and removal of the pararnn subpackage. Record updated dependency requirements and internal mypy and test-suite changes.	changelog.md
Bump package version from 0.1.2 to 0.2.0 for the new release.	Update __version__ string to 0.2.0. Ensure __version_info__ is derived from the new version string.	brainevent/_version.py

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.

[sourcery-ai]

Hey - I've reviewed your changes and they look great!

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!