AGENT SKILL

pywavelets

Routes PyWavelets users to discrete transforms, wavelet-family and CWT workflows, and wavelet-packet tree operations.

License 许可
MIT
最近更新
2026-09-03
Stars
★ 351
Forks
27

01 前言 Front Matter

name
pywavelets
license
MIT
metadata
{ "disco-role": "operating" }
description
Routes PyWavelets users to discrete transforms, wavelet-family and CWT workflows, and wavelet-packet tree operations.
disable-model-invocation
true

02 资源文件 7

03 SKILL.md 正文

PyWavelets

PyWavelets provides wavelet transforms in Python through the pywt package.

Start here

  • Install from PyPI: python -m pip install PyWavelets

  • From this checkout: python -m pip install -e . after the source build dependencies are available.

  • The current checkout's source build metadata requires Python 3.12+.

  • Core runtime only needs NumPy; Matplotlib is optional for plot-heavy demo parity.

  • Run python scripts/check_pywavelets_install.py from this skill directory for a no-network smoke check.

  • Minimal import check:

    python - <<'PY'
    import pywt
    print(pywt.__version__)
    PY
    
  • Read references/repo-provenance.md before deciding whether this skill still matches the current checkout.

  • Read references/package-overview.md for the package map and bundled sample data.

  • Read references/troubleshooting.md when imports, builds, transform parameters, or packet paths fail.

Route by task

  • Discrete wavelet transforms, inverse transforms, multilevel coefficient trees, stationary transforms, multiresolution analysis, coefficient packing, thresholding, signal extension modes, padding, or fully separable transforms: use sub-skills/discrete-transforms/SKILL.md.
  • Wavelet families, custom wavelets, wavelet functions, continuous wavelet transforms, scale/frequency conversion, or bundled demo signals/data: use sub-skills/wavelets-and-cwt/SKILL.md.
  • Wavelet packet trees in 1D/2D/ND, path-based node lookup, and packet reconstruction: use sub-skills/wavelet-packets/SKILL.md.

Common defaults

  • Prefer wavelet='db1' or wavelet='haar' for tiny smoke checks.
  • Prefer mode='symmetric' for decimated DWT/IDWT workflows unless the task says otherwise.
  • Use mode='periodization' when the workflow is explicitly SWT-based or when mra(..., transform='swt') is requested.
  • Use pywt.Modes.modes to confirm supported signal extension modes.
  • Use pywt.wavelist() and pywt.families() before choosing a custom wavelet family.
  • Use pywt.data.camera(), pywt.data.ascent(), pywt.data.aero(), pywt.data.ecg(), or pywt.data.nino() as bundled sample inputs.

What this skill does not do

  • It does not replace the focused subskills for transform, CWT, or packet workflows.
  • It does not depend on the original checkout remaining present after extraction.
  • It does not require plot windows, downloads, or accelerator backends for the core workflows in this repo.

Bundled helpers

  • scripts/check_pywavelets_install.py checks importability and representative transform, CWT, and packet smoke paths.

Reference files

  • references/package-overview.md summarizes the public package surface and source-build expectations.
  • references/data-and-demo-signals.md lists bundled example arrays and demo signals that are safe for smoke checks.
  • references/troubleshooting.md collects cross-cutting install, import, build, and transform failures.

04 同仓库更多 Skills

VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-triton-opt

Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-solve

END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-skill

PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-pattern-fusion

Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-orchestrate

Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-feedback

Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).

★ 351