PIVPy v0.2.0 Release Notes
Major release modernizing PIVPy for deep Particle Image Velocimetry (PIV) vector field post-processing, spectral turbulence analysis, vortex topology identification, out-of-core streaming, and interactive visualization.
🌟 Key Features & Highlights:
-
Kinematic Vortex Topology: Added
$\Gamma_1$ (vortex center core),$\Gamma_2$ (vortex boundary), Galilean-invariant$-criterion, Okubo-Weiss parameter {OW}$ , and noise-robust circulation vorticity. -
Advanced Turbulence Statistics & Spectra: Full Reynolds decomposition (
$\mathbf{u} = \overline{\mathbf{u}} + \mathbf{u}'$ ), 2D & radial 1D energy spectra (k)$ with exact continuous Parseval energy conservation, two-point spatial autocorrelations {uu}(r_x), R_{vv}(r_y)$, integral length scale {11}$, Taylor microscale$\lambda_T$ , and turbulent dissipation rate ($\varepsilon$ ). - Normalized Median Test & Inpainting: Westerweel & Scarano (2005) outlier detection and Laplacian harmonic inpainting.
- Velocity Gradient & Acceleration: Full strain rate tensor decomposition, maximum shear strain rate, and material acceleration \mathbf{u}/Dt$.
- Out-of-Core Streaming & IO: (1)$ memory Welford online streaming reductions (io.stream_statistics()), chunked Zarr ingestion (io.stream_directory_to_zarr()), and auto-detecting CSV reader.
- Interactive Marimo Dashboard & Visualizations: Standalone reactive Marimo explorer (pivpy.app / ds.piv.explore()) with live spectrum curves, diagnostics cards, fluid quiver animations (ds.piv.animate()), and publication figures.
- OpenPIV Skill Interoperability: Fully updated pivpy/skill.md and skills/pivpy/SKILL.md for automated LLM PIV analysis workflows.