Testing
Package
Other

plaid-ops

Standardized operations on PLAID (Physics Learning AI Datamodel) samples and datasets.

Warning

The code is still in its initial configuration stages; interfaces may change. Use with care.

1 Using the library

First create an environment using conda/mamba:

mamba create -n plaid-ops python=3.11 hdf5 "vtk>=9.4" pycgns-core muscat-core=2.5 -c conda-forge
mamba activate plaid-ops

User mode

Relies on the published PyPi package:

pip install plaid-ops

Developper mode

Installing from the sources:

pip install -e .[dev]

Note: this will install the last stable version of PLAID.

2 Core concepts

plaid-ops provides high-level utilities to manipulate meshes and fields in PLAID datasets:

• Compute dataset-wide properties (e.g., bounding boxes)
• Project fields between unstructured meshes and regular rectilinear grids
• Transfer fields from one dataset to another with consistent sample/time alignment
• Perform mesh-based feature engineering (e.g., signed-distance function)
• Visualize results with simple helpers

It builds on top of PLAID and uses Muscat FE operators under the hood for accurate interpolations.

3 Going further

See the documentation for a concise getting started guide and end-to-end examples:

• Quickstart
• Examples & Tutorials
• API reference