Quickstart¶

Everything you need to start using PLAID and contributing effectively.

1 Using the library
2 Core concepts
3 Going further

1 Using the library¶

To use the library, the simplest way is to install it from the packages available:

on conda-forge for Windows, macOS and Linux:
```
conda install -c conda-forge plaid
```
or on PyPi for Linux:
```
pip install pyplaid
```

Note

Only the conda-forge packages (all operating systems) and the Linux PyPI package include a bundled pyCGNS dependency. In other situations, which we have not tested, pyCGNS must be installed separately beforehand.
On Apple Silicon, users can force an osx-64 conda environment using CONDA_SUBDIR=osx-64, allowing installation of the existing macOS-64 builds under Rosetta.

2 Core concepts¶

Sample → API: plaid.containers.sample.Sample
Dataset → API: plaid.containers.dataset.Dataset
Problem definition → API: plaid.problem_definition.ProblemDefinition
Feature identifiers → API: plaid.types.feature_types.FeatureIdentifier
Default values mechanics
On-disk format
Interoperability

3 Going further¶

Explore example examples_tutorials for practical use cases and advanced techniques.

The API documentation provides detailed information on all available classes and methods.

Two companion libraries extend the plaid standard to support machine-learning workflows in physics:

plaid-bridges: integrations with popular ML frameworks such as PyTorch Geometric.
plaid-ops: standardized operations on PLAID samples and datasets, including advanced mesh processing (some requiring a finite-element engine) powered by muscat.