noether.core.schemas.aero

Submodules

• noether.core.schemas.aero.callbacks
• noether.core.schemas.aero.dataset
• noether.core.schemas.aero.pipeline

Classes

SurfaceVolumeEvaluationMetricsCallbackConfig	Configuration for surface/volume evaluation metrics callback.
AeroDatasetConfig	Dataset configuration for aerodynamic CFD datasets.
AeroCFDPipelineConfig	Pipeline configuration for aerodynamic CFD datasets.

Package Contents

class noether.core.schemas.aero.SurfaceVolumeEvaluationMetricsCallbackConfig(/, **data)

Bases: noether.core.schemas.callbacks.PeriodicDataIteratorCallbackConfig

Configuration for surface/volume evaluation metrics callback.

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:

data (Any)

name: Literal['SurfaceVolumeEvaluationMetricsCallback'] = 'SurfaceVolumeEvaluationMetricsCallback'

forward_properties: list[str] = []

List of properties in the dataset to be forwarded during inference.

chunked_inference: bool = False

If True, perform inference in chunks over the full simulation geometry.

chunk_properties: list[str] = []

List of properties in the dataset to be chunked for chunked inference.

batch_size: int = None

Batch size for evaluation. Currently only batch_size=1 is supported.

chunk_size: int | None = None

Size of each chunk when performing chunked inference.

sample_size_property: str | None = None

Property in the batch to determine the sample size for chunking.

validate_config()

Return type:

SurfaceVolumeEvaluationMetricsCallbackConfig

class noether.core.schemas.aero.AeroDatasetConfig(/, **data)

Bases: noether.core.schemas.dataset.StandardDatasetConfig

Dataset configuration for aerodynamic CFD datasets.

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:

data (Any)

pipeline: noether.core.schemas.aero.pipeline.AeroCFDPipelineConfig

Config of the pipeline to use for the dataset.

filter_categories: tuple[str, Ellipsis] | None = None

class noether.core.schemas.aero.AeroCFDPipelineConfig(/, **data)

Bases: noether.core.schemas.dataset.PipelineConfig

Pipeline configuration for aerodynamic CFD datasets.

Create a new model by parsing and validating input data from keyword arguments.

Raises [ValidationError][pydantic_core.ValidationError] if the input data cannot be validated to form a valid model.

self is explicitly positional-only to allow self as a field name.

Parameters:

data (Any)

num_surface_points: int

Number of surface points to sample as input for the model.

num_volume_points: int

Number of volume points to sample as input for the model.

num_surface_queries: int | None = None

Number of surface queries for the output function. If 0 or None, no query points are sampled.

num_volume_queries: int | None = None

Number of volume queries for the output function. If 0 or None, no query points are sampled.

use_physics_features: bool = False

Whether to use physics features (SDF, normals) alongside input coordinates.

dataset_statistics: noether.core.schemas.statistics.AeroStatsSchema | None = None

Dataset statistics for normalization of input features.

sample_query_points: bool = True

Whether to sample query points. If False, query points are duplicated from encoder inputs.

num_supernodes: int = 0

Number of supernodes (for UPT).

num_geometry_supernodes: int | None = None

Number of geometry supernodes (for AB-UPT).

num_geometry_points: int | None = None

Number of geometry points to sample (for AB-UPT).

num_volume_anchor_points: int | None = 0

Number of volume anchor points to sample for AB-UPT.

num_surface_anchor_points: int | None = 0

Number of surface anchor points to sample for AB-UPT.

seed: int | None = None

Random seed for sampling processes.

data_specs: noether.core.schemas.dataset.AeroDataSpecs

Data specifications for the pipeline.