neural_lam.datastore.npyfilesmeps.config

Dataclasses describing the MEPS numpy-file datastore configuration.

Classes

Dataset	Contains information about the dataset, including variable names, units,
NpyDatastoreConfig	Configuration for loading and processing a dataset, including dataset
Projection	Represents the projection information for a dataset, including the type

Module Contents

class neural_lam.datastore.npyfilesmeps.config.Dataset

Contains information about the dataset, including variable names, units, and descriptions.

Attributes:

name: The name of the dataset. var_names: A list of variable names in the dataset. var_units: A list of units for each variable. var_longnames: A list of long, descriptive names for each variable. num_forcing_features: The number of forcing features in the dataset. num_timesteps: The number of timesteps per analysis sample. step_length: The time delta between consecutive timesteps. num_ensemble_members: The number of ensemble members in the dataset. remove_state_features_with_index: Indices of state features to drop when loading the dataset.

name: str

num_ensemble_members: int

num_forcing_features: int

num_timesteps: int

remove_state_features_with_index: list[int] = []

step_length: datetime.timedelta

var_longnames: list[str]

var_names: list[str]

var_units: list[str]

class neural_lam.datastore.npyfilesmeps.config.NpyDatastoreConfig

Bases: dataclass_wizard.YAMLWizard

Configuration for loading and processing a dataset, including dataset details, grid shape, and projection information.

Attributes:

dataset: An instance of Dataset containing details about the dataset. grid_shape_state: A list representing the shape of the grid state. projection: An instance of Projection containing projection details.

dataset: Dataset

grid_shape_state: list[int]

projection: Projection

class neural_lam.datastore.npyfilesmeps.config.Projection

Represents the projection information for a dataset, including the type of projection and its parameters. Capable of creating a cartopy.crs projection object.

Attributes:

class_name: The class name of the projection, this should be a valid cartopy.crs class. kwargs: A dictionary of keyword arguments specific to the projection type.

class_name: str

kwargs: dict[str, Any]