statistics module
This module contains useful classes used to obtain metadata from the dataset generation processes.
- class src.dataset_creation.statistics.DatasetStats(stats: dict[str, Metadata])[source]
Bases:
objectClass used to store and save metadata from multiple runs to file.
- Parameters:
- statsdict[str, Metadata]
dictionary containing statistics relative to input files.
Methods
write_metadata_files(directory)Writes on disk the dataset statistics.
- class src.dataset_creation.statistics.Metadata(*, seed: int, track_type: Literal['PSS', 'AC_rail', 'subgrade'], ballast_simulation_seed: int | None, fouling_level: float, is_fouled: bool, general_deterioration: float, layer_sizes: dict[str, float], general_water_content: float, water_infiltrations: tuple[float, float, float], layer_water_ranges: tuple[tuple[float, float], tuple[float, float], tuple[float, float]], sleepers_material: Literal['wood', 'steel', 'concrete'], sleeper_positions: list[tuple[float, float, float]], fouling_material: tuple[float, float] | None = None, pss_material: tuple[float, float] | None = None, subsoil_material: tuple[float, float] | None = None)[source]
Bases:
BaseModelUtility class to keep the sample metadata information