ctf4science.data_module

Data Module for CTF models, provides loading and configuration utilities for CTF datasets.

This module handles train/test pairs, timestep generation for training and evaluation, validation splits, and dataset metadata used by models and the evaluation pipeline.

Functions

get_applicable_plots(dataset_name)	Return the list of applicable visualization types for the given dataset.
get_config(dataset_name)	Load and return the dataset configuration for the specified dataset.
get_metadata(dataset_name)	Return metadata for the given dataset.
get_prediction_timesteps(dataset_name, pair_id)	Return time steps at which prediction matrices must be evaluated after training.
get_training_timesteps(dataset_name, pair_id)	Return time steps for each training matrix for the given dataset and pair.
get_validation_prediction_timesteps(...[, ...])	Return prediction timesteps for the validation split.
get_validation_training_timesteps(...[, ...])	Return training timesteps used for the validation split.
load_dataset(dataset_name, pair_id[, transpose])	Load train and initialization data for a given dataset and pair ID.
load_validation_dataset(dataset_name, pair_id)	Load dataset and split into training, validation, and initialization.
parse_pair_ids(dataset_config)	Parse the pair_id configuration to determine which sub-datasets to process.