Quick Start

This guide shows you how to install and use chronocratic-datasets with PyTorch Lightning.

Installation

Install the package from PyPI:

pip install chronocratic-datasets

For development (with docs tooling):

pip install -e ".[docs]"

Basic Workflow

The package provides LightningDataModule subclasses for each dataset family. A typical workflow looks like this:

Forecasting Example

from pathlib import Path

from chronocratic.datasets import ForecastingMode, WeatherDataModule

weather = WeatherDataModule(
    dataset_file_path=Path("data/weather.csv"),
    mode=ForecastingMode.UNIVARIATE,
    seq_len=24,
    forecast_horizon=168,
)

weather.prepare_data()
weather.setup()
train_loader = weather.train_dataloader()

Classification Example

from pathlib import Path

from chronocratic.datasets import UCRClassificationDataModule

module = UCRClassificationDataModule(
    dataset_folder_path=Path("data/FogiDataset1"),
    target_column_name="class",
    scale_data=True,
)

module.prepare_data()
module.setup()
train_loader = module.train_dataloader()
val_loader = module.val_dataloader()
test_loader = module.test_dataloader()

Next Steps

• See the Forecasting Modules guide for in-depth forecasting dataset usage and additional datasets (ETT, Electricity).

• See the Classification Modules guide for UCR and UEA classification benchmarks.

• Browse the DataModule API Reference reference for full parameter documentation.

• Browse the Enum API Reference reference for all enum options.