Dataset Management

The datasets module provides utilities for loading and managing hypernetwork datasets.

hiper.datasets

datasets package

Provides functionality for loading, parsing, and managing hypernetwork datasets from various file formats. This package handles the conversion of external data sources into hypernetwork representations suitable for analysis and computation.

The datasets package supports standard hypernetwork data formats where each line in a dataset file represents one hyperedge by listing the node IDs that comprise it. The package automatically handles data validation, type conversion, and the construction of optimized hypernetwork objects from the parsed data.

Classes:

DataFile: Parses hypernetwork data files with robust error handling Dataset: Constructs Hypernetwork objects from parsed data files

Constants:

DATASETS: List of standard benchmark datasets available for testing

The DataFile class provides low-level file parsing with support for comments, empty lines, and flexible whitespace handling. The Dataset class builds upon this foundation to create complete hypernetwork objects ready for analysis.

Example usage:

from hiper.datasets import DataFile, Dataset, DATASETS

# Load a dataset from file datafile = DataFile(‘path/to/dataset.txt’) dataset = Dataset(‘MyDataset’, datafile)

# Access the constructed hypergraph hypernetwork = dataset.get_hypernetwork() hypernetwork.print_info()

# Work with standard benchmark datasets for dataset_name in DATASETS:

print(f’Available dataset: {dataset_name}’)

The package integrates seamlessly with the configuration system to support flexible dataset path management and standardized benchmarking workflows across different computational environments.

class hiper.datasets.DataFile(path)

Bases: object

Reads a file where each line lists node IDs for one hyperedge. Skips empty or comment lines (starting with ‘#’).

hyperedges

List of hyperedges (each a list of ints).

__init__(path)

get_hyperedges()

Return the list of parsed hyperedges.

Return type:

List[List[int]]

class hiper.datasets.Dataset(name, datafile)

Bases: object

Constructs a Hypernetwork from a DataFile.

name

Name of the dataset.

hypernetwork

Hypernetwork built from the data.

__init__(name, datafile)

get_hypernetwork()

Return the constructed Hypernetwork.

Return type:

Hypernetwork

get_name()

Return the dataset name.

Return type:

str

hiper.config

config.py

Loads project configuration from a JSON file.

hiper.config.load_config()

Load and return the project configuration from a JSON file.

Returns:

Configuration parameters (e.g., dataset path and name).

Return type:

dict