qbiocode.apps.quvine.reproducibility package#
Reproducibility Module for QuVINE
This module provides infrastructure for reproducible benchmarking: - Centralized seed management - Dataset registry - Pre-generated graph storage - Task-specific split generation and storage - Validation checks for experimental consistency
Submodules:
- qbiocode.apps.quvine.reproducibility.dataset_registry module
- qbiocode.apps.quvine.reproducibility.graph_generator module
- qbiocode.apps.quvine.reproducibility.method_adapters module
- qbiocode.apps.quvine.reproducibility.method_runner module
- qbiocode.apps.quvine.reproducibility.ppi_workflow module
- qbiocode.apps.quvine.reproducibility.seed_manager module
- qbiocode.apps.quvine.reproducibility.split_generator module
- qbiocode.apps.quvine.reproducibility.validator module
Summary#
__all__ Classes:
Central registry for all datasets used in QuVINE experiments. |
|
Generates and saves synthetic graphs for reproducible benchmarking. |
|
Unified runner for all QuVINE methods using pre-generated data. |
|
Centralized PPI preprocessing pipeline. |
|
End-to-end workflow for centralized PPI preprocessing and split generation. |
|
Validates reproducibility constraints for QuVINE experiments. |
|
Manages canonical and derived random seeds for reproducible experiments. |
|
Generates and saves task-specific splits for reproducible benchmarking. |
|
Generates and saves synthetic graphs for reproducible benchmarking. |
__all__ Functions:
Convenience constructor for the centralized PPI preprocessing workflow. |