The generic framework for solving approximate inference problem with BP

[GiggleLiu]

Ref: https://github.com/TensorBFS/TensorInference.jl

Using PyTorch.

[Morningliumengkun]

PyTorch Belief Propagation Implementation Plan

1. Core Objectives

1. Implement a generic Belief Propagation (BP) algorithm using PyTorch
2. Provide comprehensive mathematical description and code documentation

2. Technical Analysis

2.1 Algorithm Flow

Algorithm Flow: Initialize → Collect Messages → Process Messages → Damping Update → Convergence Check → Compute Marginals

2.2 Function Interface Summary

Function Name	Functionality	Input	Output
read_model_file	Parse UAI file	filepath	UAIModel
read_evidence_file	Parse evidence file	filepath	Dict[int, int]
BeliefPropagation.__init__	Construct BP object	UAIModel	BP object
initial_state	Initialize messages	BP object	BPState
collect_message	Factor→Variable message	BP, BPState	None
process_message	Variable→Factor message	BP, BPState, damping	None
belief_propagate	BP main loop	BP, parameters	(BPState, BPInfo)
compute_marginals	Compute marginal probabilities	BPState, BP	Dict[int, Tensor]
apply_evidence	Apply evidence	BP, evidence	BeliefPropagation

3. Project Structure and Testing

3.1 Project Structure

pytorch_bp_inference/
├── README.md
├── requirements.txt
├── setup.py (optional)
├── src/
│   ├── __init__.py
│   ├── uai_parser.py          # UAI file parsing
│   ├── belief_propagation.py  # BP core implementation
│   └── utils.py               # Utility functions
├── tests/
│   ├── __init__.py
│   ├── test_uai_parser.py
│   ├── test_bp.py
│   └── test_integration.py
├── examples/
│   ├── asia_network/
│   │   ├── main.py
│   │   └── model.uai
│   └── simple_example.py
└── docs/
    ├── mathematical_description.md
    ├── api_reference.md
    └── usage_guide.md

3.2 Testing

• Test parsing examples/asia-network/model.uai
• Test BP initialization and state creation
• Test message collection and processing
• Test convergence checking
• Test marginal computation
• Test evidence application
• Compare results with provided reference results

[GiggleLiu]

Make this belief propagation framework a submodule, instead of an independent package.

[GiggleLiu]

Resolved by #17