Activity

1. GiggleLiu

   added this to the v0.1: wme milestone on Jan 17, 2026
2. GiggleLiu

   assigned

   Morningliumengkun

   on Jan 17, 2026
3. Morningliumengkun commented on Jan 17, 2026

   

   Morningliumengkun

   on Jan 17, 2026

   Last edited by physsz

   Collaborator

   More actions

   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
4. GiggleLiu commented on Jan 17, 2026

   

   GiggleLiu

   on Jan 17, 2026

   Last edited by GiggleLiu

   MemberAuthor

   More actions

   Make this belief propagation framework a submodule, instead of an independent package.
5. GiggleLiu commented on Jan 19, 2026

   

   GiggleLiu

   on Jan 19, 2026

   MemberAuthor

   More actions

   Resolved by #17
6. GiggleLiu

   closed this as completedon Jan 19, 2026