Ananke ABM

Agent-Based Model for synthetic population data and activity predictions

Ananke ABM is a Python package designed to connect synthetic population data with activity predictions for agent-based modeling (ABM). It aims to use Graph Neural Networks (especially Graph Attention Models) with ODE-inspired approach to solve the issue as a spatial-temporal issue.

Features

• 🤖 Machine Learning Models: Advanced ML models including Graph Neural Networks (GNN) for population and activity modeling
• 📊 Data Generation: Tools for generating synthetic population data
• 🔍 Model Inference: Comprehensive inference capabilities for trained models, was build with Melbourne data (VISTA) in mind

Installation

Prerequisites

• Python >= 3.10
• UV package manager (recommended)

Install with UV (Recommended)

# Clone the repository
git clone https://github.com/bobkatla/ananke_abm.git
cd ananke_abm

# Install the package in development mode
uv pip install -e .

# Install with development dependencies
uv pip install -e ".[dev]"

Install with pip

# Clone the repository
git clone https://github.com/bobkatla/ananke_abm.git
cd ananke_abm

# Install the package
pip install -e .

Quick Start

Basic Usage

TODO: in working still

import ananke_abm

# Import specific modules
from ananke_abm import models, utils, data_generator

# Access model submodules
from ananke_abm.models import inference, run, gnn_embed

Command Line Interface

The package provides a command-line interface: TODO: in working

# Get version and basic info
ananke --version
ananke --help

Package Structure

src/ananke_abm/
├── __init__.py              # Main package initialization
├── models/                  # Machine learning models
│   ├── inference/          # Model inference functionality
│   ├── run/                # Model execution utilities  
│   └── gnn_embed/          # Graph Neural Network embeddings
├── utils/                   # Utility functions and helpers
└── data_generator/          # Synthetic data generation tools

Development

Setting up Development Environment

1. Clone the repository:

   git clone <repository-url>
   cd ananke_abm

2. Install UV (if not already installed):

   # On macOS/Linux
   curl -LsSf https://astral.sh/uv/install.sh | sh
   
   # On Windows
   powershell -c "irm https://astral.sh/uv/install.ps1 | iex"

3. Install in development mode:

   uv pip install -e ".[dev]"

Research Context

This package is developed as part of Duc Minh (Bob) La's PhD research at Monash University focusing on:

• Synthetic population generation
• Activity-based modeling
• Agent-based modeling applications
• Graph neural networks for population modeling

Citation

If you use this software in your research, please cite: TODO: in working

Contact

Bob La
Email: duc.la@monash.edu
Institution: Monash University

This project is part of ongoing PhD research at Monash University.