DAAG: Directed Acyclic Adjacency Graph

A specialized graph data structure for representing layered dependencies with adjacency constraints.

Overview

The Directed Acyclic Adjacency Graph (DAAG) is a graph structure with the following properties:

1. Directed: Edges flow in one direction
2. Acyclic: No cycles are allowed
3. Adjacency constraint: Edges can only connect vertices in adjacent layers

This makes DAAG particularly useful for modeling dependencies where:

• Elements are organized in distinct layers/tiers
• Dependencies can only exist between adjacent layers
• No circular dependencies are allowed

Features

• Easy creation and manipulation of layered graph structures
• Automatic enforcement of adjacency constraints
• Visualization tools for the full graph and individual subgraphs
• Interactive filtering to explore complex dependency trees

Basic Usage

from daag import DAAG

# Create a new DAAG
graph = DAAG()

# Add vertices with their respective layers
graph.add_vertex("A", 1)  # Layer 1
graph.add_vertex("B", 2)  # Layer 2
graph.add_vertex("C", 2)  # Also Layer 2
graph.add_vertex("D", 3)  # Layer 3

# Add edges (only allowed between adjacent layers)
graph.add_edge("A", "B")
graph.add_edge("A", "C")
graph.add_edge("B", "D")
graph.add_edge("C", "D")

# Visualize the graph
graph.visualize(filename="my_graph.png")

# Visualize subgraphs for each terminal node
graph.visualize_subgraphs(directory="subgraphs")

# Interactive visualization with filtering
from daag import simple_subgraph_filtering
simple_subgraph_filtering(graph)

Installation

pip install daag

Requirements

• NetworkX
• Matplotlib

Documentation

For detailed documentation, see the full technical documentation.