# Graph Theory > Graph algorithms including shortest paths, network flow, matching, connectivity, coloring, and community detection for network analysis. - Skill: `ffsshhttiikk/graph-theory` (Agent Skill) - Install (CLI): `npx skillmds@latest add ffsshhttiikk/graph-theory` - Raw SKILL.md: https://api.skillmd.com/api/skills/ffsshhttiikk/graph-theory/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - License: MIT - Author: ffsshhttiikk (https://skillmd.com/u/ffsshhttiikk) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/ffsshhttiikk/graph-theory --- # Graph Theory ## What I Do I provide comprehensive graph theory tools including shortest path algorithms, network flow, maximum matching, connectivity analysis, graph coloring, and community detection for network analysis and optimization. ## When to Use Me - Network routing and optimization - Social network analysis - Transportation and logistics - Dependency resolution - Recommendation systems - Web page ranking ## Core Concepts - **Shortest Paths**: Dijkstra, Bellman-Ford, Floyd-Warshall - **Network Flow**: Max flow, min cut, Ford-Fulkerson - **Matching**: Maximum matching, bipartite matching - **Connectivity**: Components, articulation points, bridges - **Graph Coloring**: Chromatic number, greedy coloring - **Centrality Measures**: Degree, betweenness, PageRank - **Community Detection**: Louvain, Girvan-Newman - **Planar Graphs**: Euler's formula, Kuratowski's theorem ## Code Examples ### Graph Representation ```python import numpy as np from collections import defaultdict, deque class Graph: def __init__(self, directed=False): self.adj_list = defaultdict(list) self.directed = directed def add_edge(self, u, v, weight=1): self.adj_list[u].append((v, weight)) if not self.directed: self.adj_list[v].append((u, weight)) def dijkstra(self, source): dist = {node: float('inf') for node in self.adj_list} dist[source] = 0 pq = [(0, source)] while pq: d, u = heapq.heappop(pq) if d > dist[u]: continue for v, w in self.adj_list[u]: if dist[u] + w < dist[v]: dist[v] = dist[u] + w heapq.heappush(pq, (dist[v], v)) return dist g = Graph() g.add_edge('A', 'B', 4) g.add_edge('A', 'C', 2) g.add_edge('B', 'C', 1) g.add_edge('B', 'D', 5) g.add_edge('C', 'D', 8) ``` ### BFS and DFS ```python def bfs(graph, start): visited = set() queue = deque([start]) visited.add(start) order = [] while queue: node = queue.popleft() order.append(node) for neighbor, _ in graph.adj_list[node]: if neighbor not in visited: visited.add(neighbor) queue.append(neighbor) return order def dfs(graph, start, visited=None): if visited is None: visited = set() visited.add(start) order = [start] for neighbor, _ in graph.adj_list[start]: if neighbor not in visited: order.extend(dfs(graph, neighbor, visited)) return order print(f"BFS: {bfs(g, 'A')}") print(f"DFS: {dfs(g, 'A')}") ``` ### Maximum Flow (Ford-Fulkerson) ```python def bfs_path(graph, source, sink, parent): visited = set([source]) queue = deque([source]) while queue: u = queue.popleft() for v, capacity in graph.adj_list[u]: if v not in visited and capacity > 0: visited.add(v) parent[v] = (u, capacity) queue.append(v) if v == sink: return True return False def max_flow(graph, source, sink): parent = {} max_flow = 0 while bfs_path(graph, source, sink, parent): path_flow = float('inf') v = sink while v != source: u, capacity = parent[v] path_flow = min(path_flow, capacity) v = u max_flow += path_flow v = sink while v != source: u, capacity = parent[v] graph.update_edge(u, v, -path_flow) graph.update_edge(v, u, path_flow) v = u return max_flow ``` ### PageRank ```python def pagerank(adjacency_matrix, damping=0.85, max_iter=100, tol=1e-6): n = adjacency_matrix.shape[0] out_degree = adjacency_matrix.sum(axis=1, keepdims=True) transition = adjacency_matrix / np.where(out_degree > 0, out_degree, 1) pagerank = np.ones(n) / n for _ in range(max_iter): new_pagerank = damping * (transition.T @ pagerank) + (1 - damping) / n if np.linalg.norm(new_pagerank - pagerank) < tol: break pagerank = new_pagerank return pagerank adj = np.array([[0, 1, 1], [1, 0, 1], [1, 1, 0]]) ranks = pagerank(adj) print(f"PageRank: {ranks}") ``` ### Community Detection (Louvain) ```python def modularity(adjacency, community_labels): m = adjacency.sum() / 2 n = len(community_labels) Q = 0 for i in range(n): for j in range(n): if community_labels[i] == community_labels[j]: A_ij = adjacency[i, j] k_i = adjacency[i].sum() k_j = adjacency[j].sum() Q += (A_ij - k_i * k_j / (2 * m)) return Q / (2 * m) adjacency = np.array([ [0, 1, 1, 0, 0], [1, 0, 1, 0, 0], [1, 1, 0, 1, 1], [0, 0, 1, 0, 1], [0, 0, 1, 1, 0] ]) communities = [0, 0, 1, 1, 1] mod = modularity(adjacency, communities) print(f"Modularity: {mod:.4f}") ``` ## Best Practices 1. **Graph Representation**: Use adjacency lists for sparse graphs 2. **Algorithm Selection**: Match algorithm to graph properties 3. **Memory Efficiency**: Consider sparse matrix formats 4. **Parallelization**: Some algorithms parallelize well 5. **Special Cases**: Trees, DAGs have specialized algorithms ## Common Patterns ```python import heapq # Dijkstra with path reconstruction def dijkstra_with_path(graph, source, target): dist = {node: float('inf') for node in graph.adj_list} parent = {} dist[source] = 0 pq = [(0, source)] while pq: d, u = heapq.heappop(pq) if d > dist[u]: continue if u == target: break for v, w in graph.adj_list[u]: if dist[u] + w < dist[v]: dist[v] = dist[u] + w parent[v] = u heapq.heappush(pq, (dist[v], v)) path = [] node = target while node is not None: path.append(node) node = parent.get(node) return dist[target], path[::-1] ``` ## Core Competencies 1. Shortest path algorithms 2. Maximum flow computation 3. Graph connectivity analysis 4. Centrality and ranking algorithms 5. Community detection methods