igraph

from datetime import datetime

import igraph as ig
import networkx as nx

from tool import db_util, config_util

sql1 = f"""
    select link_id,snode_id,enode_id,direction,st_astext(pts) as link_wkt
    from sd_plus_link_tmp
    where adcode = 110113 and link_type != 2
"""
dbcfg = config_util.get_cfg_by_key("gd_sdmap_20230616")
links = db_util.fetch_pg(dbcfg, sql1)
print(f"{len(links)=}")


def igraph_t():
    start = datetime.now()
    g = ig.Graph(directed=True)

    # 首先，收集所有独特的节点ID
    node_ids = set()
    for link in links:
        link_id, snode_id, enode_id, direction, link_wkt = link
        node_ids.add(snode_id)
        node_ids.add(enode_id)
    node_ids_sorted = sorted(list(node_ids))
    g.add_vertices(node_ids_sorted)
    id_to_index = {nid: index for index, nid in enumerate(node_ids_sorted)}

    edges_to_add = []
    link_ids = []
    for i, link in enumerate(links):
        link_id, snode_id, enode_id, direction, link_wkt = link
        src_index = id_to_index[snode_id]
        dest_index = id_to_index[enode_id]

        if direction == 1:
            edges_to_add.append((src_index, dest_index))
        elif direction == 2:
            edges_to_add.append((dest_index, src_index))
        link_ids.append(link_id)
    g.add_edges(edges_to_add)
    g.es['link_id'] = link_ids
    node_id_to_query = 7550147220000001902
    node_index_to_query = id_to_index[node_id_to_query]
    in_incident_edges = g.es[g.incident(node_index_to_query, mode=ig.IN)]
    out_incident_edges = g.es[g.incident(node_index_to_query, mode=ig.OUT)]

    in_incident_link_ids = in_incident_edges['link_id']
    out_incident_link_ids = out_incident_edges['link_id']
    print(f'节点 {node_id_to_query} 的所有邻边 link_ids: {in_incident_link_ids}')
    print(f'节点 {node_id_to_query} 的所有邻边 link_ids: {out_incident_link_ids}')

    print((datetime.now() - start).total_seconds())


def nx_t():
    start = datetime.now()
    digraph = nx.DiGraph()
    for i, link in enumerate(links):
        link_id, snode_id, enode_id, direction, link_wkt = link
        if direction == 2:
            digraph.add_edge(enode_id, snode_id, link_id=link_id)
        else:
            digraph.add_edge(snode_id, enode_id, link_id=link_id)

    in_edges = list(digraph.in_edges(7550147220000001902, data=True))
    in_link_ids = [in_edge[2]['link_id'] for in_edge in in_edges]
    print(f"{in_link_ids=}")
    out_edges = list(digraph.out_edges(7550147220000001902, data=True))
    out_link_ids = [out_edge[2]['link_id'] for out_edge in out_edges]
    print(f"{out_link_ids=}")
    print((datetime.now() - start).total_seconds())


if __name__ == '__main__':
    igraph_t()
    nx_t()

import igraph as ig

# 创建一个无向图示例
g = ig.Graph()
g.add_vertices(6)  # 添加6个顶点
g.add_edges([
    (0, 1), (1, 2),  # 第一个连通分量
    (3, 4), (4, 5)   # 第二个连通分量
])

# 查找图中的连通分量
clustering = g.clusters()

# 打印连通分量的数量
print(f"连通分量的数量: {len(clustering)}")

# 遍历每个连通分量并打印其中的顶点
for i, component in enumerate(clustering):
    print(f"连通分量 {i+1}: {component}")

import igraph as ig

# 创建图并添加属性
g = ig.Graph()
g.add_vertices(6)
g.add_edges([(0, 1), (0, 2), (3, 4), (4, 5)])
g.es['link_id'] = range(1, g.ecount() + 1)  # 假设link_id从1开始递增

# 找到所有连通分量
components = g.clusters()

# 对于每个连通分量，收集属于该分量的边的link_id
connected_edges_link_ids = []
for component in components:
    # 从连通分量中获取全部顶点ID
    vertex_ids = component

    # 获取该连通分量中所有边
    edges = g.es.select(_within=vertex_ids)
    
    # 收集边的link_id
    link_ids = edges['link_id']
    
    connected_edges_link_ids.append(link_ids)

# 打印每个连通分量的边的link_id
for i, link_ids in enumerate(connected_edges_link_ids):
    print(f"连通分量 {i + 1} 的边的link_id: {link_ids}")