upload code

1-框架拆解/1.3-图神经网络框架（DGL）/code/show_graph.py

+import dgl
+import dgl
+import networkx as nx
+import matplotlib.pyplot as plt
+import torch
+
+# 假设你已经有了一个DGL图对象bg
+# 如果没有，这里提供一个示例图的创建
+# u = torch.tensor([0, 1, 2, 3, 2])
+# v = torch.tensor([1, 2, 3, 4, 0])
+# bg = dgl.graph((u, v))
+
+def visualize_dgl_graph(g, node_labels=None, edge_labels=None, figsize=(32, 16)):
+    """
+    可视化DGL图
+    
+    参数:
+    g: DGL图对象
+    node_labels: 节点标签字典 (可选)
+    edge_labels: 边标签字典 (可选)
+    figsize: 图像大小
+    """
+    # 将DGL图转换为NetworkX图
+    nx_g = dgl.to_networkx(g)
+    
+    # 创建图形
+    plt.figure(figsize=figsize)
+    
+    # 设置布局
+    pos = nx.spring_layout(nx_g, seed=42)
+    
+    # 绘制节点
+    nx.draw_networkx_nodes(nx_g, pos, 
+                          node_color='lightblue', 
+                          node_size=500,
+                          alpha=0.8)
+    
+    # 绘制边
+    nx.draw_networkx_edges(nx_g, pos, 
+                          edge_color='gray',
+                          arrows=True,
+                          arrowsize=20,
+                          alpha=0.6)
+    
+    # 绘制节点标签
+    if node_labels is None:
+        node_labels = {i: str(i) for i in range(g.num_nodes())}
+    nx.draw_networkx_labels(nx_g, pos, node_labels, font_size=8)
+    
+    # 绘制边标签（如果提供）
+    if edge_labels is not None:
+        nx.draw_networkx_edge_labels(nx_g, pos, edge_labels)
+    
+    plt.title(f"DGL\n num_nodes: {g.num_nodes()}, num_edges: {g.num_edges()}")
+    plt.axis('off')
+    plt.tight_layout()
+    plt.show()
+
+# # 使用示例
+# visualize_dgl_graph(bg)
\ No newline at end of file

1-框架拆解/1.3-图神经网络框架（DGL）/code/图的构键.ipynb

+{
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 基本图"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "![图](..\\picture\\image.png)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Graph(num_nodes=4, num_edges=4,\n",
+      "      ndata_schemes={}\n",
+      "      edata_schemes={})\n",
+      "tensor([0, 1, 2, 3])\n",
+      "(tensor([0, 0, 0, 1]), tensor([1, 2, 3, 3]))\n",
+      "(tensor([0, 0, 0, 1]), tensor([1, 2, 3, 3]), tensor([0, 1, 2, 3]))\n"
+     ]
+    }
+   ],
+   "source": [
+    "import dgl\n",
+    "import torch as th\n",
+    "\n",
+    "# 边 0->1, 0->2, 0->3, 1->3\n",
+    "\n",
+    "u, v = th.tensor([0, 0, 0, 1]), th.tensor([1, 2, 3, 3])\n",
+    "g = dgl.graph((u, v))\n",
+    "print(g) # 图中节点的数量是DGL通过给定的图的边列表中最大的点ID推断所得出的\n",
+    "\n",
+    "# 获取节点的ID\n",
+    "print(g.nodes())\n",
+    "# 获取边的对应端点\n",
+    "print(g.edges())\n",
+    "# 获取边的对应端点和边ID\n",
+    "print(g.edges(form='all'))\n",
+    "\n",
+    "# 如果具有最大ID的节点没有边，在创建图的时候，用户需要明确地指明节点的数量。\n",
+    "g = dgl.graph((u, v), num_nodes=8)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(tensor([0, 0, 0, 1, 1, 2, 3, 3]), tensor([1, 2, 3, 0, 3, 0, 0, 1]))"
+      ]
+     },
+     "execution_count": 15,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "##转化成无向图\n",
+    "bg = dgl.to_bidirected(g)\n",
+    "bg.edges()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "Graph(num_nodes=6, num_edges=4,\n",
+       "      ndata_schemes={}\n",
+       "      edata_schemes={})"
+      ]
+     },
+     "execution_count": 16,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "## 节点特征和边特征\n",
+    "import dgl\n",
+    "import torch as th\n",
+    "g = dgl.graph(([0, 0, 1, 5], [1, 2, 2, 0])) # 6个节点，4条边\n",
+    "g\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "Graph(num_nodes=6, num_edges=4,\n",
+       "      ndata_schemes={'x': Scheme(shape=(3,), dtype=torch.float32)}\n",
+       "      edata_schemes={'x': Scheme(shape=(), dtype=torch.int32)})"
+      ]
+     },
+     "execution_count": 17,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "g.ndata['x'] = th.ones(g.num_nodes(), 3)               # 长度为3的节点特征\n",
+    "g.edata['x'] = th.ones(g.num_edges(), dtype=th.int32)  # 标量整型特征\n",
+    "g"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "tensor([1., 1., 1.])\n",
+      "tensor([1, 1], dtype=torch.int32)\n"
+     ]
+    }
+   ],
+   "source": [
+    "# 不同名称的特征可以具有不同形状\n",
+    "g.ndata['y'] = th.randn(g.num_nodes(), 5)\n",
+    "print(g.ndata['x'][1])                  # 获取节点1的特征\n",
+    "print(g.edata['x'][th.tensor([0, 3])])  # 获取边0和3的特征"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "Graph(num_nodes=4, num_edges=4,\n",
+       "      ndata_schemes={}\n",
+       "      edata_schemes={'w': Scheme(shape=(), dtype=torch.float32)})"
+      ]
+     },
+     "execution_count": 19,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "## 加权图\n",
+    "# 边 0->1, 0->2, 0->3, 1->3\n",
+    "edges = th.tensor([0, 0, 0, 1]), th.tensor([1, 2, 3, 3])\n",
+    "weights = th.tensor([0.1, 0.6, 0.9, 0.7])  # 每条边的权重\n",
+    "g = dgl.graph(edges)\n",
+    "g.edata['w'] = weights  # 将其命名为 'w'\n",
+    "g"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "Graph(num_nodes=100, num_edges=500,\n",
+       "      ndata_schemes={}\n",
+       "      edata_schemes={})"
+      ]
+     },
+     "execution_count": 20,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "## 外部创建接口\n",
+    "## scipy构建\n",
+    "import dgl\n",
+    "import torch as th\n",
+    "import scipy.sparse as sp\n",
+    "spmat = sp.rand(100, 100, density=0.05) # 5%非零项\n",
+    "dgl.from_scipy(spmat)                   # 来自SciPy"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 21,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "## networkx\n",
+    "import networkx as nx\n",
+    "nx_g = nx.path_graph(5) # 一条链路0-1-2-3-4\n",
+    "g=dgl.from_networkx(nx_g) # 来自NetworkX\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 22,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "## 保存dgl图\n",
+    "dgl.save_graphs(\"graph.dgl\", g)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 23,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "read_g=dgl.load_graphs('graph.dgl')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 24,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "([Graph(num_nodes=5, num_edges=8,\n",
+       "        ndata_schemes={}\n",
+       "        edata_schemes={})],\n",
+       " {})"
+      ]
+     },
+     "execution_count": 24,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "read_g"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "lbb",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.8.20"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}