小书匠 

Graph 

图论 

学过线性代数的都了解矩阵,在矩阵上的文章可做的很多,什么特征矩阵,单位矩阵等.grpah存储可以使用矩阵,比如graph的邻接矩阵,权重矩阵等,这节主要是在等到graph后,如何快速得到这些信息.详细官方文档在

目录:

---

注意:如果代码出现找不库,请返回第一个教程,把库文件导入.

12.graph和其他数据格式转换

12.1graph与字典(Dict)

1. #从字典生成图 

   ---

2. dod = {
   0: {
   1: {
   'weight': 1}}}  

   ---

3. G = nx.from_dict_of_dicts(dod) #或G=nx.Graph(dpl) 

   ---

4. plt.subplots(1,1,figsize=(6,3)) 

   ---

5. nx.draw(G, with_labels=True, font_weight='bold') 

   ---

6. plt.axis('on') 

   ---

7. plt.xticks([]) 

   ---

8. plt.yticks([]) 

   ---

9. plt.show() 

   ---

10.  

    ---

11. #图转换为字典 

    ---

12. print(nx.to_dict_of_dicts(G)) 

    ---

graph与字典(Dict)示例

输出:

{0: {1: {'weight': 1}}, 1: {0: {'weight': 1}}}

---

12.2graph与列表(List)

1. #从列表中创建graph 

   ---

2. dol = {
   0: [1,2,3]} 

   ---

3. edgelist = [(0, 1),(0,3),(2,3)] 

   ---

4.  

   ---

5. G1 = nx.from_dict_of_lists(dol) #或G=nx.Graph(dol) 

   ---

6. G2=nx.from_edgelist(edgelist) 

   ---

7.  

   ---

8. #显示graph 

   ---

9. plt.subplots(1,2,figsize=(15,3)) 

   ---

10. plt.subplot(121) 

    ---

11. nx.draw(G1, with_labels=True, font_weight='bold') 

    ---

12. plt.axis('on') 

    ---

13. plt.xticks([]) 

    ---

14. plt.yticks([]) 

    ---

15. plt.subplot(122) 

    ---

16. nx.draw(G2, with_labels=True, font_weight='bold') 

    ---

17. plt.axis('on') 

    ---

18. plt.xticks([]) 

    ---

19. plt.yticks([]) 

    ---

20. plt.show() 

    ---

21.  

    ---

22. #graph转list 

    ---

23. print(nx.to_dict_of_lists(G1)) 

    ---

24. print(nx.to_edgelist(G1)) 

    ---

graph与列表(List)示例

输出:

{0: [1, 2, 3], 1: [0], 2: [0], 3: [0]}

[(0, 1, {}), (0, 2, {}), (0, 3, {})]

---

12.3graph与numpy

1. #从numpy创建graph 

   ---

2. import numpy as np 

   ---

3. a = np.reshape(np.random.random_integers(0, 1, size=100), (10, 10)) 

   ---

4. D = nx.DiGraph(a) 

   ---

5. nx.draw(D, with_labels=True, font_weight='bold') 

   ---

6. plt.axis('on') 

   ---

7. plt.xticks([]) 

   ---

8. plt.yticks([]) 

   ---

9. plt.show() 

   ---

10.  

    ---

11. #graph返回numpy 

    ---

12. G=nx.Graph() 

    ---

13. G.add_edge(1, 2, weight=7.0, cost=5) 

    ---

14. A1 = nx.to_numpy_matrix(G) 

    ---

15. A2 = nx.to_numpy_recarray(G, dtype=[('weight', float), ('cost', int)]) 

    ---

16. print(A1,A2) 

    ---

graph与numpy示例

输出:

1. [[0. 7.] 

   ---

2. [7. 0.]] [[(0., 0) (7., 5)] 

   ---

3. [(7., 5) (0., 0)]] 

   ---

---

12.4graph与Scipy

1. #从scipy创建graph 

   ---

2. G.clear() 

   ---

3. import scipy as sp 

   ---

4. A = sp.sparse.eye(2, 2, 1) 

   ---

5. G = nx.from_scipy_sparse_matrix(A) 

   ---

6. nx.draw(D, with_labels=True, font_weight='bold') 

   ---

7. plt.axis('on') 

   ---

8. plt.xticks([]) 

   ---

9. plt.yticks([]) 

   ---

10. plt.show() 

    ---

11.  

    ---

12. #graph返回scipy 

    ---

13. A = nx.to_scipy_sparse_matrix(G) 

    ---

14. print(A.todense()) 

    ---

graph与Scipy示例

输出:

1. [[0. 1.] 

   ---

2. [1. 0.]] 

   ---

---

12.5graph与Pandas

1. #从pandas创建graph 

   ---

2. G.clear() 

   ---

3. import pandas as pd 

   ---

4. df = pd.DataFrame([[1, 1], [2, 1]]) 

   ---

5. G = nx.from_pandas_adjacency(df) 

   ---

6. nx.draw(D, with_labels=True, font_weight='bold') 

   ---

7. plt.axis('on') 

   ---

8. plt.xticks([]) 

   ---

9. plt.yticks([]) 

   ---

10. plt.show() 

    ---

11.  

    ---

12. #graph返回scipy 

    ---

13. df = nx.to_pandas_adjacency(G) 

    ---

14. print(df) 

    ---

graph与Pandas示例

输出:

1. 0 1 

   ---

2. 0 1.0 2.0 

   ---