學過線性代數的都了解矩陣,在矩陣上的文章可做的很多,什麽特征矩陣,單位矩陣等.grpah存儲可以使用矩陣,比如graph的鄰接矩陣,權重矩陣等,這節主要是在等到graph後,如何快速得到這些信息.詳細官方文檔在這裏

目錄:

• 12.graph和其他數據格式轉換
  • 12.1graph與字典(Dict)
  • 12.2graph與列表(List)
  • 12.3graph與numpy
  • 12.4graph與Scipy
  • 12.5graph與Pandas

註意:如果代碼出現找不庫,請返回第一個教程,把庫文件導入.

12.graph和其他數據格式轉換

12.1graph與字典(Dict)

1. #從字典生成圖

   ---

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. #圖轉換為字典

    ---

12. print(nx.to_dict_of_dicts(G))

    ---

輸出:

{0: {1: {‘weight‘: 1}}, 1: {0: {‘weight‘: 1}}}

12.2graph與列表(List)

1. #從列表中創建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. #顯示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轉list

    ---

23. print(nx.to_dict_of_lists(G1))

    ---

24. print(nx.to_edgelist(G1))

    ---

輸出:

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

12.3graph與numpy

1. #從numpy創建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)

    ---

輸出:

1. [[0. 7.]

   ---

2. [7. 0.]] [[(0., 0) (7., 5)]

   ---

3. [(7., 5) (0., 0)]]

   ---

12.4graph與Scipy

1. #從scipy創建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())

    ---

輸出:

1. [[0. 1.]

   ---

2. [1. 0.]]

   ---

12.5graph與Pandas

1. #從pandas創建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)

    ---

輸出:

1. 0 1

   ---

2. 0 1.0 2.0

   ---

NetworkX系列教程(11)-graph和其他數據格式轉換