CODEWITHSUNDEEP
python-3.xgraphnetworkx
4daJKong
4daJKong

Reputation: 2045

How to keep graph shape when read it by networkx

I have a file shows different points' coordinates(first 10 rows):

1 10.381090522139 55.39134945301
2 10.37928179195319 55.38858713256631
3 10.387152479898077 55.3923338690609
4 10.380048819655258 55.393938880906745
5 10.380679138517507 55.39459444742785
6 10.382474625286 55.392132993022
7 10.383736185130601 55.39454404088371
8 10.387334283235987 55.39433237195271
9 10.388468103023115 55.39536574771765
10 10.390814951258335 55.396308397998475

enter image description here

Now I want to calculate the MST(minimum spanning tree) of them so firstly I change my coordinates to weight graph(distance->weight):

n = 10
data = []
for i in range(0, n):
    
    for j in range(i + 1, n):
        temp = []
        temp.append(i)
        temp.append(j)
        x = np.array(rawdata[i, 1:3])
        y = np.array(rawdata[j, 1:3])
        temp.append(np.linalg.norm(x - y))
        data.append(temp)

Then, using networkx to load weight data:

G = nx.read_weighted_edgelist("data.txt")
T = nx.minimum_spanning_tree(G)
nx.draw(T)
plt.show()

but I cannot see the orignal shape from result: enter image description here

how to solve this problem?

Upvotes: 0

Views: 206

Answers (1)

Joel
Joel

Reputation: 23827

I'm just answering the question about the position of the nodes. I can't tell from what you've done whether the minimum spanning tree is what you're after or not.

When you plot a network, it will assign positions based on an algorithm that is in part stochastic. If you want the nodes to go at particular positions, you will have to include that information in the call in an optional argument. So define a dictionary (it's usually called pos) such that pos[node] is a tuple (x,y) where x is the x-coordinate of node and y is the y-coordinate of node.

Then the call is nx.draw(T, pos=pos).

Upvotes: 1

Related Questions