How to distinguish specific nodes in an undirected graph ?

Hello,

I have an attribute of an undirected graph called G.Nodes.Load. Where in every time step some of its values change to -inf. And I'm trying to visualize this process using Graph and Network Algorithms.

I tried the following:

figure; h = plot(G);
highlight(h, G.Nodes.Load == -inf, 'NodeColor','r' )

Kindly check the attached graph plots. Is there a clearer way than this?

EDIT : I assigned random values for "G.Nodes.Load", and then I created another attribute "G.Nodes.Capacity" :

Load = (1000-800)*rand(5000,1)+800;
Capacity = (1+0.2) * Load;
G.Nodes.Load = Load; 
G.Nodes.Capacity = Capacity;

Now I'm trying to assigne a red color for any value of Load that exceeds its corresponding capacity, a lighter red color for 80% usage of capacity (Load./capacity), light blue for 50% usage, and dark blue for 25% usage. I think in this case I don't have to highlight the nodes but instead I need to choose another way in visualizing the load./capacity ratio, am I correct?

Thanks!

Answer 1

[Steven Lord]

Create a sample graph and define some sample load and capacity data.

rng default
G = graph(bucky);
Load = (1000-800)*rand(numnodes(G),1)+800;
Capacity = (1+0.2*rand(numnodes(G), 1)) .* Load;

Set the node data to the Nodes table in G.

G.Nodes.ID = (1:numnodes(G)).';
G.Nodes.Load = Load; 
G.Nodes.Capacity = Capacity;
G.Nodes.Usage = G.Nodes.Load ./ G.Nodes.Capacity;

Plot the graph.

h = plot(G);

Select some nodes.

nodesGreaterThan97Percent = G.Nodes.Usage > 0.97;

Get info about these nodes.

info = G.Nodes(nodesGreaterThan97Percent, :)

info = 9×4 table
    ID     Load     Capacity     Usage 
    __    ______    ________    _______

     8    909.38     934.59     0.97302
     9     991.5     1021.1     0.97101
    31    941.21     955.49     0.98506
    32    806.37     815.07     0.98933
    36    964.69     989.76     0.97468
    39    990.04      992.4     0.99763
    52    935.94     951.63     0.98351
    55     823.8     848.91     0.97043
    59    917.05     931.39     0.98461

Highlight the selected nodes in the plot.

highlight(h, nodesGreaterThan97Percent, 'NodeColor', 'r')

[Waseem AL Aqqad]

Thanks, Steven

Answer 2

[Sulaymon Eshkabilov]

Here one potential err in your code is logical (==). One can determine when the value of a variable goes to infinity with this fcn: isinf().

Moreover, MATLAB plot fcn skips infinity values and thus, you'd need to determine when the values go to infinity and assign them some specific value in order to plot them. And then you can do some highlights.

[Waseem AL Aqqad]

You misunderstood my question.

I'm plotting a Graph object (Complex Network) not data points. So, I'm visulizing the nodes with this attribute G.Nodes.Load = -inf. What I did is quite correct but I'm wondering if there is a better way.

[Waseem AL Aqqad]

You might want to check this tutorial Graph and Network Algorithms