How to convert the X & Y value in a matrix format?

Hi, Given X and Y as follow:

area_length = 100;
area_width = area_length ;
x = round((area_length - 1) * (rand - 0) / (1 - 0) + 1); 
y = round((area_width - 1) * (rand - 0) / (1 - 0) + 1);

How to convert the X & Y value into matrix format? Because i want to add a weights for make i possible to perform a shortest path.

x= ;
y= ;
weights = rand ;%-ok
G = graph(x,y,weights); 
plot(G,'EdgeLabel',G.Edges.Weight);

Thanks for your help

[KSSV]

Not clear..you need to explain more...what you are trying to do?

[Asyran Abdullah]

I try to make a shortest path in existing point. But it difficult when X & Y value is not a matrix.

[KSSV]

What is your domain? I mean what area you want to define?

[Asyran Abdullah]

WSN are randomly distributed the nodes(point), then what i want to do is just find a shortest path from node to the destination (CH). The node are already distributed. But how to find the shortest path without source and destination matrix.

As shown in figure, i just want to make a connection for example from 43 to 36. Either use A path (43---36) or B path (43--46--36),which it will depend on the weights of each path.

46---43
 \   |
  \  |
   \ |
     36

However, it difficult to perform a shortest path in existing nodes(points). Hope u understand, and can give some advice or example.

Thanks.

[KSSV]

YOu need to have those points to get the shortest path........do you have those points?

[Asyran Abdullah]

Yes, this is the point:

x = round((area_length - 1) * (rand - 0) / (1 - 0) + 1); 
y = round((area_width - 1) * (rand - 0) / (1 - 0) + 1);

x&y show the coordinate of each nodes, and the value will repeatedly change. However i don't know how to change the value into matrix in order to find a shortest path.

Answer 1

[Walter Roberson]

Your x and y calculated like the above define the coordinates of a single point. To do shortest path, you need multiple points.

I suggest,

N = 50;    %nodes
cutoff = 25;

area_length = 100;
area_width = area_length ;
x = round((area_length - 1) * (rand(N,1) - 0) / (1 - 0) + 1); 
y = round((area_width - 1) * (rand(N,1) - 0) / (1 - 0) + 1);   
A = squareform( pdist([x, y]) );
A(A > cutoff) = 0;
%and do any other tweaking for the A matrix

G = digraph(A);
plot(G, 'XData', x, 'YData', y);

and you can do shortest path calculations on G

With the above code, you could also use G = graph(A) to create undirected graphs, but if you do so then you would not be able to have asymmetric distance matrix. In the real world, there are cases where the transmit and receive path might not be the same.

[Asyran Abdullah]

Thanks Sir, it works!

N =  45;    %nodes
cutoff = 25;
area_length = 100;
area_width = area_length ;

x = round((area_length - 1) * (rand(N,1) - 0) / (1 - 0) + 1); 
y = round((area_width - 1) * (rand(N,1) - 0) / (1 - 0) + 1);   
A = squareform( pdist([x, y]) );
A(A > cutoff) = 0;

%and do any other tweaking for the A matrix
G =  graph(A ); 
G.Edges.W3 = randi(10, numedges(G) , 1); %laluan
G.Edges.Weight = G.Edges.W3; %weight
h = plot(G, 'XData', x, 'YData', y, 'EdgeLabel',G.Edges.Weight);
axis square;  
grid on 

G = rmedge(G, 1:numnodes(G), 1:numnodes(G)); 
 title('Finding Shortest Path on Random Node'); 
 path = shortestpath(G,1,45);
 highlight(h,path,'NodeColor','g','EdgeColor','g')

Answer 2

[KSSV]

Simple rand..gives you a single number....try rand(N,1) with N as desired number of points.

[Asyran Abdullah]

Thanks sir!