Counting the number of elements in a network for ij matrix.

I have a network with order number assigned for each link. I want to count the number of the link such that i get the values for the number of 1s joining 2, 1s joining 3, 1s joining 4 separately. Similarly i want to calculate the count for 2s joining 3, 4. only condition is that the value direction can't be decreasing i.e. 2s to 1, 3s to 2. Finally, i want to create a matrix based upon the count. I am uploading the sample data if anyone want to give it a try in Matlab. I am also attaching the pics to clarify what i am trying to do. Thanks a lot.

[Voss]

The .xlsx file contains a column of unique HydroIDs and a column of associated Orders (1,1,1,2,1, ...). How should the HydroIDs be interpreted to ascertain which 1s are connected to which 2s, etc.? Or is that part of your question? Sorry, I may be missing something, but it's not clear how the data in the file describes connections in a network.

[Niraj Bal Tamang]

Oh. I didn't notice that. Sorry for that. I have uploaded the sample data with additional data with From_Node and TO_Node indicating the starting point and ending or meeting point ID i.e. Each line in the network starts from From_node and ends at TO_NODE. One clue can be the fact that the TO_node of upper line can be the From_node of the lower line. I am trying to create a matrix like in the table where i can have the number of elements for each combination in the sample data. Thank you once again.

Answer 1

[Aditya]

Hi Niraj,

I understand that you want to create a matrix that counts the number of links with a particular order that are connected to nodes with a higher order.

Here's a MATLAB script that could perform the task you've described using the sample data you provided.

% Read the data from a xlsx file into a table 
data = readtable('sample.xlsx'); 
 
% Find the unique orders in the network 
unique_orders = unique(data.Order); 
 
% Sort the unique orders to ensure they are in ascending order 
unique_orders = sort(unique_orders); 
 
% Initialize the matrix to store the counts of links between orders 
order_matrix = zeros(length(unique_orders)); 
 
% Loop over the data to count the links 
for i = 1:height(data) 
    from_order = data.Order(i); 
    to_node = data.TO_NODE(i); 
     
    % Find all the orders of the 'to_node' 
    to_order_rows = data(data.FROM_NODE == to_node, :); 
     
    % Iterate over each 'to_order_row' and increment the count if the order is increasing 
    for j = 1:size(to_order_rows, 1) 
        to_order = to_order_rows.Order(j); 
        if to_order > from_order 
            from_index = find(unique_orders == from_order); 
            to_index = find(unique_orders == to_order); 
            order_matrix(from_index, to_index) = order_matrix(from_index, to_index) + 1; 
        end 
    end 
end 
 
% Convert the matrix to a table for better readability 
order_table = array2table(order_matrix, 'RowNames', string(unique_orders), 'VariableNames', string(unique_orders)); 
 
% Display the resulting table 
disp(order_table);

This script iterates over all rows where the “FROM_NODE” matches the “TO_NODE” from the current row in the loop. It increments the count in the “order_matrix” only if the order is increasing, as per your requirement. The resulting “order_table” will have row and column names that correspond to the unique orders, making it more representable and easier to understand.

Hope this helps!