Grouping an Array with respect to certain elements

Hello,

I have an (Nx5) matrix. Each row corresponds to a struct of data. Data identifiers reside in the first and the second columns of matrix, columns 1 and 2.

A sample matrix:

%  srcPort | destPort | send | receive | time
%  -------------------------------------------
%  23      | 9212     | 41   | 55      | 202912.0
%  551     | 10563    | 0    | 0       | 55125.0
%  23      | 9212     | 921  | 1020    | 802912.0
%  98      | 5612     | 39   | 24      | 902912.0

I want to be able to group these data w.r.t columns 1 and 2; later want to plot 'time vs send' of each groupped data. Plotting is OK, but I am stuck on determining way to group those data?

[Sriram Tadavarty]

Hi Erdem,

May i know what kind of grouping are you referring here. Can it be bit clear with the sample matrix? Is it to group like 239212, 55110563, and so on?

[Erdem Tuna]

Hello Sriram,

Regarding the example, I want to group all matrix rows containing (srcPort=23, destPort=9121) into an identifiable structure such as matrix. I want to do this grouping with all different (srcPort, destPort) pairs. My ultimate purpose is to plot data belonging (srcPort, destPort) pairs.

[Sriram Tadavarty]

Hi Erdem,

As i get you, you have an Nx5 matrix, then you wanted to group the first two columns into a separate matrix. So, if the matrix is A (N x 5), then if you define B = A(:,[1 2]); then you will get all the indentifiers in B (N x 2). Is this the same you intended?

[Erdem Tuna]

Hello,

Indeed I want to group the rows having the same srcPort and dstPort elements, i.e. having same first and second column values.

Answer 1

[Ameer Hamza]

Try this

data = [23     9212     41      55      202912.0;
        551     10563   0       0       55125.0;
        23      9212    921     1020    802912.0;
        98      5612    39      24      902912.0];
    
[~,~,ic] = unique(data(:,[1 2]), 'rows');
seperated_data = arrayfun(@(x) data(ic==x,:), unique(ic), 'UniformOutput', 0);

Each element of cell array have same value in column 1 and 2

seperated_data{1}
seperated_data{2}
seperated_data{3}

[Erdem Tuna]

Thank you Ameer.