help in Kmeans modified clustring

Question

0 개 추천

Dear experiences,

I am new in Matlab programmation. So I used Kmeans for found the barycenter of UAV with it set of users connected.

My problem I didn't know if this is code is correct or not.

I calculate the Matrix A end Eta in another Function which called "Matching". I use the HPPP and Monte-Carlo simulation to initialisation of the position of users and UAVs.

So in this code, I update the matrix of Association A if the user is associate with drone (A(i,j)==1) then I calculate the position of drone.

Can you help me pleaase and How I show the cluster

function [cluster,center] = Placement(H,B)
global xu yu
%initialisation
e=length(B);
k = e ;%cardinal de B^A
numP = 3 ; %nombre des UAVs
center = zeros(numP, 2);
Xmax=1000; %X entre 0 et Xmax
Ymax=1000; %Y entre 0 et Ymax
y=3;
U=5;
%% Choose k data points as initial centroids
xP =  Xmax * rand(1,numP); % position des drones aleatoires
yP = Ymax * rand(1,numP);   % position des drones aleatoires
UAVspoints = [xP; yP];   %coordonées des drones
%% Repeat until number of iteration is reached or barycenter is inchangeabale
% init cluster array
cluster=zeros(5,3); % to execute rapidly I supposed we have 5 users and 3 drones
max_iteration=5;  % I chose littre number to obtain the result rapedaly
startingTime = tic;
% debut de programme Kmeans
	
	res=0;
	iteration=1;
    
	while iteration < max_iteration || res~=1   % boucle pour la convergence de Kmean
        fprintf('iteration = %d\n', iteration);
        for j=1:e-1
		
		[A,Eta]=Matching(H,B,xP(j),yP(j));   %update de Eta et Association par l'algorithme 1(Matching)
		
	       
		
			fprintf('j = %d\n', j);
			c=0;
            c=xP(j); %initialisation %% I take every point what I defined in UAVspoints
	        b=yP(j);
			
			for i=1:U-1
				if   A(i,j)==1 &&  A(i+1,j)==1 % si un utilisateur est associer a une drone alor A(ij)=1
					
					
					c(j) =i;  %je veux obtenir les indices des utilisateurs associés pour chaque drone
					% debut de recherche de barycentre (il faut drones prendre sa position au barycentre des utilisateurs connecté a lui )
					xP(j)=(xP(j)+xu(i))/length(c);
					yP(j)=(yP(j)+yu(i))/length(c);
					xP(j+1)=(xP(j+1)+xu(i+1))/length(c);
					yP(j+1)=(yP(j+1)+yu(i+1))/length(c);
				end
				% cette etape pour verifier que la drone est trouvé sa barycentre donc ses coordonnées ne changent pas
				%(d'ou sortie de la boucle While)
				if xP(j)~= yP(j+1) && yP(j)~=yP(j+1)
					res=0;
				else res =1;
				end
				
            end
            center(k, :)=[xP(j); yP(j)];
			L=length(c);
			cluster(1:L,j) = c';
		end
		iteration = iteration + 1;
	end
	
end

댓글 수: 5
이전 댓글 3개 표시 이전 댓글 3개 숨기기

Image Analyst 2020년 6월 28일

편집: Image Analyst 2020년 6월 28일

test1.m

You have a j loop inside another j loop, and the capitalization of center is not consistent -- Center is different than center. Plus you don't need semicolons after an end statement. I fixed those problems and attached it in case anyone else wants to give it a try. How long is it supposed to take? I ran it for several minutes and it never finished. I got up to iteration 12.