dbscan.m

数据挖掘中聚类算法之DBSCAN算法的matlab实现。

function [class]=dbscan(X,k,eps)
[m,n]=size(X);
class=zeros(1,m);
clusterId = 0;
X=[class',X];

for i = 1:m
  clusterId = clusterId + 1;
  if X(i,1) == 0
     [X]=expandcluster(i,X,clusterId,eps,k,n);
  end  
end

class = X(:,1)';

function [X]=expandcluster(index,X,clusterId,eps,k,n)
  Di=dist(X(index,(2:n)),X(:,(2:n)));
  D = find(Di<eps);
  if length(D) < k
    X(index,1) = -1; %noise
    return;
  else
    X(index,1) = clusterId;
    D(1)=[];  
    for i=1:length(D)
      Dmatrix(i,:)=X(D(i),:);
    end
    while length(D)~=0
      Di=dist(X(D(1),(2:n)),X(:,(2:n)));
      Result = find(Di<eps);
      if length(Result) > k
        for i = 1: length(Result)   
            if X(Result(i),1) == 0|-1
               if X(Result(i),1)==0
                  D(length(D)+1)=Result(i);
               end
               X(Result(i),1)=clusterId;
            end;
         end;
       end;
        D(1)=[];
     end;
   end;

  
%function distan = dist(x,y)
 % sum = 0;
  %nx = size(x);
  %for i = 1:nx
   %   sum = (x(i)-y(i))^2+sum;
   %end
 %distan = sqrt(sum) 
 
 %............................................
function [Di]=dist(i,x)

% function: [D]=dist(i,x)
%
% Aim: 
% Calculates the Euclidean distances between i and all objects in x	 
%								    
% Input: 
% i - an object (1,n)
% x - data matrix (m,n); m-objects, n-variables	    
%                                                                 
% Output: 
% D - Euclidean distance (m,1)


[m,n]=size(x);
Di=sqrt(sum((((ones(m,1)*i)-x).^2)'));

if n==1
   Di=abs((ones(m,1)*i-x))';
end