mykmeans.m

实现聚类K均值算法： K均值算法：给定类的个数K

% function [centres, options, post, errlog] = kmeans(...)
%
% This function has been adopted from a k means implementation
% copyright (c) Christopher M Bishop, Ian T Nabney (1996, 1997)
clear;
x=load('yeast.txt');
[m,n]=size(x);
data=x;
[ndata, data_dim] = size(data);

ncentres=10;  %聚类个数限制
niters = 50; %循环次数限制
err=0.00001;
olde=10000;

%产生若干个聚类中心
for i=1:ncentres
   ind=round(rand*ndata);
   centres(i,:)=data(ind,:);
end

% Matrix to make unit vectors easy to construct
id = eye(ncentres);

% Main loop of algorithm
for n = 1:niters

  % Save old centres to check for termination
  old_centres = centres;
  
  % Calculate posteriors based on existing centres  
  for i=1:ndata    
    for j=1:ncentres
       ss=0;
       for k=1:data_dim
         ss=ss+(data(i,k)-centres(j,k))*(data(i,k)-centres(j,k));
       end
       d2(i,j)=sqrt(ss/data_dim);
    end
  end
  
  % Assign each point to nearest centre
  [minvals, index] = min(d2', [], 1);
  post = id(index,:);

  num_points = sum(post, 1);
  % Adjust the centres based on new posteriors
  for j = 1:ncentres
    if (num_points(j) > 0)
      centres(j,:) = sum(data(find(post(:,j)),:), 1)/num_points(j);
    end
  end

  % Error value is total squared distance from cluster centres
  e = sum(minvals);
  fprintf(1, 'Cycle %4d  Error %11.6f\n', n, e);  
  if n > 1
    % Test for termination
    if  abs(old_e - e) < err     
%      return;
    end
  end
  old_e = e;
end

c=index;
[m,n]=size(x);
for i=1:ncentres                              %生成与类别数相对应的文件，用于保存各类样本
  fn=strcat('class',int2str(i));
  fn=strcat(fn,'.txt');
  fid(i)=fopen(fn,'w');
end

for i=1:m                              %将各样本存入相应的类别文件
  for j=1:n
    fprintf(fid(c(i)),'%d ',x(i,j));
  end
  fprintf(fid(c(i)),'\n');
end

for i=1:ncentres                               %关闭所有文件
  fclose(fid(i));
end