毕业设计---matlab聚簇

data_kmeans.m文件
%将多分类的letter数据进行聚簇
% load('letter_norm');
% idx=kmeans(letter_data,3);%返回每个数据属于哪个簇
% letter_label=cell2mat(letter_label);%将cell数据类型转化成char类型
% row=size(letter_label,1);
% L=['A','B','C','D','E','F','G','H','I','J','K','L','M','N','O','P','Q','R','S','T','U','V','W','X','Y','Z'];
% mat=zeros(26,3);
% for i=1:row
%     for j=1:26
%         if L(j)==letter_label(i,1)
%             mat(j,idx(i))= mat(j,idx(i))+1;
%         end
%     end 
% end
% [max_num,index]=max(mat,[],2);%index将会返回聚簇的结果
% %然后将类别重新划分，划分成为三类
% letter_label_group=[];
% for i=1:row
%     for j=1:26
%         if L(j)==letter_label(i,1)
%             letter_label_group=[letter_label_group;index(j,1)];
%         end
%     end
% end
% save letter_3group.mat letter_data letter_label_group; 

%聚簇shuttle数据：
load('shuttle_norm');
idx=kmeans(shuttle_data,3);%返回每个数据属于哪个簇
row=size(shuttle_data,1);
mat=zeros(7,3);
for i=1:row
    for j=1:7
        if j==shuttle_label(i,1)
            mat(j,idx(i))= mat(j,idx(i))+1;
        end
    end 
end
[max_num,index]=max(mat,[],2);%index将会返回聚簇的结果
%然后将类别重新划分，划分成为三类
shuttle_label_group=[];
for i=1:row
    for j=1:7
        if j==shuttle_label(i,1)
            shuttle_label_group=[shuttle_label_group;index(j,1)];
        end
    end
end
save shuttle_3group.mat shuttle_data shuttle_label_group;

%-----------------------------------------------------------------------------------------------
%重写magic数据
load('magic_norm');
row=size(magic_data,1);
magic_label_group=zeros(row,1);
for i=1:row
        if cell2mat(magic_label(i,1))=='g'
            magic_label_group(i,1)=1;
        else
            magic_label_group(i,1)=-1;
        end
end
magic=[magic_data magic_label_group];
rowrank = randperm(size(magic, 1));%z这两行的目的是打乱数据分布
magic = magic(rowrank, :);
magic_data=magic(:,1:10);
magic_label_group=magic(:,11);
save magic_2group.mat magic_data magic_label_group; 

　test.m文件

%看看重新聚簇的数据的分布情况
% load('shuttle_3group.mat')
% r=size(shuttle_data,1);
% s=zeros(3,1);
% for i=1:r
%     for j=1:3
%         if shuttle_label_group(i,1)==j
%             s(j,1)=s(j,1)+1;
%         end
%     end
% end
%结果分布s：6791，34251，2458

%-----------------------------------------------------------------------------------------
load('letter_3group.mat')
r=size(letter_data,1);
s=zeros(3,1);
for i=1:r
    for j=1:3
        if letter_label_group(i,1)==j
            s(j,1)=s(j,1)+1;
        end
    end
end
%结果分布s：2326，7019，10655

%-----------------------------------------------------------------------------------------
load('magic_norm.mat')
r=size(magic_data,1);
s=zeros(2,1);
for i=1:r
        if cell2mat(magic_label(i,1))=='g'
            s(1,1)=s(1,1)+1;
        else
            s(2,1)=s(2,1)+1;
        end
    end
%结果分布s：12332,6688