agnes算法matlab代码

AGNES（Agglomerative Nesting）算法是一种层次聚类算法，主要用于将数据点逐步合并成聚类的集合。下面是一个用MATLAB实现AGNES算法的示例代码：

function [clusters] = AGNES(data, k)
    % 初始化每个数据点为一个簇
    clusters = cell(size(data, 1), 1);
    for i = 1:size(data, 1)
        clusters{i} = i;
    end
    
    % 逐步合并簇，直到只剩下k个簇
    while length(clusters) > k
        min_dist = inf;
        merge_i = -1;
        merge_j = -1;
        
        % 计算两两簇之间的最小距离
        for i = 1:length(clusters)
            for j = (i+1):length(clusters)
                dist = min_distance(data(clusters{i}, :), data(clusters{j}, :));
                if dist < min_dist
                    min_dist = dist;
                    merge_i = i;
                    merge_j = j;
                end
            end
        end
        
        % 合并距离最近的两个簇
        clusters{merge_i} = [clusters{merge_i}, clusters{merge_j}];
        clusters(merge_j) = [];
    end
end

function [dist] = min_distance(cluster1, cluster2)
    % 计算两个簇之间的最小距离
    min_dist = inf;
    for i = 1:size(cluster1, 1)
        for j = 1:size(cluster2, 1)
            dist = norm(cluster1(i, :) - cluster2(j, :));
            if dist < min_dist
                min_dist = dist;
            end
        end
    end
    dist = min_dist;
end

这段代码首先将每个数据点初始化为一个簇，然后使用循环逐步合并簇，直到只剩下k个簇。在每次合并过程中，计算两两簇之间的最小距离，并将距离最近的两个簇合并在一起。最终返回合并后的所有簇。其中，`min_distance`函数用于计算两个簇之间的最小距离。