非参数假设性检验 Wilcoxon秩和检验 matlab

x=[0.8587,0.8591,0.8591,0.8569,0.8555,0.8390];
y=[0.8583,0.8589,0.8637,0.8589,0.8613,0.8616];
[ans,h]=ranksum(x,y)

假设存在两个分布未知的变量x和y，各有一组观测值如上所示，ranksum是matlab自带的Wilcoxon秩和检验方法，用于判断两变量是否有显著性区别。输出结果就直接是p值，为0.0823，表示：认为x与y有显著性区别，这一断言的犯错概率是8.23% (at the default 8.23%).
官网
实例