【50币悬赏】How to detect the outliers by Principal Component Analysis (PCA)

How to detect the outliers by Robust Principal Component Analysis (PCA) based on Projection Pursuit?

问题：如何通过基于投影寻踪法的主成份分析检测出数据中的异常点？
我知道R软件包pcaPP可以实现Projection Pursuit法，但通过命令PCdiagplot（）得到的正交距离OD和得分距离SD图该如何解释？
最好有R软件的相关参考资料，非常感谢！

I have solved the problem.
the R code as follows:

mydata<-read.table("b.txt",header=TRUE)
print(mydata)
newdata=mydata[,-1:-2]

library(pcaPP)
pc <- PCAproj(newdata, 12,scores=TRUE)
biplot(pc)
summary(pc)
   
pcrob=PCAgrid(newdata,k=12)
resrob=PCdiagplot(newdata,pcrob,plot=TRUE,plotbw=FALSE)

你看看