检测矢量中的异常值并绘制热图

Question

如何检测向量中的异常值？

例如，我有以下向量

color<-c("green","green","green","green","green","red","blue")
size<-c("big","small","big","big","big","small","medium")

对于每个矢量我想检测异常值并绘制热图

• color异常值位于第6和第7位
• size异常值位于2,6,7

Answer 1

像

这样的东西

color <- c("green","green","green","green","green","red","blue")
tt <- table(color)
names(tt)[tt<2]   ## values with frequency < 2
## 'blue', 'red'
ptt <- prop.table(tt)
names(ptt)[ptt<0.2]  ## values with frequency < 20%
## 'blue', 'red'