如何在 R 的矩阵列中不唯一的 select 行

Question

I have a big data set but i can explain through a simple example. For example i have a matrix "x"

x<- matrix(c(3,3,3,4,3,3,5,5,5), nrow=3, byrow=T)

now i need second row in which "x" is not unique entries. First and third rows are equal in the sense of columns.

Regards and thanks in advance,

Iftikhar Ahmad

Answer 1

You could also take advantage of the fact that equality means the standard deviation will always be 0. unfortunately we have to convert these 0s to logical either with a logical expression (below) or with as.logical .

x[apply(x, 1, sd) > 0, ]

Update

Did some benchmarking of @joran and my solutions. Mine lost:(

x <- matrix(sample(3:5,30000,T), ncol=3)    

system.time(x2 <- x[apply(x,1,sd) > 0, ])
user  system elapsed 
0.960   0.000   0.961

system.time(x2 <- x[apply(x,1,FUN=function(r){return(length(unique(r)))}) > 1,])
user  system elapsed 
0.470   0.000   0.465

But...

If we do a fully vectorised version with a similar theme, we can blow both out of the water

system.time(x2 <- x[rowSums(abs(x - rowMeans(x))) != 0, ])
user  system elapsed 
0.000   0.000   0.001

Answer 2

Is this what you're looking for:

x[apply(x,1,FUN=function(r){return(length(unique(r)))}) > 1,]

that will select rows with more than one unique value in them.