从R中的data.frame中删除每个间隔的重复行

Question

I need help to remove all repeated rows in the same interval values of data.frame.

For example, i have a data.frame like :

Time                X   Y   Z
01/01/2011 00:00    101 200 302
01/01/2011 00:05    101 200 302
01/01/2011 00:10    101 200 302
01/01/2011 00:20    100 200 303
01/01/2011 00:25    100 200 303
01/01/2011 00:30    100 200 303
01/01/2011 00:35    101 200 302
01/01/2011 00:40    100 200 303
01/01/2011 00:45    100 200 303

And after removing the repeated row values (x,y,z), i will have a result just like below :

Time                X   Y   Z
01/01/2011 00:00    101 200 302
01/01/2011 00:20    100 200 303
01/01/2011 00:35    101 200 302
01/01/2011 00:40    100 200 303

What i have tried with : unique or duplicate function, but they give the different result.

ex/ eliminate <- data[!duplicated(data[,c("X","Y","Z")]),]

This code just delete all the duplicated values in the all data.frame.

Is there somebody can help me for find the solution?

Thanks before, Regards,

Yougyz

Answer 1

Probably not the most elegant way:

data  <- within(data, C <- paste(X, Y, Z, sep = ""))
rl <- rle(data$C)$lengths
data <- data[c(1, cumsum(rl)[-length(rl)] + 1), 1:(ncol(data)-1)]

Answer 2

The following code makes your three columns of interest a single vector. Then, I just test for equality between the vector and it's offset by 1. When that is false you've had a transition to a new XYZ item.

n <- nrow(ss)
xyz <- with(ss, paste0(X, Y, Z))
sel <- xyz[1:(n-1)] !=  xyz[2:n]
ss[c(TRUE,sel),] #the first one would always be true

This is about 3x faster than Julius answer. The advantage should become greater as the dataset grows.