What's the fastest way to merge/join data.frames in R?

Question

For example (not sure if most representative example though):

N <- 1e6
d1 <- data.frame(x=sample(N,N), y1=rnorm(N))
d2 <- data.frame(x=sample(N,N),          


        

Answer 1

For simple task (unique values on both sides of join) I use match:

system.time({
    d <- d1
    d$y2 <- d2$y2[match(d1$x,d2$x)]
})

It's far more faster than merge (on my machine 0.13s to 3.37s).

My timings:

• merge: 3.32s
• plyr: 0.84s
• match: 0.12s