也许使用起来更容易table
:
x <- table(a)
x
# a
# 1 2 3 4 5
# 5 5 5 2 3
names(x)[x == max(x)]
# [1] "1" "2" "3"
which(a %in% names(x)[x == max(x)])
# [1] 1 2 3 5 6 8 10 12 13 14 15 16 17 18 20
或者,有一个类似的方法tabulate
:
x <- tabulate(a)
sort(unique(a))[x == max(x)]
以下是有关数字和字符向量的一些基准。数值数据的性能差异更为明显。
样本数据
set.seed(1)
a <- sample(20, 1000000, replace = TRUE)
b <- sample(letters, 1000000, replace = TRUE)
基准函数
t1 <- function() {
x <- table(a)
out1 <- names(x)[x == max(x)]
out1
}
t2 <- function() {
x <- tabulate(a)
out2 <- sort(unique(a))[x == max(x)]
out2
}
t3 <- function() {
x <- table(b)
out3 <- names(x)[x == max(x)]
out3
}
t4 <- function() {
x <- tabulate(factor(b))
out4 <- sort(unique(b))[x == max(x)]
out4
}
结果
library(rbenchmark)
benchmark(t1(), t2(), t3(), t4(), replications = 50)
# test replications elapsed relative user.self sys.self user.child sys.child
# 1 t1() 50 30.548 24.244 30.416 0.064 0 0
# 2 t2() 50 1.260 1.000 1.240 0.016 0 0
# 3 t3() 50 8.919 7.079 8.740 0.160 0 0
# 4 t4() 50 5.680 4.508 5.564 0.100 0 0