这个小程序可以怎么改？

g1=rnorm(30)
g2=rnorm(30,1)
g3=rnorm(30,2)
g4=rnorm(30,3)
g5=rnorm(30,4)

s1=c(mean(g1),var(g1),var(g1)/mean(g1),(var(g1))^0.5/mean(g1),length(g1))
s2=c(mean(g2),var(g2),var(g2)/mean(g2),(var(g2))^0.5/mean(g2),length(g2))
s3=c(mean(g3),var(g3),var(g3)/mean(g3),(var(g3))^0.5/mean(g3),length(g3))
s4=c(mean(g4),var(g4),var(g4)/mean(g4),(var(g4))^0.5/mean(g4),length(g4))
s5=c(mean(g5),var(g5),var(g5)/mean(g5),(var(g5))^0.5/mean(g5),length(g5))

怎么用循环的方式生成s1,s2,s3,s4,s5?

s<-list()
for(i in 1:5){
g <- rnorm(30,( i-1))
s[[i]] <- c(mean(g),var(g),var(g)/mean(g),(var(g))^0.5/mean(g),length(g))
}
s <- do.call("rbind",s)
s

list.create<-list()
result<-list()

for (i in 1:5) {
        list.create[]<-rnorm(30,(i-1))
}

bin.division<-function(x) {
        return(mean(x)/var(x))
}

bin.operation<-function(x) {
        return(mean(x)^0.5/var(x))
}

five<-function(x){
        return(list(mean(x),var(x),bin.division(x),bin.operation(x),length(x)))
}
result<-sapply(list.create, five)

result

set.seed(100)
list.create<-list()
result<-list()

for (i in 1:5) {
    list.create[[i]]<-rnorm(30,(i-1))
}

bin.division<-function(x) {
    return(mean(x)/var(x))
}

bin.operation<-function(x) {
    return(mean(x)^0.5/var(x))
}

five<-function(x){
    return(list(mean(x),var(x),bin.division(x),bin.operation(x),length(x)))
}
result<-sapply(list.create, five)

result

R中不建议用for
为什么不组成个矩阵，再算

学习学习！

给你邮件了

数据量不大用循环还是可以的，数据量一大最好用矩阵运算。

使用R语言软件尽量避免循环，

1. > gsum <- data.frame(g1=rnorm(30), g2=rnorm(30,1), g3=rnorm(30,2), g4=rnorm(30,3), g5=rnorm(30,4))
   
2. > gresult <- data.frame(gmean=apply(gsum, 2, mean), gvar=apply(gsum, 2, var), gstd <- apply(gsum,2, st))

复制代码

在R中，几百万次的循环也只比向量或矩阵运算慢几秒，在大多数的时候，完全没有必要为了使用矩阵而把非常简明的问题搞得这么复杂，当然，要是遇上数亿次的循环使用矩阵还是必要的，