《R语言实战》代码清单7-7出错

wangpeng692

[s:12]

<br />
# 代码清单7-7 使用by()分组计算描述性统计量<br />
vars <- c("mpg", "hp", "wt")<br />
dstats <- function(x)(c(mean=mean(x), sd=sd(x)))<br />
by(mtcars[vars], mtcars$am, mean)<br />
by(mtcars[vars], mtcars$am, dstats)<br />

</p>

以下为运行代码时的出错信息

<br />
mtcars$am: 0<br />
[1] NA<br />
------------------------------------------------------------------<br />
mtcars$am: 1<br />
[1] NA<br />
Warning messages:<br />
1: In mean.default(data[x, , drop = FALSE], ...) :<br />
  argument is not numeric or logical: returning NA<br />
2: In mean.default(data[x, , drop = FALSE], ...) :<br />
  argument is not numeric or logical: returning NA

</p>

<br />
Error in is.data.frame(x) :<br />
  (list) object cannot be coerced to type 'double'<br />
In addition: Warning message:<br />
In mean.default(x) : argument is not numeric or logical: returning NA<br />

</p>

Robert_Hoo

回复第1楼的 wangpeng692：

把mean改为colMeans,或者用aggregate函数

<br />
aggregate(mtcars[vars], list(mtcars$am), mean)<br />
  Group.1      mpg       hp       wt<br />
1       0 17.14737 160.2632 3.768895<br />
2       1 24.39231 126.8462 2.411000</p>
<p>by(mtcars[vars], mtcars$am, colMeans)<br />
mtcars$am: 0<br />
       mpg         hp         wt<br />
 17.147368 160.263158   3.768895<br />
--------------------------------------------------<br />
mtcars$am: 1<br />
      mpg        hp        wt<br />
 24.39231 126.84615   2.41100 </p>
<p>

</p>

wangpeng692

回复第2楼的 Robert_Hoo：谢谢

by(mtcars[vars], mtcars$am, colMeans)确实可以正确运行了

ypchen

请到这里 http://cos.name/2013/03/r-in-action/ 留言报告错误

wangpeng692

回复第4楼的 ypchen：

已留言，好像代码不怎么好复制，总会少了一段

dding2000

修改为如下即可：

vars <- c("mpg", "hp", "wt")

dstats <- function(x){

mean=sapply(x,mean)

sd=sapply(x,sd)

c(mean, sd)

}

by(mtcars[vars], mtcars$am, dstats)

shifteight

回复第6楼的 dding2000：

c(mean, sd) 最好改成 c(mean=mean, sd=sd)，否则输出中mean和sd不能区分