我正在尝试将基于多个分组的长格式数据重新整形为宽格式，但没有成功。有了这些数据：
id <- 1:20
month <- rep(4:7, 50)
name <- rep(c("sam", "mike", "tim", "jill", "max"), 40)
cost <- sample(1:100, 200, replace=TRUE)
df <- data.frame(id, month, name, cost)
df.mo.mean <- aggregate(df$cost ~ df$name + df$month, FUN="mean")
df.mo.sd <- aggregate(df$cost ~ df$name + df$month, FUN="sd")
df.mo <- data.frame(df.mo.mean, df.mo.sd)
df.mo <- df.mo[,-c(4,5)]
df.mo[3:4] <- round(df.mo[3:4],2)
head(df)
id month name cost
1 1 4 sam 29
2 2 5 mike 93
3 3 6 tim 27
4 4 7 jill 67
5 5 4 max 28
6 6 5 sam 69
我正在尝试将我的数据看作下面的内容，并尝试将其概括为未知数量的名称（但最多不超过15个）
month name1.cost.mean name1.cost.sd name2.cost.mean name2.cost.sd
1 45 4 40 6
2 ...
我已尝试reshape和do.call rbind但未成功。我能想到的另一种方法是使用循环，这意味着我做错了。我对plyr没有任何经验，并且更倾向于使用基础包来解决这个问题（出于学习目的），但如果不可能，任何其他建议都会非常有用
答案 0 :(得分：4)
set.seed(1)
library(plyr)
kk<-ddply(df,.(month,name),summarize,mean=mean(cost),sd=sd(cost))
reshape(kk,timevar="name",idvar="month",direction="wide")
month mean.jill sd.jill mean.max sd.max mean.mike sd.mike mean.sam sd.sam mean.tim sd.tim
1 4 55.3 34.62834 63.3 23.35261 57.6 22.91627 63.4 28.89906 43.3 25.42112
6 5 49.3 25.00689 51.1 27.85059 48.4 23.16223 43.0 24.33562 47.6 32.13928
11 6 60.4 23.61826 52.1 29.74503 38.6 34.39703 53.0 23.28567 52.4 20.88700
16 7 50.0 30.76073 62.7 23.98634 51.7 32.10763 52.8 32.27589 49.5 23.00845
答案 1 :(得分：1)
我不确定你要求的是什么，但也许这样的事情可能有用
> set.seed(1)
> df <- data.frame(id=1:20, month=rep(4:7, 50),
+ name=rep(c("sam", "mike", "tim", "jill", "max"), 40),
+ cost= sample(1:100, 200, replace=TRUE))
>
> DF.mean <- aggregate(cost ~ name + month, FUN=mean, data=df) ## mean
> DF.sd <- aggregate(cost ~ name + month, FUN=sd, data=df) ## sd
>
> x1 <- as.data.frame.matrix(xtabs(cost~month+name, data=DF.mean)) # reshaping mean
> colnames(x1) <- paste0(colnames(x1), ".mean")
> x2 <- as.data.frame.matrix(xtabs(cost~month+name, data=DF.sd)) # reshaping sd
> colnames(x2) <- paste0(colnames(x2), ".sd")
>
> cbind(x1, x2)
jill.mean max.mean mike.mean sam.mean tim.mean jill.sd max.sd mike.sd sam.sd tim.sd
4 55.3 63.3 57.6 63.4 43.3 34.62834 23.35261 22.91627 28.89906 25.42112
5 49.3 51.1 48.4 43.0 47.6 25.00689 27.85059 23.16223 24.33562 32.13928
6 60.4 52.1 38.6 53.0 52.4 23.61826 29.74503 34.39703 23.28567 20.88700
7 50.0 62.7 51.7 52.8 49.5 30.76073 23.98634 32.10763 32.27589 23.00845
另外，请注意@Metrics方法可以使用R基本函数完成，而无需任何额外的包：
> kk <- aggregate(cost ~ name + month, FUN=function(x) c(mean=mean(x), sd=sd(x)), data=df)
> reshape(kk,timevar="name",idvar="month",direction="wide")
month cost.jill.mean cost.jill.sd cost.max.mean cost.max.sd cost.mike.mean cost.mike.sd cost.sam.mean cost.sam.sd cost.tim.mean cost.tim.sd
1 4 55.30000 34.62834 63.30000 23.35261 57.60000 22.91627 63.40000 28.89906 43.30000 25.42112
6 5 49.30000 25.00689 51.10000 27.85059 48.40000 23.16223 43.00000 24.33562 47.60000 32.13928
11 6 60.40000 23.61826 52.10000 29.74503 38.60000 34.39703 53.00000 23.28567 52.40000 20.88700
16 7 50.00000 30.76073 62.70000 23.98634 51.70000 32.10763 52.80000 32.27589 49.50000 23.00845
答案 2 :(得分：1)
> means <- with( df, tapply(cost, list(month, name), FUN=mean) )
> sds <- with( df, tapply(cost, list(month, name), FUN=sd) )
> colnames(means) <- paste0(colnames(means), ".mean")
> colnames(sds) <- paste0(colnames(sds), ".sd")
> comb.df <- as.data.frame( cbind(means, sds) )
> comb.df <- comb.df[order(names(comb.df))]
> comb.df
jill.mean jill.mean.sd max.mean max.mean.sd mike.mean mike.mean.sd
4 62.1 22.29823 39.7 25.53016 39.6 30.11164
5 40.7 30.72838 44.4 29.12502 54.2 23.91095
6 47.3 31.54556 46.9 32.30910 65.3 30.05569
7 55.5 33.16038 45.9 28.13637 59.7 31.79815
sam.mean sam.mean.sd tim.mean tim.mean.sd
4 40.9 23.54877 58.5 21.69613
5 51.5 30.76163 34.2 32.16900
6 69.1 18.26016 55.2 32.99764
7 46.9 29.90150 55.8 27.17352
答案 3 :(得分：0)
您可以使用两个整形，然后合并结果
library(reshape2)
> dcast(df, month ~ name, mean, value.var="cost")
month jill max mike sam tim
1 4 39.5 54.6 45.6 48.4 57.4
2 5 45.1 61.7 45.4 54.5 50.8
3 6 41.9 45.7 56.4 43.1 52.1
4 7 51.6 38.6 43.6 65.1 51.5
> dcast(df, month ~ name, sd, value.var="cost")
month jill max mike sam tim
1 4 29.31154 25.25954 28.96051 31.32695 29.82989
2 5 31.02848 27.96049 34.32589 30.08599 23.95273
3 6 32.09517 32.50316 37.16988 27.03681 30.42094
4 7 19.56300 31.50026 28.65969 36.53750 26.73429