对列组

Question

在下面的数据框中我想平均所有列，然后计算每个字母类别的第一个均值之间的折叠差异，就像我在excel中做的那样：

A0<-1:4
A1<-2:5
A2<-3:6
A3<-4:7
A4<-5:8
B0<-11:14
B1<-12:15
B2<-13:16
B3<-14:17
B4<-15:18
C0<-21:24
C1<-22:25
C2<-23:26
C3<-24:27
C4<-25:28

到目前为止，我能做的是：

## step 1
df<-data.frame(A0, A1, A2, A3, A4, B0, B1, B2, B3, B4, C0, C1, C2, C3, C4)

## step 2
rbind(df, c(mean(df[,1])/sapply(df[,1:5], mean),
      mean(df[,6])/sapply(df[,6:10], mean), 
      mean(df[,11])/sapply(df[,11:15], mean)))

## step 3
data.frame(t(rbind(df, c(mean(df[,1])/sapply(df[,1:5], mean), 
           mean(df[,6])/sapply(df[,6:10], mean), 
           mean(df[,11])/sapply(df[,11:15], mean)))))

## step 4
ggplot(data.frame(t(rbind(df, c(mean(df[,1])/sapply(df[,1:5], mean), 
       mean(df[,6])/sapply(df[,6:10], mean), 
       mean(df[,11])/sapply(df[,11:15], mean))))),
       aes(1:15, X5)) + geom_bar(stat="identity")

在此之后我可以做其他方面，但是在大数据框架中第2步将非常困难。你有什么建议我怎么告诉R自动做，所以我不必自己写吗？

Answer 1

使用dplyr和tidyr：

# step 1
df <- data.frame(A0, A1, A2, A3, A4, B0, B1, B2, B3, B4, C0, C1, C2, C3, C4)

library(dplyr)
library(tidyr)

df2 <- df %>% gather(LetterNum, Value) %>% 
       separate(LetterNum, c("Letter", "Num"), sep = 1) %>% 
         group_by(Letter, Num) %>% 
          summarise(Mean1 = mean(Value)) %>% 
            mutate(Mean = first(Mean1)/Mean1)

ggplot(df2, aes(Num, Mean, group = Letter, fill = Letter)) + 
  geom_bar(stat="identity", position = "stack") + facet_wrap(~Letter)

或

# step 1
df <- data.frame(A0, A1, A2, A3, A4, B0, B1, B2, B3, B4, C0, C1, C2, C3, C4)

library(dplyr)
library(tidyr)

df2 <- df %>% gather(LetterNum, Value) %>% group_by(LetterNum) %>% 
        summarise(Mean1 = mean(Value)) %>% 
          mutate(Group = rep(LETTERS[1:(n()/5)], each = 5)) %>% group_by(Group) %>% 
             mutate(Mean = first(Mean1)/Mean1)


ggplot(df2, aes(LetterNum, Mean, fill = Group)) + geom_bar(stat="identity", 
        position = "stack") + facet_wrap(~Group, scale="free")

Answer 2

我不知道我是否理解你的问题。我尝试使用名为dplyr的库来模拟您的数据集并创建一个简短的脚本。这是一步一步的脚本。 1.创建数据集。

A0<-1:4
A1<-2:5
A2<-3:6
df <- data.frame(A0,A1, A2, A3)
df
  A0 A1 A2 A3
1  1  2  3  4
2  2  3  4  5
3  3  4  5  6
4  4  5  6  7

现在的脚本将采用每列的平均值。

mean.df <- df %>%
rbind(Mean = rowMeans(x = df[],na.rm = TRUE))
mean.df
  A0  A1  A2  A3
1    1.0 2.0 3.0 4.0
2    2.0 3.0 4.0 5.0
3    3.0 4.0 5.0 6.0
4    4.0 5.0 6.0 7.0
Mean 2.5 3.5 4.5 5.5

现在，将从第1个均值减去每个均值的脚本。这就是我的理解。

fold.diff <- mean.df %>%
rbind(Fold_Diff = mean.df[5,] - mean.df[5,1])
fold.diff
       A0  A1  A2  A3
1         1.0 2.0 3.0 4.0
2         2.0 3.0 4.0 5.0
3         3.0 4.0 5.0 6.0
4         4.0 5.0 6.0 7.0
Mean      2.5 3.5 4.5 5.5
Fold_Diff 0.0 1.0 2.0 3.0