合并两个列表组件

Question

I have a big list, but micro example would be like the following:

A <- c("A", "a", "A", "a", "A")
B <- c("A", "A", "a", "a", "a")
C <- c(1, 2, 3, 1, 4) 
mylist <- list(A=A, B=B, C= C)

expected output is merge A with B so that each component will look like AB

AA, aA, Aa, aa, Aa

better should be sorted, upper case is always first

AA, Aa, Aa, aa, Aa

Thus new list or matrix should have two columns or rows:

AA, Aa, Aa, aa, Aa
1,   2, 3,   1, 4

Now I want calculate average of C based on class - "AA", "Aa", and "aa"

Looks simple but I could not figure out easily.

Answer 1

> (ab <- paste(A, B, sep="") )
[1] "AA" "aA" "Aa" "aa" "Aa"
> (ab <- paste(A, B, sep="") )  # the joining step
[1] "AA" "aA" "Aa" "aa" "Aa"
> (ab <- sub("([a-z])([A-Z])", "\\2\\1", ab) ) # swap lowercase uppercase
[1] "AA" "Aa" "Aa" "aa" "Aa"

> rbind(ab, C)                  # matrix
   [,1] [,2] [,3] [,4] [,5]
ab "AA" "Aa" "Aa" "aa" "Aa"
C  "1"  "2"  "3"  "1"  "4" 
> data.frame(alleles=ab, count=C)  # dataframes are lists
  alleles count
1      AA     1
2      Aa     2
3      Aa     3
4      aa     1
5      Aa     4

Answer 2

I can do it if your data is arranged in a data.frame using the package plyr

> A <- c("A", "a", "A", "a", "A")
> B <- c("A", "A", "a", "a", "a")
> C <- c(1, 2, 3, 1, 4) 
> groups <- sort(paste(A, B, sep=""))
[1] "AA" "aA" "Aa" "aa" "Aa"
> my.df <- data.frame(A=A, B=B, C=C, group=groups)

> require(plyr)
> result <- ddply(my.df, "group", transform, group.means=mean(C))
> result[order(result$group, decreasing=TRUE),]
  A B C group group.means
5 A A 1    AA         1.0
3 A a 3    Aa         3.5
4 A a 4    Aa         3.5
2 a A 2    aA         2.0
1 a a 1    aa         1.0

Answer 3

With your data:

A <- c("A", "a", "A", "a", "A")
B <- c("A", "A", "a", "a", "a")
C <- c(1, 2, 3, 1, 4) 

I define a data.frame using the combination of A and B as the key column:

AB <- paste(A, B, sep='')
df <- data.frame(id=AB, C=C)

> df
  id C
1 AA 1
2 aA 2
3 Aa 3
4 aa 1
5 Aa 4

If you need to order this data.frame before the aggregation then:

df <- df[order(AB, decreasing=TRUE),]

> df
  id C
 1 AA 1
 3 Aa 3
 5 Aa 4
 2 aA 2
 4 aa 1

And with aggregate you calculate the mean for each id :

meanDF <- aggregate(C~id, data=df, mean)

> meanDF

  id   C
1 aa 1.0
2 aA 2.0
3 Aa 3.5
4 AA 1.0

But if you want to order after the aggregation, then:

df <- data.frame(id=AB, C=C)
meanDF <- aggregate(C~id, data=df, mean)
meanDF <- meanDF[order(meanDF$id, decreasing=TRUE),]