R：两个数据帧合并

Question

I want to merge two data frame, but there are some row names repeated. If the numbers of row name in two data frame are different, I hope the it can show "NA" in the less one column.

My example:

test1 <- data.frame(name = c("A", "B", "C", "C", "C", "D"), n1 = c("15", "14", "13", "12", "11", "10"))
test2 <- data.frame(name = c("A", "B", "B", "C", "C", "D"), n1 = c("30", "31", "33", "39", "38", "40")) 

Then I merge by name, I got

name n1.x n1.y

 A 15 30 B 14 31 B 14 33 C 13 39 C 13 38 C 12 39 C 12 38 C 11 39 C 11 38 D 10 40 

It will repeating What I want to is

name n1.x n1.y

 A 15 30 B 14 31 B NA 33 C 13 39 C 12 38 C 11 NA D 10 40 

What Command should I use? Thank you very much!

Answer 1

Try:

test1$indx <- with(test1, ave(1:nrow(test1), name, FUN=seq_along))
test2$indx <- with(test2, ave(1:nrow(test2), name, FUN=seq_along))
merge(test1, test2, by=c("name","indx"),all=T)[,-2]
 #   name n1.x n1.y
# 1    A   15   30
# 2    B   14   31
# 3    B <NA>   33
# 4    C   13   39
# 5    C   12   38
# 6    C   11 <NA>
# 7    D   10   40

Answer 2

I will post this before data.table.people come in with a slick, scalable and quicl solution.

Be warned, that this works for provided data set. You should examine the results of your production code carefully.

What the below code does is sticks together values for a common level. The rest is just bookkeeping.

ml <- vector("list", length(unique(test1$name)))
names(ml) <- unique(test1$name)

for (i in unique(test1$name)) {
  o1 <- test1[test1$name %in% i, , drop = FALSE]
  o2 <- test2[test2$name %in% i, , drop = FALSE]
  o.max <- max(c(nrow(o1), nrow(o2)))
  nc <- ifelse(o.max == 1, 2, o.max*2)
  out <- matrix(rep(NA, times = nc), nrow = nc/2)
  out[1:nrow(o1), 1] <- as.numeric(as.character(o1$n1))
  out[1:nrow(o2), 2] <- as.numeric(as.character(o2$n1))

  ml[[i]] <- out
}

count.each <- sapply(ml, nrow)
result <- do.call("rbind", ml)
colnames(result) <- c("n1.x", "n1.y")
data.frame(name = rep(names(ml), count.each), result)

  name n1.x n1.y
1    A   15   30
2    B   14   31
3    B   NA   33
4    C   13   39
5    C   12   38
6    C   11   NA
7    D   10   40