合并R中的数据集

Question

I am facing a trivial problem while merging 2 data.frames in R.

I am trying to merge 2 data.frames that have same column names and I would want R to merge the same name columns as one column instead of making it 2 separate columns.

Typically what happens when R encounters same name columns while merging data.frames is that it creates 2 seperate variables with suffix "x" and "y". Is there a way to specify this in the merge command to treat similar name columns in the different datasets as one column/variable?

The code that we could use as an example:

x = data.frame(id = c("a","c","d","g"), 
              maths = c(1,3,4,7),  physics = c(1,3,4,7),  chemistry = c(1,3,4,7),  
           english = c(1,3,4,7))
y = data.frame(id = c("b","c","d","e","f"),
                maths = c(5,6,8,9,7), physics = c(5,6,8,9,7), chemistry = c(5,6,8,9,7),
           english = c(5,6,8,9,7))

xy <- merge(x, y, by = "id")

Now there is a workaround for the same where we can create a new variable in the merged data set that takes the non NA values from the same name columns, but this is very inefficient if you have large number of columns.

SAS users would relate to this problem as this problem was brought to my notice by a pro SAS user, where the merge() statement combines 2 same name columns into one column.

Also, as one of the answers below mentioned, if we use:

xy <- merge(x, y, by = intersect(names(x), names(y)))

We get no intersection between the 2 data.frames. Ideally we would want there to be 4 observations here, 2 for each observation in the 2 data.frames id = c("c","d")

Would be grateful to any pro R users to help me out on this one.

Thanks!

Answer 1

Do you really want to merge or is rbind(x,y) what you are looking for? In your example this results in the same data.frame (after sorting by id ). If you want to actually merge the data.frames , you have to specify the names that you do not want to duplicate:

merge(x, y, all=TRUE)
merge(x, y, by = c("id", "maths", "physics", "chemistry", "english"), all = TRUE)

Answer 2

here is my quick solution.

Hope it helps. Note that the first column of x is my id to join

output <- merge(x[!(names(x)[2:length(names(x))] %in% names(y))], y, by.x = "id", by.y="id", all=TRUE)