如何识别数据框中相对于其他数据框的唯一列？

Question

If I have several data frames, how can I identify the columns which are unique to a certain data frame?

df1 <- data.frame(A=rnorm(5), B=rnorm(5), C=rnorm(5))
df2 <- data.frame(B=rnorm(5), C=rnorm(5), D=rnorm(5))
df3 <- data.frame(B=rnorm(5), C=rnorm(5), D=rnorm(5))

What I want to achieve is something like the unique() function, that gives me the unique columns in a data frame with respect to other data frames.

unique.columns(df1, c(df2, df3))
[1] "A"

but

unique.columns(df2, c(df1, df3))
[1] NA

since there are no unique columns in df2.

Answer 1

You could use Reduce along with setdiff to deal with any number of comparison datasets easily. The first named dataset will be compared to the remainder.

Reduce(setdiff, lapply(list(df1,df2,df3), names))
#[1] "A"

Reduce(setdiff, lapply(list(df2,df1,df3), names))
#character(0)

Answer 2

We can use setdiff and union

unique.columns <- function(df1, df2, df3) {
   setdiff(names(df1), union(names(df2), names(df3)))
}

unique.columns(df1, df2, df3)
#[1] "A"

unique.columns(df2, df1, df3)
#character(0)

---

If you are going to pass a variable number of dataframes to the function, you can change the function

unique.columns <- function(df1, ...) {
   temp <- list(...)
   setdiff(names(df1), unique(c(sapply(temp, names))))
}

unique.columns(df1, df3)
#[1] "A"

Answer 3

You could also use the "not in" using ! and %in% on the colnames of each df to get the column names that are unique to one df in comparison with other dfs.

colnames(df1)[!(colnames(df1) %in% c(colnames(df2),colnames(df3)))]
#[1] "A"
colnames(df2)[!(colnames(df2) %in% c(colnames(df1),colnames(df3)))]
#character(0)