Combine multiple data frame with coincident columns

3
votes

I would like to combine multiple DataFrames with some coincident columns, into a new DataFrame. The columns of the new DataFrame should be the coincident columns.

For example, suppose I have dataframes df1, df2, df3:

df1:
A   B   C   D
1   2   3   4

df2:
A   C   D   E
1   2  -1   5

df3:
C   D   F   G
0  -1   0   7

New dataframe
C   D 
3   4 
2  -1
0  -1

I have tried using match function in a circular way, to find the coincident columns:

match(df1,df2)

match(df2,df3)

match(df3,df1)

It takes a lot of time and lines, if I have many DataFrames. Could anyone suggest a better way to do that?

1
rdataframe

1 Answers

1
votes

An option is to get the datasets in a list and find the intersecting column name

library(tidyverse)
lst1 <- mget(paste0("df", 1:3))
nm1 <- map(lst1, names) %>%
          reduce(intersect)    
map_dfr(lst1, ~ .x %>% 
                  select(nm1))
#  C  D
#1 3  4
#2 2 -1
#3 0 -1

Or in base R

nm1 <- Reduce(intersect, lapply(lst1, names))
out <- do.call(rbind, lapply(lst1, subset, select = nm1))
row.names(out) <- NULL

data

df1 <- structure(list(A = 1L, B = 2L, C = 3L, D = 4L), class = "data.frame", row.names = c(NA, 
-1L))

df2 <- structure(list(A = 1L, C = 2L, D = -1L, E = 5L), class = "data.frame", row.names = c(NA, 
-1L))

df3 <- structure(list(C = 0L, D = -1L, F = 0L, G = 7L), class = "data.frame", row.names = c(NA, 
-1L))