How to merge data based on column of specific data in r

Question

I have three data, every data contain one column and the same rows but in diffrent orders in all the five data, i want to merge the five data based on the first column.

      column1
X       7
B        5
Y        2
D        1
Z        0


        column2
D          7 
Z          5
X          2
Y          6
B          7

       column3
Y          4
D          4
Z          8 
B          7
X          2

The result i want is like this

        column1     column2      column3
X           7          2            2
B           5          7            7
Y           2          6            4
D           1          7            4
Z           0          5            8

Answer 1

This can be easily done with the merge function.

df1 <- data.frame(rows = c("A", "B", "C", "D", "E"), column1 = c(7, 5, 2, 1, 0))
df2 <- data.frame(rows = c("D", "E", "A", "C", "B"), column2 = c(7, 5, 2, 6, 7))
df3 <- data.frame(rows = c("C", "D", "E", "B", "A"), column3 = c(4, 4, 8, 7, 2))
result <- merge(df1, df2, by = "rows", all = TRUE)
result <- merge(result, df3, by = "rows", all = TRUE)
print(result)

This should give what you want.

rows column1 column2 column3
1    A       7       2       2
2    B       5       7       7
3    C       2       6       4
4    D       1       7       4
5    E       0       5       8

Answer 2

In base R, you can use the following approach:

df_list <- list(df1, df2, df3)

Reduce(merge, lapply(df_list, \(x) 
  cbind(x, xx = row.names(x)))) |> _[-1]

#   column1 column2 column3
# 1       7       2       2
# 2       5       7       7
# 3       2       6       4
# 4       1       7       4
# 5       0       5       8

An alternative approach is to reorder by row names then put them all together. It's easier to do this when combining them in a list, then using do.call to combine everything back together. I also wrapped it in setNames to rename the columns to the desired column names:

df_list <- list(df1, df2, df3)

setNames(do.call(data.frame, lapply(df_list, \(x) x[order(rownames(x)),])), 
         paste0("column", 1:3))

#   column1 column2 column3
# 1       7       2       2
# 2       5       7       7
# 3       2       6       4
# 4       1       7       4
# 5       0       5       8

The above is identical to doing it manually, which is slower but may be easier to interpret what's going on:

df1 <- df1[order(rownames(df1)),]
df2 <- df2[order(rownames(df2)),]
df3 <- df3[order(rownames(df3)),]

data.frame(column1 = df1, column2 = df2, column3 = df3)

#   column1 column2 column3
# 1       7       2       2
# 2       5       7       7
# 3       2       6       4
# 4       1       7       4
# 5       0       5       8

Data:

df1 <- read.table(text = "column1
A        7
B        5
C        2
D        1
E        0", h = TRUE)


df2 <- read.table(text = "column2
D          7 
E          5
A          2
C          6
B          7", h = TRUE)

df3 <- read.table(text = "column3
C          4
D          4
E          8 
B          7
A          2", h = TRUE)

Answer 3

Using tidyverse:

library(tidyverse)

list(df1, df2, df3)%>%
   map(rownames_to_column) %>%
   reduce(full_join, 'rowname') %>%
   column_to_rownames()

  column1 column2 column3
A       7       2       2
B       5       7       7
C       2       6       4
D       1       7       4
E       0       5       8