bind two dataframes that have different columns names

Question

I have two data-frames

The first data

        col1  col2 col3
A1       4     11  15
A2       2      9  17
A3       3      4   4 
B1       10     5   4   
B2        6     1   8
C1       12     1  12
C2        2     5   8
D1        4     1   6 
D2        2     1   8

The second data

          meancol1   meancol2 meancol3
meanA        3          8      12
meanB        8          3       6
meanC        7          3      10
meanD        3          1       7

I want to combine the two data-frames and keep the colnames of the first dataset so the result i want is :

          col1  col2 col3
    A1       4     11  15
    A2       2      9  17
    A3       3      4   4 
    B1       10     5   4   
    B2        6     1   8
    C1       12     1  12
    C2        2     5   8
    D1        4     1   6 
    D2        2     1   8
 meanA        3     8   12
 meanB        8     3    6
 meanC        7     3   10
 meanD        3     1    7

I tried : the following function

data_all <- rbind(df1,df2)

but it didn't work

I also tried the function bind_rows from dplyr package but this one create new columns.

Thank you

Start with `names(df2) <- sub("mean", "", names(df2))` and then `rbind` the df's. — Rui Barradas, May 04 '22 at 13:30

score 4 · Answer 1 · answered May 04 '22 at 13:30

4

You could always do:

colnames(df2) <- colnames(df1)

data_all <- rbind(df1, df2)

data_all

answered May 04 '22 at 13:30

Aron Strandberg

3,040
9
15

Thank you for you answer , it works , but i'm i want to knwo if there's any way to do this in one command, – Reda May 04 '22 at 13:32

score 3 · Answer 2 · answered May 04 '22 at 13:32

A possible solution:

library(tidyverse)

df1 <- read.table(text = "        col1  col2 col3
A1       4     11  15
A2       2      9  17
A3       3      4   4 
B1       10     5   4   
B2        6     1   8
C1       12     1  12
C2        2     5   8
D1        4     1   6 
D2        2     1   8
", header=T)

df2 <- read.table(text = "meancol1   meancol2 meancol3
meanA        3          8      12
meanB        8          3       6
meanC        7          3      10
meanD        3          1       7
", header=T)

df2 %>% rename_with(~ str_remove(.x, "mean")) %>% 
  bind_rows(df1, .)

#>       col1 col2 col3
#> A1       4   11   15
#> A2       2    9   17
#> A3       3    4    4
#> B1      10    5    4
#> B2       6    1    8
#> C1      12    1   12
#> C2       2    5    8
#> D1       4    1    6
#> D2       2    1    8
#> meanA    3    8   12
#> meanB    8    3    6
#> meanC    7    3   10
#> meanD    3    1    7

Notice that the two dataframes have to have the same column names in order to have `bind_rows` working fine, @Scienceslover. That is what my solution does. — PaulS, May 04 '22 at 13:40
Yes, either way i have to select variables right for example if my first data have two other columns i should do df2[,1:3] %>% rename_with(~ str_remove(.x, "mean")) %>% bind_rows(df1, .) right ? — Reda, May 04 '22 at 13:46

score 2 · Answer 3 · answered May 04 '22 at 13:32

You could do:

rbind(df1, setNames(df2, names(df1)))

      col1 col2 col3
A1       4   11   15
A2       2    9   17
A3       3    4    4
B1      10    5    4
B2       6    1    8
C1      12    1   12
C2       2    5    8
D1       4    1    6
D2       2    1    8
meanA    3    8   12
meanB    8    3    6
meanC    7    3   10
meanD    3    1    7

score 0 · Answer 4 · answered May 04 '22 at 17:57

Another option is to use rbindlist:

library(data.table)

rbindlist(lapply(list(df1, df2), setDT, keep.rownames = TRUE))

# Or using `add_rownames` from `dplyr`
# rbindlist(lapply(list(df1, df2), add_rownames), use.names = F)

Output

       rn col1 col2 col3
 1:    A1    4   11   15
 2:    A2    2    9   17
 3:    A3    3    4    4
 4:    B1   10    5    4
 5:    B2    6    1    8
 6:    C1   12    1   12
 7:    C2    2    5    8
 8:    D1    4    1    6
 9:    D2    2    1    8
10: meanA    3    8   12
11: meanB    8    3    6
12: meanC    7    3   10
13: meanD    3    1    7

Data

df1 <- structure(list(col1 = c(4L, 2L, 3L, 10L, 6L, 12L, 2L, 4L, 2L), 
    col2 = c(11L, 9L, 4L, 5L, 1L, 1L, 5L, 1L, 1L), col3 = c(15L, 
    17L, 4L, 4L, 8L, 12L, 8L, 6L, 8L)), class = "data.frame", row.names = c("A1", 
"A2", "A3", "B1", "B2", "C1", "C2", "D1", "D2"))

df2 <- structure(list(meancol1 = c(3L, 8L, 7L, 3L), meancol2 = c(8L, 
3L, 3L, 1L), meancol3 = c(12L, 6L, 10L, 7L)), class = "data.frame", row.names = c("meanA", 
"meanB", "meanC", "meanD"))

bind two dataframes that have different columns names

4 Answers4