Question

Here is my data.

df <- data.frame(Aa = 1:10, Ab = 11:20,Ca = 41:50, Cb = 51:60, Ba = 21:30, Bb = 31:40,
Da = 51:60, Db = 61:70) # actual data has > 6000 columns
df 
   Aa Ab Ca Cb Ba Bb Da Db
1   1 11 41 51 21 31 51 61
2   2 12 42 52 22 32 52 62
3   3 13 43 53 23 33 53 63
4   4 14 44 54 24 34 54 64
5   5 15 45 55 25 35 55 65
6   6 16 46 56 26 36 56 66
7   7 17 47 57 27 37 57 67
8   8 18 48 58 28 38 58 68
9   9 19 49 59 29 39 59 69
10 10 20 50 60 30 40 60 70

mf <- data.frame (SN = 1:5, colname = c("A", "B", "C", "D", "E"))
mf 
  SN colname
1  1       A
2  2       B
3  3       C
4  4       D
5  5       E

I want to order the columns of df (actually pairs with suffix a and b) based on mf$colname. Exclude those column that is not in mf.

So output will look like:

df1 <- data.frame(Aa = 1:10, Ab = 11:20, Ba = 21:30, Bb = 31:40,Ca = 41:50, Cb = 51:60,
    Da = 51:60, Db = 61:70)   
df1
   Aa Ab Ba Bb Ca Cb Da Db
1   1 11 21 31 41 51 51 61
2   2 12 22 32 42 52 52 62
3   3 13 23 33 43 53 53 63
4   4 14 24 34 44 54 54 64
5   5 15 25 35 45 55 55 65
6   6 16 26 36 46 56 56 66
7   7 17 27 37 47 57 57 67
8   8 18 28 38 48 58 58 68
9   9 19 29 39 49 59 59 69
10 10 20 30 40 50 60 60 70

Edits:

order does not mean alphabetic order. order based on mf$colnames

thanks;

Major edits:
I think the example I provided is clear what I want to do, as I my desired column by coincidence alphabetical order. But my intention is not true. Please consider another example.

 df <- data.frame (k1a = 1:10, k1b = 11:20, a1a = 21:30, a1b = 31:40, 
j1a = 41:50, j1b = 51:60, p1a = 61:70, p1b = 71:80)
df
   k1a k1b a1a a1b j1a j1b p1a p1b
1    1  11  21  31  41  51  61  71
2    2  12  22  32  42  52  62  72
3    3  13  23  33  43  53  63  73
4    4  14  24  34  44  54  64  74
5    5  15  25  35  45  55  65  75
6    6  16  26  36  46  56  66  76
7    7  17  27  37  47  57  67  77
8    8  18  28  38  48  58  68  78
9    9  19  29  39  49  59  69  79
10  10  20  30  40  50  60  70  80

mf <- data.frame (SN = 1:5, colname = c("p1", "a1", "j1", "k1", "e1")
mf 
  SN colname
1  1      p1
2  2      a1
3  3      j1
4  4      k1
5  5      e1

this desired output

df1 <- data.frame ( p1a = 61:70, p1b = 71:80,  a1a = 21:30, a1b = 31:40,
  j1a = 41:50, j1b = 51:60, k1a = 1:10, k1b = 11:20)
df1
   p1a p1b a1a a1b j1a j1b k1a k1b
1   61  71  21  31  41  51   1  11
2   62  72  22  32  42  52   2  12
3   63  73  23  33  43  53   3  13
4   64  74  24  34  44  54   4  14
5   65  75  25  35  45  55   5  15
6   66  76  26  36  46  56   6  16
7   67  77  27  37  47  57   7  17
8   68  78  28  38  48  58   8  18
9   69  79  29  39  49  59   9  19
10  70  80  30  40  50  60  10  20

Answer 1

Changed answer:

To the updated question, I think you can do it as followed. Given dataset:

df <- data.frame (k1a = 1:10, k1b = 11:20, a1a = 21:30, a1b = 31:40, 
j1a = 41:50, j1b = 51:60, p1a = 61:70, p1b = 71:80)

mf <- data.frame (SN = 1:5, colname = c("p1", "a1", "j1", "k1", "e1"))

You can use grep() to find the indexes of variables names that correspond to each of the elements of mf$colnane

df1 <- df[unlist(sapply(as.character(mf$colname),grep,names(df)))]

This becomes:

> df1 
   p1a p1b a1a a1b j1a j1b k1a k1b
1   61  71  21  31  41  51   1  11
2   62  72  22  32  42  52   2  12
3   63  73  23  33  43  53   3  13
4   64  74  24  34  44  54   4  14
5   65  75  25  35  45  55   5  15
6   66  76  26  36  46  56   6  16
7   67  77  27  37  47  57   7  17
8   68  78  28  38  48  58   8  18
9   69  79  29  39  49  59   9  19
10  70  80  30  40  50  60  10  20