Transpose number not separated in columns using Bash

Question

I've tried the tr function to transpose in bash but it's not working. e.g. tr 'abcd' 'jkmn'...

The idea is to take a series of numbers:

92121
92911

and switch them so that they would look like this:

Here is a test dataset:

echo "92121
      92911
      29222
      22222
      22222
      22222" > ~/Desktop/output.geno

I know we can separate numbers using the cut function.

I feel that I could use

for var in 1:96
do
   tmp=$(cut -c var output.geno)
   tr $tmp
done

Possible duplicate of [An efficient way to transpose a file in Bash](http://stackoverflow.com/questions/1729824/an-efficient-way-to-transpose-a-file-in-bash) — Jose Ricardo Bustos M., Feb 20 '17 at 20:58
You have lots of syntax errors. You can't have a statement before `do`. You're missing the `$()` around the `cut` command. — Barmar, Feb 20 '17 at 20:59
in French they were saying that it's "transposition" so I thought it was the same. https://fr.wikipedia.org/wiki/Tr_(Unix) — M. Beausoleil, Feb 20 '17 at 20:59
You set the variable `tmp` then use `$temp`. Is that typo in the real code or a copying error? — Barmar, Feb 20 '17 at 21:02
@JoseRicardoBustosM., the output from the link you share is this: `92121 92911 29222 22222 22222 22222`. Is it possible to stack the numbers instead of having them in a row? Actually, it's just separating the numbers, not transposing anything — M. Beausoleil, Feb 20 '17 at 21:02
Instead of using `cut -f` use `cut -c`. And instead of `wc -w` use `wc -c`. — Barmar, Feb 20 '17 at 21:04
If it helps you `for i in {1..5}; do tmp=$(cut -c $i output.geno | tr -d '[:space:]'); echo $tmp; done` — Jose Ricardo Bustos M., Feb 20 '17 at 22:22

score 1 · Accepted Answer · answered Feb 20 '17 at 23:35

@M. Beausoleil: Try: Let's say Input_file is as follows.

cat  Input_file
92121
92911
29222
22222
22222
22222

Following code could help in same.

awk '{for(i=1;i<=length($0);i++){array[i]=array[i]?array[i] substr($0,i,1):substr($0,i,1)};MAX=MAX>i?MAX:i} END{for(j=1;j<MAX;j++){print array[j]}}'   Input_file

NON-one liner form of above solution too as follows.

awk '{
        for(i=1;i<=length($0);i++){
                                        array[i]=array[i]?array[i] substr($0,i,1):substr($0,i,1)
                                  };
        MAX=MAX>i?MAX:i
     }
     END{
                for(j=1;j<MAX;j++){
                                        print array[j]
                                  }
     }
    '   Input_file

Output will be as follows.

score 0 · Answer 2 · answered Feb 21 '17 at 02:41

Input

Output

$ awk '
{   
    n=split($0,F,r)
    for (i=1; i in F; i++)
        a[NR,i] = F[i]
}
n>p { p = n }
END {    
    for(j=1; j<=p; j++) 
    {
        str=a[1,j]
        for(i=2; i<=NR; i++)
        {
            str=str a[i,j];
        }
        print str
    }
}' file
992222
229222
192222
212222
112222

Transpose number not separated in columns using Bash

2 Answers2