Shell Programming and Scripting

Hi I'm having a problem with converting a file:

ID X
1 7
1 8
1 3
2 5
2 7
2 2

To something like this:

ID X1 X2 X3
1 7 8 3
2 5 7 2

I've tried the following loop:
for i in `cat tst.csv| awk -F "," '{print $1}'| uniq`;do grep -h $i tstb.csv | awk -F "," '{i=2;for(j=1;j<NR;j++) printf "%s,",$i;j=NR;printf "%s,",$i"\n"}';done>out.csv

But I'm getting the following
X
,7
,8,8
,3,3,3
,5
,7,7
,2,2,2

So the loop is stuffed and it looks like I'm having a problem with newlines as well. Can anyone offer a solution to help?

One way to do it with Perl:

Code:

$ 
$ cat f7
ID X
1 7
1 8
1 3
2 5
2 7
2 2
$ 
$ 
$ ##
$ perl -lane 'BEGIN {print "ID X1 X2 X3 X4"}
>             if (!/^ID/) { chomp;
>               if ($p eq "") {printf $_; $p=$F[0]}
>               elsif ($F[0]==$p) {printf " $F[1]"}
>               else {printf "\n$_"; $p=$F[0]}
>             } END {print}' f7
ID X1 X2 X3 X4
1 7 8 3
2 5 7 2
$ 
$

tyler_durden

Assuming they always come in sets of 3:

Code:

echo ID X1 X2 X3
sed 1d data | paste -d' ' - - - | cut -d' ' -f1-2,4,6

... where "data" is the name of the file to be processed, whose first line is a discarded header ("ID X").

Regards,
Alister

If you dont want the output sorted, try:

Code:

awk 'BEGIN { print "ID X1 X2 X3"; }NR>1{ A[$1]=A[$1]" "$2; } END { for (i in A) { print i A[i]; } }' file

Thanks for the suggestions. I ended up using something similar to the first suggestion, but written in python. I was hoping to do it in a bash script, but had some trouble with the other two suggestions. Thanks again for your help. If there is any interest I can put up my python script.
Cheers