Hi, I have a question that I cannot solve.
if I have a file like this (lets say "x-values.dat"):
x1
x2
x3
another file like this (lets say "y-values.dat"):
y1
y2
y3
y4
and another file like this (lets say "p-values.dat"):
p1
p2
p3
...
p12
How can I get this output?
x1 y1 p1
x1 y2 p2
x1 y3 p3
x1 y4 p4
x2 y1 p5
x2 y2 p6
x2 y3 p7
x2 y4 p8
x3 y1 p9
x3 y2 p10
x3 y3 p11
x3 y4 p12
Any help would be much appreciated
paste -d\ file1 file2 file3
If the all files have the same number of records.
No, the files doesn't have the same number of records (in my example, X has 3 values, Y has 4, and P has 12), and the code you suggested me didn't work, because it merge each line of each file, and that is not what I exactly need.
You can check my first post in order to see the output that I need.
Try this:
awk '{
while ( (getline y < "y-values.dat") > 0 ) {
s=$0 FS y
getline p < "p-values.dat"
print s FS p
}
close("y-values.dat")
}' x-values.dat
The code suggested by Franklin52 worked fine, although it prints 4 lines more. This is the output that I got:
x1 y1 p1
x1 y2 p2
x1 y3 p3
x1 y4 p4
x2 y1 p5
x2 y2 p6
x2 y3 p7
x2 y4 p8
x3 y1 p9
x3 y2 p10
x3 y3 p11
x3 y4 p12
y1 p12
y2 p12
y3 p12
y4 p12
But for the moment it's ok, I can remove these extra lines.
Thank you very much Franklin52!
You have one empty line in the file x-values.dat.
You're right, and now your script is working like a charm 
Tahnk you so much!!
$ for i in x{1..3}; do echo $i; done > x.dat
$ for i in y{1..4}; do echo $i; done > y.dat
$ for i in p{1..12}; do echo $i; done > p.dat
$ perl -nle '
use integer;
push @{$h{$ARGV}}, $_;
END {
$i=0;
$fmt = "%10.10s";
foreach $key ( sort { $#{$h{$a}} <=> $#{$h{$b}} } keys %h) {
$c[$i++] = $h{$key};
}
$j=$#{@c[2]}+1;
$k=$#{@c[1]}+1;
$l=$#{@c[0]}+1;
for($i=0, $x=0, $y=0; $i<$j; $i++) {
printf $fmt, $c[0][$x+($i*$l/$j)];
printf $fmt, $c[1][$y++];
printf "$fmt\n", $c[2][$i];
$y=0 if ($y>=$k) ;
}
print "\n";
}
' p.dat y.dat x.dat
x1 y1 p1
x1 y2 p2
x1 y3 p3
x1 y4 p4
x2 y1 p5
x2 y2 p6
x2 y3 p7
x2 y4 p8
x3 y1 p9
x3 y2 p10
x3 y3 p11
x3 y4 p12
Thank you drewk. The code posted by Franklin52 also does the job.
Does that awk script work with different number of lines in the files x y and p?
join -j 2 -o 1.1,2.1 x.dat y.dat |paste -d' ' - p.dat
A bash way
exec 3<p-values.dat
while read X
do while read Y
do read P <&3
echo "$X $Y $P"
done < y-values.dat
done < x-values.dat
exec 3>&-
Yes, the script by Franklin works with any number of lines in x (lets say, nx) and y (lets say ny). The number of lines in p (np) has to be np=nx*ny. But this is how I stated the problem..
so it's ok.
Fabulous! It is a very clever awk script btw.
I thought that the length of x, y, and p were arbitrary and that the it was just sorted right to left The perl script I wrote handles any length file x, y, and p and sorts them right to left.
Cheers,
xlength=`cat x-values.dat|wc -l`
plength=`cat p-values.dat|wc -l`
ylength=`cat y-values.dat|wc -l`
count=0
if [ $plength != $count ]
then
while read -r a
do
printf -v line '%*s' "$ylength"
echo ${line// /$a} | grep -o "$a" >> x-values.datnew
done < x-values.dat
let count=$count+1
else
break
fi
count=0
while [ $xlength != $count ]
do
cat y-values.dat>> y-values.datnew
let count=$count+1
continue
done
paste -d " " x-values.datnew y-values.datnew p-values.dat > new-values.dat
rm -f x-values.datnew y-values.datnew
cat new-values.dat
# end of script
I'm just a bit curious: I just want to know if my bash solution worked by you. (because it works by me with your sample and is independent of the numbers x and y rows.)
Yes frans, your script worked for me for any number of rows in x-values and y-values (provided np=nx*ny).
Thank you so much!