Shell Programming and Scripting

Hey fellas,

I've posted this problem a few days back and I received just one post which was in PHP that I have no idea about! (Thanks to DGPickett) It would be so nice if you can help me with this in Shell or Perl. Here is the story:

I have a big table with variables and observations. I managed to simplify the data to I/O so now it's something looks like this:


What I like to find out is the co-presence of the variables. The script has to read the table in a loop which in each round check one row with all the others (including itself). While checking 2 rows write the number of times there is number 1 for both in the same column.

Simply, it starts from row A and first check it with itself (since the first row is A!). The returning number would be the times that we have "1" in A. Then by checking row A and B it returns number of columns in which we have "1" for A and B. Like this it goes until the end of the table and writes the results to the first row of the output.

Then the output would be a table with variables as both rows and columns containing the co-presence of them based on the calculations.

It is a cool problem that I unfortunately don't have enough knowledge to solve it. If you have knowledge and time, you're a lifesaver.

Thanks for reading even if you are not going to help!

Please post the output required for this sample.

It would be something like this:


If I didn't make a mistake in my calculations!

Code:

awk 'NR>1{name[NR-1]=$1;for(i=2;i<=NF;i++) if($i==1) oneset[NR-1,i]=1}
END{
for(i=1;i<=(NR-1);i++)
{
 if(i==1)
 {
  print "*"
  for(j=1;j<=(NR-1);j++)
   print name[j]
  printf "\n"
 }
 print name[i]
 for(j=1;j<=(NR-1);j++)
 {
  n=0
  for(k=2;k<=NF;k++)
   if(oneset[i,k] && oneset[j,k])
    n++
  print n
 }
 printf "\n"
}
}' ORS='\t' file

WOW! honestly... There are a-lot here that I have to learn first! I'm going through it and try to apply it on my data. I'll let you know if I managed to make it or not.
Thanks man...

elixir_sinari is unbeatable. Still, I can see some (minor!) simplifications to his proposal:

Code:

awk     'NR>1 {name[NR-1]=$1
               for(i=2;i<=NF;i++) oneset[NR-1,i]=$i}
         END  { print "*"
                for(j=1;j<=(NR-1);j++)
                print name[j]
                printf "\n"

                for(i=1;i<=(NR-1);i++)
                {
                 print name[i]
                 for(j=1;j<=(NR-1);j++)
                         {n=0
                          for(k=2;k<=NF;k++)  n += oneset[i,k] * oneset[j,k]
                          print n
                         }
                 printf "\n"
                }
             }
        ' ORS='\t' file

Oh my! You guys are amazing! You know what Rudic I think You are right! elixir_sinari is awesome! So are you as I've seen your posts in the forum!
I was trying to understand what elixir_sinari wrote. It seems it's working what he gave me!! Then I will go through your simplification!

Thanks to you both!!