Shell Programming and Scripting

Hi,

I'm trying to reformat a file to create a new columns reflecting the previous 2 over and over. By that I mean currently each observation has two columns and I want to create a third which has a value equal to 1 minus the sum of the previous two.

This is slightly complicated as 1) I require the new column to have the same header/first row as the previous two 2) I want to skip the first three columns 3) I want to do this to the end of the line, for files with different line lengths (numbers of observations)

Just to make it clearer, this would be an example input file with just a header and one row:

Code:

Var1 Var2 Var3 ID1 ID1 ID2 ID2.... IDn IDn
Var1 Var2 Var3 0.1 0.8 0.2  0.3.... x    y

And the output would look something like this:

Code:

Var1 Var2 Var3 ID1 ID1 ID1 ID2 ID2 ID2.... IDn IDn IDn
Var1 Var2 Var3 0.1 0.8 0.1  0.2 0.3 0.5....  x   y   (1-x-y)

I imagine the solution is something similar to that described in the topic "awk or python to join alternating columns" (sorry cant post urls yet) but I'm afraid I don't have the know-how to make all the changes I need!

Any help would be greatly appreciated!

Try:

Code:

awk 'NR==1{for (i=4;i<=NF;i+=2) $(i+1)=$i" "$i}NR>1{for (i=4;i<=NF;i+=2) $(i+1)=$(i+1)" "(1-$i-$(i+1))}1' file

How about this:

Code:

awk '{printf $1" "$2" "$3}
NR==1 { for(i=4;i<=NF;i++) {
   printf " " $i;
   if(i%2) printf " "$i }}
NR>1 { for(i=4;i<=NF;i++) {
   printf " " $i;
   if (i%2) printf " "1.0-$i-$(i-1)}}
{print ""}' infile

or this

Code:

awk '{printf $1" "$2" "$3}
{ for(i=4;i<=NF;i++) {
   printf " " $i;
   if(i%2) printf " " (NR==1?$i:1.0-$i-$(i-1))
}}
{print ""}' infile

Thanks to you both! I only tried bartus11's response and it seems to have worked perfectly!

---------- Post updated at 06:28 PM ---------- Previous update was at 10:58 AM ----------

Sorry one last request as I seem to be suffering from rounding errors. Is it possible to incorporate a rule into the commands suggested that if the newly generated columns are negative (i.e. the sums of the two preceding columns used to generate the value are greater than one) they should be set instead to zero?

Thanks again!

Code:

awk 'NR==1{for (i=4;i<=NF;i+=2) $(i+1)=$i" "$i}NR>1{for (i=4;i<=NF;i+=2) {x=1-$i-$(i+1);$(i+1)=$(i+1)" "(x>0?x:0)}}1' file

Code:

awk '{for (i=4;i<=NF;i+=2) {x=1-$i-$(i+1);$(i+1)=$(i+1)" "(NR==1?$i:(x>0?x:0))}}1' file

Thanks again to you both! Very much appreciated!