My data looks like:
A 4 G 1 G 1 C 4 C 2 C 2 T 6 T 5 T 5 A 6 T 2 T 2 C 6 T 2 T 2 T 6 G 2 G 2 I am trying the command:
awk -F " " '$1==$3 {$7=$6; print $0;} $1==$5 {$7=$4; print $0;} ($1 != $3 && $1 != $5) {$7=$2; print $0}' test.txt While the data has only 5 lines the output has 7 lines and certain lines are randomly duplicated.
Somehow it happens with only this dataset and not the other datasets that I have. Can someone please help. I don't understand what is happening
You didn't describe exactly how you want this to behave - so I'm applying some guesswork here.
Look at the duplicate lines, e.g.
C 4 C 2 C 2 $1 is the same as $3 so the first block fires. $1 is the same as $5 so the second block fires.
If you only want one line of output per line of input, then only output the data in one place, e.g.
awk -F " " '$1==$3 {$7=$6;} ($1==$5) {$7=$4; } ($1 != $3 && $1 != $5) {$7=$2} ($7 != "") { print $0 }' test.txt I think this is the behaviour you are looking for, however it will produce the same or fewer lines of output than input. If you want one line of output for each input line, then remove the condition on the last block.
$7 to be "" at the line ($7 != "") ... given the logic above that line. awk -F " " '$1==$3 {$7=$6; print $0;} $1==$5 {$7=$4; print $0;} ($1 != $3 && $1 != $5) {$7=$2; print $0}' test.txt
If e.g. both $1==$3 and $1==$5 are true, both the first two blocks run and print. That's the case on lines 2 and 3. The two blocks also both set $7 from two different fields, though those are the same on the two lines it happens here.
If you want to only ever print each line a maximum of one time, you can set a flag from the branches and print based on that (or not), e.g.:
awk -F " " '{ p=0; } $1==$3 {$7=$6; p=1} $1==$5 {$7=$4; p=1} ($1 != $3 && $1 != $5) {$7=$2; p=1} p {print}' test.txt print prints $0 if no other arguments are given, and you can actually use just p without a code block in the end, as the default action is just that.
Similarly, to unconditionally print every line, you often see just a trailing 1, as in awk '/.../ { ... } 1'
You'll have to decide what to do with field $7 though, as those three branches all set them to different values.
If you want to execute only one of the blocks (at most), you can use the next statement in each to go the next line:
awk -F " " '$1==$3 {$7=$6; print; next} $1==$5 {$7=$4; print; next} ($1 != $3 && $1 != $5) {$7=$2; print; next} ' test.txt ...and actually looking at the conditions, it seems to me the last one is only ever true if and only if the first two ones are false, so we might as well write it all as an if-else:
awk -F " " '{ if ($1==$3) { $7=$6 }; else if ($1==$5) { $7=$4 }; else { $7=$2 }; print; }' test.txt $7 will already be set even without any of the conditions being true ($1 != $3 && $1 != $5) is only ever true if the first two ones are false so we might as well print unconditionally
$1==$3and$1==$5are true, both the first two blocks run and print. That's the case on lines 2 and 3. The two blocks also both set$7from two different fields, though those are the same on the two lines it happens here.-F " "is useless btw as that's setting FS to the default value it already has, just remove it.