On Tuesday 11 August 2009 10:27:26 Richard Marza wrote:
> Think of the file I'm using as a spreadsheet. The headers(column names) are
> on top and the values are below them. Each line has an item with multiple
> values under different systems.
>
>
> Item System1 System3 System4 ...
>
> nio 5.0 5.5 5.0 (these are individual
> values. They are nothing more than what they represent. The item (nio) and
> the float or integer representing its value under the different
> system...It's just a file)
>
> My goal is to take "nio" and figure out which system has a different price
> than the others. So if the script were to run through 200 lines of similar
> text it should definitely kick-out: "nio has discrepancy in System3; price
> 5.5 in line 1"
>
>
> Another thing, all systems can have different prices. This must also kick
> out. This is really a script to report discrepancies.
As I suspected, you can do the whole thing in awk only. Basically, I'm going
to assume you want to report items which don't have all the same values on all
systems. It's easy to spot those, but it might not be as easy to determine
which values are the "normal" ones and which are the discrepant ones,
especially if for example each system has a different value. You have to
provide additional logic to tell the "discrepant" values from the others. For
the moment, the script just prints out the lines where all the columns don't
have the same value.
awk 'NR==1{print;next}{for(i=3;i<=NF;i++){if($i!=$2){print;break}}}' file.txt
