getting average of one field if it matches another field

[starlite79]

Hi again.

I have a very large file (over 22000 records) for which I want to search for a field equaling 5.0, 6.0 and 7.0 (third field) and getting the count and average of the second field matching those 3 values. My code runs, but I don't trust the values I get. Could someone offer some help?

$3 == 5.0 { n = n + 1; cergares = cergares + $2 }
END	{ if ( n > 0 )
	     print " average residual for", n, "CERGA is", cergares/n
          else 
             print "No CERGA data found"
         }



$3 == 6.0 { n = n + 1; mlrs2res = mlrs2res + $2 }
END	{ if ( n > 0 )
	     print " average residual for", n, "MLRS2 is", mlrs2res/n
          else 
             print "No MLRS2 data found"
         }


$3 == 7.0 { n = n + 1; apacheres = apacheres + $2 }
END	{ if ( n > 0 )
	     print " average residual for", n, "APACHE is", apacheres/n
          else 
             print "No APACHE data found"
        }

My output to screen looks like this:
average residual for 88 CERGA is -4.17045e-06
average residual for 88 MLRS2 is 5.92045e-06
average residual for 88 APACHE is -0.002412

I highly doubt that each of the three techniques have the same number of points (88), so I know something is not right.

 

[starlite79]

I used different variables for each counter, and this seems to have solved my problem.

I used n, a, and b.

 

[Annihilannic]

Just so you know for the future... the END {} clauses will all be processed when all input files have been processed, not sequentially like you have them placed in your script, which is why your 'n' variable probably appeared to have been clobbered.

Annihilannic.