Command Line – How to Merge Files Using a Common Column

Try this,

grep -f file1 file2 | awk '!_[$2]++'

 1  C95696352 score:  -69.785 nathvy =  38 nconfs =          888
 3  C96209347 score:  -57.128 nathvy =  24 nconfs =         1188
 6  C89372772 score:  -55.728 nathvy =  22 nconfs =         3228
20  C85594749 score:  -49.780 nathvy =  44 nconfs =         4536

Explanation

• grep -f file1 file2: search in file2 for matches of patterns obtained from file1
• awk '!_[$2]++': Don't print anything if field $2 has been seen before (via)
  • _ is the array name (can be anything, e.g. "seen")
  • _[$2]++ will create an array entry with the key being the content of field $2 and add 1
  • If _[$2] was not (!) already set, print the line. The printcommand is the default action that is made by awk when the condition matches.

The key here is "at the same line" - that suggests you need to index the array with the record number in the current file, FNR, rather than with the field value $3

Ex.

awk 'NR==FNR{a[FNR]=$3; next} {print $2, $3 == a[FNR] ? "ok" : "error"}' file1 file2

Alternatively, if you know how many columns there are in the first file (let's call it N), you could just paste them side by side and then just compare $3 with $(N+3)

Ex.

paste file1 file2 | awk '{print $2, $14 == $3 ? "OK" : "error"}'

Note I omitted the -F"," because your files as posted are whitespace delimited not comma delimited - if they are actually CSV files, add it back.