linux - Subtract a constant number from a column -

i have 2 large files (~10gb) follows:

file1.csv

name,id,dob,year,age,score mike,1,2014-01-01,2016,2,20 ellen,2, 2012-01-01,2016,4,35 . . 

file2.csv

id,course_name,course_id 1,math,101 1,physics,102 1,chemistry,103 2,math,101 2,physics,102 2,chemistry,103 . . 

i want subtract 1 "id" columns of these files:

file1_updated.csv

name,id,dob,year,age,score mike,0,2014-01-01,2016,2,20 ellen,0, 2012-01-01,2016,4,35 

file2_updated.csv

id,course_name,course_id 0,math,101 0,physics,102 0,chemistry,103 1,math,101 1,physics,102 1,chemistry,103 

i have tried awk '{print ($1 - 1) "," $0}' file2.csv, did not correct result:

-1,id,course_name,course_id 0,1,math,101 0,1,physics,102 0,1,chemistry,103 1,2,math,101 1,2,physics,102 1,2,chemistry,103 

you've added column in attempt. instead set first field $1 $1-1:

 awk -f"," 'begin{ofs=","} {$1=$1-1;print $0}' file2.csv 

that semicolon separates commands. set delimiter comma (-f",") , output field seperator comma begin{ofs=","}. first command subtract 1 first field executes first, print command executes second, entire record, $0, contain new $1 value when it's printed.

it might helpful subtract 1 records not header. can add condition first command:

awk -f"," 'begin{ofs=","} nr>1{$1=$1-1} {print $0}' file2.csv 

now subtract when record number (nr) greater 1. print entire record.