CODEWITHSUNDEEP
awkdivision
Taron
Taron

Reputation: 23

awk. Rounding the result of division

I have a test file:

user1   230230  1970
user2   313221  1978
user5   212334  1924
user3   313221  1987
user4   212345  1999
user3   123456  1983
user5   121212  1903
user2   234234  1992
user3   124690  1922
user5   092343  1945

when i run this script:

 awk 'BEGIN{OFMT="%.2f"}{user_arr[$1]+=$2;user_arr_count[$1]+=1;}END{for (el in user_arr) {n=user_arr[el]/user_arr_count[el];print n}}' test

I get result

230230
273727.50
187122.33
212345
141963

but if i run

awk 'BEGIN{OFMT="%.2f"}{user_arr[$1]+=$2;user_arr_count[$1]+=1;}END{for (el in user_arr) {n=user_arr[el]/user_arr_count[el];print n" "el}}' test

I get rounds value:

230230 user1
273728 user2
187122 user3
212345 user4
141963 user5

why the result is rounded? Thanks for the explanation!

Upvotes: 2

Views: 1425

Answers (1)

Chris Seymour
Chris Seymour

Reputation: 85785

In the first script you are printing the value n and awk knows to treat it as a number however in the second script you are printing n" "el which is treated as a string because you are using string concatenation. What you want to do instead is print n,el.

I would write the script like:

$ awk '{u[$1]+=$2;c[$1]+=1}END{for(i in u)print i,u[i]/c[i]}' OFMT="%.2f" file
user1 230230
user2 273727.50
user3 187122.33
user4 212345
user5 141963

Upvotes: 3

Related Questions