How to calculate a sum of certain records? - Shell Programming and Scripting

sickboy · #1 (**permalink**) 06-10-2005

Hi,
i have a file where the records are like this.

vt100 2048 D402 MG0010 0
586 262144 D403 MG0011 1000
486 8192 D404 MG0012 270
386 8192 A423 CC0177 40
586 65536 A424 CC0182 670
486 16384 A423 CC0183 100
486 16384 A425 CC0184 80
65000 4096 B407 EE1027 80

I want firstly,
to count how many times each combination of the first 2 letters from the 4th field exist and print it like this:
3 MG or MG 3
5 CC

I tried this :

Code:

awk -F"\t" '{print $4}' pcs.txt | uniq -c

but I didn't know how to count only the first two letters of the fied.

Secondly,
I want to sum the second field for each differrent entry of the 3rd field and print them. Like this
A423 24576
A425 16384

I tried this :

Code:

awk -F"\t" '{a[$3]=$3; 
for (i = 1; i <= NR; i++){
for (i = 1; i <= NR; i++){
if(a[++i] = $3){
size+=$2}
}
print a[i]" "size
}' pcs.txt

but somewhere I have a mistake and I can't find it and also i am not sure if this is going to give me what I want.

Please help me

vino · #2 (**permalink**) 06-10-2005

This should do it for your first counting part...

Code:

sed -n -e 's/.*\([A-Z][A-Z]\)[0-9]*.*/\1/p' pcs.txt | uniq -c

For your input, it gave,

3 MG
4 CC
1 EE

Basically, the scripts strips out every character except for the two upper-case characters. And the script assumes you always have input which obey the format you provided.

Vino

vino · #3 (**permalink**) 06-10-2005

Oops...

Got the wrong message in the wrong thread.

Vino

sickboy · #4 (**permalink**) 06-10-2005

Maybe I didn't explained it right. I need to find the sum for every different record of the 3rd field. So I need to sum the two numbers of the A423 and print the total.

PS Sorry for my English

sickboy · #5 (**permalink**) 06-10-2005

Quote:

Originally Posted by vino

This should do it for your first counting part...

Code:

sed -n -e 's/.*\([A-Z][A-Z]\)[0-9]*.*/\1/p' pcs.txt | uniq -c

It is working for the part of the file i gave in the beginning but it is not working in the whole file right. Probably uniq -c is counting the same lines that are one next to the other and not spread in the file. What I did to get the result I wanted is the use of sort.

Code:

sed 's/.*\([A-Z][A-Z]\)[0-9]*.*/\1/p' pcs.txt |sort| uniq -c

Do you know if instead of sed I can use awk to get the same result?

vino · #6 (**permalink**) 06-10-2005

I think for uniq to work properly, it needs a sort'ed list.

Not sure, tho'.

Vino

r2007 · #7 (**permalink**) 06-10-2005

for 2nd question:

Code:

awk -F"\t" '{a[$3]+=$2}END{for (i in a) print i,a[i]}'

[not test]

More UNIX and Linux Forum Topics You Might Find Helpful
Thread	Thread Starter	Forum	Replies	Last Post
calculate 13 months ago	andrea_mussap	Shell Programming and Scripting	4	11-21-2007 04:14 PM
calculate 13 months ago	andrea_mussap	Shell Programming and Scripting	3	11-16-2007 11:18 AM
Count No of Records in File without counting Header and Trailer Records	guiguy	Shell Programming and Scripting	2	06-07-2007 09:15 AM
How To Calculate	krishna_sicsr	Shell Programming and Scripting	1	03-28-2007 07:24 AM
How to calculate with awk	whatisthis	Shell Programming and Scripting	4	11-09-2005 08:39 AM

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode