using uniq and awk??

amatuer_lee_3 · #1 (**permalink**) 05-16-2008

I have a file that is populated:

Code:

hits/books.hits:143.217.64.204       Thu Sep 21 22:24:57 GMT 2006
hits/books.hits:62.145.39.14         Fri Sep 22 00:38:32 GMT 2006
hits/books.hits:81.140.86.170        Fri Sep 22 08:45:26 GMT 2006
hits/books.hits:81.140.86.170        Fri Sep 22 09:13:57 GMT 2006
hits/cds.hits:67.231.144.166         Mon Sep 04 23:57:22 GMT 2006
hits/cds.hits:182.210.215.110        Tue Sep 05 22:53:53 GMT 2006
hits/cds.hits:94.83.157.230          Wed Sep 06 22:13:28 GMT 2006
hits/cds.hits:148.214.45.187         Wed Sep 06 23:25:22 GMT 2006
hits/cds.hits:221.253.17.33          Sat Sep 09 00:58:14 GMT 2006
hits/cds.hits:182.210.215.110        Sun Sep 10 05:29:28 GMT 2006
hits/cds.hits:193.140.224.143        Sun Sep 10 16:35:11 GMT 2006
hits/cds.hits:182.210.215.110        Fri Sep 15 21:08:31 GMT 2006
hits/cds.hits:199.226.220.114        Sat Sep 16 13:38:18 GMT 2006
hits/cds.hits:150.205.160.134        Mon Sep 18 19:22:45 GMT 2006
hits/contact.hits:15.111.224.138     Sat Sep 09 12:07:26 GMT 2006
hits/contact.hits:199.179.222.39     Sat Sep 09 22:30:15 GMT 2006
hits/contact.hits:199.179.222.39     Sun Sep 10 12:14:13 GMT 2006
hits/contact.hits:234.33.121.120     Sun Sep 10 22:19:39 GMT 2006

I have code that sorts this information:

Code:

echo "Page:\t \t Hits \t Unique Hits \t "

awk -F: '{print $1}' HITS | uniq
wc -l < HITS
awk -F: '{print $2}' HITS | sort -u | wc -l

My results for this code are:

Code:

Page:                            Hits                   Unique Hits
hits/books.hits
hits/cds.hits
hits/contact.hits
18
14

18 represents the total hits for the entire file
14 represents unique IP's in column two from the entire file.

However I want the output to look like this, so that each individual page name lists its own hits and unique IP's:

Code:

Page:                      Hits                       Unique Hits
hits/books.hits             4                               3
hits/cds.hits               10                              8
hits/contact.hits           4                               3

Any suggestions/solutions?

ripat · #2 (**permalink**) 05-16-2008

awk can do it all:

Code:

BEGIN {FS=":"}
{
        hits[$1]++
        ip_hits[$1 ":" substr($2, 1, 15)]
}
END {
        for (i in ip_hits){ sub(/:.+/, "", i); unique[i]++ }
        printf "%-20s %5s %s\n", "Page:", "Hits", "Unique Hits"
        for (i in hits) printf "%-20s %5s %5s\n", i, hits[i], unique[i]
}

amatuer_lee_3 · #3 (**permalink**) 05-16-2008

could you explain this a little better for me please. i like to understand code im using

thanks.

and how can i encorporate it into my code? do i just put it in as it is?

ripat · #4 (**permalink**) 05-17-2008

You just can use it instead of your code. No need for wc, uniqu or sort any more.

This is the code commented with explanation:

Code:

 # define the field separator
BEGIN{FS=":"}

# start to loop through the file. For each record, add one to
# the array hits[site.name] thix will count the total number of hits
{
        hits[$1]++

        # create one array with index: "site.name:ip_address"
        # this way there will be only one index site.name:ip_adress (unique hits)
        ip_hits[$1 ":" substr($2, 1, 15)]
}

# once all fil is processed we print array contents by looping through
# associative indexes for (i in array)....
END {
        # this line counts the number of *different* elements in array ip_hits
        for (i in ip_hits){ sub(/:.+/, "", i); unique[i]++}

        # print title
        pr_format="%-20s %5s %s\n"
        printf pr_format, "Page:", "Hits", "Unique Hits"

        # loop through array hits and array unique
        for (i in hits) printf pr_format, i, hits[i], unique[i]
}

To help you understand: here is a printout of the different arrays used above

hits (counts total number of hits per site)

Code:

[hits/cds.hits] => 10
[hits/contact.hits] => 4
[hits/books.hits] => 4

ip_hits (no value here. Just indexes)

Code:

[hits/cds.hits:182.210.215.110] => 
[hits/books.hits:143.217.64.204 ] => 
[hits/cds.hits:150.205.160.134] => 
[hits/cds.hits:221.253.17.33  ] => 
[hits/books.hits:81.140.86.170  ] => 
[hits/cds.hits:67.231.144.166 ] => 
[hits/contact.hits:15.111.224.138 ] => 
[hits/contact.hits:234.33.121.120 ] => 
[hits/cds.hits:193.140.224.143] => 
[hits/cds.hits:199.226.220.114] => 
[hits/cds.hits:94.83.157.230  ] => 
[hits/books.hits:62.145.39.14   ] => 
[hits/contact.hits:199.179.222.39 ] => 
[hits/cds.hits:148.214.45.187 ] =>

unique (counts number of elements in ip_hits)

Code:

[hits/cds.hits] => 8
[hits/contact.hits] => 3
[hits/books.hits] => 3

To use the code:

Code:

$ awk -f script-name you-input-file

radoulov · #5 (**permalink**) 05-17-2008

Another one
(use nawk or /usr/xpg4/bin/awk on Solaris):

Code:

awk -F'[: ]' 'END {
fmt = "%-20s\t%s\t%s\n"
printf fmt, "Page:", "Hits", "Unique Hits"
for (p in h)
  printf fmt, p, h[p], u[p]
}
!_[$1,$2]++ { u[$1]++ }
{ h[$1]++ }' file

amatuer_lee_3 · #6 (**permalink**) 05-17-2008

Code:


awk -F'[: ]' 'END {
fmt = "%-20s\t%s\t%s\n"
printf fmt, "Page:", "Hits", "Unique Hits"
for (p in h)
  printf fmt, p, h[p], u[p]
}
!_[$1,$2]++ { u[$1]++ }
{ h[$1]++ }' file

Again could you explain this for me please?

ripat · #7 (**permalink**) 05-17-2008

Quote:

Originally Posted by radoulov

Another one
(use nawk or /usr/xpg4/bin/awk on Solaris):

Code:

awk -F'[: ]' 'END {
fmt = "%-20s\t%s\t%s\n"
printf fmt, "Page:", "Hits", "Unique Hits"
for (p in h)
  printf fmt, p, h[p], u[p]
}
!_[$1,$2]++ { u[$1]++ }
{ h[$1]++ }' file

That's neat and terse! Exactly the way I like it! I should have thought to use a regex as FS.

More UNIX and Linux Forum Topics You Might Find Helpful
Thread	Thread Starter	Forum	Replies	Last Post
uniq command???	skyineyes	UNIX for Dummies Questions & Answers	2	05-28-2008 06:27 AM
uniq options	dhanamurthy	Shell Programming and Scripting	0	05-08-2008 08:08 AM
Uniq using only the first field	Digby	UNIX for Dummies Questions & Answers	8	01-16-2008 06:25 AM
help on UniQ	vishal_ranjan	HP-UX	0	06-21-2007 10:33 AM
sort/uniq	jimmyflip	UNIX for Dummies Questions & Answers	3	10-17-2002 05:09 AM

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes	Rate This Thread
Switch to Linear Mode Hybrid Mode Switch to Threaded Mode	Rate This Thread:


	Powered by