Shell Programming and Scripting

I'm running this command:

awk -v INPUT=2 -f test.awk test

isn't this correct. Also I just copied your code into test.awk and test is a file with numbers, like so:

1
2
3
4
5
6

The error I get is this:

awk: syntax error at source line 1 source file test.awk
context is
awk -v >>> INPUT= <<< $INPUT '
awk: bailing out at source line 18

Right now I'm sitting in Mac OS X, but I need it to run on Solaris. Maybe this is a part of the problem? Also I see that the last row in the first column you get the average value of the column above, that is really nice. Could it be done for every column?

copy the below code to the file test.awk and run like this ./test.awk

Code:

 awk -v INPUT=4 '
BEGIN { ORS=""}
{

print $0 " "
if (( NR % INPUT ) == 0) {
print "\n"
count++
}

if (( NR % INPUT ) == 1) 
sum += $0

}

END { 
 op = sum/count
print op }' test

Alright I ran what you suggested and it worked. How can you make it so that you can provide a variable for how many columns you would like, as you showed me in in the previous program? And then how to take to average value of each column and make a nice looking formated outprint?

My previous code was parameterized. $INPUT is the variable that defines how many columns to be created.

Code:

awk -v INPUT=$INPUT '
BEGIN { ORS=""}
{

print $0 " "
if (( NR % INPUT ) == 0) {
print "\n"
count++
}

if (( NR % INPUT ) == 1) 
sum += $0

}

END { 
 op = sum/count
print op }' Filename

So I changed to INPUT=$INPUT like in the original, but I still isn't able to execute the program, the command I'm running is:

awk -v INPUT=2 -f test.awk

and changed the filename to test.txt (which is my list). But is it possible to also make the the input-file a variable just like INPUT? And also the issue with adding each element in a column and then dividing it with the number of elements.

you need to run the script as i have shown before. Yes even the filename can be parameterized. $Filename is the filename

try this:

Code:

awk -v INPUT=$INPUT '
BEGIN { ORS=""}
{

print $0 " "
if (( NR % INPUT ) == 0) {
print "\n"
count++
}

if (( NR % INPUT ) == 1) 
sum += $0

}

END { 
 op = sum/count
print op }' $(echo $Filename)

Is there any nice command with awk to count the elements in an column? Not an entire program but some nice short way to count the elements like so:

1
2
3
4

Then the "counter" should say 4 since there is 4 elements in the column.