awk without printing newline

awk '{sum+=$3}; END {printf "%f",sum/NR}' ${file}_${f}_v1.xls >> to-plot-p.xls

print will insert a newline by default. You dont want that to happen, hence use printf instead.

---

The ORS (output record separator) variable in AWK defaults to "\n" and is printed after every line. You can change it to " " in the BEGIN section if you want everything printed consecutively.

---

_{I guess many people are entering in this question looking for a way to avoid the new line in awk. Thus, I am going to offer a solution to just that, since the answer to the specific context was already solved!}

In awk, print automatically inserts a ORS after printing. ORS stands for "output record separator" and defaults to the new line. So whenever you say print "hi" awk prints "hi" + new line.

This can be changed in two different ways: using an empty ORS or using printf.

Using an empty ORS

awk -v ORS= '1' <<< "hello
man"

This returns "helloman", all together.

The problem here is that not all awks accept setting an empty ORS, so you probably have to set another record separator.

awk -v ORS="-" '{print ...}' file

For example:

awk -v ORS="-" '1' <<< "hello
man"

Returns "hello-man-".

Using printf (preferable)

While print attaches ORS after the record, printf does not. Thus, printf "hello" just prints "hello", nothing else.

$ awk 'BEGIN{print "hello"; print "bye"}'
hello
bye
$ awk 'BEGIN{printf "hello"; printf "bye"}'
hellobye

Finally, note that in general this misses a final new line, so that the shell prompt will be in the same line as the last line of the output. To clean this, use END {print ""} so a new line will be printed after all the processing.

$ seq 5 | awk '{printf "%s", $0}'
12345$
#    ^ prompt here

$ seq 5 | awk '{printf "%s", $0} END {print ""}'
12345

---

one way

awk '/^\*\*/{gsub("*","");printf "\n"$0" ";next}{printf $0" "}' to-plot.xls

---

You can simply use ORS dynamically like this:

awk '{ORS="" ; print($1" "$2" "$3" "$4" "$5" "); ORS="\n"; print($6-=2*$6)}' file_in > file_out

---

If Perl is an option, here is a solution using fedorqui's example:

seq 5 | perl -ne 'chomp; print "$_ "; END{print "\n"}'

Explanation:
chomp removes the newline
print "$_ " prints each line, appending a space
the END{} block is used to print a newline

output: 1 2 3 4 5

---

Here's the awk way without having to printf and END (assuming your input is less than, say,500 MB):

seq 199 | mawk NF=NF RS='^$' FS='\n' OFS=

123456789101112131415161718192021222324252627282930313233343
536373839404142434445464748495051525354555657585960616263646
566676869707172737475767778798081828384858687888990919293949
596979899100101102103104105106107108109110111112113114115116
117118119120121122123124125126127128129130131132133134135136
137138139140141142143144145146147148149150151152153154155156
157158159160161162163164165166167168169170171172173174175176
177178179180181182183184185186187188189190191192193194195196
197198199

(reformatted for readability - it's one single line, with \n at its end)

And if you simply wanna place a space in between :

seq 19 | mawk 'NF && --NF' RS='^$' FS='\n'

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

if you don't mind 1 extra trailing space, then it's even simpler

mawk NF=NF RS='^$' FS='\n'

---