In a bash shell, I am trying to write a short one-liner to divide the line count of one file by the line count of another. I would like show at least two decimal places from floating-point division.
In other words, I would like a one-liner to print the percentage of one file's line count of another.
For example, if I have a first file (first.txt) with 25 lines and a second file (second.txt) with 100 lines, the one-liner would output .25.
The shortest I could come up with made use of command substitution and redirection
$ echo "scale=2; $(wc -l <first.txt) / $(wc -l <second.txt)" | bc
.25
Let me know if you have questions.
First some test material:
$ for i in {1..25} ; do echo $i >> first ; done
$ for i in {1..100} ; do echo $i >> second ; done
Then some awk:
$ awk 'END{print(NR-FNR)/FNR}' first second
0.25
NR is the total number of records in both files. At the END FNRis the number of records in the latter file, so (NR-FNR)/FNR is the record count of the first file divided by the record count of the second.
Other than the solutions provided here, you need to take your question to the code golf course .
alternatively, you could use gawk as:
gawk 'NR==FNR{a+=1;next} {b+=1} END{printf "%.2f\n", a/b}' first.txt second.txt
The following stores the line counts of the two files in two variables. The filenames and leading whitespace are removed from the output of wc using sed and cut. The division is then performed on the two variables using bc.
count1=$(wc -l first.txt | sed -e 's/^[[:space:]]*//' | cut -f 1 -d ' ') && count2=$(wc -l second.txt | sed -e 's/^[[:space:]]*//' | cut -f 1 -d ' ') && echo "scale=2; $count1/$count2" | bc
Shortest version using bc and grep
echo "$(grep -c '$' file1) / $(grep -c '$' file2)" | bc -l
An alternate GNU awk answer, using a bunch of builtin variables :
gawk '
ENDFILE {nr[FILENAME] = FNR}
END {printf "%.2f\n", nr[ARGV[1]] / nr[ARGV[2]]}
' first.txt second.txt
One time use
echo `wc -l <first.txt`/`wc -l <second.txt`|bc -l
One script use
e(){ echo `wc -l <$1`/`wc -l <$2` | bc -l;}
e first.txt second.txt
One user use: add function to .bashrc
Everyone use: add function to /etc/bash.bashrc
bash-4.4$ a=$(perl -e "print $(cat file1|wc -l) /$(cat file2|wc -l)")
bash-4.4$ echo $a
0.333333333333333
Edited the answer to support decimal values
An alternative to all the bc answers:
dc <<< "2k $(wc -l < file1) $(wc -l < file2) /p"
The 2k sets the precision, here the output will always have two decimal places.
