# sum numbers in multiple files **URL:** \ **Category:** Shell Programming and Scripting\ **Created:** [July 21, 2008, 2:23pm UTC](https://community.unix.com/t/sum-numbers-in-multiple-files/205674 "2008-07-21T14:23:10Z")\ **Posts on this page:** 6\ **Page:** 1
**Author:** ![pattywac](https://community.unix.com/letter_avatar/pattywac/32/6_22734026db12803ebb3e059622ac3534.png) [@pattywac](https://community.unix.com/u/pattywac)\ **Post date:** [July 21, 2008, 2:23pm UTC](https://community.unix.com/t/sum-numbers-in-multiple-files/205674/1 "2008-07-21T14:23:10Z")
I have 11 directories with around 200 files in each. In each directory the files are labeled out.0 through out.201 . Each file has around 118 numbers in a single column. I need to sum the files in each directory so each directory will have a resultant vector that is 118 numbers long. I then need to combine the vectors into a single file that is 11\*118 (1298) numbers long. Someone mentioned using 'cat' but since i only wrote my first csh script over the weekend I have no idea where to start. thanks in advance ---
**Author:** ![joeyg](https://community.unix.com/user_avatar/community.unix.com/joeyg/32/780_2.png) [@joeyg](https://community.unix.com/u/joeyg)\ **Post date:** [July 21, 2008, 3:37pm UTC](https://community.unix.com/t/sum-numbers-in-multiple-files/205674/2 "2008-07-21T15:37:58Z")
```nohighlight DirA out.0 contains 118 rows, each with a number out.1 contains 118 rows, each with a number ... DirB repeats ``` Fairly straight-forward to hav a loop add up all of the numbers. But, not sure what you mean by resulting vector. ---
**Author:** ![pattywac](https://community.unix.com/letter_avatar/pattywac/32/6_22734026db12803ebb3e059622ac3534.png) [@pattywac](https://community.unix.com/u/pattywac)\ **Post date:** [July 21, 2008, 3:47pm UTC](https://community.unix.com/t/sum-numbers-in-multiple-files/205674/3 "2008-07-21T15:47:38Z")
you have the directory structure right. Im not trying to sum up each number within the file. Within each dir all 200 of the first entries would be summed, same for the second ..... through the 118th entry of each file. This would result in each directory having a summed column that is 118 numbers long. I then need to combine those numbers from each dir into one file so it would contain the 118 numbers \* 11 dir = 1298 numbers in the final data file. so inside a dir: out.1: 0 5 10 out.2 0 5 100 sum= 0 10 110 ---
**Author:** ![joeyg](https://community.unix.com/user_avatar/community.unix.com/joeyg/32/780_2.png) [@joeyg](https://community.unix.com/u/joeyg)\ **Post date:** [July 21, 2008, 4:34pm UTC](https://community.unix.com/t/sum-numbers-in-multiple-files/205674/4 "2008-07-21T16:34:36Z")
```nohighlight DirA out.0 contains 118 rows, each with a number out.1 contains 118 rows, each with a number ... DirB repeats ``` ```nohighlight inside DirA open out.0 add up all rows in the file, get a number 31543 write that to a temp file call DirA_sums open out.1 add up all rows in the file, get a number 19548 write that to a temp file call DirA_sums DirA_sums now contains 31543 19548 continue through all of the files inside DirB process the out.? files in same manner continue through DirK (11th directory) ``` After the above, I would have eleven files, each containing 118 numbers (which are the sums from the files within the directory) Do I got that right? ---
**Author:** ![pattywac](https://community.unix.com/letter_avatar/pattywac/32/6_22734026db12803ebb3e059622ac3534.png) [@pattywac](https://community.unix.com/u/pattywac)\ **Post date:** [July 21, 2008, 4:46pm UTC](https://community.unix.com/t/sum-numbers-in-multiple-files/205674/5 "2008-07-21T16:46:10Z")
not quite..... should be more like this: ```nohighlight inside DirA open out.0 store all rows in the file to array open out.1 store all rows in the file to temparray set array = array + temparray open out.2 store all rows in the file to temparray set array = array + temparray continue through all of the files Write array to tempfile out.DirA inside DirB process the out.0-200 files in same manner continue through DirK (11th directory) after 11th directory Write out.DirA to finaloutput.file append out.DirB to finaloutput.file ... append out.DirK to finaloutput.file ``` ---
**Author:** ![joeyg](https://community.unix.com/user_avatar/community.unix.com/joeyg/32/780_2.png) [@joeyg](https://community.unix.com/u/joeyg)\ **Post date:** [July 21, 2008, 5:21pm UTC](https://community.unix.com/t/sum-numbers-in-multiple-files/205674/6 "2008-07-21T17:21:30Z")
given a sample structure (2 dirs, 4 files in each) each file has five numbers in it, one per line ./DirA/out.0 ./DirA/out.1 ./DirA/out.2 ./DirA/out.3 ./DirB/out.0 ./DirB/out.1 ./DirB/out.2 ./DirB/out.3 Start of the script programming ```nohighlight > cat manip #! /usr/bin/bash find . -name out* | sort >file_list rm *.arr 2>/dev/null while read zf do echo $zf tfile=$(echo $zf | cut -d "/" -f2) arr_txt=$(cat $zf | tr "\n" " ") echo "$arr_txt" >>"$tfile".arr done manip ./DirA/out.0 ./DirA/out.1 ./DirA/out.2 ./DirA/out.3 ./DirB/out.0 ./DirB/out.1 ./DirB/out.2 ./DirB/out.3 ``` see the resultant two files created ```nohighlight > cat DirA.arr 2 4 6 8 10 1 3 5 7 9 21 22 31 32 33 12 13 14 15 16 > cat DirB.arr 3 4 5 6 7 6 7 19 10 11 5 10 9 6 14 19 20 2 3 10 ``` I know that the files were read properly, to create each line in my .arr files. So, if this looks to be on-course, it is fairly straight-forward to combine my .arr files into one larger file. So...