UNIX for Dummies Questions & Answers

Hello. I've read a few threads on how to sort on multiple fields, but I still can't get my file to sort correctly.

I have a comma delimited .csv file will over a hundred fields. I want to sort it by field 2, field 62 and then field 61 (integer fields).

input looks like this well swap field 61 before 62 (sorry):
field 2 field 62 field 61
4609 0 100
48223 -9567 3304
48289 0 3023
48339 -9525 3008
48491 -9754 3054
51610 -7716 3008
51610 \N (this is NULL) \N (this is NULL)
51610 -7716 3008
51610 \N (this is NULL) \N (this is NULL)
53033 -12234 4769

i used:

sort -n -t ',' -k 2,2 -k 62,62 -k 61,61 -T /tempdirectory/temp input.csv -o output.csv

in the results field 2 sorted correctly, but the others are not sorting the nulls correctly. It plops it right in the middle of the same data. For example, it will be 123, 123, null,null, 123, 123 instead of null, null, 123, 123, 123, 123.

Help anyone?

tah might be because you are using -n option.. if you observe man page of sort you will find this
so choose the correct sort option.

Thank you.

I added the -b option to ignore the blanks. This option worked. It sorted the numeric fields and the null fields in the correct order.