# Unexpected Interaction Of Features

Subscribe!

My latest posts can be found here:
Previous blog posts:

# An Unexpected Interaction of Features - 2018/06/07

I've been dealing with some data, and using my usual technique of using command-line tools to play with it for a while before writing a program to do the full analysis.

But something was wrong, and it took me a while to work it out.

I was sorting a file:

 which aerodynamically electroencephalogram exotically aerodynamically a differentiation -> a aerodynamically aerodynamically differentiation electroencephalogram exotically which

But my file has as the first field a count:

 5 which 15 aerodynamically 20 electroencephalogram 10 exotically 15 aerodynamically 1 a 15 differentiation -> 10 exotically 15 aerodynamically 15 aerodynamically 15 differentiation 1 a 20 electroencephalogram 5 which

That's not what I wanted, but this was a game I'd played before. The utility sort is working on the data as text, so it's alphabetical. I need to sort using -n to get it to sort numerically:

 5 which 15 aerodynamically 20 electroencephalogram 10 exotically 15 aerodynamically 1 a 15 differentiation -> 1 a 5 which 10 exotically 15 aerodynamically 15 aerodynamically 15 differentiation 20 electroencephalogram

Excellent, but now I realise there are repeated lines, and I need to de-duplicate. So I use sort -u to do that:

 5 which 15 aerodynamically 20 electroencephalogram 10 exotically 15 aerodynamically 1 a 15 differentiation -> 10 exotically 15 aerodynamically 15 differentiation 1 a 20 electroencephalogram 5 which

The duplication is gone, but the screwy ordering is back, because I forgot the "numerical" flag, so sort -nu is what I need:

 5 which 15 aerodynamically 20 electroencephalogram 10 exotically 15 aerodynamically 1 a 15 differentiation -> 1 a 5 which 10 exotically 15 aerodynamically 20 electroencephalogram

Spot the difference.

Yes, the "differentiation" line has gone, and I can only assume that when both the n and u flags are set, it only takes the numbers into account when deciding if there are duplicates. I haven't explored whether, for a given number, it (a) sorts and keeps the first, (b) sorts and keeps the last, (c) keeps the first in the input then sorts, (d) keeps the last in the input then sorts, or (e) something else.

But it's certainly not what I expected.

So now it's back to using "sort -n | uniq" rather than "sort -nu".

For reference: "sort --version" returns "sort (GNU coreutils) 8.21"

 <<<< Prev <<<< Archimedes Hat Box Theorem : >>>> Next >>>> Why Top Posting Has Won ...

You can follow me on Mathstodon.

## Send us a comment ...

 You can send us a message here. It doesn't get published, it just sends us an email, and is an easy way to ask any questions, or make any comments, without having to send a separate email. So just fill in the boxes and then

 Your name : Email : Message :