List without duplicates?

[Jeremy Hylton]

Wolfgang Grafen <Wolfgang.Grafen at marconicomms.com> writes:

> > Is there a built in Python function that eliminates duplicates in a list?

> Try this ...
> 
> old_list = [1, 2, 3, 2, 1, 4, 3, 5, 2]
> 
> 
> unique_list = []
> 
> for member in old_list:
>     if not member in unique_list:
>         unique_list.append(member)
> 

This is a bad way to do it.  The "not member in unique_list" test
involves a linear search of unique_list.  This gives your uniquify
function O(n^2) behavior.  The dictionary version posted earlier uses
the dictionary's hash table and achieves constant time for the test.
It will perform much better.

Jeremy