Python list of tuples

Question

I'm a newbie with python and programming in general and need some help: I have a list that I've created earlier in the program by appending from a loop (i.e. I can't just redefine my list now to solve my problem), of 24 4-tuples:

elementary = [(23, 1, 18, 4), (23, 1, 6, 16), (23, 1, 4, 18), (23, 2, 18, 3), (23, 2, 12, 9), (23, 2, 9, 12), (23, 2, 3, 18), (23, 3, 18, 2), (23, 3, 2, 18), (23, 4, 18, 1), (23, 4, 1, 18), (23, 5, 14, 7), (23, 5, 7, 14), (23, 6, 16, 1), (23, 6, 9, 8), (23, 6, 8, 9), (23, 6, 1, 16), (23, 7, 14, 5), (23, 7, 5, 14), (23, 8, 9, 6), (23, 8, 6, 9), (23, 9, 12, 2), (23, 9, 8, 6), (23, 9, 6, 8), (23, 9, 2, 12), (23, 12, 9, 2), (23, 12, 2, 9), (23, 14, 7, 5), (23, 14, 5, 7), (23, 16, 1, 6), (23, 18, 4, 1), (23, 18, 3, 2), (23, 18, 2, 3), (23, 18, 1, 4)]

but now would like to get rid of the tuples that are just re-arranged... in other words, after the first tuple ((23,1,18,4)) I would let get rid of (23,1,4,18), (23,4,1,18), etc..., and if possible, I'd like to do this throughout the list, so that I only end up with 6 completely distinct 4-tuples. Is there any way to do this without going back and doing something differently earlier in my program? Any help would be greatly appreciated. Thanks!

• tacaswell: Welcome to SO. What have you tried?

Answer 1

How about:

{tuple(sorted(t)): t for t in elementary}.values()

Answer 2

As a 1-liner, this sorts each 4-tuple, then creates a set of the result, shich has the effect of removing the duplicates. I'm assuming your 4-tuples are allowed to have the order of elements changed.

set(tuple(sorted(i)) for i in elementary)
>>> set((5, 7, 14, 23), (6, 8, 9, 23), (2, 3, 18, 23), (1, 4, 18, 23), (1, 6, 16, 23), (2, 9, 12, 23))

• wim: This destroys the original tuples ordering, which may or may not be a problem ...

Answer 3

Do you really have only 24? If so, a slow-ish solution with some unnecessary memory allocations may work fine here, and save you programming time:

elementary_unique = set(tuple(sorted(t)) for t in elementary)

Now elementary_unique is a set rather than a list - if it matters, you can use

elementary_unique = list(set(tuple(sorted(t)) for t in elementary))

instead, but this will be a little slower than the first version.

• user2006083: Thank you all very much! I ended up using the elementary_unique set version but I'll give all of the suggestions a shot as a learning exercise. Thanks again!

Answer 4

You can recognize the equivalent combinations whenever sorted(tuple1) == sorted(tuple2).

The code is short and sweet:

>>> set(map(tuple, map(sorted, elementary)))
set([(5, 7, 14, 23), (6, 8, 9, 23), (2, 3, 18, 23), 
 (1, 4, 18, 23), (1, 6, 16, 23), (2, 9, 12, 23)])

If you need to preserve the ordering of each of the first distinct tuples, it takes a little more work:

>>> uniq = set()
>>> for t in elementary:
 s = tuple(sorted(t))
 if s not in uniq:
 uniq.add(s)
 print t
(23, 1, 18, 4)
(23, 1, 6, 16)
(23, 2, 18, 3)
(23, 2, 12, 9)
(23, 5, 14, 7)
(23, 6, 9, 8)

Answer 5

If you really want to get down and dirty with the comparisons:

In [1028]: elementary = [(23, 1, 18, 4), (23, 1, 6, 16), (23, 1, 4, 18), (23, 2, 18, 3), (23, 2, 12, 9), (23, 2, 9, 12), (23, 2, 3, 18), (23, 3, 18, 2), (23, 3, 2, 18), (23, 4, 18, 1), (23, 4, 1, 18), (23, 5, 14, 7), (23, 5, 7, 14), (23, 6, 16, 1), (23, 6, 9, 8), (23, 6, 8, 9), (23, 6, 1, 16), (23, 7, 14, 5), (23, 7, 5, 14), (23, 8, 9, 6), (23, 8, 6, 9), (23, 9, 12, 2), (23, 9, 8, 6), (23, 9, 6, 8), (23, 9, 2, 12), (23, 12, 9, 2), (23, 12, 2, 9), (23, 14, 7, 5), (23, 14, 5, 7), (23, 16, 1, 6), (23, 18, 4, 1), (23, 18, 3, 2), (23, 18, 2, 3), (23, 18, 1, 4)]
In [1029]: for e in elementary: 
 add = True 
 for a in answer: 
 if all(_e in a and e.count(_e)==a.count(_e) and len(e)==len(a) for _e in e):
 add = False
 if add:
 answer.append(e) 
In [1030]: answer
Out[1030]: 
[(23, 1, 18, 4),
 (23, 1, 6, 16),
 (23, 2, 18, 3),
 (23, 2, 12, 9),
 (23, 5, 14, 7),
 (23, 6, 9, 8)]