Iterate over pairs in a list (circular fashion) in Python

Question

The problem is easy, I want to iterate over each element of the list and the next one in pairs (wrapping the last one with the first).

I\'ve thought about two unpyth

Answer 1

def pairs(lst):
    i = iter(lst)
    first = prev = item = i.next()
    for item in i:
        yield prev, item
        prev = item
    yield item, first

Works on any non-empty sequence, no indexing required.

Answer 2

This might be satisfactory:

def pairs(lst):
    for i in range(1, len(lst)):
        yield lst[i-1], lst[i]
    yield lst[-1], lst[0]

>>> a = list(range(5))
>>> for a1, a2 in pairs(a):
...     print a1, a2
...
0 1
1 2
2 3
3 4
4 0

If you like this kind of stuff, look at python articles on wordaligned.org. The author has a special love of generators in python.

Answer 3

def pairs(ex_list):
    for i, v in enumerate(ex_list):
        if i < len(list) - 1:
            print v, ex_list[i+1]
        else:
            print v, ex_list[0]

Enumerate returns a tuple with the index number and the value. I print the value and the following element of the list ex_list[i+1]. The if i < len(list) - 1 means if v is not the last member of the list. If it is: print v and the first element of the list print v, ex_list[0].

Edit:

You can make it return a list. Just append the printed tuples to a list and return it.

def pairs(ex_list):
    result = []
    for i, v in enumerate(ex_list):
        if i < len(list) - 1:
            result.append((v, ex_list[i+1]))
        else:
            result.append((v, ex_list[0]))
    return result

Answer 4

Of course, you can always use a deque:

from collections import deque
from itertools import *

def pairs(lst, n=2):
    itlst = iter(lst)
    start = list(islice(itlst, 0, n-1))
    deq = deque(start, n)
    for elt in chain(itlst, start):
        deq.append(elt)
        yield list(deq)

Answer 5

i=(range(10))

for x in len(i):
    print i[:2]
    i=i[1:]+[i[1]]

more pythonic than this is impossible

Answer 6

I've coded myself the tuple general versions, I like the first one for it's ellegant simplicity, the more I look at it, the more Pythonic it feels to me... after all, what is more Pythonic than a one liner with zip, asterisk argument expansion, list comprehensions, list slicing, list concatenation and "range"?

def ntuples(lst, n):
    return zip(*[lst[i:]+lst[:i] for i in range(n)])

The itertools version should be efficient enough even for large lists...

from itertools import *
def ntuples(lst, n):
    return izip(*[chain(islice(lst,i,None), islice(lst,None,i)) for i in range(n)])

And a version for non-indexable sequences:

from itertools import *
def ntuples(seq, n):
    iseq = iter(seq)
    curr = head = tuple(islice(iseq, n))
    for x in chain(iseq, head):
        yield curr
        curr = curr[1:] + (x,)

Anyway, thanks everybody for your suggestions! :-)