Pop multiple items from the beginning and end of a list

Question:

Suppose I have a list of items like this:

mylist=['a','b','c','d','e','f','g','h','i']

I want to pop two items from the left (i.e. a and b) and two items from the right (i.e. h,i). I want the most concise an clean way to do this. I could do it this way myself:

for x in range(2):
    mylist.pop()
    mylist.pop(0)

Any other alternatives?

Asked By: Ehsan88

||

Answers:

You could slice out a new list, keeping the old list as is:

mylist=['a','b','c','d','e','f','g','h','i']
newlist = mylist[2:-2]

newlist now returns:

['c', 'd', 'e', 'f', 'g']

You can overwrite the reference to the old list too:

mylist = mylist[2:-2]

Both of the above approaches will use more memory than the below.

What you’re attempting to do yourself is memory friendly, with the downside that it mutates your old list, but popleft is not available for lists in Python, it’s a method of the collections.deque object.

This works well in Python 3:

for x in range(2):
    mylist.pop(0)
    mylist.pop()

In Python 2, use xrange and pop only:

for _ in xrange(2):
    mylist.pop(0)
    mylist.pop()

Fastest way to delete as Martijn suggests, (this only deletes the list’s reference to the items, not necessarily the items themselves):

del mylist[:2]
del mylist[-2:]

First 2 elements: myList[:2]
Last 2 elements: mylist[-2:]

So myList[2:-2]

Answered By: zer0uno

If you don’t want to retain the values, you could delete the indices:

del myList[-2:], myList[:2]

This does still require that all remaining items are moved up to spots in the list. Two .popleft() calls do require this too, but at least now the list object can handle the moves in one step.

No new list object is created.

Demo:

>>> myList = ['a','b','c','d','e','f','g','h','i']
>>> del myList[-2:], myList[:2]
>>> myList
['c', 'd', 'e', 'f', 'g']

However, from your use of popleft I strongly suspect you are, instead, working with a collections.dequeue() object instead. If so, *stick to using popleft(), as that is far more efficient than slicing or del on a list object.

Answered By: Martijn Pieters

From a performance point of view:

  • mylist = mylist[2:-2] and del mylist[:2];del mylist[-2:] are equivalent
  • they are around 3 times faster than the first solution for _ in range(2): mylist.pop(0); mylist.pop()

Code

iterations = 1000000
print timeit.timeit('''mylist=range(9)nfor _ in range(2): mylist.pop(0); mylist.pop()''', number=iterations)/iterations
print timeit.timeit('''mylist=range(9)nmylist = mylist[2:-2]''', number=iterations)/iterations
print timeit.timeit('''mylist=range(9)ndel mylist[:2];del mylist[-2:]''', number=iterations)/iterations

output

1.07710313797e-06

3.44465017319e-07

3.49956989288e-07

Answered By: mxdbld

To me, this is the prettiest way to do it using a list comprehension:

>> mylist=['a','b','c','d','e','f','g','h','i']
>> newlist1 = [mylist.pop(0) for idx in range(2)]
>> newlist2 = [mylist.pop() for idx in range(2)]

That will pull the first two elements from the beginning and the last two elements from the end of the list. The remaining items stay in the list.

Answered By: Chris Hubley

Python3 has something cool, similar to rest in JS (but a pain if you need to pop out a lot of stuff)

mylist=['a','b','c','d','e','f','g','h','i']
_, _, *mylist, _, _ = mylist
mylist == ['c', 'd', 'e', 'f', 'g']  # true
Answered By: mastercoder8
mylist = ['a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o']

new_list = [mylist.pop(0) for _ in range(6) if len(mylist) > 0]

>>> new_list
['a', 'b', 'c', 'd', 'e', 'f']

new_list = [mylist.pop(0) for _ in range(6) if len(mylist) > 0]
>>> new_list
['g', 'h', 'i', 'j', 'k', 'l']

new_list = [mylist.pop(0) for _ in range(6) if len(mylist) > 0]
>>> new_list
['m', 'n', 'o']
Answered By: Jeremy Palmer
Categories: questions Tags: , ,
Answers are sorted by their score. The answer accepted by the question owner as the best is marked with
at the top-right corner.