You can try this
import random
class foo:
def __init__(self):
self.bar = random.randint(1,100)
def getbar(ls,i):
ls.append(foo())
ls[i].bar = ls[i].bar * 3
ls = []
[getbar(ls,i) for i in range(10)]
On Thu, Jan 22, 2009 at 4:45 AM, Diez B. Roggisch <de...@nospam.web.de>wrote:
> MRAB schrieb:
>
>> Diez B. Roggisch wrote:
>>
>>> Dr Mephesto wrote:
>>>
>>> Hi,
>>>> Im new to python, and OOP, and am trying to get a handle on list
>>>> comprehension.
>>>>
>>>> Say I have a class Foo with a property called bar:
>>>>
>>>> class Foo:
>>>> def __init__(self):
>>>> self.bar = random.randint(1,100)
>>>>
>>>> and then I make a list of these objects:
>>>>
>>>> Newlist = []
>>>> for x in range(10):
>>>> Newlist.append(Foo())
>>>>
>>>> Now, suppose I wanted to triple the value of 'bar', I could always do:
>>>>
>>>> for x in range(10):
>>>> Newlist[x].bar = Newlist[x].bar * 3
>>>>
>>>> but can I do this using list comprehension? Thanks in Advance!
>>>>
>>>
>>> No, as such, because list-comprehensions require you to have an
>>> *expression*
>>> in front of the iteration:
>>>
>>> resultlist = [<expr> for <variable(s)> in <iterable>]
>>>
>>> Now what you of course can do is this:
>>>
>>> def multiply(item):
>>> item.bar = item.bar * 3
>>>
>>> [multiply(i) for i in items]
>>>
>>> However, doing this will make python produce a list of None-references -
>>> which is a waste. It's up to you if you care about that, but generally it
>>> is frowned upon because of that, and the fact that the conciseness of the
>>> list-comp here isn't really helping with the readability.
>>>
>>> If you had:
>>
>> def multiply(item):
>> item.bar = item.bar * 3
>> return item
>>
>> then:
>>
>> [multiply(i) for i in items]
>>
>> would return items. Still a bad idea, though, because you're using a list
>> comprehension for its side-effect.
>>
>
> And redundant, which was the reason I ommited it.
>
>
> Diez
> --
> http://mail.python.org/mailman/listinfo/python-list
>
--
http://mail.python.org/mailman/listinfo/python-list
