On 11/27/13 6:14 AM, Chris Angelico wrote:
On Wed, Nov 27, 2013 at 10:07 PM, Amjad Syed <amjad...@gmail.com> wrote:
Thanks Chris for the reply. But i would like sliding function to be scalable,
as input string can be of 100 letters.
A hundred isn't much to work with, and your code will be fairly
simple. Give it a try with small strings, see how it goes; then try it
on your full-size - I doubt there'll be a problem.
Now, if you were talking about a hundred million letters, then maybe
there'd be a problem. But even there, I'd start with the simple and
clean approach, and only optimize for performance when it becomes
obviously necessary (like when my MUD client was able to saturate one
core of my CPU, just by me typing a lot of commands very rapidly -
that needed fixing!).
ChrisA
Using ChrisA's idea:
def sliding_window(iterable, n, fill=None):
values = list(iterable)
num_values = len(values)
if fill is not None:
values.extend([fill]*(n-1))
need_more = (2*num_values-1) - len(values)
values.extend(values[:need_more])
for start in range(num_values):
yield values[start:start+n]
l = list("LEQNABC")
for n in range(2, len(l)+1):
print [''.join(x) for x in sliding_window(l, n)]
for n in range(2, len(l)+1):
print [''.join(x) for x in sliding_window(l, n, fill="_")]
Produces:
['LE', 'EQ', 'QN', 'NA', 'AB', 'BC', 'CL']
['LEQ', 'EQN', 'QNA', 'NAB', 'ABC', 'BCL', 'CLE']
['LEQN', 'EQNA', 'QNAB', 'NABC', 'ABCL', 'BCLE', 'CLEQ']
['LEQNA', 'EQNAB', 'QNABC', 'NABCL', 'ABCLE', 'BCLEQ', 'CLEQN']
['LEQNAB', 'EQNABC', 'QNABCL', 'NABCLE', 'ABCLEQ', 'BCLEQN', 'CLEQNA']
['LEQNABC', 'EQNABCL', 'QNABCLE', 'NABCLEQ', 'ABCLEQN', 'BCLEQNA',
'CLEQNAB']
['LE', 'EQ', 'QN', 'NA', 'AB', 'BC', 'C_']
['LEQ', 'EQN', 'QNA', 'NAB', 'ABC', 'BC_', 'C__']
['LEQN', 'EQNA', 'QNAB', 'NABC', 'ABC_', 'BC__', 'C___']
['LEQNA', 'EQNAB', 'QNABC', 'NABC_', 'ABC__', 'BC___', 'C____']
['LEQNAB', 'EQNABC', 'QNABC_', 'NABC__', 'ABC___', 'BC____', 'C_____']
['LEQNABC', 'EQNABC_', 'QNABC__', 'NABC___', 'ABC____', 'BC_____',
'C______']
100 elements is really nothing, don't try to over-optimize it. And if
your inputs are really strings, not general iterables, you can use the
same logic but with string operations, and you'll likely have better
performance anyway. Less general, true, but better for your actual problem.
--Ned.
--
https://mail.python.org/mailman/listinfo/python-list
