See the timing difference between using for loops and itertools

timeit_sample.py

from itertools import ifilter
import sys
import timeit

REPEAT = 3
EXECUTIONS = 10000
VALUES = range(0, 10000)

def double(x):
    return x*2

def with_loop(values,):
    for x in values:
        x = double(x)

def with_itertools(values):
    for x in values:
        _ = x


if __name__ == '__main__':
    print "\n=== timeit tests REPEAT={0} EXECUTIONS={1} TOTAL={2} ===".format(
            REPEAT, EXECUTIONS, REPEAT*EXECUTIONS)

    t = timeit.repeat('with_loop(VALUES)',
                      setup="from __main__ import"
                            "        double, with_loop, VALUES",
                      repeat=REPEAT,
                      number=EXECUTIONS)
    print "with_loop:      avg = %.8f" % (sum(t)/len(t))

    t = timeit.repeat('with_itertools(ifilter(double, VALUES))',
                      setup="from __main__ import"
                            "        double, with_itertools, VALUES, ifilter",
                      repeat=REPEAT,
                      number=EXECUTIONS)
    print "with_itertools: avg = %.8f" % (sum(t)/len(t))

=== timeit tests REPEAT=3 EXECUTIONS=10000 TOTAL=30000 ===
with_loop: avg = 9.94407606
with_itertools: avg = 12.94683170

Interesting that itertools adds a 30% overhead?!

Note: Python 2.7.6 on MacBook PRO