dict update vs. assignment speed

dupdate.py

import time

class TimeIt(object):
    def __init__(self, name):
        self.name = name

    def __enter__(self, *args, **kwargs):
        self.start_t = time.time()

    def __exit__(self, *args, **kwargs):
        self.end_t = time.time()
        print "{}: took {}s".format(self.name, self.end_t - self.start_t)

d = {}

if __name__ == "__main__":
    for n in [1000, 10000, 100000]:
        with TimeIt("d.update() x {}".format(n)):
            for i in range(n):
                d.update({"x": i})

        with TimeIt("d[] =  x {}".format(n)):
            for i in range(n):
                d["x"] = i

dupdate.py-results

d.update() x 1000: took 0.000370025634766s
d[] =  x 1000: took 0.000133991241455s
d.update() x 10000: took 0.00342798233032s
d[] =  x 10000: took 0.00124788284302s
d.update() x 100000: took 0.0354580879211s
d[] =  x 100000: took 0.013514995575s

OK