Why assigning a function as variable is faster in Python

Question

Today I observed an interesting behavior using Python.

def f():
    ls1 = []
    for x in range(1000000):
        ls1.append(x)
%timeit f()
77.2 ms ± 1.83 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)

And tested a second function.

def f2():
    ls1 = []
    lsa = ls1.append
    for x in range(1000000):
        lsa(x)        
%timeit f2()
56 ms ± 566 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)

Assigning append to a variable was faster than using append inside the loop.

Why the second function is faster?

Maarten Fabré · Accepted Answer · 2018-09-06 13:53:50Z

2

Python is a dynamic language. Attribute lookup is comparable to a dict lookup, so in the first scenario, you first lookup ls1 in the local scope and then lookup append on the ls1-object each of 1000000 iterations. In the second scenario you lookup the append on ls1 only once, and then 1 million iterations where you lookup lsa in the local scope

answered Sep 6, 2018 at 13:53

Maarten Fabré

7,0781 gold badge19 silver badges37 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

blhsing · Accepted Answer · 2018-09-06 13:51:10Z

2

Because in the second function, you have the reference to the method already and can call it directly in every iteration, whereas in the first function, it has to first find the append method of the ls1 object first in every iteration before making the call.

answered Sep 6, 2018 at 13:51

blhsing

109k9 gold badges88 silver badges132 bronze badges

Collectives™ on Stack Overflow

Why assigning a function as variable is faster in Python

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related