Timeit and Understanding how fast different approaches can be

ex_1.py

$ python3 -m timeit "'foo' + 'bar'"
50000000 loops, best of 5: 5.31 nsec per loop

ex_2.py

$ python3 -m timeit -r 10 "'foo' + 'bar'"
50000000 loops, best of 10: 5.29 nsec per loop

ex_3.py

$ python3 -m timeit -r 10 -n 100000 "'foo' + 'bar'"
100000 loops, best of 10: 5.28 nsec per loop

ex_4.py

$ python3 -m timeit -r 10 -n 100000 "''.join(('foo', 'bar'))"
100000 loops, best of 10: 42.5 nsec per loop

ex_5.py

$ python3 -m timeit -r 10 -n 100000 "'%s, %s' % ('foo', 'bar')"
100000 loops, best of 10: 63.1 nsec per loop

ex_6.py

def test_str_1() -> str:
    return 'foo' + 'bar'

def test_str_2() -> str:
    return ''.join(('foo', 'bar'))

def test_str_3() -> str:
    return '%s, %s' % ('foo', 'bar')

if __name__ == "__main__":
    import timeit
    r = 10
    n = 100000
    print("test_str_1: ", timeit.repeat(setup="from __main__ import test_str_1", stmt="test_str_1()", number=n, repeat=r))
    print("test_str_2: ", timeit.repeat(setup="from __main__ import test_str_2", stmt="test_str_2()", number=n, repeat=r))
    print("test_str_3: ", timeit.repeat(setup="from __main__ import test_str_3", stmt="test_str_3()", number=n, repeat=r))

# $ python3 ex_6.py    
# test_str_1:  [0.003694088023621589, 0.0037375990068539977, 0.003711616969667375, 0.0036614019772969186, 0.0036590969539247453, 0.003655606007669121, 0.003692480968311429, 0.0037061700131744146, 0.0036806080024689436, 0.003687592048663646]
# test_str_2:  [0.008561848953831941, 0.008489075000397861, 0.00847748201340437, 0.00858063600026071, 0.008465468999929726, 0.008496129012200981, 0.008449406013824046, 0.008478877949528396, 0.008490820997394621, 0.008434668998233974]
# test_str_3:  [0.013133066997397691, 0.013086761988233775, 0.013070349988993257, 0.013110997970215976, 0.01309297897387296, 0.013118749018758535, 0.013096819049678743, 0.013113581982906908, 0.013256893958896399, 0.013094236026518047]

ex_6_with_matplotlib.py

def test_str_1() -> str:
    return 'foo' + 'bar'

def test_str_2() -> str:
    return ''.join(('foo', 'bar'))

def test_str_3() -> str:
    return '%s, %s' % ('foo', 'bar')


if __name__ == "__main__":
    import matplotlib.pyplot as plt
    import timeit
    r = 10
    n = 1000000
    test_str_1 = timeit.repeat(setup="from __main__ import test_str_1", stmt="test_str_1()", number=n, repeat=r)
    test_str_2 = timeit.repeat(setup="from __main__ import test_str_2", stmt="test_str_2()", number=n, repeat=r)
    test_str_3 = timeit.repeat(setup="from __main__ import test_str_3", stmt="test_str_3()", number=n, repeat=r)

    f = plt.figure()
    f.set_figwidth(10)
    f.set_figheight(5)

    plt.plot(test_str_1, marker='o', label="sum")
    plt.plot(test_str_2, marker='o', label="join")
    plt.plot(test_str_3, marker='o', label="% interpolation")
    plt.title(f"List Extension Methods and Best Times from {n} Executions in {r} Rounds")
    plt.xlabel("Rounds")
    plt.ylabel("Units of Seconds")
    plt.xticks(range(0, r))
    plt.legend()
    plt.show()

ex_7.py

def list_concat_1():
    l = ["a", "b", "c"]
    return l + ["d", "e", "f"]

def list_concat_2():
    l = ["a", "b", "c"]
    l += ["d", "e", "f"]
    return l

def list_concat_3():
    l = ["a", "b", "c"]
    l.extend(["d", "e", "f"])
    return l

if __name__ == "__main__":
    import timeit
    r = 10
    n = 100000
    print("list_concat_1: ", timeit.repeat(setup="from __main__ import list_concat_1", stmt="list_concat_1()", number=n, repeat=r))
    print("list_concat_2: ", timeit.repeat(setup="from __main__ import list_concat_2", stmt="list_concat_2()", number=n, repeat=r))
    print("list_concat_3: ", timeit.repeat(setup="from __main__ import list_concat_3", stmt="list_concat_3()", number=n, repeat=r))

# $ python3 ex_7.py
# list_concat_1:  [0.013129231985658407, 0.013070706045255065, 0.012952743971254677, 0.012848541024141014, 0.012713467003777623, 0.012749644985888153, 0.012814807007089257, 0.0128303820383735, 0.012878363020718098, 0.012793435947969556]
# list_concat_2:  [0.01266855897847563, 0.01276836299803108, 0.012766197032760829, 0.013836722995620221, 0.012670096999499947, 0.012525454978458583, 0.0126781280268915, 0.01254989899462089, 0.012621487025171518, 0.012537747039459646]
# list_concat_3:  [0.01478160498663783, 0.016013532993383706, 0.015955634997226298, 0.014748011017218232, 0.014714905992150307, 0.01476882299175486, 0.015143102034926414, 0.015066416002810001, 0.014712321979459375, 0.014758766046725214]

ex_7_with_matplotlib.py

def list_concat_1():
    l = ["a", "b", "c"]
    return l + ["d", "e", "f"]

def list_concat_2():
    l = ["a", "b", "c"]
    l += ["d", "e", "f"]
    return l


def list_concat_3():
    l = ["a", "b", "c"]
    l.extend(["d", "e", "f"])
    return l

if __name__ == "__main__":
    import matplotlib.pyplot as plt 
    import timeit
    r = 10
    n = 1000000
    list_concat_1 = timeit.repeat(setup="from __main__ import list_concat_1", stmt="list_concat_1()", number=n, repeat=r)
    list_concat_2 = timeit.repeat(setup="from __main__ import list_concat_2", stmt="list_concat_2()", number=n, repeat=r)
    list_concat_3 = timeit.repeat(setup="from __main__ import list_concat_3", stmt="list_concat_3()", number=n, repeat=r)

    f = plt.figure()
    f.set_figwidth(10)
    f.set_figheight(5)

    plt.plot(list_concat_1, marker='o', label="sum")
    plt.plot(list_concat_2, marker='o', label="incremented")
    plt.plot(list_concat_3, marker='o', label="extended")
    plt.title(f"List Extension Methods and Best Times from {n} Executions in {r} Rounds")
    plt.xlabel("Rounds")
    plt.ylabel("Nanoseconds (nsecs)")
    plt.xticks(range(0, r)) 
    plt.legend()
    plt.show()