EP 9 Considérer les expressions de générateur pour les grandes compréhensions

  • LIst comprehensions can cause problems for large inputs by using too much memory.

Effective python

Generator is lazy evaluation and comprehension is eager evaluation.

>>> comprehension = [x ** x for x in range(10)]
>>> comprehension
[1, 1, 4, 27, 256, 3125, 46656, 823543, 16777216, 387420489]
>>> generator = (x ** x for x in range(10))
>>> generator
<generator object <genexpr> at 0x10afeea98>
>>> next(generator)
1
>>> next(generator)
1
>>> next(generator)
4
>>> next(generator)
27
>>> next(generator)
256
>>> next(generator)
3125
>>> next(generator)
46656
>>> next(generator)
823543
>>> next(generator)
16777216
>>> next(generator)
387420489
>>> next(generator)
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
StopIteration

yakshaving Different type of evaluation strategy

Generator can make a chian of iterator but be careful when you execute it more than once

>>> it = (x * x for x in range(10))
>>> roots = ((x, x ** 0.5) for x in it)
>>> next(it)
0
>>> next(it)
1
>>> next(it)
4
>>> next(it)
9
>>> next(roots)
(16, 4.0)
>>> next(roots)
(25, 5.0)
>>> next(roots)
(36, 6.0)
>>> next(roots)
(49, 7.0)
>>> next(it)
64
>>> next(roots)
(81, 9.0)

Copy iter: itertools.tee

Recommended Posts

EP 9 Considérer les expressions de générateur pour les grandes compréhensions
EP 16 Envisager un générateur au lieu de renvoyer des listes