Like map() and filter(), zip can create and return a generator. generator is an implementation of iterator with lazy evaluation features, which can reduce memory usage very effectively.
zip() can combine elements of the same index from multiple lists so the combined object is in the tuple state.
a = [1, 2, 3]
b = [4, 5, 6]
c = [*zip(a, b)]
# c = [(1, 4), (2, 5), (3, 6)]We can use the zip(*zip_object) syntax to achieve the unzip.
c = [(1, 4), (2, 5), (3, 6)]
print(list(zip(*c))) # [(1, 2, 3), (4, 5, 6)]When the list is not the same length, the shortest list will be selected as the base.
a = [1, 2, 3]
b = [4, 5, 6, 8]
c = [100, 777, 999, 1213, 4456]
d = [*zip(a, b, c)]
# d = [(1, 4, 100), (2, 5, 777), (3, 6, 999)]If you want to use the longest list as a base and add padding to other list elements, use itertools.zip_longest() instead of zip().
from itertools import zip_longest
a = [1, 2, 3]
b = [4, 5, 6, 8]
c = [100, 777, 999, 1213, 4456]
d = [*zip_longest(a, b, c)] # empty will be filled with None
# d = [(1, 4, 100), (2, 5, 777), (3, 6, 999), (None, 8, 1213), (None, None, 4456)]
d = [*zip_longest(a, b, c, fillvalue=0)] # filled with 0
# d= [(1, 4, 100), (2, 5, 777), (3, 6, 999), (0, 8, 1213), (0, 0, 4456)]In machine learning model training, it is often necessary to shuffle the dataset, which can be achieved using the zip() function as described below.
import random
X = [1, 2, 3, 4, 5, 6]
y = [0, 1, 0, 0, 1, 1]
zipped_data = list(zip(X, y))
# [(1, 0), (2, 1), (3, 0), (4, 0), (5, 1), (6, 1)]
random.seed(42)
random.shuffle(zipped_data)
# [(4, 0), (2, 1), (3, 0), (5, 1), (1, 0), (6, 1)]
shuffled_zipped_data = map(list, zip(*zipped_data))
# map([4, 2, 3, 5, 1, 6], [0, 1, 0, 1, 0, 1])
new_X, new_y = shuffled_zipped_data
# new_X = [4, 2, 3, 5, 1, 6]
# new_y = [0, 1, 0, 1, 0, 1]