prodius

Calculating cartesian product is not easy if not using itertools.product(). Prodius library allows to perform the same but also allowing iterables to be callables that return iterable. Provided callables will be called when iterable is needed to calculate cartesian product.

itertools.product() ofsen raise MemoryError when one of iterables is too large. No matter matter the size of iterable, prodius will never write whole iterator into memory.

Issues

• Iterables cannot be more than 11.
• Maybe be slower than itertools.product().

Installing

Prodius can be installed with pip in your command-line.

pip install prodius

Usage

Prodius is just easy just like builtin itertools.product().

Realise that this wont offer any benefit as iterator are passed direcly.

import prodius

iterables = [range(10), range(10)]
for item in prodius.product(*iterables, repeat=2):
    print(item)

Prodius also allows iterables to contain callable that returns iterable.

The difference is that functions were used as iterables which provide iterables on demand.

import prodius

iterables = [lambda: range(10), lambda: range(10)]
for item in prodius.product(*iterables, repeat=2):
    print(item)

This compares itertools.product() with prodius.product().

import prodius

# prodius.product() is used to calculate cartesian product.
# No exception or MemoryError expected even if iterables are large.
iterables = [lambda: range(10000000000), lambda: range(10000000000)]
product = prodius.product(*iterables) # works as espected

import itertools

# itertools.product() is used to calculate cartesian product.
# MemoryError expected as iterables are too large.
iterables = [range(10000000000), range(10000000000)]
product = itertools.product(*iterables) # MemoryError

Prodius also gives control of over items of iterable to be used in cartesian product. That could be accomplised using itertools.cycle() by returning different iterable each time function is called.

import prodius
import itertools

foods_numbers_list = [[1,2], ["apple", "orange"]]
foods_numbers_cycle = itertools.cycle(foods_numbers_list)

def numbers_foods():
    return next(foods_numbers_cycle)

for item in prodius.product([100, 200], numbers_foods):
    print(item)
# (100, 1)
# (100, 2)
# (200, 'apple')
# (200, 'orange')

Notice that 100 was matched with numbers (1, 2) while 200 matched with ('apple', 'orange').

Thats a fantastic trick with prodius even if it does not have realworld application but who knows.

License

MIT license