itertools

A lightning talk by Daniel Nephin

... standardizes a core set of fast, memory efficient tools that are useful by themselves or in combination. Together, they form an “iterator algebra” making it possible to construct specialized tools succinctly and efficiently in pure Python.

Slicing

>>> seq = range(200)
>>> itertools.islice(seq, 5, 10)) 
5, 6, 7, 8, 9
            

Infinite

>>> itertools.count(start=3, step=2)
3, 5, 7, 9, 11, 13, 15, 17, 19, 21, ...

>>> itertools.cycle(range(3))
0, 1, 2, 0, 1, 2, 0, 1, 2, 0, 1, 2, ...

>>> itertools.repeat('a')
'a', 'a', 'a', 'a', 'a', 'a', 'a', ...
                    

Filtering

>>> eq_to_six = functools.partial(operator.eq, 6)
>>> filter(eq_to_six, range(200))
6

>>> itertools.filterfalse(eq_to_six, range(200))
0, 1, 2, 3, 4, 5, 7, 8, 9, 10, ...

>>> lt_six = functools.partial(operator.gt, 6)
>>> itertools.dropwhile(lt_six, range(200))
6, 7, 8, 9, 10, 11, 12, 13, 14, 15, ...

>>> itertools.takewhile(lt_six, range(200))
0, 1, 2, 3, 4, 5
                    

Splitting

>>> div_by_three = lambda v: v // 3
>>> itertools.groupby(range(10), key=div_by_tree)
(0, [1, 2]), (1, [3, 4, 5]), (2, [6, 7, 8]), ...

>>> itertools.tee(range(4), 3)
(0, 1, 2, 3), (0, 1, 2, 3), (0, 1, 2, 3)
                    

Combining

>>> itertools.chain(['a', 'b', 'c'], [1, 2, 3])
'a', 'b', 'c', 1, 2, 3 

>>> itertools.zip_longest(range(3), ascii_lowercase)
(0, 'a'), (1, 'b'), (2, 'c'), (None, 'd'), ...
                    

Combinatorics

>>> itertools.product('ABCD', 'xy')
('A', 'x'), ('A', 'y'), ('B', 'x'), ('B', 'y'), ...

>>> itertools.combinations('abc', 2)
('a', 'b'), ('a', 'c'), ('b', 'c')

>>> itertools.permutations('abc')
('a', 'b', 'c'), ('a', 'c', 'b'), ('b', 'a', 'c'),...
                    

itertool gems

Period

import itertools, operator

period = 200
cycles = map(operator.not_,
             itertools.cycle(range(period)))

# iterate forever
for cycle_is_complete in cycles:
    if cycle_is_complete:
        # Do occasionally
        ...
    # Do this every time
    ...
                    

Flatten

import itertools

nested_lists = [
    [1, 2, 3],
    [12, 19, 20],
    [20, 40, 50],
]
itertools.chain.from_iterable(nested_lists)
1, 2, 3, 12, 19, 20, 20, 40, 50
                    

Segment

import itertools, string

def build_groups(size, iterable, fill=None):
    seq = itertools.repeat(iter(iterable), size)
    return itertools.zip_longest(*seq, fillvalue=fill)

build_groups(3, string.ascii_lowercase)
('a', 'b', 'c'), ('d', 'e', 'f'), ('g', 'h', 'i'),...