PART 3 · ADVANCED PYTHON · CHAPTER 10

itertools, functools and sorting

FreeLesson 34 of 78

The standard library already has the building blocks most coding rounds ask you to write by hand. Know them, then be ready to write them yourself.

GROUPBY ONLY GROUPS NEIGHBOURSPYTHON 3.12

from itertools import groupby

statuses = ["delivered", "cancelled", "delivered", "delivered"]

print([(k, len(list(g))) for k, g in groupby(statuses)])
print([(k, len(list(g))) for k, g in groupby(sorted(statuses))])

Output

[('delivered', 1), ('cancelled', 1), ('delivered', 2)]
[('cancelled', 1), ('delivered', 3)]

THE FIVE YOU WILL USE MOSTPYTHON 3.12

from itertools import (batched, islice, chain,
                       accumulate, pairwise)

ids = range(101, 108)
print(list(batched(ids, 3)))            # Python 3.12+
print(list(islice(ids, 2, 5)))          # slice any iterable
print(list(chain([1, 2], (3,), "ab")))  # glue iterables
print(list(accumulate([1198, 796, 1499])))   # running total
print(list(pairwise(["01-10", "02-02", "03-05"])))

Output

[(101, 102, 103), (104, 105, 106), (107,)]
[103, 104, 105]
[1, 2, 3, 'a', 'b']
[1198, 1994, 3493]
[('01-10', '02-02'), ('02-02', '03-05')]

MIXED SORT DIRECTIONS, PARTIAL, REDUCEPYTHON 3.12

from functools import partial, reduce
from operator import itemgetter

rows = [("Delhi", 747), ("Pune", 3498),
        ("Delhi", 846), ("Pune", 1598)]

# city descending, amount ascending:
# two stable sorts, least important key first
rows.sort(key=itemgetter(1))
rows.sort(key=itemgetter(0), reverse=True)
print(rows)

parse_bin = partial(int, base=2)
print(parse_bin("1010"), reduce(lambda a, b: a * b, [1, 2, 3]))

Output

[('Pune', 1598), ('Pune', 3498), ('Delhi', 747), ('Delhi', 846)]
10 6

Tool	Does	Hand-written version asked in rounds
`collections.Counter`	Counts items	dict with get(k, 0) + 1
`heapq.nlargest(k, xs)`	Top k without a full sort	A size-k min-heap (P2)
`collections.deque(maxlen=k)`	Fixed-size sliding window	List with pop(0), which is O(n)
`functools.lru_cache`	Memoize a function	dict cache, OrderedDict for LRU (P16)
`bisect.insort`	Keep a list sorted on insert	Binary search by hand

Practice before you move on

Group the rows of orders.csv by status with itertools.groupby (sort first) and print the count per group.
Sort the orders by status ascending and by amount descending in a single sorted() call.