itertools and random

Why this matters

Two more useful corners of the standard library: itertools builds combinations without nested loops, and random adds chance to games, simulations and sampling. Both have a trick worth knowing.

The idea

from itertools import combinations, permutations, product, chain

list(combinations("abc", 2))      # [('a','b'), ('a','c'), ('b','c')]  order ignored
list(permutations("ab"))          # [('a','b'), ('b','a')]             order matters
list(product("ab", [1, 2]))       # [('a',1), ('a',2), ('b',1), ('b',2)]
list(chain([1, 2], [3]))          # [1, 2, 3]

These return iterators (lazy sequences), so wrap them in list(...) to see everything.

import random

rng = random.Random(42)           # your own generator, with a seed
rng.randint(1, 6)                 # a whole number from 1 to 6, both included
rng.choice(["a", "b", "c"])       # one item
rng.shuffle(some_list)            # reorders the list IN PLACE

The same seed always gives the same sequence. That makes code that uses chance testable: take the generator as a parameter instead of calling random directly. For passwords and tokens use the secrets module instead; random is not secure.

Try it

Print list(combinations([1, 2, 3, 4], 3)). Count the results, then use permutations and compare. Run random.Random(1).randint(1, 6) twice. Why is the answer the same?

Your turn

Write five functions:

• all_pairs(items): every pair of different items as a list of tuples, without repeats in a different order: all_pairs(["a", "b", "c"]) is [("a", "b"), ("a", "c"), ("b", "c")].
• menu_combos(mains, sides): every main with every side as text like "soup + bread", grouped by main.
• roll_dice(count, rng): a list of count dice rolls (1 to 6) using the generator rng.
• shuffled_copy(items, rng): a new shuffled list. The original must not change.
• pick_winner(names, rng): one random name from the list.

Recap

• combinations, permutations and product replace nested loops; wrap results in list().
• A seeded random.Random(seed) repeats exactly; pass it in so your code can be tested.
• shuffle works in place and returns None.

Common mistakes

• shuffled = random.shuffle(items), which sets shuffled to None.
• randint(1, 6) includes 6, but range(1, 6) stops at 5.
• Reading an iterator twice: after one pass over combinations(...) it is empty.