将项目列表随机分组，但某些项目不能在同一组中

Question

I have a list of names that I want to put into n-number of groups. I can do that easily. However, my issue is, that some of the items cannot be grouped together no matter what. Is there a simple solution to this ?

This is how I do it now:

import random

name_list = ["Item1", "Item2", "Item3", "Item4", "Item5", "Item6", "Item7"]
n = 2

random.seed(4)
random.shuffle(name_list)

test = [name_list[i:i + n] for i in xrange(0, len(name_list), n)]

print(test)

So from this, I will get 4 groups, randomly shuffled, with only 1 in the last group (that is okay). However, let's say that Item1 and Item5 can't end up together. How would I dot that ?

Answer 1

Here we get groups or "chunks" for randomly distributed items.

Given

import random
import itertools as it

import more_itertools as mit


random.seed(4)
n = 2
name_list = ["Item1", "Item2", "Item3", "Item4", "Item5", "Item6", "Item7"]
blacklist = ["Item1", "Item5"]

Code

def get_random_groups(lst, size=2, blacklist=None):
    """Return distributed items among `m` groups of lists."""
    if blacklist is None:
        blacklist = []

    # Compute number of chunks
    length = (len(lst))
    if length % 2 == 0:
        m = length//size
    else:
        m = (length+1)//size

    # Edge cases
    if len(blacklist) > m:
        raise ValueError("Black-list is too long.  Must be < number of chunks.")
    if not set(blacklist).issubset(set(lst)):
        raise ValueError("Black-list is not a subset of the main list.")

    # Resize list
    remnant_list = list(set(lst) - set(blacklist))
    random.shuffle(remnant_list)

    # Build chunks and distribute (blacklist first)
    chunks = [list(c) for c in mit.distribute(m, blacklist)]
    random.shuffle(chunks)

    for chunk in chunks[:]:
        while len(chunk) < size:
            try:
                chunk += [remnant_list.pop(-1)]
            except IndexError:
                return chunks

Demo

# Without blacklist
get_random_groups(name_list, size=n)
# Out: [['Item5', 'Item1'], ['Item3', 'Item6'], ['Item7', 'Item2'], ['Item4']]

# With blacklist
get_random_groups(name_list, size=n, blacklist=blacklist)
# [['Item2', 'Item6'], ['Item1', 'Item3'], ['Item5', 'Item7'], ['Item4']]

# Invalid blacklist
blacklist = "Item1 Item3 Item 5 Item6 Item7".split()  
get_random_groups(name_list, size=n, blacklist=blacklist)
# ValueError: Blacklist is too long.  Must be < number of chunks.

# Invalid blacklist
blacklist = "Item1 Item9".split()  
get_random_groups(name_list, size=n, blacklist=blacklist)
# ValueError("Black-list is not a subset of the main list.")

---

Details

• The number of chunks is computed based on lists of even or odd length.
• Edge cases are tested (optional).
• Items in the blacklist will be distributed among chunks first. Therefore, those items must be removed from the list of remaining elements ( remnant_list ).
• more_itertools.distribute is a third-party tool that facilitates chunking and distributing items among m number of chunks. You can optionally implement your own algorithms to accomplish these two tasks.
• A random remnant item is added to a shuffled chunk unless the size of the chunk is satisfied. This double shuffling ensure items are placed randomly.