permutations with unique values

Question

itertools.permutations generates where its elements are treated as unique based on their position, not on their value. So basically I want to avoid duplicates like this:

Answer 1

You can make a function that uses collections.Counter to get unique items and their counts from the given sequence, and uses itertools.combinations to pick combinations of indices for each unique item in each recursive call, and map the indices back to a list when all indices are picked:

from collections import Counter
from itertools import combinations
def unique_permutations(seq):
    def index_permutations(counts, index_pool):
        if not counts:
            yield {}
            return
        (item, count), *rest = counts.items()
        rest = dict(rest)
        for indices in combinations(index_pool, count):
            mapping = dict.fromkeys(indices, item)
            for others in index_permutations(rest, index_pool.difference(indices)):
                yield {**mapping, **others}
    indices = set(range(len(seq)))
    for mapping in index_permutations(Counter(seq), indices):
        yield [mapping[i] for i in indices]

so that [''.join(i) for i in unique_permutations('moon')] returns:

['moon', 'mono', 'mnoo', 'omon', 'omno', 'nmoo', 'oomn', 'onmo', 'nomo', 'oonm', 'onom', 'noom']