multiset-key-dict

This small library implements a mapping-like object whose keys are taken to be unordered multisets (Counters, bags), regardless of the type of the actual indexes used.

Example

Create a MultisetKeyDict and insert an element for the set {1, 3, 5}.

from multiset_key_dict import MultisetKeyDict
mskd = MultisetKeyDict()
mskd[[1, 3, 5]] = 19

Access the element just added.

# In a different order.
print(mskd[[3, 5, 1]]) # 19
# With a different type of collection.
print(mskd[{3, 5, 1}]) # 19

__iter__ iterates over items in the MultisetKeyDict, viewed as frozensets (i.e., not accounting for multiplicity).

# Add more data.
mskd[[1]] = 9
mskd[[5, 6]] = 12
print(list(mskd)) # [(frozenset({1, 3, 5}), 19), (frozenset({1}), 9), (frozenset({5, 6}), 12)]

To iterate over key-value pairs where the keys are the FrozenMultiset objects that act as the true keys in the structure, use multiset_iter or items methods instead.

mskd[[1, 1]] = 90
print(list(mskd))
# [(frozenset({1, 3, 5}), 19), (frozenset({1}), 9), (frozenset({5, 6}), 12),
# (frozenset({1}), 90)]

print(list(mskd.multiset_iter()))
# [(FrozenMultiset({(3, 1), (1, 1), (5, 1)}), 19), 
#  (FrozenMultiset({(1, 1)}), 9),
#  (FrozenMultiset({(6, 1), (5, 1)}), 12),
#  (FrozenMultiset({(1, 2)}), 90)]

To get the keys as frozenset objects, use set_keys. To get the keys as FrozenMultiset objects, use multiset_keys or keys.

It's sometimes useful to get the set of elements of the keys. For that, you can use the key_elements method.

print(mskd.key_elements()) # frozenset({1, 3, 5, 6})