Say I have these two lists:
top_levels = ["a", "b", "c"]
sub_levels = ["d", "e", "f"]
How do I create:
nested_dict = {
"a": {"d": 0, "e": 0, "f": 0},
"b": {"d": 0, "e": 0, "f": 0},
"c": {"d": 0, "e": 0, "f": 0},
}
Add a comment
|
2 Answers
You can use a comprehension:
>>> top_levels = ["a", "b", "c"]
>>> sub_levels = ["d", "e", "f"]
>>> {t: {s: 0 for s in sub_levels} for t in top_levels}
{'a': {'d': 0, 'e': 0, 'f': 0}, 'b': {'d': 0, 'e': 0, 'f': 0}, 'c': {'d': 0, 'e': 0, 'f': 0}}
5 Comments
DYZ
This answer is incorrect. It creates three references to the same dictionary. Try modifying one of the dictionaries:
d['a']['d']=10 - and see when happens to the other two. The list/dict comprehension solution is correct.
ddejohn
Good catch. Fixed.
ddejohn
Now I'm sitting here thinking "wtf is the point of
fromkeys()?" I never actually use it but it was the first thing I thought of trying when I read the OP. Now I'm actually wondering, what is its use-case?
ShadowRanger
@blorgon: It's dramatically faster for the cases where it can be used. Among other things, back before
set existed, and now, when dict can be used to make a set-like thing that maintains insertion order, it's a quick drop in replacement for the set constructor. And it's perfectly safe whenever the values aren't used, or are immutable types like int. In your code, you could safely do {t: dict.fromkeys(sub_levels, 0) for t in top_levels}, you just can't use it for the outer dict due to aliasing.ddejohn
Right, that was actually the first iteration of my solution in REPL, but at that point it seems silly to not just resort back to a pure comprehension version. I'd be curious about the performance for larger sequences but am too lazy to do any testing.
tops =['a','b','c']
subs = ['d','e','f']
sub_dict= {i:0 for i in subs}
nested = {t:sub_dict.copy() for t in tops}
print(nested)
{'a': {'d': 0, 'e': 0, 'f': 0}, 'b': {'d': 0, 'e': 0, 'f': 0}, 'c': {'d': 0, 'e': 0, 'f': 0}}
create the subs dict and make copies of it.
Comments
