Get particular Nested level Items from Dictionary – Python

Our task is to extract all items at a particular depth level from a given nested dictionary,. This means retrieving only the key-value pairs that exist at a specific depth within the dictionary. For example: d = {‘a’: {‘b’: {‘c’: 5, ‘d’: 6}, ‘e’: 7}, ‘f’: 8} and if we want all key-value pairs at level 2 then the output should be: {‘b’: {‘c’: 5, ‘d’: 6}, ‘e’: 7}.

Using reduce() from functool

In this method we use reduce() to traverse the dictionary level by level extracting key-value pairs at the required depth.

from functools import reduce

d = {'a': {'b': {'c': 5, 'd': 6}, 'e': 7}, 'f': 8}

lvl = 2
res = reduce(lambda d, _: {k: v for v in d.values() if isinstance(v, dict) for k, v in v.items()}, range(lvl - 1), d)

print(res)  

{'b': {'c': 5, 'd': 6}, 'e': 7}

Explanation:

• reduce() iterates lvl – 1 times and this ensures we extract key-value pairs at the correct level without going too deep.
• d.values() selects only nested dictionaries: we ignore non-dictionary values and focus on deeper levels.
• Dictionary comprehension {k: v for k, v in v.items()} this flattens the nested dictionaries while preserving structure.

Using update() with Iteration

We use a stack to traverse the dictionary iteratively and extract key-value pairs at the required depth.

data = {'a': {'b': {'c': 5, 'd': 6}, 'e': 7}, 'f': 8}

lvl, res = 2, {}
stk = [(data, 0)]

while stk:
    d, depth = stk.pop()
    if depth == lvl - 1:
        res.update({k: v for k, v in d.items() if isinstance(v, dict)})
    for v in d.values():
        if isinstance(v, dict):
            stk.append((v, depth + 1))

print(res) 

{'b': {'c': 5, 'd': 6}}

Explanation:

• stk stores dictionaries along with their depth.
• If depth == lvl – 1 then we use update() to add key-value pairs where values are dictionaries.
• We traverse values adding nested dictionaries to stk with increased depth.

manjeet_04

Follow

Improve

Article Tags :

• Python
• Python Programs
• Python dictionary-programs

Practice Tags :

• python