CODEWITHSUNDEEP
pythondictionary
user15051990
user15051990

Reputation: 1895

Remove nan's from the nested dictionary in python

Remove NaN values from the nested dictionary in python. My nested dictionary looks like below:

my_dict = {'abc':{'a':0.987, 'b':0.765, 'c': numpy.nan}, 'mda':{'a':0.145, 'b':0.584, 'e':numpy.nan}, 'fop':{'a':0.145, 'b': numpy.nan, 'c':0.485}}

I want to drop Nan values from the nested dictioanry, expected result looks like:

my_dict = {'abc':{'a':0.987, 'b':0.765}, 'mda':{'a':0.145, 'b':0.584}, 'fop':{'a':0.145, 'c':0.485}}

i have written below code to remove Nan's from the nested dictionary:

def remove_nan_from_dict(my_dict):
    new_dict = {}
    original_dict = collections.defaultdict()
    from math import isnan
    for k, v in my_dict.items():
        for k1, v1 in v.items():
            print(v1)
            if not isnan(v1):
                new_dict[k1] = v1
            print(new_dict)
        original_dict[k] = new_dict
        # print(original_dict.items())
    return original_dict

But getting below dictionary:

defaultdict(None, {'abc': {'a': 0.145, 'b': 0.584, 'c': 0.485}, 'mda': {'a': 0.145, 'b': 0.584, 'c': 0.485}, 'fop': {'a': 0.145, 'b': 0.584, 'c': 0.485}})

Upvotes: 0

Views: 1275

Answers (3)

Susana Isabel Santos
Susana Isabel Santos

Reputation: 559

I had a 3 level nested-dictionary so I used the following: 

import numpy as np    
for k, v in your_dict.items():
    for kk, vv in your_dict[k].items():
        for kkk, vvv in your_dict[k][kk].items():
            get_key = your_dict[k][kk][kkk]
            print(get_key)
            for kkkk, vvvv in get_key.copy().items():
                if np.isnan(vvvv):
                    del get_key[kkkk]

Upvotes: 0

C.Nivs
C.Nivs

Reputation: 13106

You can just use a dict comprehension here and check if the value is an np.nan:


new_dict = {k: {a: b for a, b in v.items() if not np.isnan(b)} for k, v in my_dict.items()}

Upvotes: 1

Hello.World
Hello.World

Reputation: 740

You can use 2 simple loops like this:

my_dict = {'abc':{'a':0.987, 'b':0.765, 'c': numpy.nan}, 'mda':{'a':0.145, 'b':0.584, 'e':numpy.nan}, 'fop':{'a':0.145, 'b': numpy.nan, 'c':0.485}}

for items in my_dict:

    get_key = my_dict[items]

    for key, values in get_key.copy().items():
        if numpy.isnan(values):
            del get_key[key]
my_dict
> {'abc': {'a': 0.987, 'b': 0.765},
 'mda': {'a': 0.145, 'b': 0.584},
 'fop': {'a': 0.145, 'c': 0.485}}

Upvotes: 1

Related Questions