如果python中的值为nan/null/empty，如何从字典列表中删除键和值

Question

如果值在 python 中为 nan/null/empty，如何从字典列表中删除键和值。我试图使用下面的代码删除，但它没有被删除。 代码有什么问题吗。

如果 value 是 nan/null/empty，我们需要键值对。

代码尝试

尝试了第一种方法

for key,value in address_dic.items():
    if value is 'nan':
        del address_dic[key]
        

尝试了第二种方法

for key,value in dict(address_dic[address]).items():
    if value is 'nan':
        del address_dic[address][key]
        

在address_dic中输入数据

{'source_id': 4402350, 'address': [{'address_type': 'Primary', 'address_line_1': '11815', 'address_line_2': nan}, {'address_type': 'work', 'address_line_1': nan, 'address_line_2': 123}]}

预期输出

{'source_id': 4402350, 'address': [{'address_type': 'Primary', 'address_line_1': '11815'}, {'address_type': 'work', 'address_line_2': 123}]}

Answer 1

在迭代字典或列表的值时，您不应修改它。创建一个新的地址列表，其中包含您需要的值并用它替换现有列表。

输入数据

address_dic = {
    "source_id": 4402350,
    "address": [
        {
            "address_type": "Primary",
            "address_line_1": "11815",
            "address_line_2": float("nan")
        },
        {
            "address_type": "work",
            "address_line_1": float("nan"),
            "address_line_2": 123
        }
    ]
}

主逻辑

new_address = []

for addr in address_dic["address"]:
    # iterate through the dictionary and add the key-value pair
    # into a new dictionary if the value is not nan
    new_addr = {key: val for key, val in addr.items() if val == val} 
    new_address.append(new_addr)

address_dic["address"] = new_address

输出

{
    "source_id": 4402350,
    "address": [
        {
            "address_type": "Primary",
            "address_line_1": "11815",
        },
        {
            "address_type": "work",
            "address_line_2": 123
        }
    ]
}

单线

address_dic["address"] = [{key: val for key, val in addr.items() if val == val} for addr in address_dic["address"]]

---

您可以通过多种方式检查值是否为 nan：

val = float("nan") # setting val as nan

熊猫

import pandas as pd
pd.isna(val) # returns True if val is nan

Numpy

import numpy as np
np.isnan(val) # returns True if val is nan

没有外部库

val == val # returns False if val is nan

Answer 2

喜欢下面的

data = {'source_id': 4402350,
        'address': [{'address_type': 'Primary', 'address_line_1': '11815', 'address_line_2': 'nan'},
                    {'address_type': 'work', 'address_line_1': 'nan', 'address_line_2': 123}]}

for entry in data['address']:
    for k in list(entry.keys()):
        if entry[k] == 'nan':
            del entry[k]
print(data)

输出

{'source_id': 4402350, 'address': [{'address_type': 'Primary', 'address_line_1': '11815'}, {'address_type': 'work', 'address_line_2': 123}]}

Answer 3

for addr in address_dic[address]:
    new_addr = {key: val for key, val in addr.items() if val == val}
    new_address.append(new_addr)
print(new_address)
address_dic["address"] = new_address

Answer 4

这个问题有两个问题。首先，您的dict 中有一些values 是字典列表的问题。其次，您不能在循环时更改字典，因此您需要制作副本。

试试：

d = {'source_id': 4402350, 'address': [{'address_type': 'Primary', 'address_line_1': '11815', 'address_line_2': 'nan'}, {'address_type': 'work', 'address_line_1': 'nan', 'address_line_2': 123}]}



for k,v in d.copy().items():
    if type(v) == list:
        for i, dd in enumerate(v):
            for k1,v1 in dd.copy().items():
                if v1 == 'nan':
                    dd.pop(k1)
                    v[i] = dd
    else:
        if v == 'nan':
            d.pop(k)

print(d)