我有一个JSON，它通过http://api.worldweatheronline.com/的有用API发出请求。
我正在努力将JSON转换为表格格式，例如pandas数据框。我认为问题是由于JSON的嵌套结构造成的。
我尝试过System.DllNotFoundException: 'Unable to load DLL 'mydll.dll': The specified module could not be found. (Exception from HRESULT: 0x8007007E)'
，但是由于“天气每小时”跨越多行，而初始行跨越了单行，因此这种格式无法正确处理嵌套结构，因此无法正确格式化。
我还尝试了导出为JSON并以pd.DataFrame(json)的形式读取，但是这也遇到了类似的问题。
非常感谢您的帮助！
JSON如下：
pd.read_json我想要的示例csv：
{'data': {'request': [{'type': 'UK Postcode', 'query': 'E4'}],
'weather': [{'date': '2018-11-28',
'astronomy': [{'sunrise': '07:39 AM',
'sunset': '03:57 PM',
'moonrise': '09:46 PM',
'moonset': '12:21 PM',
'moon_phase': 'Last Quarter',
'moon_illumination': '69'}],
'maxtempC': '13',
'maxtempF': '56',
'mintempC': '10',
'mintempF': '51',
'totalSnow_cm': '0.0',
'sunHour': '3.1',
'uvIndex': '0',
'hourly': [{'time': '0',
'tempC': '9',
'tempF': '48',
'windspeedMiles': '4',
'windspeedKmph': '7',
'winddirDegree': '212',
'winddir16Point': 'SSW',
'weatherCode': '296',
'weatherIconUrl': [{'value': 'http://cdn.worldweatheronline.net/images/wsymbols01_png_64/wsymbol_0033_cloudy_with_light_rain_night.png'}],
'weatherDesc': [{'value': 'Light rain'}],
'precipMM': '1.6',
'humidity': '93',
'visibility': '11',
'pressure': '1010',
'cloudcover': '100',
'HeatIndexC': '9',
'HeatIndexF': '48',
'DewPointC': '8',
'DewPointF': '46',
'WindChillC': '8',
'WindChillF': '46',
'WindGustMiles': '8',
'WindGustKmph': '12',
'FeelsLikeC': '8',
'FeelsLikeF': '46'},
{'time': '300',
'tempC': '9',
'tempF': '48',
'windspeedMiles': '7',
'windspeedKmph': '11',
'winddirDegree': '174',
'winddir16Point': 'S',
'weatherCode': '266',
'weatherIconUrl': [{'value': 'http://cdn.worldweatheronline.net/images/wsymbols01_png_64/wsymbol_0033_cloudy_with_light_rain_night.png'}],
'weatherDesc': [{'value': 'Light drizzle'}],
'precipMM': '0.5',
'humidity': '92',
'visibility': '15',
'pressure': '1009',
'cloudcover': '100',
'HeatIndexC': '9',
'HeatIndexF': '48',
'DewPointC': '8',
'DewPointF': '46',
'WindChillC': '7',
'WindChillF': '45',
'WindGustMiles': '11',
'WindGustKmph': '17',
'FeelsLikeC': '7',
'FeelsLikeF': '45'},
{'time': '600',
'tempC': '11',
'tempF': '51',
'windspeedMiles': '11',
'windspeedKmph': '18',
'winddirDegree': '175',
'winddir16Point': 'S',
'weatherCode': '266',
'weatherIconUrl': [{'value': 'http://cdn.worldweatheronline.net/images/wsymbols01_png_64/wsymbol_0033_cloudy_with_light_rain_night.png'}],
'weatherDesc': [{'value': 'Light drizzle'}],
'precipMM': '0.7',
'humidity': '95',
'visibility': '14',
'pressure': '1008',
'cloudcover': '100',
'HeatIndexC': '11',
'HeatIndexF': '51',
'DewPointC': '10',
'DewPointF': '50',
'WindChillC': '8',
'WindChillF': '47',
'WindGustMiles': '19',
'WindGustKmph': '31',
'FeelsLikeC': '8',
'FeelsLikeF': '47'},
{'time': '900',
'tempC': '12',
'tempF': '54',
'windspeedMiles': '12',
'windspeedKmph': '19',
'winddirDegree': '208',
'winddir16Point': 'SSW',
'weatherCode': '296',
'weatherIconUrl': [{'value': 'http://cdn.worldweatheronline.net/images/wsymbols01_png_64/wsymbol_0017_cloudy_with_light_rain.png'}],
'weatherDesc': [{'value': 'Light rain'}],
'precipMM': '0.8',
'humidity': '93',
'visibility': '14',
'pressure': '1008',
'cloudcover': '100',
'HeatIndexC': '12',
'HeatIndexF': '54',
'DewPointC': '11',
'DewPointF': '52',
'WindChillC': '10',
'WindChillF': '51',
'WindGustMiles': '20',
'WindGustKmph': '32',
'FeelsLikeC': '10',
'FeelsLikeF': '51'},
{'time': '1200',
'tempC': '13',
'tempF': '56',
'windspeedMiles': '16',
'windspeedKmph': '26',
'winddirDegree': '209',
'winddir16Point': 'SSW',
'weatherCode': '266',
'weatherIconUrl': [{'value': 'http://cdn.worldweatheronline.net/images/wsymbols01_png_64/wsymbol_0017_cloudy_with_light_rain.png'}],
'weatherDesc': [{'value': 'Light drizzle'}],
'precipMM': '0.2',
'humidity': '87',
'visibility': '15',
'pressure': '1008',
'cloudcover': '100',
'HeatIndexC': '13',
'HeatIndexF': '56',
'DewPointC': '11',
'DewPointF': '52',
'WindChillC': '11',
'WindChillF': '52',
'WindGustMiles': '25',
'WindGustKmph': '41',
'FeelsLikeC': '11',
'FeelsLikeF': '52'},
{'time': '1500',
'tempC': '13',
'tempF': '56',
'windspeedMiles': '19',
'windspeedKmph': '31',
'winddirDegree': '205',
'winddir16Point': 'SSW',
'weatherCode': '266',
'weatherIconUrl': [{'value': 'http://cdn.worldweatheronline.net/images/wsymbols01_png_64/wsymbol_0017_cloudy_with_light_rain.png'}],
'weatherDesc': [{'value': 'Light drizzle'}],
'precipMM': '0.5',
'humidity': '84',
'visibility': '15',
'pressure': '1007',
'cloudcover': '100',
'HeatIndexC': '13',
'HeatIndexF': '56',
'DewPointC': '11',
'DewPointF': '51',
'WindChillC': '11',
'WindChillF': '52',
'WindGustMiles': '30',
'WindGustKmph': '48',
'FeelsLikeC': '11',
'FeelsLikeF': '52'},
{'time': '1800',
'tempC': '10',
'tempF': '51',
'windspeedMiles': '19',
'windspeedKmph': '31',
'winddirDegree': '216',
'winddir16Point': 'SW',
'weatherCode': '122',
'weatherIconUrl': [{'value': 'http://cdn.worldweatheronline.net/images/wsymbols01_png_64/wsymbol_0004_black_low_cloud.png'}],
'weatherDesc': [{'value': 'Overcast'}],
'precipMM': '0.0',
'humidity': '82',
'visibility': '15',
'pressure': '1007',
'cloudcover': '100',
'HeatIndexC': '12',
'HeatIndexF': '54',
'DewPointC': '11',
'DewPointF': '51',
'WindChillC': '10',
'WindChillF': '51',
'WindGustMiles': '30',
'WindGustKmph': '49',
'FeelsLikeC': '10',
'FeelsLikeF': '51'},
{'time': '2100',
'tempC': '6',
'tempF': '44',
'windspeedMiles': '15',
'windspeedKmph': '23',
'winddirDegree': '219',
'winddir16Point': 'SW',
'weatherCode': '353',
'weatherIconUrl': [{'value': 'http://cdn.worldweatheronline.net/images/wsymbols01_png_64/wsymbol_0025_light_rain_showers_night.png'}],
'weatherDesc': [{'value': 'Light rain shower'}],
'precipMM': '0.5',
'humidity': '82',
'visibility': '15',
'pressure': '1008',
'cloudcover': '70',
'HeatIndexC': '8',
'HeatIndexF': '47',
'DewPointC': '10',
'DewPointF': '50',
'WindChillC': '6',
'WindChillF': '44',
'WindGustMiles': '24',
'WindGustKmph': '39',
'FeelsLikeC': '6',
'FeelsLikeF': '44'}]}]}}
答案 0 :(得分：1)
我将您的JSON保存到了一个文件（顺便说一下，“ test.json”，其中的撇号需要换成逗号，以便json模块可以解析它）并读取它json模块。您想以pandas DataFrame结尾，我们将需要defaultdicts来使扁平化变得容易：
import json
import pandas as pd
from collections import defaultdict
由于您要使用的嵌套结构在每个新上下文中都会看起来有所不同，因此首先需要一个函数来展平由多个dict和list组成的常规嵌套对象，直到您使用得到浅dict：我的建议：
def flatten_json(json):
flattened_dict = defaultdict(list)
def flatten(structure):
global key
if type(structure) is dict:
for key in structure:
flatten(structure[key])
elif type(structure) is list:
for item in structure:
flatten(item)
else:
flattened_dict[key].append(structure)
flatten(json)
return flattened_dict
快到了。我们唯一需要解决的问题是，如果所有dict值都具有相同的长度，则pandas只能使用字典构建Dataframe。因此，我们需要一个函数来将较短的值填充到最长的长度：
def fill_up_dict(dicti):
max_length = max([len(liist) for liist in dicti.values()])
for value in dicti.values():
remaining_length = max_length - len(value)
value.extend([""]*remaining_length)
return dicti
完成。只需读取您的JSON，将其展平，填充并转储到Dataframe中即可：
with open('test.json', 'r') as f:
json_data = json.load(f)
flat_dict = flatten_json(json_data)
homogeneous_dict = fill_up_dict(flat_dict)
df = pd.DataFrame(homogeneous_dict)
出站df的外观如下：
DewPointC DewPointF FeelsLikeC FeelsLikeF HeatIndexC HeatIndexF WindChillC
0 8 46 8 46 9 48 8
1 8 46 7 45 9 48 7
2 10 50 8 47 11 51 8
3 11 52 10 51 12 54 10
4 11 52 11 52 13 56 11
5 11 51 11 52 13 56 11
6 11 51 10 51 12 54 10
7 10 50 6 44 8 47 6
8
9
10
11
12
13
14
15
WindChillF WindGustKmph WindGustMiles ... totalSnow_cm \
0 46 12 8 ... 0.0
1 45 17 11 ...
2 47 31 19 ...
3 51 32 20 ...
4 52 41 25 ...
5 52 48 30 ...
6 51 49 30 ...
7 44 39 24 ...
8 ...
9 ...
10 ...
11 ...
12 ...
13 ...
14 ...
15 ...
type uvIndex value \
0 UK Postcode 0 http://cdn.worldweatheronline.net/images/wsymb...
1 Light rain
2 http://cdn.worldweatheronline.net/images/wsymb...
3 Light drizzle
4 http://cdn.worldweatheronline.net/images/wsymb...
5 Light drizzle
6 http://cdn.worldweatheronline.net/images/wsymb...
7 Light rain
8 http://cdn.worldweatheronline.net/images/wsymb...
9 Light drizzle
10 http://cdn.worldweatheronline.net/images/wsymb...
11 Light drizzle
12 http://cdn.worldweatheronline.net/images/wsymb...
13 Overcast
14 http://cdn.worldweatheronline.net/images/wsymb...
15 Light rain shower
visibility weatherCode winddir16Point winddirDegree windspeedKmph \
0 11 296 SSW 212 7
1 15 266 S 174 11
2 14 266 S 175 18
3 14 296 SSW 208 19
4 15 266 SSW 209 26
5 15 266 SSW 205 31
6 15 122 SW 216 31
7 15 353 SW 219 23
8
9
10
11
12
13
14
15
windspeedMiles
0 4
1 7
2 11
3 12
4 16
5 19
6 19
7 15
8
9
10
11
12
13
14
15
[16 rows x 40 columns]
祝您的项目好运！
D。