如何将嵌套的 json 文件加载到 pandas DataFrame

Question

please help I cannot seem to get the json data into a Dataframe.

loaded the data data =json.load(open(r'path'))#this works fine and displays: json data

{'type': 'FeatureCollection', 'name': 'Altstadt Nord', 'crs': {'type': 'name', 'properties': {'name': 'urn:ogc:def:crs:OGC:1.3:CRS84'}}, 'features': [{'type': 'Feature', 'properties': {'Name': 'City-Martinsviertel', 'description': None}, 'geometry': {'type': 'Polygon', 'coordinates': [[[6.9595637, 50.9418396], [6.956624, 50.9417382], [6.9543173, 50.941603], [6.9529869, 50.9413664], [6.953062, 50.9408593], [6.9532873, 50.9396289], [6.9533624, 50.9388176], [6.9529333, 50.9378373], [6.9527509, 50.9371815], [6.9528367, 50.9360659], [6.9532122, 50.9352884], [6.9540705, 50.9350653], [6.9553258, 50.9350044], [6.9568815, 50.9351667], [6.9602074, 50.9355047], [6.9608189, 50.9349165], [6.9633939, 50.9348827], [6.9629433, 50.9410622], [6.9616236, 50.9412176], [6.9603898, 50.9414881], [6.9595637, 50.9418396]]]}}, {'type': 'Feature', 'properties': {'Name': 'Gereonsviertel', 'description': None}, 'geometry': {'type': 'Polygon', 'coordinates': [[[6.9629433, 50.9410622], [6.9629433, 50.9431646], [6.96 11408, 50.9433539], [6.9601752, 50.9436649], [6.9588234, 50.9443409], [6.9579651, 50.9449763], [6.9573213, 50.945801], [6.9563128, 50.9451926], [6.9551756, 50.9448546], [6.9535663, 50.9446518], [6.9523432, 50.9449763], [6.9494464, 50.9452602], [6.9473435, 50.9454495], [6.9466998, 50.9456928], [6.9458415, 50.946531], [6.9434168, 50.9453954], [6.9424726, 50.9451926], [6.9404342, 50.9429888], [6.9404771, 50.9425156], [6.9403269, 50.9415016], [6.9400479, 50.9405281], [6.9426228, 50.9399872], [6.9439103, 50.9400143], [6.9453051, 50.9404875], [6.9461634, 50.9408931], [6.9467427, 50.941096], [6.9475581, 50.9410013], [6.9504227, 50.9413191], [6.9529869, 50.9413664], [6.9547464, 50.9416368], [6.9595637, 50.9418396], [6.9603898, 50.9414881], [6.9616236, 50.9412176], [6.9629433, 50.9410622]]]}}, {'type': 'Feature', 'properties': {'Name': 'Kunibertsviertel', 'description': None}, 'geometry': {'type': 'Polygon', 'coordinates': [[[6.9629433, 50.9431646], [6.9637129, 50.9454917], [6.9651506, 50.94792 52], [6.9666097, 50.9499124], [6.9667599, 50.9500882], [6.9587777, 50.9502504], [6.9573213, 50.945801], [6.9579651, 50.9449763], [6.9588234, 50.9443409], [6.9601752, 50.9436649], [6.9611408, 50.9433539], [6.9629433, 50.9431646]]]}}, {'type': 'Feature', 'properties': {'Name': 'Nördlich Neumarkt', 'description': None}, 'geometry': {'type': 'Polygon', 'coordinates': [[[6.9390331, 50.9364418], [6.9417153, 50.9358738], [6.9462214, 50.9358062], [6.9490109, 50.9355628], [6.9505129, 50.9353329], [6.9523798, 50.9352924], [6.9532122, 50.9352884], [6.9528367, 50.9360659], [6.9527509, 50.9371815], [6.9529333, 50.9378373], [6.9533624, 50.9388176], [6.9532381, 50.9398222], [6.9529869, 50.9413664], [6.9504227, 50.9413191], [6.9475581, 50.9410013], [6.9467427, 50.941096], [6.9453051, 50.9404875], [6.9439103, 50.9400143], [6.9424663, 50.9399574], [6.9400479, 50.9405281], [6.9390331, 50.9364418]]]}}]}

now i cannot seem to fit it into a Dataframe // pd.DataFrame(data) --> ValueError: Mixing dicts with non-Series may lead to ambiguous ordering. full error

I tried to flatten with json_flatten but ModuleNotFoundError: No module named 'flatten_json' even though I installed json-flatten via pip install

also tried df =pd.DataFrame.from_dict(data,orient='index')

df Out[22]: 0 type FeatureCollection name Altstadt Nord crs {'type': 'name', 'properties': {'name': 'urn:o... features [{'type': 'Feature', 'properties': {'Name': 'C...

df Out[22]

Answer 1

The json data contains elements with different datatypes and these cannot be loaded into one single dataframe.

View datatypes in the json:

[type(data[k]) for k in data.keys()]
# Out: [str, str, dict, list]

data.keys()
# Out: dict_keys(['type', 'name', 'crs', 'features'])

You can load each single chunk of data in a separate dataframe like this:

df_crs = pd.DataFrame(data['crs'])
df_features = pd.DataFrame(data['features'])

data['type'] and data['name'] are strings

data['type']
# Out 'FeatureCollection'

data['name']
# Out 'Altstadt Nord'

Answer 2

I think you can use json_normalize to load them to pandas.

test.json in this case is your full json file (with double quotes).

import json
from pandas.io.json import json_normalize

f = open('test.json')
data = json.load(f)
df = json_normalize(data, record_path=['features'], meta=['name']) 

print(df)

This results in a dataframe as shown below.

You can further add record field in the normalize method to create more columns for the polygon coordinates.

You can find more documentation at https://pandas.pydata.org/pandas-docs/version/1.2.0/reference/api/pandas.json_normalize.html

Hope that helps.