Python 来自pd.io.json.json\u normalize()数据帧行的原始dict/json
我使用Python 来自pd.io.json.json\u normalize()数据帧行的原始dict/json,python,json,pandas,dictionary,Python,Json,Pandas,Dictionary,我使用pd.io.json.json\u normalize()创建了一个pandas数据框,其中包含从dicts创建的行。dataframe中的值(不是键/列名)已被修改。我想从数据帧的一行中检索一个dict,其嵌套格式与原始dict相同 sample = { "A": { "a": 7 }, "B": { "a": "name", "z":{ "dD": 20 ,
pd.io.json.json\u normalize()
创建了一个pandas数据框,其中包含从dicts创建的行。dataframe中的值(不是键/列名)已被修改。我想从数据帧的一行中检索一个dict,其嵌套格式与原始dict相同
sample = {
"A": {
"a": 7
},
"B": {
"a": "name",
"z":{
"dD": 20 ,
"f_f": 3 ,
}
}
}
df = pd.io.json.json_normalize(sample, sep='__')
正如预期的那样,df.columns
返回我:
Index(['A__a', 'B__a', 'B__z__dD', 'B__z__f_f'], dtype='object')
我现在想“逆转”这个过程
我可以保证原始dict(键或值)中没有任何字符串具有作为子字符串的“\uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu