DataFrame访问字典中列表中的嵌套键

Question

The following data is part of a much larger dataframe with a lot of nested keys. Say I want to access "humidity" or "windSpeed" How would I do that?

df = pd.DataFrame({"data":[{"time":1422802800,"humidity":0.62,"windSpeed":2.62}]})

The purpose is to select only certain keys and append them to a CSV file, rather then appending the entire dataframe to the CSV file.

Answer 1

You'd need to use apply with a lambda and index into the dict:

In[69]:
df['data'].apply(lambda x: x['time'])

Out[69]: 
0    1422802800
Name: data, dtype: int64

and like-wise for humidity:

In[71]:
df['data'].apply(lambda x: x['humidity'])

Out[71]: 
0    0.62
Name: data, dtype: float64

I'd advise against storing non-scalar values in a df, it's non-performant as you lose any vectorised advantages of using dataframes