问题
I have a pandas DataFrame where one column contains lists and would like to select the rows where the lists are not empty.
Example data:
df = pd.DataFrame({'letter': ["a", "b", "c", "d", "e"],
'my_list':[[0,1,2],[1,2],[],[],[0,1]]})
df
letter my_list
0 a [0, 1, 2]
1 b [1, 2]
2 c []
3 d []
4 e [0, 1]
What I'd like:
df
letter my_list
0 a [0, 1, 2]
1 b [1, 2]
4 e [0, 1]
What I'm trying:
df[df.my_list.map(lambda x: if len(x) !=0)]
... which returns an invalid syntax error. Any suggestions?
回答1:
Empty lists evaluate to False in a boolean context
df[df.my_list.astype(bool)]
letter my_list
0 a [0, 1, 2]
1 b [1, 2]
4 e [0, 1]
回答2:
Or you can follow your own logic using length of the list to determine keep or not .
df[df.my_list.str.len().gt(0)]
Out[1707]:
letter my_list
0 a [0, 1, 2]
1 b [1, 2]
4 e [0, 1]
来源：https://stackoverflow.com/questions/49700794/selecting-rows-of-pandas-dataframe-where-column-is-not-empty-list