I have the following pandas dataframe:
Col
0 []
1 []
2 [(foo, bar), (foo, bar)]
3 []
4 []
5 []
6 []
7 [(foo, bar), (foo, bar)]
I would like to remove all the empty lists (*):
Col
2 [(foo, bar), (foo, bar)]
7 [(foo, bar), (foo, bar)]
For the above I tried:
df = df.loc[df.Col != '[]']
df
and
df.pipe(lambda d: d[d['Col'] != '[]'])
However, none of them worked. So, my question is how can I remove all the empty lists from the dataframe like (*)?.
Slicing through the data frame as though the values were strings instead of lists may work:
df[df.astype(str)['Col'] != '[]']
df.Col.str., I also tried that one before posting this question.
astype to declare the series as a str in this case. .str implies you're planning to manipulate the variable after the assignment, or more simply, do something with it. It's pretty much there as an indicator for the preceding function to work with. e.g. str.replace() wouldn't work correctly if used without the column having already been made up of str values or having not been declared upon using astype.Series.astype: Cast object to input numpy.dtype Return a copy when copy = Truedf.Col.str- dtype, being the str, hasn't been 'casted' :) Hope this helps. Have a good day!You can check the length of the lists with .str accessor:
df[df.Col.str.len() != 0]
# Col
#3 [(foo, bar), (foo, bar)]
#6 [(foo, bar), (foo, bar)]
