Home Append list element in pandas dataframe column depending on value of another column

Questions

Append list element in pandas dataframe column depending on value of another column

June 17, 2023

Whenever I try to append to a list given some condition, the value ends up being None instead of the list with the appended value. Example:

# sample data
data = {"bool_col": [True, False, True, True, False]}
my_df = pd.DataFrame.from_dict(data)

# instantiate column of empty lists
my_df["list_col"] = [[] for r in range(len(my_df))]

# append value to list_col when bool_col is True
my_df["list_col"] = my_df.apply(lambda x: x["list_col"].append("truth!") if x["bool_col"] else [], axis=1)

my_df

I’ve tried wrapping x["list_col"] in list() prior to calling append() to no avail. I’m not sure how to do this while retaining whatever list values may already be present and appending a new one.

>Solution :

In your case: append() appends item to list in place and returns None, which is allocated to my_df["list_col"]. You want to return the list in your apply lambda.

Try this: my_df["list_col"] = my_df.apply(lambda x: x["list_col"] + ["truth!"] if x["bool_col"] else [], axis=1)

Output:

   bool_col  list_col
0      True  [truth!]
1     False        []
2      True  [truth!]
3      True  [truth!]
4     False        []

This should work for your usecase, since if I repeat the apply twice, I get

   bool_col          list_col
0      True  [truth!, truth!]
1     False                []
2      True  [truth!, truth!]
3      True  [truth!, truth!]
4     False                []