Append element to column with numpy.ndarray datatype in Pandas

Question

I want to append a string to a column which is a numpy.ndarray object.The following code is not working:

def filter_by_player(df, players, team):
    filtered_df = df[df['player'].isin(players)]
    filtered_df['league'] = filtered_df['league'].apply(lambda x: x + [team])
    return filtered_df

the league column looks like this ['barca','real','sevilla']. I want to add to it but the code above is not working.

players = ['messi', 'benzema', 'busquets']

league_df
player      | team
messi       | ['barca']
lewandowski | ['dortmund', 'bayern', 'barca']

when i call the function filter_by_player(players, 'psg')

the new dataframe league_df should become this as messi is in the list of players:

player      | team
messi       | ['barca', 'psg']

What do you want to be in the league column after filtering? — The Photon
– The Photon, Commented Jan 11, 2023 at 0:47
Can you provide some sample data and the desired output for that sample data? — BrokenBenchmark
– BrokenBenchmark, Commented Jan 11, 2023 at 0:56
so you want to append the psg to the list of players tohat you pass as parameter? and then filter the data by that list as well? Its unclear if you are performing 2 tasks (addition of team name and filtering ) or you are just filtering. — Akshay Sehgal
– Akshay Sehgal, Commented Jan 11, 2023 at 1:13
the filtering is straight forward. it is the appending I am having issues with. league column is an numpy.ndarray object and so appending is not straight forward — user872009
– user872009, Commented Jan 11, 2023 at 1:21

Akshay Sehgal · Accepted Answer · 2023-01-11 01:41:15Z

1

IIUC, each of the elements in the team or league column is a numpy array. So, use np.append instead of appending lists. Try this -

import numpy as np

def filter_by_player(df, players, team):
    filtered_df = df[df['player'].isin(players)]
    filtered_df['team'] = filtered_df['team'].apply(np.append, args=([team]))
    return filtered_df

new_df = filter_by_player(league_df, ['messi'], 'psg')
print(new_df)

  player          team
0  messi  [barca, psg]

Just to double check if the initial datatype and post datatypes are numpy -

print(type(league_df.iloc[0,1]))
print(type(new_df.iloc[0,1]))

<class 'numpy.ndarray'>
<class 'numpy.ndarray'>

EDIT: If you arent not strict about keeping it as a numpy array, try this -

def filter_by_player(df, players, team):
    filtered_df = df[df['player'].isin(players)]
    filtered_df['league'] = filtered_df['league'].apply(lambda x: list(x)+[team])
    return filtered_df

new_df = filter_by_player(league_df, ['messi'], 'psg')
print(new_df)
print(type(league_df.iloc[0,1]))
print(type(new_df.iloc[0,1]))

  player        league
0  messi  [barca, psg]

<class 'numpy.ndarray'>
<class 'list'>

edited Jan 11, 2023 at 1:41

answered Jan 11, 2023 at 1:22

Akshay Sehgal

19.4k3 gold badges26 silver badges57 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

user872009 Over a year ago

This works! More elegant than mine! Thank you!!

Akshay Sehgal Over a year ago

have also added the list approach (modified your approach), since you mentioned you are not too strict about keeping it as numpy. Glad to help!

Collectives™ on Stack Overflow

Append element to column with numpy.ndarray datatype in Pandas

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related