Adding column to pandas DataFrame containing list of other columns' values

Question

I have a DataFrame to which I need to add a column. The column needs to be a list of two values:

Current table:

    lat  long  other_value
0   50   50    x
1   60   50    y
2   70   50    z
3   80   50    a

Needed table:

    lat  long  other_value  new_column
0   50   50    x            [50, 50]
1   60   50    y            [60, 50]
2   70   50    z            [70, 50]
3   80   50    a            [80, 50]

I know this is super simple, but the documentation doesn't seem to cover this (at least not apparently).

Alex Riley · Accepted Answer · 2015-02-23 20:21:58Z

26

One way is to use tolist():

>>> df['new_column'] = df[['lat', 'long']].values.tolist()
>>> df
   lat  long other_value new_column
0   50    50           x   [50, 50]
1   60    50           y   [60, 50]
2   70    50           z   [70, 50]
3   80    50           a   [80, 50]

In general though, I'd be very wary of using lists in DataFrames since they're more difficult to manipulate in columns and you don't get many of the performance benefits that come with integers/floats.

edited Feb 23, 2015 at 20:21

answered Feb 23, 2015 at 20:08

Alex Riley

178k46 gold badges274 silver badges247 bronze badges

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

4 Comments

Liam Foley Over a year ago

I've wonder about lists in dataframes. Any info on why you can't check if a location in a df is an empty list ?

Eric Miller Over a year ago

I updated my problem to include an additional column. That's why .values won't work. That is, unless I can say .values and specify the columns to use. Can I do that you think?

Alex Riley Over a year ago

@LiamFoley: I'm not sure to be honest. I expect that the Pandas developers never really had lists (or other Python data structures) in mind for df values and so they're not really supported. Perhaps that will change for future releases. In the mean time, some of the str methods work alright with lists, e.g. df['new_column'].str.len() == 0 will check if a list is empty.

Eric Miller Over a year ago

Also, I believe lists are converted into numpy array objects if you need to access the column individually. (not 100% sure though)

Haleemur Ali · Accepted Answer · 2015-02-23 20:11:40Z

2

you could use zip

df['new_column'] = list(zip(df.lat, df.long))

answered Feb 23, 2015 at 20:11

Haleemur Ali

28.6k6 gold badges67 silver badges89 bronze badges

Collectives™ on Stack Overflow

Adding column to pandas DataFrame containing list of other columns' values

2 Answers 2

4 Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

4 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related