python-Pandas df.sum() across multiple columns unexpected arg 'axis' error

Question

here I have a pandas Dataframe df like:

     A    B    C
 0   1    2    3
 1   1    2    3
 3   1    2    3

Then I have a list

x=['B','C']

I want to get the sum of each number of each row under B & C columns. So I write like:

df[x].sum(axis=1).values

However, I get an error

TypeError: f() got an unexpected keyword argument 'axis'

I don't understand why this will give an error. My code is run in ipython notebook. Could you give any suggestions? Thanks.

UPDATE: the real df is like:

                  Date      Ayotte  Hassan
Date                                 
2016-06-29     2016-06-29    46.8    45.3
2016-06-30     2016-06-30    46.8    45.3
2016-07-01     2016-07-01    46.8    45.3
2016-07-02     2016-07-02    46.8    45.3
2016-07-03     2016-07-03    46.8    45.3
2016-07-04     2016-07-04    46.8    45.3
2016-07-20     2016-07-20    45.8    45.2
2016-07-21     2016-07-21    45.8    45.2
2016-07-22     2016-07-22    45.8    45.2
   ...            ...         ...     ...
2016-10-09     2016-10-09    48.0    44.5
2016-10-10     2016-10-10    48.0    44.5
2016-10-11     2016-10-11    46.7    44.7
2016-10-16     2016-10-16    46.3    44.0
2016-10-17     2016-10-17    46.3    44.0
2016-10-18     2016-10-18    46.0    44.3
2016-10-19     2016-10-19    45.7    45.3
2016-10-20     2016-10-20    44.0    46.0
2016-10-21     2016-10-21    44.0    46.0
2016-10-22     2016-10-22    44.0    46.0
2016-10-23     2016-10-23    44.0    46.0

The dtypes of df is

Date      datetime64[ns]
Ayotte           float64
Hassan           float64
dtype: object

Then, I did

df = df.resample('D')

The df shown above is the data before resample. The list x is

x=['Ayotte','Hassan']

Then the error comes when I run this code

print df[x].sum(axis=1).values

Did you mean to have your index go 0, 1, 3 in your DataFrame? Because I did the exact same thing (with the proper indicies) and it works. — elPastor
– elPastor, Commented Nov 18, 2016 at 2:08
.values is the issue Edit, I take that back....It works either way for me — anshanno
– anshanno, Commented Nov 18, 2016 at 2:15
@anshanno I tried without .values, but still get the same error. — Mr_Pi
– Mr_Pi, Commented Nov 18, 2016 at 2:19

Paul H · Accepted Answer · 2016-11-18 02:09:19Z

1

It's hard to diagnose w/o a continuous example that demonstrate the error.

If I start out with:

import numpy
import pandas
df = pandas.DataFrame(
    numpy.arange(9).reshape(3, 3),
    index=['a', 'b', 'c'],
    columns=['X', 'Y', 'Z']
)
print(df)

Which gives:

I can then do:

df[['X', 'Y']].sum(axis=1)

To get:

a     1
b     7
c    13
dtype: int64

answered Nov 18, 2016 at 2:09

Paul H

68.7k23 gold badges165 silver badges139 bronze badges

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

1 Comment

Mr_Pi Over a year ago

Thanks for reply. I also try something myself and there is no error. My actual dataframe df has three columns: date, person1, person2. The index of df is also date. The date is like 2016-11-18. Under person1/2, there are float64 numbers. when I try to calculate the sum as above, then I will get the error.

anshanno · Accepted Answer · 2016-11-18 02:49:35Z

0

In:

,Date,Ayotte,Hassan
2016-06-29,2016-06-29,46.8,45.3
2016-06-30,2016-06-30,46.8,45.3
2016-07-01,2016-07-01,46.8,45.3
2016-07-02,2016-07-02,46.8,45.3
2016-07-03,2016-07-03,46.8,45.3
2016-07-04,2016-07-04,46.8,45.3
2016-07-20,2016-07-20,45.8,45.2

Code:

import pandas as pd

df = pd.read_csv('file.txt', index_col=0)
df['Total'] = df[['Ayotte', 'Hassan']].sum(axis=1)

print df

Out:

                  Date  Ayotte  Hassan  Total
2016-06-29  2016-06-29    46.8    45.3   92.1
2016-06-30  2016-06-30    46.8    45.3   92.1
2016-07-01  2016-07-01    46.8    45.3   92.1
2016-07-02  2016-07-02    46.8    45.3   92.1
2016-07-03  2016-07-03    46.8    45.3   92.1
2016-07-04  2016-07-04    46.8    45.3   92.1
2016-07-20  2016-07-20    45.8    45.2   91.0

answered Nov 18, 2016 at 2:49

anshanno

3544 silver badges21 bronze badges

3 Comments

Mr_Pi Over a year ago

Thanks. I tried this but still get the same error. However, I found if I remove the line df = df.resample('D'), then it will work. Do you have any idea about that?

anshanno Over a year ago

@Mr_Pi not sure exactly, I have never used resample. Are you using that resample just to get the date into Year-Month-Day? From the docs it looks like you are supposed to call it on a per series basis series.resample(). But if all you want is the date in Y-m-d you should be using something like df['in_day'] = df.my_datetime.dt.strftime('%Y-%m-%d')

Paul H Over a year ago

resample doesn't return a dataframe. it returns a grouping object, you need to aggregate the resample object before treating it like a normal dataframe.

Collectives™ on Stack Overflow

python-Pandas df.sum() across multiple columns unexpected arg 'axis' error

2 Answers 2

1 Comment

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related