how to remove index column from multi index dataframe

Question

I've seen a lot of answers on flattening multi-level index. I just want to remove the very first index column on the left of this multi-index dataframe.

From this:

Attributes       Date    Close    Close               
Symbols                   AMZN      ^DJI  
0          2020-12-01  3220.08  29823.92  
1          2020-11-30  3168.04  29638.64  
2          2020-11-27  3195.34  29910.37  
3          2020-11-25  3185.07  29872.47

I'm looking for this outcome:

       Date    Close    Close               
               AMZN      ^DJI  
2020-12-01  3220.08  29823.92  
2020-11-30  3168.04  29638.64  
2020-11-27  3195.34  29910.37  
2020-11-25  3185.07  29872.47

Is this possible?

if I export to CSV -- df.to_csv("my_data.csv", index=False) -- I'm able to drop the index, but I'm not sure how to do it without exporting to csv. — Adam Schroeder
– Adam Schroeder, Commented Dec 2, 2020 at 6:06
index is fundamental to a dataframe and will always be there. However, you can do, df = df.reset_index(drop=True) , but an index will still be there of 0,1,2,3,4, etc. — David Erickson
– David Erickson, Commented Dec 2, 2020 at 6:07
I thought so too, but your code is not dropping that attirbutes/symbols part. — Adam Schroeder
– Adam Schroeder, Commented Dec 2, 2020 at 6:09
This is what I get when I print(df.columns) MultiIndex([( 'Date', ''), ( 'Close', 'AMZN'), ( 'Close', '^DJI'), ( 'High', 'AMZN'), ( 'High', '^DJI'), ( 'Low', 'AMZN'), ( 'Low', '^DJI'), ( 'Open', 'AMZN'), ( 'Open', '^DJI'), ('Volume', 'AMZN'), ('Volume', '^DJI')], names=['Attributes', 'Symbols']) — Adam Schroeder
– Adam Schroeder, Commented Dec 2, 2020 at 6:10
You can rename it to '' (blank) if that is what you mean: pandas.pydata.org/docs/reference/api/pandas.Index.rename.html — David Erickson
– David Erickson, Commented Dec 2, 2020 at 6:10

jezrael · Accepted Answer · 2020-12-02 06:17:33Z

2

If need set first MultiIndex column to index use DataFrame.set_index with rename index and columns names by DataFrame.rename_axis - so output is DataFrame with DatetimeIndex:

df = df.set_index([('Date', '')]).rename_axis(index=None, columns=('Date',''))
print (df)
Date          Close          
               AMZN      ^DJI
2020-12-01  3220.08  29823.92
2020-11-30  3168.04  29638.64
2020-11-27  3195.34  29910.37
2020-11-25  3185.07  29872.47

edited Dec 2, 2020 at 6:17

answered Dec 2, 2020 at 6:13

jezrael

868k103 gold badges1.4k silver badges1.3k bronze badges

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

4 Comments

Adam Schroeder Over a year ago

Hi @jezrael, I tried that but I get: raise TypeError("Must pass list-like as names.") TypeError: Must pass list-like as names.

jezrael Over a year ago

@AdamSchroeder - My typo, sorry, need columns=('Date','')

ira Over a year ago

@jezrael how about removing the AMZN without altering the structure?

jezrael Over a year ago

@ira - Remove not possible, only rename like df = df.rename(columns={'AMZN':''})

Flair · Accepted Answer · 2022-01-09 11:53:35Z

0

If I've understood correctly, this is easiest way to do it:

df.reset_index().drop(columns="Attributes", axis = 1)

edited Jan 9, 2022 at 11:53

Flair

2,9572 gold badges33 silver badges45 bronze badges

answered Jan 4, 2022 at 19:59

Andrew Little

333 bronze badges

Collectives™ on Stack Overflow

how to remove index column from multi index dataframe

2 Answers 2

4 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

4 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related