Python: dataframe.get() with multi-index

Question

I have a dataframe df structured as follows:

  from       rate   to
0  CHF   1.000000  CHF
1  CHF  19.673256  MXN
2  CHF   0.000000  ZAR
3  CHF   0.000775  XAU
4  CHF  32.961405  THB

and I've re-indexed it to be as follows:

               rate
from to            
CHF  CHF   1.000000
     MXN  19.675255
     ZAR   0.000000
     XAU   0.000775
     THB  32.961068
...

using df = df.set_index(['from','to'],drop=True, append=False).

I want to simply call get on the dataframe using a pair of indices. For example, I would expect df[['CHF','MXN']] to return 19.675255 but instead an error is raised stating "['CHF' 'MXN'] not in index".

It seems like this should be very straightforward -- what am I doing wrong?

Thanks

cs95 · Accepted Answer · 2018-01-08 22:04:55Z

1

Option 1
pd.IndexSlice

idx = pd.IndexSlice

df.loc[idx['CHF', 'MXN']]

rate    19.673256
Name: (CHF, MXN), dtype: float64

If you want the result as a scalar, call .item -

df.loc[idx['CHF', 'MXN']].item()
19.673256

Option 2
xs

df.xs(['CHF', 'MXN'])

rate    19.673256
Name: (CHF, MXN), dtype: float64

answered Jan 8, 2018 at 22:04

cs95

406k106 gold badges744 silver badges797 bronze badges

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

5 Comments

Blake Over a year ago

Awesome thank you. This works. Off-hand, do you know which option should run fastest, or are they the same?

cs95 Over a year ago

@Blake You're operating on an index, so both methods would probably be equally fast. It would depend on the individual overhead each function entails. My guess is loc would be slightly faster.

BENY Over a year ago

Let me asking a question :-)

cs95 Over a year ago

@Wen I have one for you to answer: stackoverflow.com/questions/48122394/…

BENY Over a year ago

@cᴏʟᴅsᴘᴇᴇᴅ added an answer to that question :-)

Collectives™ on Stack Overflow

Python: dataframe.get() with multi-index

1 Answer 1

5 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related