CODEWITHSUNDEEP
pandas
kjo
kjo

Reputation: 35331

"Reindexing only valid with uniquely valued Index objects"

When I run code like this:

import pandas as pd

A = pd.DataFrame([('a', -1.374201, 35),
                  ('b', 1.415697, 29),
                  ('a', 0.233841, 18),
                  ('b', 1.550599, 30),
                  ('a', -0.178370, 63),
                  ('b', -1.235956, 42),
                  ('a', 0.088046, 2),
                  ('b', 0.074238, 84)], columns='key value other'.split())

B = A.groupby('key')['value'].mean()

C = pd.DataFrame([('a', 0.469924, 44),
                  ('b', 1.231064, 68),
                  ('a', -0.979462, 73),
                  ('b', 0.322454, 97)], columns='key value other'.split())

D = C.set_index('key')
D['value'] -= B

...the last line fails with the error:

Exception: Reindexing only valid with uniquely valued Index objects

What am I doing wrong?

Upvotes: 1

Views: 3038

Answers (1)

Paul H
Paul H

Reputation: 68216

If I follow your example correctly (thanks for adding it, BTW), I believe what you need is as simple as:

D.sub(B, axis='index')

Which gives me:

In [29]: D.sub(B, axis='index')
Out[29]: 
        value      other
key                     
a    0.777595  44.307671
a   -0.671791  73.307671
b    0.779919  67.548856
b   -0.128690  96.548856

As you can see, this messes up the other column. If that's a problem, you're back in the same duplicate index situation, unfortunately.

Upvotes: 3

Related Questions