Python Pandas-基于单个索引的具有多个索引的逻辑索引数据框

Question

I'm still pretty new to Pandas but I've searched around quite a bit and can't quite find what I'm looking for.

So here is my problem:

I have two dataframe - one with mutliple indexes and another with just one index

df1=

               value1  value2
ind1 ind2
a      1          1.1     7.1
b      2          2.0     8.0
c      3          3.0     9.0
a      4          4.0    10.0
b      5          5.0    11.0
c      6          6.0    12.0


df2=

           value1  value2
ind1
a           8.0     7.0
b           9.0     8.0
c           3.0     9.0
d           11.0   10.0
e           12.0    11.0
f           1.0    12.0

I would like to index data from df1 based on df2 where value1 > value2 .

df2['value1'] > df2['value2']

I know that I can get the data from df2 with

df2.loc[df2['value1'] > df2['value2']]

But how would I get data from df1? I tried:

df1.loc[df2['value1'] > df2['value2']]

But it fails with

*** IndexingError: Unalignable boolean Series key provided

Any suggestions will be much appreciated, thank you!

Answer 1

Get the indices from df2 and then select df1 on those indices:

indices = df2.loc[df2['value1'] > df2['value2']]
>>>indices
Index([u'b', u'd', u'e'], dtype='object')
>>>df1.ix[indices]

ind1 ind2   val1    val2
b    2      2.0     8.0
b    5      5.0     11.0