CODEWITHSUNDEEP
pythonpandas
msakya
msakya

Reputation: 9801

Merging dataframes with duplicate indices

In the following example in pandas package of python, when merging two dataframes with duplicate/common indices, these indices are http://pandas.pydata.org/pandas-docs/dev/10min.html#join

Is it possible to merge it like following instead:

     key  lval    rval
  0  foo     1,2     4,5

and will it make a difference if the 1,2,4,5s are string?

Upvotes: 0

Views: 962

Answers (1)

Andy Hayden
Andy Hayden

Reputation: 375375

I don't recommend using this datastructure, I expecte there is a better way depending on what the next stage of your analysis is...

Saying that, here's one way to do it. Take Series of lists for each key:

In [11]: l = left.groupby('key')['lval'].apply(list)

In [12]: l.name = 'lval'

In [13]: l
Out[13]:
key
foo    [1, 2]
Name: lval, dtype: object

In [14]: r = right.groupby('key')['rval'].apply(list)

In [15]: r.name = 'rval'

And then concat these into a DataFrame:

In [16]: pd.concat([l, r], axis=1)
Out[16]:
       lval    rval
key
foo  [1, 2]  [4, 5]

Upvotes: 2

Related Questions