CODEWITHSUNDEEP
pythonpandas
Dror
Dror

Reputation: 13071

Row-wise union of dataframe in pandas

Say that I have two DataFrames's:

df1 = pd.DataFrame([('A', 0.3), ('B', 0.4)], columns = ('ID', 'Buy'))
df2 = pd.DataFrame([('B', 3), ('A', 4)], columns = ('ID', 'Sell'))

That yield:

    ID  Buy
0   A   0.3
1   B   0.4

and

    ID  Sell
0   B   3
1   A   4

respectively.

Now, I want to obtain a single DataFrame that collects the data, namely:

    ID  Buy  Sell
0   A   0.3  4
1   B   0.4  3

Note that the order of the lines in df1 and df2 may not be the same. Furthermore, there might ID's that appear only in one frame and not in the other --- in this case the missing value should be filled with NaN I guess.

How can I do it?

I tried something like

pd.concat([df1, df2], join = 'outer', axis = 1)

but doesn't return the desired result.

Upvotes: 1

Views: 1685

Answers (1)

EdChum
EdChum

Reputation: 394209

I think you want to merge on ID column:

In [12]:

df1 = pd.DataFrame([('A', 0.3), ('B', 0.4)], columns = ('ID', 'Buy'))
df2 = pd.DataFrame([('B', 3), ('A', 4)], columns = ('ID', 'Sell'))
df1.merge(df2, on='ID', how='outer')
Out[12]:
  ID  Buy  Sell
0  A  0.3     4
1  B  0.4     3

Upvotes: 1

Related Questions