How to do intersection match between 2 DataFrames in Pandas? - pandas

Question

Assume exists 2 DataFrames A and B like following
A:
a A
b B
c C
B:
1 2
3 4
How to produce C DataFrame like
a A 1 2
a A 3 4
b B 1 2
b B 3 4
c C 1 2
c C 3 4
Is there some function in Pandas can do this operation?

Answer 1

First all values has to be unique in each DataFrame.
I think you need product:
from itertools import product
A = pd.DataFrame({'a':list('abc')})
B = pd.DataFrame({'a':[1,2]})
C = pd.DataFrame(list(product(A['a'], B['a'])))
print (C)
0 1
0 a 1
1 a 2
2 b 1
3 b 2
4 c 1
5 c 2
Pandas pure solutions with MultiIndex.from_product:
mux = pd.MultiIndex.from_product([A['a'], B['a']])
C = pd.DataFrame(mux.values.tolist())
print (C)
0 1
0 a 1
1 a 2
2 b 1
3 b 2
4 c 1
5 c 2
C = mux.to_frame().reset_index(drop=True)
print (C)
0 1
0 a 1
1 a 2
2 b 1
3 b 2
4 c 1
5 c 2
Solution with cross join with merge and column filled by same scalars by assign:
df = pd.merge(A.assign(tmp=1), B.assign(tmp=1), on='tmp').drop('tmp', 1)
df.columns = ['a','b']
print (df)
a b
0 a 1
1 a 2
2 b 1
3 b 2
4 c 1
5 c 2
EDIT:
A = pd.DataFrame({'a':list('abc'), 'b':list('ABC')})
B = pd.DataFrame({'a':[1,3], 'c':[2,4]})
print (A)
a b
0 a A
1 b B
2 c C
print (B)
a c
0 1 2
1 3 4
C = pd.merge(A.assign(tmp=1), B.assign(tmp=1), on='tmp').drop('tmp', 1)
C.columns = list('abcd')
print (C)
a b c d
0 a A 1 2
1 a A 3 4
2 b B 1 2
3 b B 3 4
4 c C 1 2
5 c C 3 4