根据另一个DataFrame填充Pandas列

Question

I have two data frames, and want to know how to add a column to one of them using certain values from the other. Specifically, I have data frames that look like:

foo = pd.DataFrame( np.random.rand(3,3))
foo.columns = ['col_1','col_2','col_3']

      col_1     col_2     col_3
0  0.661546  0.554032  0.753549
1  0.063641  0.490173  0.998119
2  0.370046  0.424208  0.125751


bar = pd.DataFrame( [[1, 2], [1,1], [3,3], [1,2], [2,1], [3,2]])

   0  1
0  1  2
1  1  1
2  0  3
3  1  2
4  2  1
5  0  2

I want to add a column to bar whose value is the value of foo at the location given by the columns of bar . So, the desired result would be:

   0  1  anything
0  1  2  0.490173
1  1  1  0.063641
2  0  3  0.753549
3  1  2  0.490173
4  2  1  0.370046
5  0  2  0.554032

My application for this involves very large data frames, so I don't think iterating through is a good option. Any help would be appreciated.

Answer 1

Try this

foo['Index']=foo.index
df=pd.melt(foo,id_vars=['Index'],value_vars=[1,2,3])
df
Out[563]: 
   Index variable     value
0      0        1  0.178661
1      1        1  0.065537
2      2        1  0.926429
3      0        2  0.139027
4      1        2  0.502449
5      2        2  0.971156
6      0        3  0.161616
7      1        3  0.530899
8      2        3  0.420385



bar.merge(df,left_on=[0,1],right_on=['Index', 'variable'],how='left')\
    .drop(['Index', 'variable'],axis=1)

   0  1     value
0  1  2  0.502449
1  1  1  0.065537
2  0  3  0.161616
3  1  2  0.502449
4  2  1  0.926429
5  0  2  0.139027