如何基于共享公用密钥的另一个数据框的值“更新” df？ 蟒蛇

Question

how do i 'update' a12 based on the values of another dataframe that shares a common key? In the example below, the common key is column a.

a12 =

  a  b  c
  0  1  1
  1  na na

try10 =

  a  b  c
  1  1  1

when i use a merge, I get something like this.

pd.merge(a12, try10) =

  a  b  c  b_y  c_y
  0  1  1   na  na
  1  na na  1   1

I decided to do it manually, but i'm thinking there must be a more pythonic way than the below. I'd appreciate your help.

for i, val in a12.iterrows():
    for x, xval in try10.iterrows():
        if xval['Firm1'] == val['Firm']: 
            try10.ix[x]['AMranking'] =  val['AMranking']
            try10.ix[x]['numlawyers'] = val['numlawyers']
            try10.ix[x]['grossprofits'] = val['grossprofits']  

del try12['firm']

Answer 1

There is an combine_first method that you can use. See here. You'll need to set the desired key in both dataframes as index.

In [128]: a12.set_index ('a').combine_first(try10.set_index('a'))
Out[128]: 
   b  c
a      
0  1  1
1  1  1

Answer 2

You can just apply dropna to a12 before merge :

In [53]:

a12 = pd.DataFrame({'a':[0, 1], 'b':[1, np.nan], 'c':[1, np.nan]})
try10= pd.DataFrame({'a':[1], 'b':[1], 'c':[1]})
In [54]:

print pd.merge(a12, try10, how='outer', left_on='a', right_on='a')
   a  b_x  c_x  b_y  c_y
0  0    1    1  NaN  NaN
1  1  NaN  NaN    1    1
In [55]:

print pd.merge(a12.dropna(0), try10, how='outer')
   a  b  c
0  0  1  1
1  1  1  1

If a is the index:

In [57]:

print pd.merge(a12.dropna(0).reset_index(), try10.reset_index(), how='outer').set_index('a')
   b  c
a      
0  1  1
1  1  1