使用combine-first函数或update实现
df1 = pd.DataFrame([[np.nan, 3, 5], [4, np.nan, np.nan],[np.nan, 6, np.nan]])
df2 = pd.DataFrame([[7, np.nan, 8], [9, 10, 666]],index=[1, 2])
print(df1)
print('--------------------------------')
print(df2)
print('--------------------------------')
# 通过combine_first来讲df2填充到df1里空值的位置上，并且是根据index来做的。
print(df1.combine_first(df2))
print('--------------------------------')
# 如果df3的index比df1的多，那么就直接更新到df1上。
df3 = pd.DataFrame([[7, np.nan, 8], [9, 10, 666]],index=['a',1])
print(df1.combine_first(df3))
print('--------------------------------')
# update，根据index的位置，将相同位置的数值用df2直接覆盖df1。
df1.update(df2)
print(df1)
print('--------------------------------')
运行结果：
0 1 2
0 NaN 3.0 5.0
1 4.0 NaN NaN
2 NaN 6.0 NaN
--------------------------------
0 1 2
1 7 NaN 8
2 9 10.0 666
--------------------------------
0 1 2
0 NaN 3.0 5.0
1 4.0 NaN 8.0
2 9.0 6.0 666.0
--------------------------------
0 1 2
0 NaN 3.0 5.0
1 4.0 10.0 666.0
2 NaN 6.0 NaN
a 7.0 NaN 8.0
--------------------------------
0 1 2
0 NaN 3.0 5.0
1 7.0 NaN 8.0
2 9.0 10.0 666.0
--------------------------------
