连接两个索引长度不同的 Pandas DataFrame 列

Question

How do I add a merge columns of Pandas dataframe to another dataframe while the new columns of data has less rows? Specifically I need to new column of data to be filled with NaN at the first few rows in the merged DataFrame instead of the last few rows. Please refer to the picture. Thanks.

Answer 1

Use:

df1 = pd.DataFrame({
        'A':list('abcdef'),
        'B':[4,5,4,5,5,4],
})

df2 = pd.DataFrame({
        'SMA':list('rty')
})

df3 = df1.join(df2.set_index(df1.index[-len(df2):]))

Or:

df3 = pd.concat([df1, df2.set_index(df1.index[-len(df2):])], axis=1)
print (df3)
   A  B  SMA
0  a  4  NaN
1  b  5  NaN
2  c  4  NaN
3  d  5    r
4  e  5    t
5  f  4    y

How it working:

First is selected index in df1 by length of df2 from back:

print (df1.index[-len(df2):])
RangeIndex(start=3, stop=6, step=1)

And then is overwrite existing values by DataFrame.set_index :

print (df2.set_index(df1.index[-len(df2):]))
  SMA
3   r
4   t
5   y