根据列名称将pandas DataFrame列连接在一起

Question

I have a pandas Dataframe that looks something like this:

df
A.1  B.1  C.1  A.2  B.2  C.2  A.3  B.3  C.3
1    a    m    4    d    p    7    g    s
2    b    n    5    e    q    8    h    t
3    c    o    6    f    r    9    i    u

and I would like to merge all of the columns containing similar names in to the same column. So the output would be a DataFrame like this:

df
A  B  C  
1  a  m
2  b  n
3  c  o
4  d  p
5  e  q
6  f  r
7  g  s
8  h  t
9  i  u

What is the most efficient/elegant way to do this?

Answer 1

Here's one approach, turning your columns into a multi-index and unstacking.

In [220]: mult_index = [tuple(x.split('.')) for x in df.columns]
In [221]: df.columns = pd.MultiIndex.from_tuples(mult_index)

In [222]: df.stack()
Out[222]: 
     A  B  C
0 1  1  a  m
  2  4  d  p
  3  7  g  s
1 1  2  b  n
  2  5  e  q
  3  8  h  t
2 1  3  c  o
  2  6  f  r
  3  9  i  u