Pandas迭代索引并附加剩余的行

Question

I have the following frame:

df = pd.DataFrame(data={'id':[1,2,3],'x':[3,2,1]}).set_index('id')
print df
    x
id   
1   3
2   2
3   1

I need to iterate the rows, so that for each id, we append the remaining rows, eg the output should be:

    x
id   
1   3
    2
    1
2   2
    1
3   1

Ideally O(n) solution that generalizes to m columns.

Answer 1

Use list comprehension with slice by positions with iloc for list of DataFrame s and then concat for join all together, last remove second level of MultiIndex :

comp = [df.iloc[i:] for i in range(len(df.index))]
df = pd.concat(comp, keys=df.index).reset_index(level=1, drop=True)

print (df)
    x
id   
1   3
1   2
1   1
2   2
2   1
3   1

If need both level s is possible add rename_axis for set MultiIndex name s:

df = pd.concat(comp, keys=df.index).rename_axis(('a','b'))
print (df)
     x
a b   
1 1  3
  2  2
  3  1
2 2  2
  3  1
3 3  1

EDIT:

First level dont display repeated values, but no values are lost:

#change default option multi_sparse=True
with pd.option_context('display.multi_sparse', False):
    print (df)

     x
a b   
1 1  3
1 2  2
1 3  1
2 2  2
2 3  1
3 3  1

Answer 2

Something like

df.groupby(level=0).x.apply(lambda y : (np.arange(y)+1)[::-1]).apply(pd.Series).stack()
Out[320]: 
id   
1   0    3.0
    1    2.0
    2    1.0
2   0    2.0
    1    1.0
3   0    1.0
dtype: float64