Pandas DataFrame groupby + apply did the first function twice

In [1]: import pandas as pd
   ...: import numpy as np
   ...: 

In [2]: pd.__version__
Out[2]: '0.20.3'

In [3]: 

In [3]: a = pd.DataFrame( np.random.random(100).reshape(10,10) )

In [4]: a.iloc[:,0] = "a_str"

In [5]: a.iloc[3:10,0] = "b_str"

In [6]: def f(group):
   ...:     print( type( group) )
   ...:     print( group.shape )
   ...:     print( "done this function" )
   ...:     

In [7]: a.groupby(0 , axis = 0).apply(f)
<class 'pandas.core.frame.DataFrame'>
(3, 10)
done this function
<class 'pandas.core.frame.DataFrame'>
(3, 10)
done this function
<class 'pandas.core.frame.DataFrame'>
(7, 10)
done this function
Out[7]: 
Empty DataFrame
Columns: []
Index: []

In [8]: a.iloc[:,0]
Out[8]: 
0    a_str
1    a_str
2    a_str
3    b_str
4    b_str
5    b_str
6    b_str
7    b_str
8    b_str
9    b_str
Name: 0, dtype: object

In [9]: set(a.iloc[:,0].values)
Out[9]: {'a_str', 'b_str'}

Problem description

the DataFrame is grouped by the first coloum, and then it applied a function. The first apply will perform twice. Is it a bug or something wrong with my code?

I have also tried different str named column it will do it twice as well.

Best wishes, Yilang.

Comment From: jreback

see the big read box:http://pandas.pydata.org/pandas-docs/stable/groupby.html#flexible-apply this is an implementation detail.

Comment From: cyrushu

Thanks a lot~