Applying aggregate function on columns of Pandas pivot table

Question

I generated the following pivot table via taking maximum of values in Z column:

   val
X   x1     x2
Y   y1  y2 y1  y2
ID
a    9   1  5  11
b    8  10  7   6

After taking max on Z values, I need to report mean(y1,y2). The desired table is:

    val
X    x1          x2
Y    mean(y1,y2) mean(y1,y2)
ID
a    5           8
b    9           6.5

How can I achieve this using pandas?

My MWE:

#!/usr/bin/python
from pandas import DataFrame
import pandas as pd
import numpy as np

data=pd.read_table('data.txt')
pv=data.pivot_table(index=['ID'], columns=['X','Y'], values=['val'], aggfunc=np.max )
print pv

data.txt:

ID  X   Y   Z   val
a   x1  y2  z1  1
b   x1  y1  z2  2
a   x2  y2  z2  3
a   x1  y1  z4  4
a   x2  y1  z1  5
b   x2  y2  z3  6
b   x2  y1  z2  7
b   x1  y1  z3  8
a   x1  y1  z3  9
b   x1  y2  z3  10
a   x2  y2  z2  11

Answer 1

You can pass a level param and axis to calc the mean for the desired axis level:

In [142]:
pv.mean(level='X', axis=1)

Out[142]:
X   x1   x2
ID         
a    5  8.0
b    9  6.5