import pandas as pd
import numpy as np

In [24]: s = pd.Series(np.random.randn(10))

In [25]: norm = colors.Normalize(s.min() + .2, s.max() - .2)

In [26]: s
Out[26]:
0   -0.767090
1   -0.236357
2    0.031890
3    1.339422
4    0.994139
5   -1.481244
6   -0.253904
7   -0.032531
8    0.777265
9   -0.315098
dtype: float64

# normalize numpy array, doesn't modify
In [27]: norm(s.values)
Out[27]:
masked_array(data = [ 0.21240206  0.43165251  0.54246791  1.08262188  0.93998197 -0.08262188
  0.42440363  0.51585521  0.85038937  0.39912391],
             mask = False,
       fill_value = 1e+20)

In [28]: s
Out[28]:
0   -0.767090
1   -0.236357
2    0.031890
3    1.339422
4    0.994139
5   -1.481244
6   -0.253904
7   -0.032531
8    0.777265
9   -0.315098
dtype: float64

In [29]: norm(s)  # normalize series...
Out[29]:
masked_array(data = [ 0.21240206  0.43165251  0.54246791  1.08262188  0.93998197 -0.08262188
  0.42440363  0.51585521  0.85038937  0.39912391],
             mask = False,
       fill_value = 1e+20)

In [30]: s  # does modify.
Out[30]:
0    0.212402
1    0.431653
2    0.542468
3    1.082622
4    0.939982
5   -0.082622
6    0.424404
7    0.515855
8    0.850389
9    0.399124
dtype: float64

In [43]: matplotlib.__version__
Out[43]: '1.5.0'