When one tries to print numpy scalars the printing does not respect np.set_printoptions.

import numpy as np
c = np.float64(0.128920820982908)
print(c)
with np.printoptions(precision=2):
    print(c)
    print(np.array([c]))

gives

0.128920820982908
0.128920820982908
[0.13]

This seems to be in contradiction to the docstring:

These options determine the way floating point numbers, arrays and other NumPy objects are displayed.

(emphasis is mine)