I know set_printoptions determines the way floating point numbers are displayed.
set_printoptions
np.set_printoptions(precision=2) np.random.uniform(size=(2,3)) <