objectif

Masquer des valeurs (NaN-like) avec numpy.ma.

code minimal

import numpy as np
m = np.ma.array([1, -1, 3], mask=[False, True, False])
print(m.mean().item())

utilisation

import numpy as np
m = np.ma.masked_less([1,2,3], 2)
print(m.filled(0).tolist())

variante(s) utile(s)

import numpy as np
x = np.ma.array([1,2,3], mask=[0,1,0])
print(np.ma.is_masked(x[1]))

notes

• Idéal quand on ne veut pas propager des NaN.