masked arrays and views (Trac #1305) #1903

Original ticket http://projects.scipy.org/numpy/ticket/1305 on 2009-11-26 by trac user tgirod, assigned to @pierregm.

when creating a view on a masked array with mask=False, changing the mask in the view won't propagate to the mask in the original array.

However, when the mask is already an array before the view is created, altering the mask in the view alters the mask in the original array.

this difference of behaviour can lead to troubles

import numpy as np
import numpy.ma as ma

x = ma.zeros(25)
xv = x.reshape((5,5))

print x.mask

> False

print xv.mask

> False

xv.mask = np.identity(5)

print x.mask

> False

print xv.mask

> [[ True False False False False]
>  [False  True False False False]
>  [False False  True False False]
>  [False False False  True False]
>  [False False False False  True]]