using masks and numpy record arrays

[Robert Kern]

On 5/25/11 3:27 PM, Catherine Moroney wrote:
> Hello,
>
> I am trying to work with a structured array and a mask, and am encountering some
> problems.

You will want to ask numpy questions on the numpy mailing list:

   http://www.scipy.org/Mailing_Lists

> For example:
>
>  >>> xtype = numpy.dtype([("n", numpy.int32), ("x", numpy.float32)])
>  >>> a = numpy.zeros((4), dtype=xtype)
>  >>> b = numpy.arange(0,4)
>  >>> a2 = numpy.zeros((4), dtype=xtype)
>  >>> mask = numpy.where(b%2 == 0)
>  >>> a2[:]["n"] += b ! this changes the values of a2
>  >>> a[mask]["n"] += b[mask] ! this does not change the values of a
>  >>> a2
> array([(0, 0.0), (1, 0.0), (2, 0.0), (3, 0.0)],
> dtype=[('n', '<i4'), ('x', '<f4')])
>  >>> a
> array([(0, 0.0), (0, 0.0), (0, 0.0), (0, 0.0)],
> dtype=[('n', '<i4'), ('x', '<f4')])
>
> Why do the values of a2 get updated, and yet the values of a do not?

Only the final [] on the left-hand side of the assignment actually turns into a 
.__setitem__() call to the object that is the result of the expression to its 
left. a[mask] makes a copy while a2[:] makes a view.

You could do

   a["n"][mask] += b[mask]

since a["n"] will also make a view and the .__setitem__() on it will propagate 
back to the original memory.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
  that is made terrible by our own mad attempt to interpret it as though it had
  an underlying truth."
   -- Umberto Eco