[Numpy-tickets] [NumPy] #989: masked_array results differ between inplace and regular operators

NumPy numpy-tickets@scipy....
Mon Jan 26 11:18:50 CST 2009


#989: masked_array results differ between inplace and regular operators
----------------------------+-----------------------------------------------
 Reporter:  danielmcdonald  |       Owner:  pierregm
     Type:  defect          |      Status:  new     
 Priority:  normal          |   Milestone:          
Component:  numpy.ma        |     Version:  1.1.1   
 Severity:  normal          |    Keywords:          
----------------------------+-----------------------------------------------
 The result of a masked array addition differs between inplace math
 operators and the regular math operators. The unexpected behavior results
 in the regular operator, it appears that the 'mask' variable is not being
 set in the same fashion. Below is interpreter output. NumPy 1.0.3 produces
 the expected result, NumPy 1.1.1 and NumPy 1.2.1 do not.

 The following output was produced using Python 2.5.2, NumPy 1.2.1 on an
 OSX 10.5.6 x86 machine:

 >>> from numpy import ma, array
 >>> a = ma.array([1,2,3],mask=array([False,False,False]))
 >>> a
 masked_array(data = [1 2 3],
      mask = [False False False],
      fill_value=999999)

 >>> a *= array([2,3,4])
 >>> a
 masked_array(data = [2 6 12],
      mask = [False False False],
      fill_value=999999)

 >>> a = ma.array([1,2,3],mask=array([False,False,False]))
 >>> a * array([2,3,4])
 masked_array(data = [ 2  6 12],
      mask = False,
      fill_value=999999)

-- 
Ticket URL: <http://scipy.org/scipy/numpy/ticket/989>
NumPy <http://projects.scipy.org/scipy/numpy>
The fundamental package needed for scientific computing with Python.


More information about the Numpy-tickets mailing list