[NumPy-Tickets] [NumPy] #1431: accessing multiple fields in a recarray gives fields in wrong order

[NumPy Trac]

#1431: accessing multiple fields in a recarray gives fields in wrong order
------------------------+---------------------------------------------------
 Reporter:  samtygier   |       Owner:  somebody
     Type:  defect      |      Status:  new     
 Priority:  normal      |   Milestone:          
Component:  numpy.core  |     Version:          
 Keywords:              |  
------------------------+---------------------------------------------------
 when using the list in list notation to access multiple columns of a
 record array the columns are returned in the order that they are stored in
 the array, not in the order asked for.
 {{{
 >>> x = np.array([(1.5,2.5,(1.0,2.0)),(3.,4.,(4.,5.)),(1.,3.,(2.,6.))],
         dtype=[('x','f4'),('y',np.float32),('value','f4',(2,2))])
 >>> x[['x','y']]
 array([(1.5, 2.5), (3.0, 4.0), (1.0, 3.0)],
      dtype=[('x', '<f4'), ('y', '<f4')])
 >>> x[['y','x']]
 array([(1.5, 2.5), (3.0, 4.0), (1.0, 3.0)],
      dtype=[('x', '<f4'), ('y', '<f4')])
 }}}

 this different to what happens with a 2d array, where they are returned in
 the order asked for:

 {{{
 >>> a = array([0,0.1,0.2,0.3,0.4])
 then
 >>> a[[0,1,4]]
 array([ 0. ,  0.1,  0.4])
 >>> a[[4,1,0]]
 array([ 0.4,  0.1,  0. ])
 }}}

 I suggest that the it would be more consistent and predictable for the
 order to be the one asked for in both cases.

 The current behaviour for the recarray was undocumented until today, so
 there can't be many people using it. therefore it might be worth break
 backwards compatibility to make the interface better.

 Please see discussion at
 http://thread.gmane.org/gmane.comp.python.numeric.general/36933

 a simple patch is attached.

-- 
Ticket URL: <http://projects.scipy.org/numpy/ticket/1431>
NumPy <http://projects.scipy.org/numpy>
My example project