[Numpy-discussion] Summing indices of heterogeneous shapes

[Michael Hoffman]

In the following example I can sum up lists of column indices:

 >>> x = numpy.arange(30)
 >>> x.shape = (6, 5)
 >>> x
array([[ 0,  1,  2,  3,  4],
        [ 5,  6,  7,  8,  9],
        [10, 11, 12, 13, 14],
        [15, 16, 17, 18, 19],
        [20, 21, 22, 23, 24],
        [25, 26, 27, 28, 29]])
 >>> cols = [[0, 1], [2, 3]]
 >>> x[:, cols].sum(2)
array([[ 1,  5],
        [11, 15],
        [21, 25],
        [31, 35],
        [41, 45],
        [51, 55]])

Is there a way to do this if the shapes of the lists are heterogeneous? 
For example, what if I wanted to set cols = [[0, 1], [2, 3], [1, 2, 3]]? 
What's the best way to get a sensible result?

Thanks!
-- 
Michael Hoffman