At the moment it is not possible to transpose a distributed matrix. The
following two improvements would be nice:
1) Transposing of distributed array with any number of dimensions should be
implemented. See http://www.scipy.org/Numpy_Example_List#transpose
2) Lazy transposing. That is, use a transpose-flag instead of actually doing
the transposing. A physical transposing could then be avoided if the array
is only used in a BLAS operation.
Original issue reported on code.google.com by madsbk on 6 Jan 2010 at 3:06
Original issue reported on code.google.com by
madsbk
on 6 Jan 2010 at 3:06