On Wed, Jun 12, 2013 at 1:24 PM, Sebastian Berg <[email protected]> wrote: > On Tue, 2013-06-11 at 09:24 -0700, Jaime Fernández del Río wrote: >> I noticed today that the documentation for np.transpose states, for >> the return value, that "A view is returned whenever possible." >> > I guess a subclass could cause a copy (the code looks like subclassing > doing something fancy is not be anticipated, there may be a bug here), > other then that, I don't see a reason for this comment. Transposing is > always possible for strided memory patterns and a copy will never occur.
Indeed, there is no codepath in PyArray_Transpose() that even contemplates a memory copy. -- Robert Kern _______________________________________________ NumPy-Discussion mailing list [email protected] http://mail.scipy.org/mailman/listinfo/numpy-discussion
