On 23 June 2010 16:13, Alan Bromborsky <abro...@verizon.net> wrote: > Alan Bromborsky wrote: >> In the transpose function we have transpose(a,axis) where axis can be a >> list of integers. But exactly what to the integers mean? If axis = >> [i1,i2] switching axis i1 with axis i2 is obvious, but what if axis = >> [i1,i2,i3]. Does this describe a cyclic permutation where >> i1->i2->i3->i2 or what does it describe?
The integers represent previous dimension positions, and the position of the integers in the transpose argument represent the new positions. >>> a = np.ones((5,6,7)) >>> a.shape (5, 6, 7) >>> np.transpose(a, (1,0,2)).shape (6, 5, 7) i.e. the dimension 1 in a is moved to the 0th position, dimension 0 in a is in the 1th position, and dimension 2 stays the same. HTH, Angus. -- AJC McMorland Post-doctoral research fellow Neurobiology, University of Pittsburgh _______________________________________________ NumPy-Discussion mailing list NumPy-Discussion@scipy.org http://mail.scipy.org/mailman/listinfo/numpy-discussion