Stefan van der Walt wrote: > On Tue, Jun 19, 2007 at 12:35:05PM +0200, Sturla Molden wrote: > >> On 6/19/2007 12:14 PM, Sturla Molden wrote: >> >> >>> h[0,:,numpy.arange(14)] is a case of "sdvanced indexing". You can also >>> see that >>> >>> >>> h[0,:,[0,1,2,3,4,5,6,7,8,9,10,11,12,13]].shape >>> (14, 4) >>> >> Another way to explain this is that numpy.arange(14) and >> [0,1,2,3,4,5,6,7,8,9,10,11,12,13] is a sequence (i.e. iterator). So >> when NumPy iterates the sequence, the iterator yields a single integer, >> lets call it I. Using this integer as an index to h, gives a = h[0,:,I] >> which has shape=(4,). This gives us a sequence of arrays of length 4. In >> > > If you follow this analogy, > > x = N.arange(100).reshape((10,10)) > x[:,N.arange(5)].shape > > should be (5, 10), while in reality it is (10, 5). >
No, in this case, there is no ambiguity regarding where to put the sub-space, so it is put in the "expected" position. It could be argued that when a single integer is used in one of the indexing dimensions then there is also no ambiguity --- but the indexing code does not check for that special case. There is no bug here as far as I can tell. It is just perhaps somewhat unexpected behavior of a general rule about how "indirect" or "advanced" indexing is handled. You can always do h[0][:,arange(14)] to get the result you seem to want. -Travis > Cheers > Stéfan > _______________________________________________ > Numpy-discussion mailing list > Numpy-discussion@scipy.org > http://projects.scipy.org/mailman/listinfo/numpy-discussion > > > _______________________________________________ Numpy-discussion mailing list Numpy-discussion@scipy.org http://projects.scipy.org/mailman/listinfo/numpy-discussion
