Thank you so much for your prompt answer, Stéfan. It's a very very interesting method. I will keep it for future. :)
But, i tested it with a few examples and the speed of execution is just a tiny bit slower than what i told you i was using. So it's not faster, it's about the same speed. Thank you again. I will play with your method a little more. --- On Thu, 6/18/09, Stéfan van der Walt <[email protected]> wrote: From: Stéfan van der Walt <[email protected]> Subject: Re: [Numpy-discussion] Advanced indexing advice? To: "Discussion of Numerical Python" <[email protected]> Date: Thursday, June 18, 2009, 2:16 AM Hi Cristi 2009/6/18 Cristi Constantin <[email protected]>: > I have a question about advanced indexing. > > I have 2 matrices : > >>>> > a=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]]) > > b=array([[1, 0, 1], > [0, 2, 0], > [0, 0, 3]]) >>>> > > I want to put all NON-zero elements from array B into array A, but use > offset! Here's a solution using views: offset = np.array([1,1]) slices = [slice(*x) for x in zip(offset, offset + b.shape)] c = a[slices] mask = (b != 0) c[mask] = b[mask] Regards Stéfan
_______________________________________________ Numpy-discussion mailing list [email protected] http://mail.scipy.org/mailman/listinfo/numpy-discussion
