Re: [Numpy-discussion] aa.astype(int) truncates and doesn't round

2011-01-06 Thread Robert Kern
On Thu, Jan 6, 2011 at 04:14, wrote: > just something I bumped into and wasn't aware of > aa > array([ 1.,  1.,  1.,  1.,  1.]) aa.astype(int) > array([0, 1, 0, 0, 0]) aa - 1 > array([ -2.22044605e-16,   2.22044605e-16,  -2.22044605e-16, >        -3.33066907e-16,  -3.33066907e-16])

Re: [Numpy-discussion] aa.astype(int) truncates and doesn't round

2011-01-06 Thread Robert Kern
On Thu, Jan 6, 2011 at 09:40, Keith Goodman wrote: > On Thu, Jan 6, 2011 at 2:14 AM,   wrote: >> just something I bumped into and wasn't aware of >> > aa >> array([ 1.,  1.,  1.,  1.,  1.]) > aa.astype(int) >> array([0, 1, 0, 0, 0]) > aa - 1 >> array([ -2.22044605e-16,   2.22044605e-16

Re: [Numpy-discussion] aa.astype(int) truncates and doesn't round

2011-01-06 Thread josef . pktd
On Thu, Jan 6, 2011 at 10:40 AM, Keith Goodman wrote: > On Thu, Jan 6, 2011 at 2:14 AM,   wrote: >> just something I bumped into and wasn't aware of >> > aa >> array([ 1.,  1.,  1.,  1.,  1.]) > aa.astype(int) >> array([0, 1, 0, 0, 0]) > aa - 1 >> array([ -2.22044605e-16,   2.22044605e

Re: [Numpy-discussion] aa.astype(int) truncates and doesn't round

2011-01-06 Thread Keith Goodman
On Thu, Jan 6, 2011 at 2:14 AM, wrote: > just something I bumped into and wasn't aware of > aa > array([ 1.,  1.,  1.,  1.,  1.]) aa.astype(int) > array([0, 1, 0, 0, 0]) aa - 1 > array([ -2.22044605e-16,   2.22044605e-16,  -2.22044605e-16, >        -3.33066907e-16,  -3.33066907e-16

[Numpy-discussion] aa.astype(int) truncates and doesn't round

2011-01-06 Thread josef . pktd
just something I bumped into and wasn't aware of >>> aa array([ 1., 1., 1., 1., 1.]) >>> aa.astype(int) array([0, 1, 0, 0, 0]) >>> aa - 1 array([ -2.22044605e-16, 2.22044605e-16, -2.22044605e-16, -3.33066907e-16, -3.33066907e-16]) >>> np.round(aa).astype(int) array([1, 1, 1, 1, 1])

Re: [Numpy-discussion] NEP for faster ufuncs

2011-01-06 Thread Francesc Alted
2011/1/5, Mark Wiebe : > On Wed, Jan 5, 2011 at 4:26 AM, Francesc Alted wrote: > >> Sorry for the naive question, but I use the numpy.fromiter() iterator >> quite a few in my projects. and I'm curious on whether this new >> iterator would allow numpy.fromiter() to go faster (I mean, in Python >> s