[Numpy-discussion] Re: Exact representable difference between for two arrays.
On Fri, 20 Dec 2024 at 18:23, Andrew Nelson wrote: > Wow, that would be a pretty serious optimization bug to override Kahan >> summation. Are you speaking of some theoretical future version of the >> interpreter, or of the current state of things? >> > > I just wanted to know how the current interpreter worked. > (For an augmentation of finite differences in scipy.optimize) ___ NumPy-Discussion mailing list -- numpy-discussion@python.org To unsubscribe send an email to numpy-discussion-le...@python.org https://mail.python.org/mailman3/lists/numpy-discussion.python.org/ Member address: arch...@mail-archive.com
[Numpy-discussion] Re: Exact representable difference between for two arrays.
> > Wow, that would be a pretty serious optimization bug to override Kahan > summation. Are you speaking of some theoretical future version of the > interpreter, or of the current state of things? > I just wanted to know how the current interpreter worked. ___ NumPy-Discussion mailing list -- numpy-discussion@python.org To unsubscribe send an email to numpy-discussion-le...@python.org https://mail.python.org/mailman3/lists/numpy-discussion.python.org/ Member address: arch...@mail-archive.com
[Numpy-discussion] Re: Exact representable difference between for two arrays.
On Fri, Dec 20, 2024 at 4:48 AM Andrew Nelson via NumPy-Discussion wrote: > > Or is there a risk that the Python interpreter would prematurely optimize > that to give: > > ``` > dx = h > ``` > > > -- > _ > Dr. Andrew Nelson Wow, that would be a pretty serious optimization bug to override Kahan summation. Are you speaking of some theoretical future version of the interpreter, or of the current state of things? Matti ___ NumPy-Discussion mailing list -- numpy-discussion@python.org To unsubscribe send an email to numpy-discussion-le...@python.org https://mail.python.org/mailman3/lists/numpy-discussion.python.org/ Member address: arch...@mail-archive.com
[Numpy-discussion] Exact representable difference between for two arrays.
Hi all, Let's assume we have two float arrays, `x0` and `h`, both of which have the same shape. Let's calculate the sum of those two arrays: ``` x1 = x0 + h ``` The true representable difference between `x1` and `x0` is: ``` dx = x1 - x0 ``` Is it also possible to get that exact representation by doing the following: ``` dx = (x0 + h) - x0 ``` Or is there a risk that the Python interpreter would prematurely optimize that to give: ``` dx = h ``` -- _ Dr. Andrew Nelson _ ___ NumPy-Discussion mailing list -- numpy-discussion@python.org To unsubscribe send an email to numpy-discussion-le...@python.org https://mail.python.org/mailman3/lists/numpy-discussion.python.org/ Member address: arch...@mail-archive.com
