Re: [Numpy-discussion] Disallow Accelerate as a LAPACK backend for NumPy

[Matti Picus]

   On Tue, Nov 12, 2019 at 12:41 AM Matti Picus 
   wrote:

   Apple has dropped support for Accelerate. It has bugs that have not been
   fixed, and is closed source so we cannot fix them ourselves. We have
   been getting a handful of reports from users who end up building NumPy
   on macOS, and inadvertently link to Accelerate, then end up with wrong
   linalg results. In PR 14880https://github.com/numpy/numpy/pull/14880  I
   propose to disallow finding it when building NumPy. At this time it will
   remain in distutils as one of the backends to support users, but how do
   people feel about a future PR to totally remove it?

Someone pointed out that Apple has not officially dropped support as far as it can be 
determined. Sorry for the bad information. However, I still stand by the "has bugs 
that have not been fixed, and is closed source". An alternative to dropping 
automatic support for it would be to find a channel for engaging with Apple to report and 
fix the bugs.

Matti

___
NumPy-Discussion mailing list
NumPy-Discussion@python.org
https://mail.python.org/mailman/listinfo/numpy-discussion

Re: [Numpy-discussion] Disallow Accelerate as a LAPACK backend for NumPy

[Ralf Gommers]

On Fri, Nov 15, 2019 at 5:27 AM Matti Picus  wrote:

> On Tue, Nov 12, 2019 at 12:41 AM Matti Picus 
> wrote:
>
> Apple has dropped support for Accelerate. It has bugs that have not
> been
> fixed, and is closed source so we cannot fix them ourselves. We have
> been getting a handful of reports from users who end up building NumPy
> on macOS, and inadvertently link to Accelerate, then end up with wrong
> linalg results. In PR 14880https://github.com/numpy/numpy/pull/14880
> I
> propose to disallow finding it when building NumPy. At this time it
> will
> remain in distutils as one of the backends to support users, but how do
> people feel about a future PR to totally remove it?
>
> Someone pointed out that Apple has not officially dropped support as far
> as it can be determined. Sorry for the bad information. However, I still
> stand by the "has bugs that have not been fixed, and is closed source". An
> alternative to dropping automatic support for it would be to find a channel
> for engaging with Apple to report and fix the bugs.
>

That's been tried, repeatedly. I would suggest not to spend time on that.
Apple knows, they have just decided it's not important to them.

Spending time on contributing to either OpenBLAS or BLIS/libFLAME seems
like a more useful activity.

Cheers,
Ralf
___
NumPy-Discussion mailing list
NumPy-Discussion@python.org
https://mail.python.org/mailman/listinfo/numpy-discussion

Re: [Numpy-discussion] Disallow Accelerate as a LAPACK backend for NumPy

[Ilhan Polat]

We have a wiki page with all the details on scipy repo for the rationale of
why we wanted to drop it. There is no need to discuss further about the
situation of Accelerate.

On Fri, Nov 15, 2019, 15:41 Ralf Gommers  wrote:

>
>
> On Fri, Nov 15, 2019 at 5:27 AM Matti Picus  wrote:
>
>> On Tue, Nov 12, 2019 at 12:41 AM Matti Picus 
>> wrote:
>>
>> Apple has dropped support for Accelerate. It has bugs that have not
>> been
>> fixed, and is closed source so we cannot fix them ourselves. We have
>> been getting a handful of reports from users who end up building NumPy
>> on macOS, and inadvertently link to Accelerate, then end up with wrong
>> linalg results. In PR 14880https://github.com/numpy/numpy/pull/14880
>> I
>> propose to disallow finding it when building NumPy. At this time it
>> will
>> remain in distutils as one of the backends to support users, but how
>> do
>> people feel about a future PR to totally remove it?
>>
>> Someone pointed out that Apple has not officially dropped support as far
>> as it can be determined. Sorry for the bad information. However, I still
>> stand by the "has bugs that have not been fixed, and is closed source". An
>> alternative to dropping automatic support for it would be to find a channel
>> for engaging with Apple to report and fix the bugs.
>>
>
> That's been tried, repeatedly. I would suggest not to spend time on that.
> Apple knows, they have just decided it's not important to them.
>
> Spending time on contributing to either OpenBLAS or BLIS/libFLAME seems
> like a more useful activity.
>
> Cheers,
> Ralf
>
> ___
> NumPy-Discussion mailing list
> NumPy-Discussion@python.org
> https://mail.python.org/mailman/listinfo/numpy-discussion
>
___
NumPy-Discussion mailing list
NumPy-Discussion@python.org
https://mail.python.org/mailman/listinfo/numpy-discussion

[Numpy-discussion] State of numpy-financial

[Marcelo Gasparian Gosling]

Hi everyone!

So, np-financial is being phased out of mainline Numpy, right? I have a
patch for the IRR function that I'd like to submit, is it just as easy as
making a PR on Github?

Cheers,

Marcelo
___
NumPy-Discussion mailing list
NumPy-Discussion@python.org
https://mail.python.org/mailman/listinfo/numpy-discussion

Re: [Numpy-discussion] State of numpy-financial

[Warren Weckesser]

On 11/15/19, Marcelo Gasparian Gosling  wrote:
> Hi everyone!
>
> So, np-financial is being phased out of mainline Numpy, right? I have a
> patch for the IRR function that I'd like to submit, is it just as easy as
> making a PR on Github?


Hi Marcelo,

The new home for the NumPy financial functions is
https://github.com/numpy/numpy-financial

Pull requests are welcome!

Warren


>
> Cheers,
>
> Marcelo
>
___
NumPy-Discussion mailing list
NumPy-Discussion@python.org
https://mail.python.org/mailman/listinfo/numpy-discussion

Re: [Numpy-discussion] a new grant for NumPy and OpenBLAS!

[Evgeni Burovski]

Congratulations!

пт, 15 нояб. 2019 г., 2:42 Ralf Gommers :

> Hi all,
>
> I'm very pleased to announce that NumPy and OpenBLAS have received a joint
> grant for $195,000 from the Chan Zuckerberg Initiative.
>
> In summary, this grant is for high-level documentation, website
> development and graphic design, governance activities and community
> building for NumPy, and for technical work on OpenBLAS (which is one of
> NumPy's key dependencies). I wrote a blog post on what this grant is about
> and some background at [1], and published the full proposal at [2]. The
> program managers wrote a blog post about the whole program [3] which is
> also well worth reading.
>
> I'm looking forward to what we'll be able to do with this grant. The work
> is planned to start quite soon, Dec 1st, and run for one year.
>
> Questions and ideas on any aspect of this grant are very welcome!
>
> Cheers,
> Ralf
>
> [1] https://labs.quansight.org/blog/2019/11/numpy-openblas-CZI-grant/
> [2]
> https://figshare.com/articles/Proposal_NumPy_OpenBLAS_for_Chan_Zuckerberg_Initiative_EOSS_2019_round_1/10302167
> [3]
> https://medium.com/@cziscience/the-invisible-foundations-of-biomedicine-4ab7f8d4f5dd
>
> ___
> NumPy-Discussion mailing list
> NumPy-Discussion@python.org
> https://mail.python.org/mailman/listinfo/numpy-discussion
>
___
NumPy-Discussion mailing list
NumPy-Discussion@python.org
https://mail.python.org/mailman/listinfo/numpy-discussion

[Numpy-discussion] Cleanup of travisCI tests.

[Charles R Harris]

Hi All,

I think there are some travisCI tests that we can eliminate, see  tests

for
the current proposed set. I think we can eliminate the following

INSTALL_PICKLE5=1  # Python3.8 has this
NUMPY_EXPERIMENTAL_ARRAY_FUNCTION=0

And possibly one or both of

NPY_RELAXED_STRIDES_CHECKING=0
NPY_RELAXED_STRIDES_CHECKING_DEBUG=1

Thoughts?

Chuck
___
NumPy-Discussion mailing list
NumPy-Discussion@python.org
https://mail.python.org/mailman/listinfo/numpy-discussion

[Numpy-discussion] welcome Shaloo Shalini to the NumPy team

[Ralf Gommers]

Hi all,

Please welcome Shaloo Shalini to the NumPy team. Shaloo is an
experienced technology
marketing consultant with a passion for open source. She has been
contributing graphical content, information mapping and technical writing
since the summer. Most of that still needs to be integrated into the docs,
the website, and the user survey that Inessa is designing, but I'm really
excited about it. Nice examples are
https://github.com/numpy/numpy-surveys/pull/3 and
https://github.com/numpy/numpy.org/pull/23.

The NumPy Steering Council has decided to use part of our new Tidelift
income to contract Shaloo part-time for the next three months to help us
create content for the new website and create an information map that will
aid us in, e.g., restructuring the documentation and consolidating
duplicate content. Both of these activities will be very helpful in our
efforts to make NumPy more accessible and serve especially beginning users
and contributors better.

I'm really looking forward to Shaloo's continued contributions!

Cheers,
Ralf
___
NumPy-Discussion mailing list
NumPy-Discussion@python.org
https://mail.python.org/mailman/listinfo/numpy-discussion

Re: [Numpy-discussion] Cleanup of travisCI tests.

[Stephan Hoyer]

We still support explicitly opting-out of __array_function__, so I think we
should still keep that test configuration.

On Fri, Nov 15, 2019 at 1:50 PM Charles R Harris 
wrote:

> Hi All,
>
> I think there are some travisCI tests that we can eliminate, see  tests
> 
>  for
> the current proposed set. I think we can eliminate the following
>
> INSTALL_PICKLE5=1  # Python3.8 has this
> NUMPY_EXPERIMENTAL_ARRAY_FUNCTION=0
>
> And possibly one or both of
>
> NPY_RELAXED_STRIDES_CHECKING=0
> NPY_RELAXED_STRIDES_CHECKING_DEBUG=1
>
> Thoughts?
>
> Chuck
>
> ___
> NumPy-Discussion mailing list
> NumPy-Discussion@python.org
> https://mail.python.org/mailman/listinfo/numpy-discussion
>
___
NumPy-Discussion mailing list
NumPy-Discussion@python.org
https://mail.python.org/mailman/listinfo/numpy-discussion

Re: [Numpy-discussion] a new grant for NumPy and OpenBLAS!

[Ralf Gommers]

On Thu, Nov 14, 2019 at 9:35 PM Stefan van der Walt 
wrote:

> Hi Ralf,
>
> On Thu, Nov 14, 2019, at 15:42, Ralf Gommers wrote:
>
> I'm very pleased to announce that NumPy and OpenBLAS have received a joint
> grant for $195,000 from the Chan Zuckerberg Initiative.
>
>
> Congratulations on receiving this grant!  I am very happy to see more
> funded involvement in NumPy, and am pleased that you will get to focus some
> time on increasing participation and diversifying roles.  It is absolutely
> crucial that we succeed in this if we want to remain a healthy project---so
> thank you for all your hard work on that front.
>

Thanks Stéfan and everyone else!

Cheers,
Ralf
___
NumPy-Discussion mailing list
NumPy-Discussion@python.org
https://mail.python.org/mailman/listinfo/numpy-discussion