[Numpy-discussion] Re: NumPy security roadmap proposal
On Fri, Jun 13, 2025 at 11:13 AM Andrew Nelson via NumPy-Discussion < numpy-discussion@python.org> wrote: > > On Fri, 13 Jun 2025 at 16:43, Ralf Gommers via NumPy-Discussion < > numpy-discussion@python.org> wrote: > >> >> For 2FA and repository/PyPI access, we'll start making changes soon. Note >> that GitHub has recently made changes to its 2FA settings that ask for >> action from many people: on https://github.com/orgs/numpy/people you can >> see that under "Two-factor authentication" the options increased; there is >> now a Secure/Insecure distinction instead of only Enabled/Disabled. If you >> want to move yourself from Insecure to Secure, you have to disable the >> SMS/mobile recovery option in your personal settings under "Password and >> authentication". A large majority of the 94 people with permissions are >> currently marked as Insecure. >> > > Having just visited this page I can't see any Two-factor authentication, > or secure/insecure properties listed. > It may only be visible to org owners then. > Remember that 2FA isn't just SMS, it could be an Authenticator app, > Physical key (yubikey), etc. > Yes indeed. The other methods are considered secure by GitHub, just SMS/mobile is not. Cheers, Ralf ___ 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: Allow matrix multiplication in structured arrays
Hi Abel,
As long as your x,y,z are next to each other, you can transform from
your structure to an unstructured array via a view, which has very
little cost. Though you need to be a bit careful with offsets, etc., if
there are also other elements in the structured dtype.
Example, with some extra fields:
dtype = np.dtype([("i", np.int64), ("x", np.float64), ("y", np.float64), ("z",
np.float64), ("j", np.int64)])
atoms = np.array(
[
(1, 0.0, 0.0, 0.0, -1),
(2, 1.0, 0.0, 0.0, -2),
(3, 0.0, 1.0, 0.0, -3),
(4, 1.0, 1.0, 1.0, -4),
],
dtype=dtype,
)
dt2 = np.dtype([("i", np.int64), ("xyz", np.float64, (3,)), ("j", np.int64)])
xyz = atoms.view(dt2)["xyz"]
xyz
# array([[0., 0., 0.],
#[1., 0., 0.],
#[0., 1., 0.],
#[1., 1., 1.]])
xyz[:] = 9.
atoms
array([(1, 9., 9., 9., -1), (2, 9., 9., 9., -2), (3, 9., 9., 9., -3),
(4, 9., 9., 9., -4)],
dtype=[('i', ' I'm using structured arrays to store atoms data produced by LAMMPS (I'm using
> a structured array that follows its format). I need to rotate the positions:
>
> ```
> import numpy as np
>
> transform = np.array([[1, 0, 0], [0, 1, 0], [0, 0, 1]], dtype=np.float64)
> dtype = np.dtype([("x", np.float64), ("y", np.float64), ("z", np.float64)])
> # real case with more fields, integers, bools, strings
>
> atoms = np.array(
> [
> (0.0, 0.0, 0.0),
> (1.0, 0.0, 0.0),
> (0.0, 1.0, 0.0),
> (1.0, 1.0, 1.0),
> ],
> dtype=dtype,
> )
>
> atoms[["x", "y", "z"]] = atoms[["x", "y", "z"]] @ transform.T
> ```
>
> But this produces:
>
> ```
> Traceback (most recent call last):
> File "c:\Users\acgc99\Desktop\rotation.py", line 16, in
> atoms[["x", "y", "z"]] = atoms[["x", "y", "z"]] @ transform.T
> ~~~^
> numpy._core._exceptions._UFuncNoLoopError: ufunc 'matmul' did not contain a
> loop with signature matching types (dtype([('x', ' ' None
> ```
>
> I can convert to unstructured arrays, but I guess that doing that change
> multiple times is not efficient when working with tens of millions of atoms.
> ___
> 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: m...@astro.utoronto.ca
___
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] Allow matrix multiplication in structured arrays
I'm using structured arrays to store atoms data produced by LAMMPS (I'm using a
structured array that follows its format). I need to rotate the positions:
```
import numpy as np
transform = np.array([[1, 0, 0], [0, 1, 0], [0, 0, 1]], dtype=np.float64)
dtype = np.dtype([("x", np.float64), ("y", np.float64), ("z", np.float64)]) #
real case with more fields, integers, bools, strings
atoms = np.array(
[
(0.0, 0.0, 0.0),
(1.0, 0.0, 0.0),
(0.0, 1.0, 0.0),
(1.0, 1.0, 1.0),
],
dtype=dtype,
)
atoms[["x", "y", "z"]] = atoms[["x", "y", "z"]] @ transform.T
```
But this produces:
```
Traceback (most recent call last):
File "c:\Users\acgc99\Desktop\rotation.py", line 16, in
atoms[["x", "y", "z"]] = atoms[["x", "y", "z"]] @ transform.T
~~~^
numpy._core._exceptions._UFuncNoLoopError: ufunc 'matmul' did not contain a
loop with signature matching types (dtype([('x', ' None
```
I can convert to unstructured arrays, but I guess that doing that change
multiple times is not efficient when working with tens of millions of atoms.
___
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: Bumping CPU baseline to x86-64-v2
On Fri, Jun 13, 2025 at 11:07 AM Jerome Kieffer wrote: > Hi Matti, > > Sorry for the delay ... > > In one of my project I am working on, we use based Avoton server > (Intel C2350) for CI/CD which can be rented today (2025) for less than > 5€/month > at online.net (a french provider). Switching to more recent generation > of processor (E3 1245v5) would imply at least 30€/month budget which is > much more expensive. > That CPU was released in Q3 2013 and does support the new x86-64-v2 baseline we are proposing (as you already said above). So there is no problem here, is there? Cheers, Ralf ___ 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: Bumping CPU baseline to x86-64-v2
Hi Matti, Sorry for the delay ... In one of my project I am working on, we use based Avoton server (Intel C2350) for CI/CD which can be rented today (2025) for less than 5€/month at online.net (a french provider). Switching to more recent generation of processor (E3 1245v5) would imply at least 30€/month budget which is much more expensive. Concerning the energy cost, I believe Online has well optimized their cost and if they still offer this kind of server at such low price, it probably means this processor is still doing its job (which I can confirm) and that the manufacturing cost has already been paid of. It is like the switch to electric cars: it is not because all car owners would (miraculously) switch to electric cars that the climate issue would be (miraculously) addressed. Cheers, -- Jérôme Kieffer On Sun, 18 May 2025 13:00:37 +0300 matti picus via NumPy-Discussion wrote: > Interesting. Could you give some more information that might convince NumPy > to continue supporting these old machines? Renting implies you do not own > them and are paying for the service. Are the energy/speed tradeoffs worth > continuing with them, rather than asking the hosting service for a more > modern machine? Do they use Numpy2.x in the CI/CD pipeline? ___ 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: NumPy security roadmap proposal
On Fri, Jun 13, 2025 at 9:40 AM Ralf Gommers via NumPy-Discussion wrote: > > ... > For 2FA and repository/PyPI access, we'll start making changes soon. Note > that GitHub has recently made changes to its 2FA settings that ask for action > from many people: on https://github.com/orgs/numpy/people you can see that > under "Two-factor authentication" the options increased; there is now a > Secure/Insecure distinction instead of only Enabled/Disabled. If you want to > move yourself from Insecure to Secure, you have to disable the SMS/mobile > recovery option in your personal settings under "Password and > authentication". A large majority of the 94 people with permissions are > currently marked as Insecure. > > Cheers, > Ralf > You may need member acess to see the "Two-factor authentication" dropdown selector, but in any case it seems disabling the SMS/mobile recovery option is now recommended practice. Be sure to download and keep your recovery codes safe, as that will now be the only accepted mode to regain access if you loose your 2fa access (i.e. loose or change your phone number). ___ 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: NumPy security roadmap proposal
On Fri, 13 Jun 2025 at 16:43, Ralf Gommers via NumPy-Discussion < numpy-discussion@python.org> wrote: > > For 2FA and repository/PyPI access, we'll start making changes soon. Note > that GitHub has recently made changes to its 2FA settings that ask for > action from many people: on https://github.com/orgs/numpy/people you can > see that under "Two-factor authentication" the options increased; there is > now a Secure/Insecure distinction instead of only Enabled/Disabled. If you > want to move yourself from Insecure to Secure, you have to disable the > SMS/mobile recovery option in your personal settings under "Password and > authentication". A large majority of the 94 people with permissions are > currently marked as Insecure. > Having just visited this page I can't see any Two-factor authentication, or secure/insecure properties listed. Remember that 2FA isn't just SMS, it could be an Authenticator app, Physical key (yubikey), etc. ___ 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
