[Numpy-discussion] Tensor Typing presentation and discussion on Wednesday! (As part of the NumPy Community Meeting)

ALFONSO LUIS CASTAÑO MARIN alfonsoluis.castanom at um.es
Tue Jun 15 14:12:49 EDT 2021


  Hi Stéfan,

  To what potential PEP are you referring exactly? Type arithmetic or  
automatically interpret `1` as `Literal[1]` wherever it makes sense?
Indeed to have a common repository with an initial document would be a  
good way to go. So far we did not start anything because the PEP on  
Variadics was still waiting for appoval. However, we also wanted to  
make sure that Numpy folks believe that this approach towards type  
arithmetic makes sense for their use cases.
Maybe we could sync on this on the next Numpy Development meeting?

Best,
Alfonso.

Quoting Stefan van der Walt <stefanv at berkeley.edu>:

> Hi Alfonso, Matthew,
>    
>   On Tue, Jun 15, 2021, at 06:59, ALFONSO LUIS CASTAÑO MARIN wrote:
>
>>  First of all, when talking about type arithmetic I suggest to  
>> refer to the latest presentation for the summit rather than from  
>> the Tensor Typing Meetings  
>> (https://docs.google.com/presentation/d/1NZzZfbUCXi78LDDe8C3hAaK9on8ZiX3Z2k4BlZXwsJ8/edit). Here I intentionally refer to the +,-,*,// syntax rather than the Add[] one. Moreover, I also write 1 + A rather than Literal[1] + A, assuming that we could write a PEP for enabling this behavior if needed, and as Matthew said, it seems feasible based on what was said in the  
>> summit.
>
>    
>   This PEP sounds like it is of some interest to many of us.  Would  
> it make sense to start a repository with an initial document that we  
> can iterate on, or how do you typically develop these?
>    
>   Stéfan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.python.org/pipermail/numpy-discussion/attachments/20210615/98db9144/attachment-0001.html>


More information about the NumPy-Discussion mailing list