C++ linear algebra proposal

C++ standard has been moving forward really fast in recent years.

2 Likes

Apologies for such late reply to this thread. On the fortran side of things, I came across this paper in the J3-fortran website numbered 22-184. Which mentions features as such:

  1. Rank Agnostic Feature
  • Enable additional usage of assumed rank arguments

  • Deferred rank entities

  • Rank agnostic nested loops

  • Mathematics of Arrays

  1. Further support for containers
  • Provide looping mechanism for containers

  • Provide analog of ELEMENTAL

From a cursory reading all of them seems useful to me so, I hope they goes through for f202y.

3 Likes

Interesting - I have been working with @Lenore on the catenation operation that is one important part of Mathematics of Arrays. We have now gathered a lot of empirical material on the performance of various implementations and are ready to write an article about it.

4 Likes

The above paper is from the Fortran generics subgroup, @everythingfunctional is on this forum, maybe others too. We are trying to prototype it in LFortran.

@Arjen are you in touch with the generics subgroup? They are moving quite fast with approving the papers for the standard, so if you have any suggestions, let’s get them in.

1 Like

@certik, I get the emails but I am afraid I have not participated actively. This should be a good moment to get in touch again!

3 Likes

Yesterday we had some discussions about starting work on use case papers for the deferred/assumed rank features. Any help is definitely welcome.

2 Likes

Count me in :slight_smile:

2 Likes

We meet Mondays at 1 US Eastern time, so if you can attend a meeting that would be great.

1 Like

That would be 6 o’clock my time. A bit tricky, but I should be able to arrange it.

1 Like

Do you have a Zoom or other invitation for this?

1 Like

I just sent you the link in a private message.

Anybody else interested to join? Just ping me and I"ll share the link. The more people we have the better!

1 Like