Federico Perini

icon indicating a website link http://www.federicoperini.info

icon company Wisconsin Engine Research Consultants icon location Madison, WI icon location I like keyboards, flames, and stuff that runs fast.

Results 172 comments of Federico Perini

Add GitHub Actions CI for running tests

Thanks, this is a good idea. The original tests aren't really testing much so my plan is to gradually implement them with the object-oriented interface, and know what data should...

Building and testing stdlib on WSL1 with two fails

All the failing tests seem to involve 80-bit `xdp` precision. I have no idea why that may be. Maybe @jalvesz has experience with running stdlib with WSL?

Building and testing stdlib on WSL1 with two fails

@jalvesz our Windows CI builds do not support `xdp`, I wonder if there was a reason that was left out?

Adding matrix norms to the `stdlib_linalg` module.

Great idea @loiseaujc. Rn I'm trying to tackle the decompositions (pseudo-inverse, Cholesky, QR) and I was planning to address norms and condition number next. So, your contribution would be very...

Adding matrix norms to the `stdlib_linalg` module.

Thanks @loiseaujc for the detailed comments. I very mostly agree with your ideas and I would summarize the consensus so far: - let's have two separate functions for vector (`norm`)...

Adding matrix norms to the `stdlib_linalg` module.

> I prefer the `matrix_form = "Hessenberg"` Absolutely agree, and it would be more in line with the `norm_type` argument

Adding matrix norms to the `stdlib_linalg` module.

> Here is a draft/hacky implementation of `norm` I'd like to discuss a bit before drafting the PR. looks good so far! > * At the moment, we have the...

`lstsq`: return correct array size

Thanks a lot @jvdp1. I will wait another couple of days and then merge, if there are no further comments.

Fixes for extended and quad precision checking. Add CI cheks with fpm

I agree with @jvdp1: great work @jalvesz, thank you for finding this out! I don't want to have you rewrite your PR but here is some thought. Elsewhere because LAPACK...

linalg: Singular Value Decomposition

`ifx` has an internal compiler error on the test: will need to investigate