Ever wondered whether fast direct solvers are compatible with Quadrature by Expansion, a method for the evaluation of singular layer potentials? Wonder no more 🙂 In https://arxiv.org/abs/2504.13809, we offer an algorithmic recipe, analysis, an end-to-end error model, and some weighting tricks, the latter two applicable beyond QBX, along with numerical experiments. #layerpot #fastalg #fastalgorithm #numpde #numerics #scicomp #paper 🎓 📖
#SciComp
I'm releasing lecture slides and related demos (Jupyter notebooks) for my classes in open-source form on Github:
https://andreask.cs.illinois.edu/lecture-slides-and-demos-released-as-open-source/ 🗨️ 🎓 📖 🧑🏫 #teaching #scicomp #opensource #lecturenotes
I like to do math, small prototypes and other experimenting with Python in Jupyter notebooks. NumPy, SciPy, SymPy and Matplotlib have been very handy.
But recently Python was too slow for what I was doing and I had trouble juggling Numpy arrays like I often have (it is required for speed). I know some #Rust so I decided to give #evcxr a go because it has a Jupyter kernel. And omg, I wasn't disappointed! It was mindbogglingly fast compared to what I'm used to and it also made the hard part of my code way easier to get right.
However, I only needed math functions provided by std this time. Now I'm trying to figure out if there are science Rust libs that are convenient enough to replace most Python in my notebooks. I know it's a fragmented space, but maybe one could already gather a comprehensive collection of crates manually.
Any recommendations?
![End of a long Jupyter notebook cell in VSCodium with Rust code in it. Running it took 6.3s. The uncommented code that is visible is computing something in parallel using rayon:
x.par_iter_mut().enumerate().for_each(|(n, xn)| {
let n = n as i32;
for k in n-N+1..=n {
*xn += a(k, lam, t1, t4, sr)
* (-lam/(sr as f64)*((n-k) as f64)).exp()
* (omega[k.max(0) as usize]/(sr as f64)
* ((n-k) as f64)
+ phi[k.max(0) as usize]).cos();
}
});](https://files.mastodon.social/cache/media_attachments/files/113/714/005/444/099/240/original/80213e89d26a4084.png)
How busy is the #HPC cluster? Maybe the real question is "how long do I have to wait?". Our new blog post goes into some of these questions we get asked sometimes, and the high-level picture probably applies other places too.
https://aaltoscicomp.github.io/blog/2024/how-busy-is-the-cluster/
#SciComp #HPC #ComputationalScience #Slurm
Interested in how we do our online teaching: production values, streaming, everyone can attend, videos and materials available immediately? The @coderefinery Train the Trainer workshop will go over it all, 4 Tuesdays starting tomorrow EU mornings. Online, you can drop by for the days that interest you.
https://fosstodon.org/@coderefinery/112913487426684269
#OnlineTeaching #HPC #SciComp #RSEng #CodeRefinery
We are live! Currently the Intro to #SciComp that talks about resources needed for computing. Check out this behind the scenes view...
#SciComp / #HPC Kickstart day 1 is for a broad audience. (livestream starting today at 12:00, times in EEST):
- 11:50: icebreakers
- 12:00 intro
- 12:10: From data storage to your science
- 12:25: Computational reproducibility and open science
- 13:00: Interview, "Humans of SciComp"
- 13:25: What can you do with a cluster?
- 14:00: Connecting to a cluster
- 15:00: How to ask for help with (super)computers
- 15:25: VS Code on HPC
https://twitch.tv/coderefinery
https://scicomp.aalto.fi/training/scip/kickstart-2024/
Our #SciComp / #HPC kickstart course starts tomorrow at 12:00 EEST / 11:00 CEST. It is livestream, so anyone can watch with no questions asked. Register for the Q&A document, even if you aren't at #AaltoUniversity (and we teach so that it's useful to others too).
Day 1: basic background and examples. Day 2-3: Using a cluster.
https://scicomp.aalto.fi/training/scip/kickstart-2024/
#ScientificComputing #teaching
Our "#HPC Summer Kickstart" course is 4-6 June 2024 (EU afternoons). Everyone is invited: it's #livestream so yes, *you* can attend.
- Day 1: general #SciComp concepts and practices
- Day 2-3: Running stuff on a cluster (Slurm, batch jobs, interactive, parallel, etc).
It's a first course that everyone should be able to follow. You can re-advertise to your colleagues and be their local guide to your own cluster.
https://scicomp.aalto.fi/training/scip/kickstart-2024/
#RSEng #OpenScience #ReproducibleResearch
TTT4HPC today was excellent, and we had a great presentation on containers (and one of our best lessons, period): If you've wondered what they are and how to use them for #SciComp / #RSEng on #HPC/desktop, this is the starting point you need.
Video is processed: https://youtu.be/9nHhB3Tn_BU
Material: https://coderefinery.github.io/ttt4hpc_containers/
Tuesday Tools and Techniques for #HPC is now live on #livestream. Today's lesson are about containers on clusters, and lasts about two hours starting from now (with exercises in the afternoon).
Stream: https://twitch.tv/coderefinery
Course: https://scicomp.aalto.fi/training/scip/ttt4hpc-2024/
#SciComp #RSEng #ReproducibleResearch #CodeRefinery
Would you like to be part of the next #CodeRefinery workshop?
- We are looking for new instructors, contributors and helpers! You'll get onboarding, mentoring and an exciting opportunity to share your passion!
Join us for planning today, 06.05 at 14 CEST at https://coderefinery.org/join/meetings/
Are you planning to attend our first in person conference? **Please sign up by May 5th**! 👈 We are in the process of finalizing reservations of rooms and meals and we cannot ensure space for everyone after that date! https://nordic-rse.org/events/2024-in-person-conference/
You aren't sure and would like to know what is on the menu? The book of accepted abstracts has been published! 📖 👀 https://nordic-rse.org/events/2024-in-person-conference/nordic-2024-book-of-abstracts.pdf
"Tuesday Tools and Techniques for HPC" (with @coderefinery people) is starting: it's not a basic course but more advanced cool stories about how we actually use and think about #HPC clusters. As usual, it's a #livestream so anyone can watch: https://scicomp.aalto.fi/training/scip/kickstart-2024/ (and register for interaction) #teaching #SciComp #RSEng #carpentries #CodeRefinery
Many of our partners are also providers of High Performance Computing services.
In April/May we will gather for 4 Tuesdays to explore how #HPC can be used in daily research computing work:
"Tuesday Tools and Techniques for HPC" organized by #AaltoSciComp
More info and registration: https://scicomp.aalto.fi/training/scip/ttt4hpc-2024/
#TTT4HPC #SciComp
The call for abstracts for the first Nordic RSE in person conference is closing at midnight (Eastern Europe Summer Time)! We look forward to seeing your contributions!
Do you have questions about applying #RSEng concepts learned in the #CodeRefinery workshop last week to your own work?
-> 'Bring your own code' today or next week's Tuesday at 13-15 CET and discuss with our specialists (zoom link in wrap up email)
It's a wrap!
On Thursday we concluded our 9th large online #CodeRefinery workshop.
It seemed our updated #git lesson was well received, which makes the hard work worth it!
Thanks to everyone involved, this would not have been possible without you <3
All materials, recordings, and Q&A will remain available indefinitely (and are suitable for independent study) : https://coderefinery.github.io/2024-03-12-workshop/
Are you ready for our first in person conference? We just extended the abstract deadline to April 5th! Join us! https://nordic-rse.org/events/2024-in-person-conference/
We look forward to meeting you on May 30-31 2024, in Aalto! #NordicRSE #RSEng #SciComp #opensource #HPC
Client Info
Server: https://mastodon.social
Version: 2025.04
Repository: https://github.com/cyevgeniy/lmst