Our yearly #SciComp and #HPC kickstart course is open for reg: 6-8 June, EU afternoons. As usual, #livestream for everyone in the world + extra local help - we welcome anyone to watch.
Why register: see the map of your career, brush up on some skills, and get connected to a community of others. The course (especially day 1) is designed around what we see young and old researchers struggle with. Day 2-3 go into cluster usage.
https://scicomp.aalto.fi/training/scip/kickstart-2023/ #RSEng #coderefinery #openscience
This course on parallel programming by @JukkaSuomela at Aalto is absolutely fantastic:
https://ppc.cs.aalto.fi/
It covers everything: how the hardware works, thread based parallelism, pipelining, vectorization, even GPU programming! The lecture notes are super clear and detailed.
In about one hour, Richard Darst & co will talk about how ASC runs its Research Software Engineer group: target audience is people interested in RSE groups in Nordics. #RSE #RSEng #SciComp
We have a new "#git the way you need it" #cheatsheet. Most projects are small and simple, so our presentation should is organized around that. We shouldn't try to make every project perfect, but make sure that git is used at all - and we do that by guiding learners along their actual path. #RSEng #CodeRefinery #HPC #ResearchSoftware
https://aaltoscicomp.github.io/cheatsheets/git-the-way-you-need-it-cheatsheet.pdf
We started for the day! git-intro all day
- now: motivation
- soon: basics of making a repository, committing changes
- later on: branching
This thread will be used for #CodeRefinery 2023 March schedule updates. #teaching #livestream #RSEng #git #VersionControl #OpenScience #SciComp
https://twitch.tv/coderefinery
https://coderefinery.github.io/2023-03-21-workshop/
look ⬇️
> If someone had just show me this, I could have saved 3 months from my #PhD. — someone ~2015 (not about #CodeRefinery)
There is no better investment than saving your future time. Is this week's minor deadline more important than your future career? aybe ask your supervisor and their answer might surprise you.
Attend the #CodeRefinery workshop this week and next, look ⬆️ thread to see more.
#RSEng #SciComp #git #github #openscience #ResearchSoftware #AcademicMastodon
https://coderefinery.github.io/2023-03-21-workshop/
What is this #CodeRefinery thing we've been talking about, and why might you attend?
- Workshop, for scientists, not about programming but the skills to use it well for #SciComp and #DataScience
- If you, or someone you know, has to use programming to do your work, but you know you could be more efficient - come!
> This was the best course I ever had, just make up something and say I said it. -- participant, ~2019
Starts Tue, online&in-person at partners: https://coderefinery.github.io/2023-03-21-workshop/
We met the #Aalto University ITS cloud team. Relevant points for our users:
- Most uses have been high-level APIs such as speech recognition. Virtual machines, HPC, etc are better made locally. #HPC/AI devs stay where you are.
- They can make cloud (Azure for now) subscriptions for you https://www.aalto.fi/en/services/it-services-for-research
- They pay basic usage, above some limit project pays
- If you need help using these, Aalto #RSEng can help with the software side. https://scicomp.aalto.fi/rse/
Our "Summer Scientific Computing Kickstart" course has a page now - dates are 6-8 June 2023. It's basically our #HPC cluster introduction course for new researchers, combined with the necessary #SciComp background so everyone can reach the level they need.
It uses the standard #CodeRefinery style: Livestream so everyone can attend, local support for those from #Finland and partners. We welcome anyone to register, we try to be useful for outsiders. #OpenScience #RSEng
https://scicomp.aalto.fi/training/scip/kickstart-2023/
@davidjamesweir We usually have them twice a year. According to my calendar the following seems to be 19-21&26-28 September 2023.
It's completely fine to attend just the first week, too. First week now + (review&)second week in the autumn would be a great way to learn, too!
We had an internal development day to talk about our status and future goals. We settled on these three broad topics for special mention:
1. Teaching: continue focus on big, basic courses for new SciComp community members, and support (#SciComp garage, RSE) for advanced self-study course support.
2. Triton v3: we have big software/OS updates coming which will take lots of time, both doing it and supporting changes.
3. Help our largest users use #LUMI HPC (#EuroHPC).
https://aaltoscicomp.github.io/blog/2023/march-development-day/
In about one hour, Richard Darst will talk about what new groups leaders (who may or may not be computational specialists) need to know about starting their own research group. There is some presentation material, but hopefully practical Q&A takes up most of the time.
Registration is still possible:
https://www.aalto.fi/en/events/managing-computational-and-data-based-research-at-aalto
Join @eglerean from our team in ~2h from now:
Sharing and re-using personal data in research seems often impossible for researchers in Europe with the limitations imposed by GDPR or ethical committees. However it is possible to do so, but it requires planning in advance and good security measures in proportion to the actual risks. Join @eglerean and the Italian #Reproducibility #Network today at 3pm CET: https://www.itrn.org/initiatives/onlineseminars"
We attended an #AaltoUniversity Data Agents meeting about #ResearchIntegrity. Notes:
- Everyone knows falsification and plagiarism -type cases, but authorship (misappropriation, not being credited if you contributed, honorary authorship) are issues, too
- The formal process is very heavy, but you can confidentially ask RI advisors for advice/ideas, too
- ... especially if you are a junior researcher
- We aren't RI advisors but can help direct you to them.
https://www.aalto.fi/en/services/research-ethics-and-research-integrity
We've got stickers and patches, if that's your thing. You can find them from around our usual departments, or ask us.
But the story is more than that. We hope to use these to start a community around #SciComp, #ResearchSoftware, #HPC, etc: we hope the black-background ones can be used as an "Ask me anything" sign - to recognize people who want to help others. Does anyone else do something like this?
Read more: https://aaltoscicomp.github.io/blog/2023/stickers/ #RSEng
![Black background vs white background?
You may notice that for the patches some have a black background and some have a white background. Black-background means “Ask me anything about the tools of scientific computing, I am happy to help or at least point you the right direction (as much as I can)!”
[side note: we don't have black-background stickers yet, we will try to get some].](https://cdn.fosstodon.org/media_attachments/files/109/919/105/164/044/936/original/371b782487ae2a21.png "Black background vs white background?
You may notice that for the patches some have a black background and some have a white background. Black-background means “Ask me anything about the tools of scientific computing, I am happy to help or at least point you the right direction (as much as I can)!”
[side note: we don't have black-background stickers yet, we will try to get some].")
It's #CodeRefinery time! CodeRefinery is one our main #training courses for #scientists, covering all the skills such as #VersionControl (#git), #documentation, #ReproducibleResearch, #testing, and more
Free to attend to anyone in the world: 21-23 & 28-30 March, EU mornings: https://coderefinery.github.io/2023-03-21-workshop/
We recommend everyone attend this course at least once.
#RSEng #HPC #SciComp #Research #AcademicMastodon #phdstudents #ResearchSoftware #OpenScience
![[caption contains additional info]
Core CodeRefinery lessons
Also see all of our online lessons: Lesson index
These are the lessons which are included in the "standard" CodeRefinery workshop, and what we believe that every computational scientist (even those with very small amounts of computation) should know.
- Introduction to version control (day 1) [git, basics and solo work]
- Collaborative distributed version control (day 3) [How to work together with git and github]
- Documentation (day 5) [in-code, readmes, Sphinx]
- Jupyter notebooks (day 5) [basics + intermediate tricks to help people get on the same page]
- Automated testing (day 6) [how to make your code more reliable]
- Social coding and open software (day 4) [what does it take to share and reuse, and why?]
- Reproducible research (day 4) [How to make your work reusable - either by yourself in a few weeks, or others]
- Modular code development type-along (day 6) [Put it all together and see how we make a real code fit together]](https://cdn.fosstodon.org/media_attachments/files/109/908/400/951/904/222/original/52d6814c6816ae60.png "[caption contains additional info]
Core CodeRefinery lessons
Also see all of our online lessons: Lesson index
These are the lessons which are included in the \"standard\" CodeRefinery workshop, and what we believe that every computational scientist (even those with very small amounts of computation) should know.
- Introduction to version control (day 1) [git, basics and solo work]
- Collaborative distributed version control (day 3) [How to work together with git and github]
- Documentation (day 5) [in-code, readmes, Sphinx]
- Jupyter notebooks (day 5) [basics + intermediate tricks to help people get on the same page]
- Automated testing (day 6) [how to make your code more reliable]
- Social coding and open software (day 4) [what does it take to share and reuse, and why?]
- Reproducible research (day 4) [How to make your work reusable - either by yourself in a few weeks, or others]
- Modular code development type-along (day 6) [Put it all together and see how we make a real code fit together]")
Check out our crash course on minimum interactive #CommandLine #shell for scientists: https://scicomp.aalto.fi/scicomp/shell/
Are you a scientist who has to connect programs together or run some stuff on a #Linux cluster (#HPC)? You should know a few basic command line commands - maybe not everything, but you'll definitely see some around. Check out those curated tutonial based on what people really need.
On a #HPC cluster, array jobs let you parallelize things without parallelizing your code - parallelize an easy script instead. For many tasks, this is enough! The basic idea is same code, slightly different data, and #ShellScript:s connect it all. Our array tutorial explains the concepts and provides copy-and-paste examples, and works on any #Slurm cluster. #RSEng #SciComp #tip
What do Aalto Research Software Engineers - or your local RSEs - do? I made this quick chart based on a discussion last night. See anything that could be relevant to you?
Our queue is short right now so let us know. https://scicomp.aalto.fi/rse/
