Skip to content

2025-11-27

2025-11-24

I will have my colleague Sahar observing this day, so I will prepare are bit more explicitly.

However, first I start with items from the previous reflection:

  • [ ] Schedule at 4 hours to fix 'Hello little turtles'
  • [ ] Consider to schedule 4 hours to fix 'Graphics' instead
  • [ ] Add one of these to the morning sessions schedule
  • [x] Update goal: to be able to run a Python script on an HPC cluster, that runs a graphical library
  • [x] Make more explicit: when done, we go home. When working together, we will go home sooner
  • [x] Make goal of second half clearer: to be able to read/write a variable from/to a file
  • [ ] Merge sessions 'Variables, expressions and statements' into 1 page
  • [x] Put in non-goals: 'do complex Python'
  • [x] Refer to HPC Python course more
  • [ ] Think on how to implement this, i.e. work more independently instead of in a group
  • [ ] Think on how to implement to make exercises more collaborative, e.g. let them write their own code (as a group) based on the book?

Also:

  • [ ] Allow the faster learners to skip the simpler chapter(s) of the book

2025-11-25

Taking a look at the course content, the book seems to slow the learners down:

Lets update the schedule:

  • First session goal: run Python on HPC
  • Second session goal: run a Python script on HPC
  • Third session goal: run a Python script with something graphical on HPC

I note that this still follows all the same learning outcomes, except for using a book; but that one is used after 12:00. The things after 12:00 I delay preparing for now.

Status:

  • Introduction: needs to be checked
  • First session goal: good, has video
  • Second session goal: good, needs video
  • Third session goal: needs work, needs video

My goal is to focus on beginners. This means that if the evaluation results state that the course went too slow, I am happy. This means that if the evaluation results state that the course went too fast, I am unhappy.

To save time, I will only prepare videos for the HPC clusters that do have a registered learner:

  • 3x Unknown: I assume this is Dardel or Tetralith
  • 2x C3SE: I assume Alvis, as this is a NAISS HPC cluster
  • 5x HPC2N: 5x Kebnekaise
  • 7x LUNARC: 7x COSMOS
  • 1x COSMOS-SENS: I assume that the course coordinator has contacted this person
  • 4x NSC: 4x Tetralith
  • 1x PDC: 1x Dardel
  • 5x UPPMAX
  • 1x NSC and UPPMAX
  • 1x PDC and NSC

This means I can skip Bianca and LUMI :-)

Status update:

  • Introduction: done
  • First session goal: done
  • Second session goal: done, except for cropping the raw videos and uploading them
  • Third session goal: needs work, needs video

Time to work on the third session, on Pelle first.

Done! Now only the raw videos need to be cropped and the session before lunch is ready!

Observation questions

Here I my predictions:

  • The group may be more experienced than the challenge level of the content. As this is a course for beginners, I see that as a win

Observation of first hour

This day, I want to experiment using variable group sizes, where people are allowed to decide in a group of 1 or 2-3. How do you think that that worked out?

  • Was this setup clearly explained?
  • Did I do what I promised regarding this setup?
  • If I see that something can be improved, was this reasonable judgment that was communicated well enough?
  • Would you judge that this setup is to be preferred over random assignment of 2-3 people in each room?

Observation of fourth hour

This day, I want to experiment letting the more experienced Python users work on things at their level. How do you think that that worked out?

  • Was this setup clearly explained?
  • Did I do what I promised regarding this setup?
  • If I see that something can be improved, was this reasonable judgment that was communicated well enough?
  • Would you judge that this setup is to be preferred over all following the same basic theory?