Pre-Installation #9
Description
It would be good to have a simstack instance on jupyter where the first 12 points of installation would be done.
Hi,
good, maybe we can discuss later.
For the .bashrc I see various options:
- there is an option to show hidden files in the file explorer,
- we provide a base image with vi/nano or
- we use a hacky workaround with the available shell commands ( I can think of at least 3 ;) right away)
Best Jörg
On 26. Feb 2026, at 16:26, Wenzel, Wolfgang (INT) <wolfgang.wenzel@kit.edu> wrote:
Hi all
Lessons from todays onboarding ( 2 students, fit in python)
- Good news: we got through to run WF from the UI, make new ones, run the new ones from the UI
- Into slides were missing (I had some in the previous lecture, but I think we need more concrete ones)
- 15 minutes to install simstack, get 1 runner up and run “add”
- 30 minutes to make a new add_and_multiply node (mostly because of lack of IDE support in jupyter)
- 15 minutes to get a pandas dataset loaded
- How do we edit the .bashrc in a jupyter notebook
- Things that are missing:
- Model inspector (which models are available)
- Dataset Creation in the UI (if there is no fitting dataset type, you cannot do anything).
- Our datasets are bad for large data, it would be super inefficient to convert a csv with 1000 rows of 3 floats into the present dataset structure.
- Lookup of and creation of “named” csv, pandas datasets (which potentially store the data on file)
Overall I am happy that it mostly worked