Frequently Asked Questions

Answers to common questions

Unanswered Questions

If you have a question that is not answered here, please let me know by sending me an e-mail to vlcek@beyondsimulations.com or by creating an issue on GitHub.

Contribution and Mistakes

If you have found a mistake in the course material or if you have any suggestion on how to improve the course, please let me know by sending me an e-mail to vlcek@beyondsimulations.com or by creating an issue on GitHub.

FAQs

How can I download PDF slides from the lecture?

1. First, open the lecture you want to download the slides from.
2. Then, click on the RevealJS button in the top right corner.
3. Now, click on the three stacked bars in the lower left corner.
4. Then, click on Tools in the upper left corner.
5. Now you can select PDF Export Mode and then save the slides as a PDF.

Unfortunately, this method does not work perfectly on all browsers. If you have a Chrome based browser, you should be fine.

Can I upload my downloaded .py back into the browser?

No. This is a one-way door. The browser editor cannot open an uploaded file. Your download is for two things: submitting checkpoints and keeping a safety copy. If you want to keep working in the browser, reloading the page is fine, but treat the downloaded file as your real safety net (see the next question).

Where is my notebook work saved?

Your progress lives in the browser tab: reloading the page keeps it, but closing the tab and opening the course link again later starts you from a clean notebook. There is no cross-device sync in the browser, and clearing browser data or private mode also wipes it. The only guaranteed copy is the download: press Cmd/Ctrl+S, then menu → Download → Download Python code. Without the save, the downloaded file is empty or misses your last edits. Make that download a habit at the end of every session.

How are checkpoints graded?

The live checks under the tasks of a checkpoint are provisional: they tell you you’re on track. Final grading re-runs your submitted .py against a reference test suite on my side. Upload your download to the matching Moodle assignment before the 40 minutes are up; Moodle’s clock decides what is on time, so start the download a couple of minutes early. An upload up to 5 minutes late still scores in full, one between 5 and 10 minutes late is capped at half the points, and after 10 minutes Moodle no longer takes the file. Arriving late does not extend your 40 minutes, and an uploaded file cannot be replaced. Session II has an ungraded 15-minute practice run (Checkpoint 0) before the lab, so the routine is familiar by Checkpoint 1.

What may I use during a checkpoint?

Slides, the cheatsheet and your own notebooks may be open. Other people may not help, in the room or by message. Checkpoints 1 to 3 are AI-free, and that includes the course chatbot. In Checkpoints 4 and 5 any AI tool is allowed.

How do I open and work on a lab?

The lab is done in class, right after the lecture and a short break; whatever is left, you finish at home. Each session has its own section in the sidebar, with a Tutorial page. Labs are not graded and nothing is handed in. In Sessions I to VII, click Open in browser: the notebook runs in that tab, nothing to install. A notebook is a page of cells; click into a cell and press Cmd/Ctrl+Enter (or the ▶ button) to run it, and the output appears below. The check cells under each exercise turn green when your answer is right. Before you leave: Cmd/Ctrl+S, then top-right notebook menu → Download → Download Python code.

From Session VIII the lab is a script instead of a notebook. Download the script and its data file from the tutorial page into your python-labs folder, and run it in Zed’s terminal with uv run python followed by the file name. The script checks your tasks in order and stops at the first one whose answer is missing or has the wrong form. Once every check passes, it prints a block that you show in class. Showing it is how you get feedback on your numbers, not a condition for leaving, and a lab you finish at home does not have to be shown later.

What if I miss a checkpoint?

There are no make-ups, and the arithmetic is built for one miss: four checkpoints (48 points) plus the project (40 points) still clear the 60-point pass mark. A second miss makes passing hard, so tell me early if something is going wrong.

If you are ill, hand in a medical certificate. The missed checkpoint is then left out of the count, and you pass with 60 percent of the points that remain: 52.8 of 88 after one certified miss. The certificate also excuses the session for the attendance rule.

How is attendance counted?

I record it in Moodle in every session. You need 75 percent to pass the course, which is 10 of the 13 sessions. A session missed with a medical certificate counts as excused.

Are there solutions for the labs?

I do not publish solutions. If you are stuck on an exercise or want to see how it is solved, ask me, in class or by e-mail.

How does the final project work?

Groups of three are formed in Session X, in class (a pair only where the numbers don’t come out even). A group needs at least one laptop that runs Zed, where the agent lives, so three iPads cannot form a group. A partner with an iPad can commit from a GitHub codespace (a Linux computer in a browser tab), or works on a partner’s laptop. Each group works in one private GitHub repository and hands in its link on Moodle before its own demo. A demo takes 7 minutes plus 3 minutes of questions. All demos happen in Session XIII, the order is drawn at the start of the session, and everyone in the group has to be there. Project and presentation count 40 points, judged on scope, work and presentation. Everyone in a group gets the same points, and an unfinished project can be presented too.

When can I ask questions?

In class, or by e-mail (address at the top of this page). There are no separate office hours.

How does the chatbot work?

The course AI is a chatbot that uses Mistral models and some custom code hosted on Hetzner in Germany. If you have any questions about the course, feel free to ask the AI. In Part I it explains and hints rather than handing you finished code; from Session VI on it gives full code on request. Note that the AI is not perfect and sometimes the answers might be incorrect. For more information about how the data is processed, please refer to the privacy policy.