How to do an assignment¶
Assignments are marimo notebooks that check your work as you go and let you submit each question from inside the notebook. This page walks through the whole loop once; after that it is the same for every assignment.
There is a practice assignment on this page. Press Assignment in the header, or Open assignment on the card at the end of the page, and work it as you read. It does not count for anything and every answer is written out in the question — it runs the same sign-in, checking and submission machinery as a real assignment, so you can watch each step described below actually happen. Do it once now: if your Dartmouth sign-in is going to give you trouble, this is a far better place to find that out than an hour before something is due.
1. Open the assignment¶
Each assignment belongs to a page and opens inside it — the graded ones to their chapter, the practice one to this page. Go to the page — the Programming chapter, say — and press Assignment in the header, or Open assignment on the card at the end of the page. The assignment opens in a drawer along the bottom of the window and starts running in your browser; the first time, it spends a few seconds installing its packages. The page stays where it is above the drawer, so you can read it while you work — try it here and these instructions stay in view. Minimize tucks the drawer down to its bar without stopping the notebook.
Your work is saved in this browser as you type. The drawer is your own copy of the assignment: close the tab, come back tomorrow, and your answers are where you left them. History in the drawer's bar shows every version of that copy — checkpoints as you work, a version each time you submit, and any you name yourself — and lets you look at one, compare it with your current copy, and restore it. Because the copy lives in this browser, work on one machine, in one browser, and press Download .py if you want a backup of your own.
If you would rather work outside the browser, press Download .py in the drawer and run:
Everything below works the same way there.
2. Work and check¶
Each question has a place for your code or your written answer, followed by a Check cell. Check runs instantly, as often as you like, and tells you whether your answer passes the visible tests. A cell that says Complete the code cell above first is simply waiting for your answer.
Checks are for learning. They are not your grade, and some questions run extra hidden tests when you submit. The practice assignment's second question is built to show this: its check runs one test, and submitting runs two more that are not in your notebook.
3. Sign in and submit¶
Press Grader · sign in in the drawer's bar, or Sign in with Dartmouth near the top of the notebook. A tab opens on Dartmouth's login page (with Duo); when it says you can close the tab, the notebook shows you as signed in. Your password never enters the notebook. One sign-in covers every assignment on this site and lasts eight hours.
Then press Submit under a question. Autograded questions come back with a score and feedback within about a minute; written questions show waiting for grade until an instructor reads them. You can submit again; the assignment says which attempt counts. Each submission is also kept in History, so you can always see what you sent.
4. See your feedback¶
Under each question there is a Feedback cell. It loads on its own when you open the notebook and shows your latest attempt for that question: the score, the maximum, any comments, and how many attempts you have made. Refresh checks again without re-running the cell.
Autograded questions fill it in within about a minute of submitting. Written questions show your attempt until an instructor reads it, then the comments appear there too — so come back to this cell rather than waiting for an email.
Your sign-in lasts eight hours. Coming back the next day, the Feedback cell will say Sign in to see your feedback with a button right there; one click and your score appears. That is normal, not an error.
Everything is also on the grader's website at grader.dartbrains.org, where you can see every assignment and every attempt after signing in.
5. Your data¶
Some chapters — the single-subject GLM is the first — start with a short section called Where the data lives and a Sign in with Dartmouth button. Everything in the chapter works without it: the datasets are public on Hugging Face, and that is where the files come from if you just run the notebook. Signing in switches the chapter to the course's own storage, which gives you two things the public copy cannot:
- The class copy of the data, including files that are only for this course.
- Storage that is yours. Anything you save is there the next time you open any notebook — a different chapter, a new molab session, your laptop — and nobody but you, the instructor and the TAs can see it.
Sign in once per notebook; after that it remembers you. In your code it looks like this:
from dartbrains_tools import storage
course = storage.course() # the class copy, read-only
private = storage.private() # yours, read-write
path = course.local_path("localizer/sub-S01/func/sub-S01_task-localizer_events.tsv")
private.put("week3/betas.pkl", betas) # pickle, .npy, .csv, .json, .nii.gz by extension
betas = private.get("week3/betas.pkl")
private.ls("week3")
local_path() gives you an ordinary file path, so nibabel, nilearn and nltools load it like any other file. Once you are signed in, molab's Files panel (the folder icon on the left) also lists the class copy and your own storage under Remote storage, where you can browse and download files.
Two things to know:
- Where you run matters. Storage works in molab and on your own machine. In the edit view on this site — the notebook running inside the page — it is not available yet; the chapter falls back to the public data and says so.
- Not at Dartmouth? Skip the button. Every chapter works the same way from the public data.
If the assignment is updated¶
When we publish a new version of an assignment, the drawer's bar says update available and offers Update…. Updating replaces your copy with the new version and keeps the old one in History, so nothing is lost either way — and you can undo it right after. Your submitted attempts are on the grader and are not affected.
If something looks wrong¶
- A red traceback in a Check cell after you have written an answer is a Python error in your code; read its last line.
- Could not reach the grader means your network is blocking the grading server; try another network or submit later.
- Nothing appears when you press Assignment: the assignment needs a laptop or desktop browser — on a phone the chapter is readable but the notebook does not run.
- Your answers are gone after clearing your browser's site data or working in a private window; that is where the copy lives. Use Download .py to keep a backup.
Assignment: Practice: how submitting works
- Q1. Write down a number
- Q2. A function, and the tests you cannot see
- Q3. A question a person reads