Learn

The curriculum and the tutor. It keeps your courses, tracks what you have actually learned, and answers with a question instead of a patch.

What you can do

• Browse a course as modules and lessons, with your place kept between sessions
• Work through a course you downloaded, or one you wrote yourself in a folder
• Predict what a cell will print before the run button unlocks
• Open hints one rung at a time, from a nudge down to an explained solution
• Ask the tutor about a cell you are stuck on and get a question back
• See which concepts you have mastered, which you are learning, and which you are stuck on
• Find out when a cell gave the right answer for the wrong reason
• Switch to teaching mode to write the hints, set the gates, and tag the concepts
• Look at which concepts a whole class is stuck on

Getting started

Open the Learn window (Learn ↗ in the app bar) to see the courses in your workspace, then pick a lesson. The tutor panel is on the right. To write a lesson rather than take one, use the mode switch in the lesson header.

A lesson is an ordinary Jupyter notebook, so the same file still opens as a plain notebook anywhere else.

Requirements

• Notebook Lab, to run the lessons.
• Course Market, optionally, to download published courses. Without it you write your own lessons, which is how every course starts anyway.
• A connected model, for the tutor. Without one the lessons still work: the gates, the authored hints and the cells all run. Only the conversation and the diagnosis go quiet.

Good to know

• Progress belongs to whoever has the workspace open. There is no learner identity yet, so it is single-user.
• The cohort view shows sample data and says so. A real cohort service is the next step, and the view does not change when it arrives.
• Hints and solutions live inside the lesson file, so a determined student can read them in a text editor. That is fine for classroom use, and it is the other thing the cohort service fixes.