Mini Projects
There are two mini projects, each worth 10% of your grade. Both may be completed individually or in pairs.
Unlike the homework, a mini project is not a set of numbered problems. You are given a question and a required method, and you decide how to build and present the argument. What you hand in is a short report with required elements, not answers in order.
| Out | Due | Repository | |
|---|---|---|---|
| MP1 — Regression and Simulation | Mon 3/01 | Wed 3/17 | mp01 |
| MP2 — Extremes and Model Comparison | Mon 4/05 | Fri 5/07 | mp02 |
The linked pages are the full instructions, public so you can read them before the project is released. If you are enrolled, work from a fork of the repository — that is where the data and the package environment are.
General information
- Fork the project repository. When a project is released, a link will be posted on Ed Discussion. Click Fork on GitHub to create a copy under your own account, then clone your fork. This gives you the package environment and a repository you can point at if you need help with your code.
- Clone your fork, not the original. You cannot push to
bee-envdata-cornell/mp01; look for<your-username>/mp01. If you clone the wrong one, re-clone your fork and move any edited files across — nothing is lost. - Submit a PDF to Gradescope by 9:00pm Eastern on the due date, with pages tagged. Untagged submissions take a non-negotiable 10% deduction.
- Pairs: exactly one member submits, and it is the submitter’s responsibility to add the other member in Gradescope. Put both names and NetIDs on the first page. See group work for what is expected of a pair.
- You do not have to use the provided notebook, and you do not need to submit code. Your notebook or scripts should be in your repository in case there are questions.
How they are graded
Each project is graded by Prof. Srikrishnan against the standard rubrics, applied to the required elements rather than to sections. The report is graded on whether each element is present and convincing, not on whether it appears under a particular heading.
Two things worth knowing before you start:
- A model with no assumptions stated earns nothing for that element. This is the single thing the course most wants you to stop doing.
- A check that fails, honestly reported and diagnosed, earns full credit. A check that fails and is not noticed earns none. You are not being graded on having a model that works.
Full logistics are on the project policies page.