Mini Project and Term Project Policies

This page covers logistics for the computational work: two mini projects and the term project.

Mini Projects

There are two mini projects, each computational and unit-scale. Together they are worth 20% of your grade.

Dates are on the project schedule.

These may be completed individually or in pairs. See Group Work for expectations if you work in a pair.

Accepting and Submitting

  • Mini projects are distributed as GitHub repositories that you fork. When one is released, a link to the repository will be posted on Ed Discussion. Click Fork on GitHub to create a copy under your own account.
  • Clone your fork, not the original. You cannot push to the original, so look for <your-username>/mp01, not bee-envdata-cornell/mp01. If you clone the wrong one, re-clone your fork and move any edited files across — nothing is lost.
  • Open the local repository folder (use File -> Open Folder, not File -> Open) so that all paths point to the right location.
  • Run the initial notebook cell (starting with Pkg.activate()) to load the Julia environment and install packages.
  • Commit and push frequently. This helps us answer questions, and gives you a backstop if something goes wrong close to the deadline — the timestamp of your commit is evidence that you finished on time even if we cannot render your PDF until the next day.
  • Submit a PDF to Gradescope with pages tagged. Failure to tag pages results in a non-negotiable 10% deduction.
  • Due at 9:00pm Eastern. Late submissions accepted within 24 hours at a 50% penalty.

What We Are Looking For

Both mini projects are a full pass through the loop, and both are graded against the rubrics. MP1 is deliberately incomplete in ways later weeks expose — Gaussian errors that maximum entropy will question, residual independence that the time series week will test. That is not a flaw in the assignment; we will come back to it in lecture, and noticing it yourself is worth saying out loud in your writeup.

NoteMP2 Has Gaps In It, On Purpose

The record you are given for MP2 has missing months, because real gauge records do. Computing an annual maximum from a year with a three-month gap is a genuine open problem, not an exercise — is that year’s maximum real, or an artifact of what was not recorded? Missing-data mechanisms are covered in lecture during MP2’s window, which is deliberate. Say what you did about the gaps and why.

BEE 5850 Requirements

  • MP1: the Bayesian treatment is required, not an optional extension.
  • MP2: WAIC must be computed from your own posterior, not from the provided posterior sample.

Term Project

The term project is 25% of your grade, and is a question and dataset of your own choosing. It may be completed individually or in pairs.

There are three deliverables — a proposal, a check-in, and a report with presentation. Dates are on the project schedule.

The Proposal

No more than 2 pages (11 point font, 1 inch margins), not including figures or references, covering:

  • The underlying science question and any associated hypotheses;
  • The dataset(s) you will analyze;
  • Statistical model(s) for the hypotheses and your planned strategy for analysis.
ImportantThis Is Where Problem Formulation Gets Assessed

The proposal is the only place in this course where you decide what question is worth asking of what data. Everywhere else, I hand you the problem. It is graded accordingly — substantively, not as an administrative step.

The test I will apply: can you say what would count as an answer? A proposal that cannot does not yet describe a project, however interesting the dataset. Come talk to me early if you are not sure.

The Check-In

A ten-minute conversation in office hours, not a written deliverable. Pass/fail on having something to show. Its job is to catch a project going wrong while there is still time to fix it, which a conversation does better than a document.

The Report

No more than 5 pages (11 point font, 1 inch margins), not including references and figures. Detailed rubrics will be provided later in the semester.

Every report must include a section stating what a conventional canned analysis of your data would have concluded, and why yours differs. If it does not differ, say so and explain why — that is a legitimate finding, and sometimes the standard analysis is the right one. What is not acceptable is not knowing.

BEE 5850 Requirements

Students in BEE 5850 must additionally defend their specification against a named alternative specification, including a simulation study establishing the conditions under which the two diverge and what evidence would distinguish them. “I chose a GEV” is a specification; “I chose a GEV rather than a GPD, and here is the sample size at which that choice stops mattering” is a defense.

Presentations

Presentations run during the university finals period, at a date and time assigned by the Office of the University Registrar. Expect to devote approximately 6 hours to the project during the exam period.