Data Modeling for Environmental Science
PBIO 6940 (3 credits)




Spring 2024; Online & Asynchronous
Instructor: Brian Beckage (Brian.Beckage@uvm.edu)
Office Hours (remote by appointment)


Course Description

Course Goals

Grading

Books & Media

Class Schedule (coming soon)

Resources

Policies


Course Description

Back to top

This course provides an introduction to data modeling using the R statistical computing language and likelihood, information theoretic, Bayesian and machine learning approaches to inference. The course includes a focus on the R language as a tool for data modeling and emphasizes examples and case studies from ecological and environmental sciences.


Class Structure. This class is an online asynchronous class, meaning that it is an online class that has no scheduled meeting time but allows students to work on class material on their own schedule. The class component includes a variety of activities, including videos, quizzes, discussions of class material and assignments.

  1. Readings. This is a reading intensive class: We will read 1-2 chapters per week from the assigned text. The chapter readings will provide the basis for online discussions in Yellowdig. There will be weekly quizzes in Blackboard over the assigned readings.

  2. Supporting videos. Each week students will watch lecture videos, where the author (Richard McElreath) of our text presents each chapter. These will reinforce our chapter readings and provide an additional perspective on the assigned readings. These video lectures will also provide the basis for class discussions in Yellowdig, e.g., please ask questions, make comments, and respond to your classmates' questions in Yellowdig. We will also watch a series of short R videos that will introduce key concepts of R.

  3. Exercises. Students will receive a set of problems that incorporate the statistical and/or R programming components of the assigned chapters for each week. Students can work on these in small groups in or outside of class. I encourage students to work together and share information and approaches as a means of learning the material, but each student should produce their own solutions (i.e., should write/type their own R code).


Course Goals

Back to top

    Students will

  1. gain proficiency in the R statistical programming language
  2. become familiar with Bayesian approaches to statistical inference.


Achieving these goals will require:


Grading

Back to top

Student grades will be based on the following components:

  • Participation in discussions through Yellowdig: 30% of course grade.
  • Quizzes on assigned readings: 15% of course grade.
  • Homework assignments as Complete/Incomplete: 30% of course grade.
  • Final exam: 25% of course grade.
  • Grading Scale:

  • Percentage

    Grade

    • 98-100
    • 93-97
    • 90-92
    • 88-89
    • 83-87
    • 80-82
    • 78-79
    • 73-77
    • 70-72
    • A+
    • A
    • A-
    • B+
    • B
    • B-
    • C+
    • C
    • C-

    Grades less than 70 are considered failing for grad courses.



    Books and Media

    Back to top

    Students will also be assigned readings from the following set of books (below). These books will be required for this class and are available in the university bookstore or as otherwise noted below.

    1. Primary text for the statistics portion of this class: McElreath, Richard. 2020. Statistical Rethinking: A Bayesian Course with Examples in R and STAN (2nd Edition). ISBN-13: 978-0367139919 Link to book.
    2. Reference texts for R:
      1. The Book of R by Tilman M. Davies. Base R perspective. A good place to start.
      2. R for Data Science by Hadley Wickham and Garrett Grolemund. R from a Tidyverse perspective.
      3. Advanced R by Hadley Wickham. Great book for understanding the language.

    We will also watch a variety of videos inside of class on both statistics and R.

    Resources

    Back to top


    We will build data models in the free, open source programming language R. Below are some additional resources for installing and using R as well as for our assigned text.


    R project for statistical computing

    RStudio IDE

    Posit Cloud

    Statistical Rethinking website

    Statistical Rethinking github

    Installation of RStan/Stan & Rethinking package

    Author's Statistical Rethinking R code

    Alt Implementation

    Tidyverse/BRMS Implementation

    Video: Install R: Windows

    Video: Install R: Mac

    Video: Install RStudio

    Video: Set Working Dir on Windows

    Video: Set Working Dir on Mac

    R search engine

    Example: Setting up Rethinking

    Instructions for installing Rethinking

    Introduction to Statistical Thought

    MCMC: Learning Metropolis-Hastings and Hamiltonian MC



    Policies

    Back to top