QERM 598: Computational Methods in Quantitative Ecology
- This course is designed to introduce students to computational methods in Ecological Modeling. Commonly used techniques in Applied Statistics will be covered, as well as some deterministic and stochastic models encountered in Mathematical Biology. While we cover the mechanics of doing various statistical analyses, the goal of the course is to develop the ability to use computational tools creatively to help visualize and grasp statistical concepts and modeled processes. There will be lots of emphasis on simulation and graphical visualization, since these are such an important tools in practice but scantily touched upon in other courses.
- Our primary tool will be R.
- Each week will consist of a lecture, an R lab and an assignment. One of the practical goals is to create a body of labs, assignments and datasets that can be used as a basis for a continuation of this course in years to come. All of these materials will be stored on this website.
- There will be a final project involving some combination of analysis and presentation using R. These will be presented on the last day of class, aka: Annual QERM Miniconference on Computational Methods in Quantitative Ecology.
Practical Details
- Location: Odegaard Undergraduate Library Computer Lab 102
- Times: Wednesday - 1:30-3:30
- Instructor: Eli Gurarie with the possible participation of guest lecturers, under the benevolent auspices of Dr. Tilmann Gneiting.
- Office Hours: all day Wednesday in Loew 328 and by appointment.
- Target Audience: First-year QERM students and other qualified students interested in Quantitative Ecology.
- Prerequisites: Statistical Inference (STAT 512), Mathematical biology (AMATH 420)
Week by Week: Homeworks, Labs, Data and Notes for QERM 598
Week 0: Basic Introduction to R
If you don't have R already, you can download it from this website. You'll need to download the executable binaries: For Windows users they are here; for MacOS users they are here. Install R and run it.
R is an open source program, which means it's free, efficient and no-frills. It also means that documentation is patchy at best and learning how to use it involves poking around the internet, blindly groping around the internal help file, and, whenever possible, stealing other people's code.
Thankfully, there is lots of expertise at QERM! To start off, download or open the file below, print it out, read it, and walk through the examples. There is a brief assignment at the end of the lab for practice.
- Lab/Homework 0 (due Wednesday, January 9).
Week 1: Distributions and the Central Limit Theorem
Obtaining quantiles and p-values from distributions. Plotting cdfs and pdfs. Generating random data. Introducing plotting functions and loops. Illustrating the Central Limit Theorem numerically.
- Lab 1: Distributions and the Central Limit Theorem (due Wednesday, January 9).
- Homework 1 (due Wednesday, January 16).
Week 2: Comparing two samples: Monte Carlo and t-tests
You will need to download the following data files to perform the lab and the homeworks:
- Ant Weight data (Single column of weight data)
- ants.csv (Comma-delimited data file)
- ants.dat (Space-delimited data file)
You will also need to read Mike Keim's notes on non-parametric tests.
Week 3: Introduction to Analysis of Variance
- Intro to ANOVA lecture
- Intro to Anova lab
- Homework 3
- Sea Lion data for homework
- Math Facts for ANOVA
Week 4: Linear Regression
- Linear Regression lecture
- In-class R lab
- Data for lecture/lab
- Homework 4
- Sea Lion inter-parturition data for homework
Week 5: Chi-squared Contingency Tables
- Lecture on contingency tables and Chi-squared tests
- In-class R lab
- Data files for lecture/lab:
- Homework
There is, by the way, a great explanation/derivation of the theory behind Chi-Squared contingency tests and other applications of the central limit theorem here.
Also, visit this link on making mosaic plots of categorical data: R tips/Contingency Tables.
Week 6: Introduction to Stochastic Modelling
- Introductory lecture on Stochastic Modeling
- In-class R codes:
- Homework
Week 7: Population Dynamics Models and MCMC
- Lecture presentation
- R code
- No homework
Week 8: Project Discussions
- a super-brief intro to times series: SuperBriefIntroToTimeSeries.r
Week 9: Fitting models and optimization
Week 10: Final project presentations
- Please visit the Final Projects page to browse through final project presentations and R code.