Ecological models and data
ZOO 6927, Winter/Spring 2000
PLEASE NOTE THAT THESE ARE NOTES FROM A PREVIOUS YEAR. The material will be similar to the planned Fall 2001 course but the details will (of course) be different.

Section 3770
3 credits

Meeting time: 11:45-12:35 pm (MWF)

Meeting place: Psych 287 (at least initially)

Instructor: Ben Bolker
bolker@zoology.ufl.edu
Executive summary: This course will focus on methods for fitting relatively realistic ecological models to lab or field data. The main source will be The Ecological Detective by Mangel and Hilborn (Princeton University Press, 1997), and the main activity will be learning to fit models to data using a statistical package and programming language called S-PLUS (or its clone, called R).

Course summary

Syllabus

Readings

General information on R

Tutorials and problem sets

Errata for the Ecological Detective: HTML or PDF

Quiz answers

Relationships between distributions (HTML, PDF)

Notes on the math from ED:

pp. 74-75 (HTML, PDF)
pp. 84-85 (HTML, PDF)

preliminary notes on basic calculus and probability

Notes on lectures:

Lecture 1: overview of ecological detection
Lecture 2: philosophy of stats; intro to likelihood (nest predation example)
Lecture 3: a few details on R (short; spent most of the class discussing philosophy of stats.)
Lecture 4 (and some of 5): a bestiary of useful functions; how to deal with functions; Taylor series
lecture 5: a little bit about Bayes' rule
lecture 6: averaging, Jensen's inequality, and the "delta method"
lecture 7: ruminations on randomness; a bestiary of useful distributions, emphasizing their origins and relationships
lecture 8: ruminations on modelling; some examples of ecological models
lecture 9 (notes not av. yet): statistical power and simulations
lecture 10: the differences between raw data and frequency tables; basics of optimization
lecture 11: two-dimensional optimization and confidence regions
lecture 12: alternative optimization approaches
lecture 15: confidence limits on parameters
lecture 16: some details on sums of squares and process vs. measurement error
lecture n-1: generalized linear models

If you have more questions you can send e-mail, come find me (318 Carr, office hours Tues-Thurs afternoons [before Zoo seminars and faculty meetings]). You can also look at some old lecture notes from an earlier version of this course.

Prerequisites

Statistics, calculus, and/or permission of instructor (advanced undergraduates are encouraged to take the class, with permission).

The course summary has a bit more information on the level of preparation I will be expecting. You can also get ahold of a copy of The Ecological Detective by Mangel and Hilborn, and take a look at it (especially Chapter 3). If you don't understand every bit, that's fine; if it all looks like complete gibberish you might consider taking a basic stats. course instead.

Last modified 21 Feb 2000 (lecture 12, new R code)
List of recent changes

Ecological models and data ZOO 6927, Winter/Spring 2000 PLEASE NOTE THAT THESE ARE NOTES FROM A PREVIOUS YEAR. The material will be similar to the planned Fall 2001 course but the details will (of course) be different.

Prerequisites

Ecological models and data
ZOO 6927, Winter/Spring 2000
PLEASE NOTE THAT THESE ARE NOTES FROM A PREVIOUS YEAR. The material will be similar to the planned Fall 2001 course but the details will (of course) be different.