Skip to content
 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

58 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Fitting hierarchical models in Template Model Builder

Course materials for a virtual TMB workshop at the Universidad de Concepción, January 2022.

Tenative syllabus:



Topic

Morning Topic (Theory and Mechanisms)

1.5 hour

10am-11:30 CST

5am-6:30am PST

Definitely recorded

Late Morning - Case-Studies / Projects 1 hour [C&N]

12-1pm CST

7-8am PST

Potentially recorded

Afternoon Lab

3 hours [C&N]

2-5pm CST

9-12pm PST

Live –virtual

Monday

Course introduction; Review model fitting; introduce TMB

Numerical optimization using derivatives; Finite differences vs autodiff; Refresher on Rstudio & R. [Cole]

Maximum likelihood of single parameter Normal model (Excel and R) [Cole]


Poisson model

Project time: proposals and scope

Tuesday

Building linear models in TMB

TMB/C++ syntax, workflow, factors, debugging, factors in TMB [Cole]

Maximum likelihood estimation; Fit LM in R by hand, using lm() and in TMB [Cole]

Exploring Beverton-Holt TMB model


Wednesday

GLMs

GLM theory [To do] [Noble]

Estimating uncertainty: standard errors, Delta method [Noble]

likelihood profile, bootstrapping


Poisson GLM

Project time

Thursday

Random effects in TMB

Types of mixed effects models; Marginal maximum likelihood [Noble]

Project time

VBG for pooled samples and individual ones;

Hierarchical VBG;


Friday

Beyond GLMMs

Advanced TMB features: mapping, REML, simulation, functions
sdmTMB and glmmTMB? [To do] [Cole?]

Project time

Project presentations

Some material was modified with permission from a course by Jim Thorson at https://github.com/James-Thorson/2016_Spatio-temporal_models.

Ideas for next time

Data sets

  • Just have one data set that we keep analyzing over and over
    • Has fixed effects
    • Has random effects
    • Has a slope
  • First day - Start with just a mean and normal likelihood
  • Second day - add the slope
  • Third day - add random effects and fit with LME
  • Fourth day - add GLM

Advantages of recording

  • Was good for presenting ideas and could be used in the future

Uses of channels

  • Could create multiple rooms for groups of 2
  • Use them in the labs so that 2 students can share a room
  • Professores can move among rooms and check in on the students working on the labs and projects

Topics on which to focus an entire class

  • CPUE analysis
  • glmmTMB and sdmTMB?

About

Univ Concepción (Chile), 15-19 Jan 2018. Fitting hierarchical models in TMB. Instructor: Cole Monnahan

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages