Introduction to Graphical Models in RevBayes (Week 9)

Topics

• Overview of Graphical Models
• RevBayes Notation for Implementing Graphical Models
• Running MCMC in RevBayes

Readings

• Höhna S, et al. 2014. Probabilistic graphical model representation in phylogenetics. 63:753-771. [REQUIRED]
• Höhna S, et al. 2016. RevBayes: Bayesian phylogenetic inference using graphical models and an interactive model-specification language. 65:726-736. [REQUIRED]
• Jordan MI. 2004. Graphical models. Statistical Science. 19:140-155. [OPTIONAL]

Homework

• Write out your own graphical model on paper for the Normal model
• Take a picture of your graphical model
• Now, write out the Rev code to specify this model
• Set up an MCMC analysis for this model using the built-in MCMC machinery in RevBayes
• Save your commands in a text file that can be run from the command line
• Run your analysis
• Open your log file in Tracer and examine the results. Do you feel confident in your estimates?
• Set up a new analysis for the Normal, but with more informative priors
• Rerun your analysis
• Examine your logs again in Tracer. How have things changed?
• Save your (1) graphical model image, (2) analysis files, (3) log files, and (4) descriptions of your results in a folder with your name. Add this folder to the homework folder. Submit a pull request to the class page with your files.