/statrethinking_winter2019

Statistical Rethinking course at MPI-EVA from Dec 2018 through Feb 2019

Statistical Rethinking: A Bayesian Course Using R and Stan

Winter 2018/2019

Instructor: Richard McElreath

Location: Max Planck Institute for Evolutionary Anthropology, 4th floor seminar room

When: 10am-11am Mondays & Fridays (see calendar below)

Text: Statistical Rethinking, 1st ed and 2nd ed draft chapters https://xcelab.net/rm/statistical-rethinking/

Materials

R package

We'll use the Experimental (development) branch of my R package, because it has new features to go with the new course notes. First, you need to install rstan. Go to http://mc-stan.org/ and find the instructions for your platform. Then you can install the rethinking package:

install.packages(c("devtools","mvtnorm","loo","coda"),dependencies=TRUE)
library(devtools)
install_github("rmcelreath/rethinking",ref="Experimental")

The code is all on github: https://github.com/rmcelreath/rethinking/tree/Experimental

You can also run everything from the RStudio Cloud in your web browser. This is up-to-date with rethinking 1.80 and rstan 2.17.3 (as of 4 Dec 2018): https://rstudio.cloud/project/56157

Draft chapters

We'll use draft chapters for the 2nd edition of my book, Statistical Rethinking. If you have the password that I give out in lecture (attend class or watch the lectures), you can download them at this link: https://xcelab.net/rm/sr2/.

Homework and solutions

I will also post problem sets and solutions here. Check the folders at the top.

Lectures

I will add links to slides and videos in the calendar below. The slides will be posted here: https://speakerdeck.com/rmcelreath. The videos will be here: https://www.youtube.com/playlist?list=PLDcUM9US4XdNM4Edgs7weiyIguLSToZRI.

Calendar & Topical Outline

Week 1:
Dec 3: The Golem of Prague <slides> <video>
Dec 7: Garden of Forking Data <slides> <video>

Dec 10: Geocentric Models <slides> <video>
Dec 14: Wiggly Orbits <slides> <video>

Jan 7: Spurious Waffles <slides> <video>
Jan 11: Haunted DAG [slides] [video]

Jan 14/18: Ulysses Compass (Overfitting and information criteria)

Jan 21/25: Conditional Manatees (Interactions)

Jan 28/Feb 1: Markov chain Monte Carlo / Maximum entropy & GLMs

Feb 4/8: Generalized Linear Models

Feb 11/15: Models with Memory (Multilevel models 1)

Feb 18/22: Adventures in Covariance (Multilevel models 2)

Feb 25/Mar 1: Missing Values & Measurement Error / Generalized Linear Madness (models beyond the GLM(M) framework)