Stat 341
Computational Bayesian Statistics
Spring 2021
[calendar] [test info] [resources] [from class] [homework]
Instructor
Randall Pruim
office: North Hall 284
phone: (616) 526-7113
E-mail: rpruim@calvin.edu
Time/Location
Monday, Wednesday, Friday 1:30-2:20, in SB 010
Office Hours
Office hours will be conducted via teams. Times still TBD.
You can also send me an email or Teams message with some times that work for you.
Course Objectives
Upon completion of Stat 341, students will be able to
- Explain the logic of the Bayesian inference framework and use this framework to interpret Bayesian analyses.
- Create and describe Bayesian models appropriate for a given application.
- Use computational tools to apply Bayesian methods to data sets.
- Derive posterior distributions analytically for simple situations.
In addition, we will encounter the application of statistics in a wide range of interesting applications.
Internet Resources
I will maintain an email list of all students registered in this class and will occasionally use it to distribute information and reminders of various things pertaining to this course. If you prefer to read your email from an account other than your Calvin student account and do not have it set up to forward automatically, send me email with the email address you prefer.
Please check your email daily. You are responsible for any information communicated via email.
Teams
We will use teams for remote work. This includes students who are taking the entire course remotely and for days when we do not meet in our classroom. Emails sent to the course email list will also be posted in the General channel in Teams.
Web Pages
In addition to this home page, I will also maintain a list of web resources pertaining to this course. You are responsible for any information appearing on the course web pages. Items I have prepared and maintain online include
- a calendar of daily readings, lecture topics, exams, etc.
- a list of homework assignments and due dates.
- information about tests and exams (appearing shortly before each test date).
For quick access to these and other resources, see the navigation bar at the top of this page.
Other Important Information
See me
If you are having difficulty with any portion of the course, do not hesitate to see me. Do this as soon as possible, certainly well in advance of any deadlines (like tests) so that we can work to fix the problem.
Textbook
The required text for this course is
Statistical Rethinking by Richard McElreath, 2nd edition.
McElreath is a good and interesting writer, but I don’t partcularly like many of his coding choices. For code, I recommend you look at
This book I’m creating that will include the use of {ggformula} for graphics and other code more to my preferences.
This is based on Solomon Kurz’s Statisical Rethinking with brms, ggplot2, and the tidyverse.
I’ll be updating things as we go along, so if you look to far forward, you will see things that I’ve not yet converted.
The main difference will be my use of {ggformula} and perhaps some other R packages that help make it easier to work with Bayesian models in R. (I’ll know more once I get farther into the conversion process.)
Grading
Grading will be based on the following approximate weighting:
- Homework, etc: 15%
- Tests: 60% (but see note under Tests, below)
- Final Exam: 25%
Tests
Tests should be taken when they are scheduled. I do not generally offer make-up, alternate or late tests. Instead, if you miss one test (for any reason) or if your final exam score is better than your worst test, then your final exam score will be substituted for that test.
Technology
When we need a statistics package in this course to ease our calculations or visualizations, we will use a program called R. R is a very powerful statistical tool and programming language and is being actively developed by statisticians from all over the world who contribute to the main program, its interface, or the many add-on modules (called packages) that are available to handle specialized tasks. R is free and available for Mac, PC, or Linux. It has also been installed on the computers in the Mathematics and Statistics Computer Lab (basement of North Hall) and on some other machines around campus (CS lab and Engineering lab, I think).
The RStudio company has provided an excellent integrated development environment for R. This is the current best and easiest way to use R. Furthermore, we have set up an RStudio server on campus that allows you to run R in a web browser without any need to install the software yourself. Your session is restored each time you return, and you can work on multiple computers without losing your work when you move from one to another.
If you prefer to install R and RStudio on your own machine, the software is free and easy to install on Macs, PCs, and linux boxes. You can get R at http://cran.r-project.org/ and and RStudio at http://rstudio.org/download/desktop
You will need to install a number of additional R packages, and later in the course you will need to have Stan installed and configured to work with R.
Special Circumstances
Occasionally there are special circumstances that require that the rules and guidelines above be adjusted for a particular student. In such cases, it is the responsibility of the student to inform me of the situation as soon as possible, so that the appropriate arrangements can be made. This includes, but is not limited to, students with documented disabilities.