GCSR - Example 3.7

Attached is the Python source code for GCSR Example 3.7, the bioassay problem. The code is based on Helle Sorensen's code at KU. It samples from alpha, beta posteriors using grid approximation and displays a contour plot, and then, samples from the joint distribution and displays an X-Y plot of the sampled values. The plots duplicate Figure 3.4 (a) and (b) from GCSR.

Code

GCSR

I will publish sample code (Python), and solutions for selected questions found in Andrew Gelman's excellent book Bayesian Data Analysis 2nd Edition (GCSR). Each posting will be solution for one question, if required, will include mathematical formulas (using mathurl.com), and Python, R code. R code will be copy and paste from Gelman's own solutions PDF. We might also include code from Guilherme Rocha (another Bayesian Statistics lecturer), or from Jeff Hart (Texas A&M), or from Brian Junker (Carnegie Mellon University).

To install R on ubuntu, apt-get install r-base and r-base-dev.

You can run the R code from command line using "R -f [file]". If there are plots displayed in the R code, they will be dumped into a file called Rplots.pdf in your current directory.

Github