search

StatsWithR / statsr

Companion package for the Coursera Statistics with R specialization
Other
71 stars 83 forks source link

MCMC to generate samples under the Cauchy prior using Jeffrey's Zellner-Siow Cauchy prior #13

Closed KenneySnell closed 3 years ago

KenneySnell commented 4 years ago

Version .1 of the StatsR package has options for call function bayes_inference(y=tthm,data = tapwater, statistic = "mean", mu_0 = 35, rscale = 1, prior = "JZS", type = "ci", method = "sim")

with the above parameters.

Using the latest version the following parameters are not available a) mu_0 = 35 b) prior = "JZS", c) rscale = 1, and d) method = "sim"

When I call the updated version of Statsr (2.1), I get approximately the same answer, but the following outputs if saved in an output variable (MCMC) are missing

a) mu b) samples (mu, sig2, delta, g)

Can you explain what happened to the Jeffrey's Zellner-Siow Cauchy and how to get this called in the latest version of Stats?

Thank you,

Kenney

jaekookang commented 3 years ago

I had the same problem as "unused arguments" using the bayes_inference function when following examples in chapter 5 of An Introduction to Bayesian Thinking book.

KenneySnell commented 3 years ago

Jaffrey

You have to go back to a previous version of R to get the function to work.

Regards

Kenney

Sent from my iPhone

On Nov 28, 2020, at 9:47 PM, Jaekoo Kang notifications@github.com wrote:

 I had the same problem as "unused arguments" using the bayes_inference function when following examples in chapter 5 of An Introduction to Bayesian Thinking book.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

merliseclyde commented 3 years ago

fixed in version 0.3.0 (see issue #15 ). will be on CRAN shortly