[This article was first published on Xi'an's Og » R, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.
We are making slow progress on the normal and regression chapters as we decided to write the package at the same time we revise the chapters… Jean-Michel transformed the variable selection and model choice R codes of the regression chapter into generic functions that will fit within the package. I rewrote the section on testing, removing all connections to classical testing, 0-1 loss functions and 5% errors, towards a purely model-choice perspective. This does not modify the computational aspects but it makes the discourse more coherent and bypasses the debate about whether or not posterior probabilities are commensurable to p-values and all that. Last morning, I went running to the two creeks (locally called calanques) of Morgiou and Sugiton on a very nice costal trail.