Errata

Chapter 4

Table 4.8, page 120. The results in this table are WRONG. However if you run the R/BUGS script in the scrbook package you do get the right answer, it just doesn't match up with the Table 4.8 results. (Thanks to Bob Dorazio for pointing this out).

Chapter 6

In the block of R code at the bottom of the page, the line:

> y2d <- apply(y3d, c(1,2), sum)

Is incorrecct for the current definition of the SCR23darray() function. The correct line should be:

> y23 <- apply(y3d, c(1,3), sum)

Thanks to Mike Meredith for pointing this out. This type of error occurs in a few places in the book but not always in the help files for the R function. This is because we standardized that the 3-d array should be nind x J x K (as in chapter 7) but failed to correct some of the code bits.

Most of the "intlik" functions have an error in how the state-space area is calculated. Version 0.23 of the scrbook package has them all fixed but very small differences will exist between the results you obtain and those shown in Ch. 6 of the book.

NOTE: 10/27/2013 --- I AM IN THE PROCESS OF UPDATING THE ENTIRE SCRBOOK PACKAGE TO BE CONSISTENT WITH THE 'nind x J x K' CONVENTION

Chapter 8:

The SCRgof() function contained in the scrbook R package had an error in it (Thanks to Mike Meredith for that). As a result, several of the Bayesian p-values reported for the wolverine data were incorrect although the general conclusions do not appear to be. The SCRgof function has been fixed (we hope) as of v. 0.12 of the scrbook package.

Chapter 9

p. 258, panel 9.1, BUGS code for multinomial SCR model. The prior for alpha0 should be something like dnorm(0,0.1) or dnorm(0,0.01) but definitely not dnorm(0,10). The example code in the simMnSCR help file of the scrbook package has been updated and will appear in the January 2014 update of the package.

Chapter 10

p. 291 equation 4 lines from bottom: should not be an "nG" in the fraction out front

p. 293 Eq. 10.5.3: this equation is slightly wrong -- it is missing the 2nd hunk of variance from the iterated conditional variance formula!

should be everything shown there + (1/pbar^2)*N*pbar*(1-pbar)

Chapter 19,

p. 518. The JAGS code for the spatial mark-resight model with known number of marks and telemetry data is incorrect. Details can be found here.