Issue in Gibbs sampling for Ising models (page 501), version 2023-08-15

[davideC00]

Hi,

there is an issue in the highlighted phrase:

If $x_i \in \{0,1\}$, the conditional distribution in the equation 12.27 is $σ(2J(a_i-d_i))$. I think the problem is that the definition $\eta_i=x_i(a_i-d_i)$ works only if $x_i\in \{-1,1\}$ but it is has been used also for the case $x_i \in \{0,1\}$

[davideC00]

Also here:

It should be a plus sign

[murphyk]

thanks! Fixed.