In this file we give important typos that were corrected after the lecture. 
The changes are ordered by chapter and then date. 
(Thanks to all pointing these and other grammatical ones out!) 


Chapter 8:
- Figure 8.9: IGMRF -> GMRF 

- P145: Link to INLA download had to be updated

Chapter 7:
- Section 7.4.1: page 120: the hyperparameters are wrongly given in the text (exchange t and y).

- R-Code 7.3: The prior of kappa_t is wrongly specified. It should read:
  ```postshape <- priorshape + c(n/2, (n-2)/2)```
  As a result, the top row of Figure 7.4 slightly changes: worse mixing and higher dependency than with the previously wrongly specified prior.

- R-Code 7.2: (no error but) loop was slightly rewritten to be better aligned with text.  


Chapter 6:
- Table 6.1: all X,x should be replaced with Y,y


Chapter 5:
- R-Code 5.14: fitW2 <- tm2$par[2]*W1 + tm2$par[3]*W2   (no minus!)

- R-Code 5.10: we are minimizing the neg-2 loglikelihood. Hence, we have to divide
   the Hessian with 2. The uncertainties at the end of the code are now much more
   similar.

- equation (5.6) page 74: it should read
     \hat\y = \1\hat\beta + \hat\lambda\W(\y-\1\hat\beta)


Chapter 2:
- (2.37):  Y_{n+1} should read Y_{n+m}

- a Comment about the automatically included mean has been added in Example 2.4.


Chapter 1:
- Property 1.6: formula for the determinant should read:  det(A)=det(A_11)det(C)