Chapter 4
Section 4.4.2, p 81
Figure 4.22: some values of
a,
b and
c are incorrect.
Here is the figure with the corrected values.
Section 4.3.1, p 69
Example 4.1: formula for logit is incorrect.
$$
{\rm logit}(p_{ij}) = \log \left( \frac{p_{ij}}{1-p_{ij}} \right)
$$
Section 4.4.3, p 82
Unwanted parenthesis ")" should be removed in Model 1:
$$
\begin{aligned}
P(y_{ij}=1 | \psi_i ) &= \pi_i \\\
P(y_{ij}=0 | \psi_i ) &= 1 -\pi_i
\end{aligned}
$$
Section 4.3.1, p 70 p and
1-p should be interchanged in the formula for the negative binomial distribution:
$$P(y=k ; r,p) = \frac{\Gamma(k+r)}{k!\, \Gamma(r)}(1-p)^k p^r $$
$$P(y=k ; r,p) = {k+r-1 \choose k} (1-p)^k p^r $$
Section 4.3.2, p 73 Parenthesis are missing in the formula for the variance of the generalized Poisson distribution:
$${\rm var}(y) = \frac{\lambda}{(1-\delta)^3}$$
Section 4.5.4, p 91
The correct formula for the Gompertz hazard function is
$$
h(t,\psi_i) = b_i \, \eta_i \, e^{b_i \, t}
$$
Section 4.6.5, p 101
Model 1: the Mlxtran code for
R is
R = S0 + Emax*C/(EC50+C),
Model 2: it should be
ke instead of
\ke in the Mlxtran code
Chapter 5
Section 5.2.1, p 105
Figure 5.1: the vertical color bar on the right should have different shades of grey.
Here is the right figure.
Section 5.2.2, p 107
The parameters of the distribution of
\( \log(\psi) \)
should be its mean and its variance
\( \omega^2 \)
(instead of the standard deviation
\( \omega \))
$$
\log(\psi) \sim N(\mu, \omega^2)
$$
Section 5.3.2, p 115
Figure 5.7 is wrong.
Here is the right one.
Then, the last comment concerning example 5.2 should be
"Notice that the two proposed covariate models give similar predictions for well chosen values of
\( \beta \)".
Section 5.3.4, p 119
The definition of the indicator function is missing:
"with
\( \beta_{k^*} = 0 \) and where
\( 1 \) is the indicator function
(
\( 1_A = 1 \) if
A is true, 0 otherwise).
Section 5.3.5, p 121
In the Mlxtran code
covariate1_model.txt,
the distribution of
V and
Cl should be log-normal (instead of normal):
V = {distribution=lognormal, reference=V_pop, sd=omega_V, covariate=lw70, coefficient=beta_V}
Cl = {distribution=lognormal, reference=Cl_pop, sd=omega_Cl, covariate={lw70,gender}, coefficient={beta1_Cl, beta2_Cl}}
Chapter 6
Section 6.1.3, p 145
Figures 6.2 and 6.3: the vertical color bars on the right should have different shades of grey.
Here are the right
Figure 6.2 and
Figure 6.3.
Chapter 7
Section 7.2.4
Results presented in the Table page 177 and in Figure 7.4 page 178 were obtained with
\( \gamma = (0, 0.02, 0.04, 0.1, 0.3, +\infty) \)
Section 7.3.2.2
Figure 7.7, p 184: x-labels and y-labels are inverted
(x-labels should be predictions and y-labels should be observations).
p 185,
Individual weighted residuals: a subscript
i is missing in the correlation matrix
\( \hat{R}_i \).
Section 7.3.4, p199
Results presented in the book with models M2 and M3 were obtained assuming a residual error model of the form
\(
y= f + a\,e_1+b\,f\,e_2
\),
and not of the form
\(
y = f + (a+b\,f)e
\).
Download
dataAndCodes.zip and run the Monolix projects
warfarinPK_M2a, warfarinPK_M3a, warfarinPK_M2b and
warfarinPK_M3b to see the results obtained with both residual error models.
Chapter 9
Section 9.2.3, p 240
Item 4. (MCEM): the reference (Fort and Moulines, 2003) is missing.
Section 9.5.1, p 265
Figure 9.11 (a): y-tick labels should be (14385, 14386, 14387, 14388, 14389)
Appendix B
Equation (B.2), p 297
A "-" before the sum is missing in the second line.
Appendix C
Model 3, p 309
the list of inputs is:
input = {F0, Tk0, ka, V, Cl)
elimination(cmt=1, Cl) should be replaced by
elimination(cmt=1, k=Cl/V).
Appendix D
tumor2_simulx.R, p 322 and tumor3_simulx.R, p 323
"
administration" is a deprecated syntax. It should be replaced by "
treatment".
Bibliography
Reference "O'Connell and Krause" should be "Krause and O'Connell" (my humblest apologies to the authors!)