Errata (second edition first printing)

The page numbers indicated between parentheses correspond to the printed

Chapter 2
page 43 (44):
expresion 2.1 should be

θ ∼ Beta(α, β)
y ∼ Bern(p = θ)

or equivalently

θ ∼ Beta(α, β)
y ∼ Binom(n = 1, p = θ)

page 50 (51):
• As we can see, the result looks somewhat similar for lossf_a is θ̂ = 0.32
page 53 (54):
• As I am not a [not a] ethologist [. . . ]
page 59 (60):
• second paragraph: “I(t’)s the theoretical distribution [. . . ]”
• last paragraph, third sentence: replace “for ν > 2” with “for ν <= 2”
page 63 (64):
• second paragraph: “how a few of the predictive samples look[s] very flat.”
page 64 (65):
• first paragraph: “to the value[s] estimated”
page 65 (66):
• second paragraph: “help their kids grow[n] stronger”
page 66 (67):
• the pooled standard deviation should have a plus (+) instead of a minus
(-) sign between the group standard deviations. That is:
µ − µ1
q2
σ2 2 +σ1 2
2

1
page 73 (74):
• first paragraph: “i(t’)s also possible”
page 74 (75):
• first bullet point: “We have [have] defined two [. . . ]”
page 83 (84):
• first paragraph: “And that is, ladies, gentlem[a] (e) n”

Chapter 3
page 95 (97):
• first paragraph: “this constrain(t) is relaxed”
page 100 (102):
• second to last paragraph: “interval [-1, 1][. It does not matter about](,
regardless of) the scale of the data.”
• last paragraph: “how much y change(s)”
page 109 (111):
• code block, 4th and 3rd line from bottom: f-strings missing the “f”.
page 116 (118):
• first paragraph: “with an[d] increasing amount”
page 119 (121):
• second paragraph: “They are just (k)nobs”
page 120 (122):
• first paragraph: “Well that’s the subject of Chapter [6] (5), Model Com-
parison”
page 121 (123):
Here, β is a vector of coefficients of length m, that is, the number of (in)dependent
variables.
page 128 (130):
• first paragraph: “Using a fo(rest plot)”
• Figure 3.22 have been updated
page 133 (135):
• Figure 3.26 have been updated
page 135 (137):

2
 • In all of the examples we have seen so far, the (in)dependent variables
contribute additively to the predicted variable.
page 136 (138):
• For those cases, we may want to consider the variance as a (linear) function
of the (in)dependent variable.
Page 141 (143):
• Exercise 6: “ArviZ functions like plot_trace and plot_pair”
Page 142 (144):
• Exercise 14: This time (a)dd uncertainy to the linear plot
• Exercise 14: Exercise should reference Figure 3.17, not Figure 3.18

Chapter 4
page 154 (156):
• second paragraph: “we take advantage[s]”
page 160 (162):
• third paragraph: “and 50 virgini[n] (c) as”
page 163 (165):
• bottom paragraph: “Chapter 5, [Modeling with Linear Regression] (Model
Comparison)”
page 167 (169):
• last line: "x! is the factorial of x, that is, x! = x×(x−1)×(x−2)×· · ·×2×1.
page 170 (172):
• equation (4.25):
µxi e−µ
p(yj = ki ) = ψ
xi !
page 175 (177):
• “Extensions such (as) the ones we [we] saw”

Chapter 5
page 185 (187):
• first paragraph: “shocked or even disappoint[ing] (ed) by”
• equation 5.1: p(Tsim > Tobs |y)
Page 216 (218):

3
 • "KL diverge(nce) is useful because it is a way of measuring how close to
distributins are
page 219 (221):
• Bayes factors are problematic to use, given that they are very sensitive[ly]
to prior specification,

Chapter 6
page 230 (232):
• last paragraph: “does not necessary depend(s) on data”
page 237 (239):
• third paragraph: “what we do to turn logistic[s] regression into a”
page 238 (240):
• first paragraph: “someone already decide(d) the name”
page 239 (241):
• third paragraph: “br[ake] (eak) the stick”
page 241 (243):
• first paragraph: “represent(s) how confiden[ce] (t) we are”
page 245 (247):
• first paragraph: “This model also show(s) a less smooth”
page 247 (249):
• last paragraph: “the latent variable before doing inference, wi[n]ch may
lead to”
page 248 (250):
• second paragraph:
– “thus i[n] (t) may be convenient”
– “th[is] (ese) models can lead”
• last paragraph: “can be interpreted as continuous mixture model(s)”

Chapter 7
page 253 (255):
• last paragraph:
– “we express the first [one] function”
– “for the second [one] function”
page 255 (258):

4
 • equation 7.4: second term on the RHS should be: (x2 − x02 )2
• third paragraph: “covariance matrix looks [appears] for different inputs”
• info box: “Thus, the close[st] (r) two points are on the x axis[;] (,) the
mo[st] (re) similar we expect their values to be on the y axis”
page 259 (262):
• last paragraph: “and [this is not the exception with] Gaussian processes
(are no exception)”
page 261 (264):
• second paragraph: “x is the independent variable[s], and y is the dependent
variable[s]”
• third paragraph: “module, [Often, ]for length-scale parameters, priors
avoiding zero (often) work better”
page 267 (270):
• first paragraph: “their geographical similar[ly] (ity)”
page 271 (274):
• second paragraph: “counteracting the effect of it(’s)[ over] close neighbors”
page 279 (282):
• last paragraph: “to the time a disaster[s] happen[s] (ed)”
page 280 (283):
• first paragraph: “Let’s load [at] the data”
page 285 (288):
• last paragraph: “We may imag[e] (ine) that”

Chapter 8
page 295 (298):
• second paragraph: “[Also is] (It’s also) one of the building block(s)”
page 303 (306):
• third paragraph: “The rule to decide whether to accept or reject is known
as the Metropolis criteri[a] (on)”
page 304 (307):
• step 2, it should read Choose a new parameter value xi+1 rather than
Choose a new parameter value xi + 1.
• The uppercase Q’s should be lowercase q’s q(xi+1 | xi )
page 318 (321):

5
 • second paragraph: “samples from the noncentered model ha[s] (ve) almost
no autocorrelation”
page 319 (322):
• second paragraph: " we will need a [more] (larger) effective sample size"
page 327 (331):
• first paragraph:
– “One book that is generally refere[e]d (to) as”
– “You may also want to check (out) the book”