Typos: Chapter 4 - Githubissues

kenjioman commented 5 years ago

(Still present as of the 2019.02.19 version)

Section 4.1, Pg 72, last sentence of 1st paragraph of section

... of the thousand people who are standing ...
- I'm not a grammar expert so I may be wrong on this, but it "sounds better" to me without the who.
  - An alternative rearrangement that also sounds weird: "Can you predict, of the thousand people, what proportion who are standing ..." (again, removing the who makes this sound better, in my opinion).
    - An alternative to "who" that sounds ok: "Can you predict the proportion of the thousand people who are standing ...".

Section 4.3.5, Pg 88, Model definition equations (between code block 4.26 and 4.27)

Left equation says $\mu$ ~ Normal(178, 20), but the R code on the right says mu ~ dnorm(156, 10)
- For consistency with the code blocks, should be 178, 20

Section 4.4, Pg 92, Rethinking "What is regression"? box, 2nd paragraph

Not a typo, but I felt this could use a bit more detail (it wasn't clear to me how one uses the father's height vs the population of fathers, what is meant by individual measurements being assigned a common distribution -- compared to what we are doing in this chapter -- etc). Perhaps add an endnote?

Section 4.4.1.3, Pg 96, 1st full paragraph on page

?? Defining $\beta$ as ...
- Broken link, or something else?

Section 4.4.3.4, Pg 106, paragraph after code block 4.55, last sentence

... you saw in the right-hand plot in Figure 4.7.
- There are 2 right-hand plots in Figure 4.7, but the consequence of the amount of $\mu$ depending on the weight is observable in any of these plots.

Section 4.5.1, Pg 113 or 114, code block 4.69 or the model specification (math) right above it

In the prior specification of $\beta_2$ and $\beta_3$, the model specification (math) and the R-code next to it in the main text does not match what is in the code block. Specifically, the text uses Normal(0, 1), but the code block uses Normal(0, 10).

mcol commented 5 years ago

Section 4.4.1.3, p.95

"Now we have 200 pairs of \alpha and \beta values": however, in R code 4.38 N was set to 100, so it should say "100 pairs".

mcol commented 5 years ago

Section 4.4.3, p.99

"knowledgable" should be "knowledgeable"

PabloIsCoding commented 5 years ago

Section 4.3.2, p.83

It reads "It is true is this case that we have so much data that even this silly prior is harmless." Should probably be "It is true that in this case we have so much data that even this silly prior is harmless."

PabloIsCoding commented 4 years ago

Section 4.4.1, p.95-96

Inconsistent label and explanation: The linear model of height says "[likelihood]", but in the explanations of the labels on page 96 it says "probability of the data"

Side comment: "probability of the data" doesn't seem like the best way of expressing this. The probability of observing a given sample on a continuous distribution is zero. As far as I recall there is no discussion in previous chapter of the simplification of saying "probability of the data".

downey205 commented 4 years ago

Hi Richard,

I am following Pablo’s lead and posting typos in hopes that they are helpful.

-Sean

AUGUST 2 VERSION

P130 - "here I’ve chosen R for marriage rate" --> "here I’ve chosen M for marriage rate"

P211 - "In previous chapters, when I forced us to revise the priors until the prior predictive distribution produced only reasonable outcomes. " --> Delete "when"?

P218 - "To see how the WAIC calculations actually work, consider a simple regression fit with map:" --> "map"--"quap"

R code 7.28

n <- length(waic_m6.6) --> waic_m6.6 is not defined, so I think this should be n <- length(waic_m6.7)

234 - mentions the 'ensemble' function, but I do not this it has been used in the chapter.

270 - the chain very slow makes -- the chain very slowly makes

287 - the chains is much clear, --> the chains is much clearer,

SWITCHED TO READING THE SEPT 24 VERSION

P326 The problem is that deviance is part normalizing constant. Missing word?

P338 - "You can reproduce is with this code:" is -> it

P418 - "The values α ̄ = 1.4 ..." 1.4->1.5

P419 - "Now to fit the model to the simulated data, using map2stan [Ulam]. I’ll use a single long chain in this example, but keep in mind that you need to use multiple chains to check convergence to the right posterior distribution. In this case, it’s safe. But don’t get cocky." Needs editing because the code example uses 4 chains

P425 - "I’ve down this on" --> down -> done

P434 - "the same simulate chimpanzee," simulate->simulated

Dr. Sean S. Downey Associate Professor, Department of Anthropology Core Faculty, Sustainability Institute The Ohio State University 4034 Smith Laboratory 174 W. 18th Avenue Columbus OH, 43210-1106

Office: 614-688-3904 Mobile: 240-392-0220 Belize: +501-664-9012 SkypeId: seandowney1234 https://osu-cfaes.zoom.us/j/3613727623

On 29Oct, 2019, at 10:42 AM, PabloIsCoding notifications@github.com<mailto:notifications@github.com> wrote:

Section 4.4.1, p.95-96

Inconsistent label and explanation: The linear model of height says "[likelihood]", but in the explanations of the labels on page 96 it says "probability of the data"

Side comment: "probability of the data" doesn't seem like the best way of expressing this. The probability of observing a given sample on a continuous distribution is zero. As far as I recall there is no discussion in previous chapter of the simplification of saying "probability of the data".

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://github.com/rmcelreath/statrethinking_winter2019/issues/10?email_source=notifications&email_token=AMSFJQTCYLQ4TFPZAOGG4PLQRBDXDA5CNFSM4GZYI3U2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOECQX77I#issuecomment-547454973, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AMSFJQXXFQZDWW6A26HE6ZDQRBDXDANCNFSM4GZYI3UQ.

PabloIsCoding commented 4 years ago

Hi @downey205 , I noticed that you answered an email so you may not have noticed that there are open issues on github for each chapter of the book. As a matter of fact I didn't start this thread!

I have no idea if Richard cares about where we post the typos, but maybe some of yours were already spotted in some other thread (I haven't checked though). I'd suggest that you take a look here: https://github.com/rmcelreath/statrethinking_winter2019/issues

By the way, it's a great idea to post the date of download of the book just in case they change, so I'll start doing it myself!

rmcelreath / statrethinking_winter2019

Typos: Chapter 4 #10

R code 7.28