rpruim/fastr2

Foundations and Applications of Statistics: An Introduction Using R

fastr2's Issues

Questions for LittleSurvey

Is there a source for the survey questions behind the LittleSurvey data? (Would be useful for Problem 1.21.)

Example 2.2.3: "binom.R" not found by snippet

snippet("binom") runs several snippets containing the string "binom", eventually throwing an error at binom-grid.R, but never runs binom.R.

Current workaround is snippet("^binom\\.R$") .

broken link to additional materials

In the Preface to the Second Edition a link is given to additional materials:

http://www.ams.org/bookpages/amstext-28

It appears to be broken.

Textbook Exercise 4.2

It seems that the problem should be to explain why the first moment is not useful. After all:

\begin{aligned}
  \sigma^2 &\approx s^2 \frac{n}{n-1},\\
  \frac{(\theta - - \theta)^2}{12} &\approx s^2 \frac{n}{n-1},\\
  \frac{4\theta^2}{12} &\approx s^2 \frac{n}{n-1},\\
  \theta &\approx \sqrt{\frac{3n}{n-1}} s,\\
  \hat{\theta} &= \sqrt{\frac{3n}{n-1}} s.
\end{aligned}

Wilson CI or Wald CI

Hi - just want to check if the wilson.ci() function returns Wilson score method? It looks like the plus-4 adjustment to Wald (aka "normal approximation" aka "asymptotic" method).

The Wilson method is discussed by Agresti (1998) "Approximate is Better than Exact for Interval Estimation of Binomial Proportions", The American Statistician. Link here: https://math.unm.edu/~james/Agresti1998.pdf. There is also an adjustment for cases where x (number of successes)=1 or x=n-1, see last paragraph of section 4.

document Concrete data set and add reference

document Unemployment data

Suggested package "Devore6" has been removed from CRAN

I installed 'fastR' with the suggestions turned on and it issued a warning that it couldn't find 'Devore6'. It's been archived - http://cran.r-project.org/web/packages/Devore6/index.html

snippet death-penalty01: defendant, not victim

Given its role in the text (p.25), the snippet "death-penalty01" probably should be:

tally(death ~ defendant, data = DeathPenalty)

Problems with updated fastR2 and or updated ggplot2/ggformula

I updated the three packages listed above and ran into error messages with
snippet(“gf-dist”). In particular, the third of these reports an error because of a missing aesthetic in gf_dist(“binom”, params = list(size = 4, prob = 0.5),
kind = “histogram”, binwidth = 1)
Problem eliminated on my MacBook only after restoring the Library from a prior incarnation on backup. Sorry I cannot be more specific. It took a long time to figure out which Library contained the former package state.

References to geyser in package MASS

In Exercise 1.20, perhaps refer to geyser as MASS::geyser (as you do on page 14 when the data is first referenced).

library(fastR2) enables access to the help file on geyser, but geyser itself does not get on the search path. (Because fastR2 depends on mosaic which in turn imports MASS?) This could stymie students.

Missing Data in pitching2005

The BAOpp data column in the pitching2005 data set, which records the opposing hitters' batting average, is missing all of its data.

remove/fix email addresses

CRAN has detected:

fastR2: icyeh at [chu.edu.tw](http://chu.edu.tw/)
fastR2: kern at [mathcs.duq.edu](http://mathcs.duq.edu/)

Make fastR2 compatible with ggplto2 version 2.3

See https://github.com/tidyverse/ggplot2/blob/master/NEWS.md#ggplot2-2219000 for info about breaking changes.

R 3.5.1??

I just received a copy of the 2nd edition...looks great!!
I tried to install.packages("fast2R") but its not available for R 3.5.1...
So...? Will it be? Or workarounds?

Thanks