projectmosaic/littlebooks

Materials for the MOSAIC "Teaching Statistics with R and RStudio"

---

---

Project MOSAIC Little Books

You can find PDFs of the Little Books here:

• Start Teaching Statistics Using R [view] [download]

  This book presents instructors with an overview of our approach to teaching statistics with R and an introduction to our primary R toolkit.

• A Student's Guide to R [view] [download]

  This book is organized by analysis method and demonstrates how to perform all of the statistical analyses typically covered in an Intro Stats course. It can serve a good reference for both students and faculty.

  This was formerly known as A Compendium of Commands to Teach Statistics With R but has been reworked to make it more student friendly.

• Start R in Calculus [Amazon]

This book describes the use of R in calculus based on the successful redesign of the first semester calculus course at Macalester College.

There is a spanish language translation of the Student's Guide to R available at https://github.com/jarochoeltrocho/MOSAIC-LittleBooks-Spanish (kudos to Francisco Javier Jara Ávila, https://github.com/jarochoeltrocho)

littlebooks's Issues

add updated Student Guide plus Teaching Stats to CRAN

Facet Histogram in ggformula

I'm not sure how to facet histograms so that the histograms are vertically facetted. Using ggformula, I can only get the facets to be horizontal.
Thanks

review FrontMatter.Rnw for Teaching Statistics using R

There was a considerable amount of reference to instructors as the primary audience in the CommonMaterial/FrontMatter.Rnw. I made this more generic, which probably necessitates adding a section to StartTeaching with discussion of audience. Several parts commented out include:

%Nevertheless, those who are unfamiliar with \R\ or who have never used \R\ for teaching are understandably cautious about using it with students. If you are in that category, then this book is for you. Our goal is to reveal some of what we have learned teaching with \R\ and to make teaching statistics with \R\ as rewarding and easy as possible -- for both students and faculty. We will cover both technical aspects of \R\ and \RStudio\ (e.g., how do I get \R\ to do thus and such?) as well as some perspectives on how to use computation to teach statistics. The latter will be illustrated in \R\ but would be equally applicable with other statistical software.

%Others have used \R\ in their courses, but have perhaps left the course feeling
like there must have been better ways to do this or that topic. If that
sounds more like you, then this book is for you, too. As we have been working
on this book, we have also been developing the \pkg{mosaic}

add example of horizontal=FALSE into the bwplot discussion of factors in the student guide

Randy suggests:

library(mosaic)
bwplot( birthmonth ~ length, data = KidsFeet)
bwplot( length ~ birthmonth, data = KidsFeet, horizontal = FALSE)

Add link to Start R in Calculus

Just need to determine the correct URL and add it to the README file.

Update CRAN version of LittleBooks

https://cran.r-project.org/other-docs.html

Require more requires

Hi all, I was suggesting some MOSAIC resources to people and noticed that I got an error while doing this command in RStudio (in https://github.com/ProjectMOSAIC/LittleBooks/blob/master/Starting/EarlyRExamples.Rnw):

<<births1,out.width=".8\\textwidth",fig.width=5>>=
xyplot(births ~ dayofyear, data=Births78)
@

The reason is, of course, that I didn't load the mosaicData package like is done a few lines down. So I have two suggestions:

• Naturally, make this (first) example work properly

• More importantly, make it really clear why this is necessary there to the beginner - there is just a comment in the code itself! For instance, in the sentence

  The \dataframe{Births78} data set contains the number of births in the United States for each day of 1978.

it could be also explained how this works within R.

My apologies if this is in another issue - I didn't see it, anyway. And maybe you ask for it to be loaded earlier in the text? I didn't see it, though (granted, I was skimming); then again, people might start using such a book anywhere, n'est pas? Maybe every example should have the load or require, though that might be too tedious... Looking good!

Broken link to http://www.mosaic-web.org/go/teachingRlicense.html ?

As far as I can see, the link to http://www.mosaic-web.org/go/teachingRlicense.html is broken, isn't it?

For a possible german translation of the book, we need to know exactly is on the referred webpage, as it might be part of the copyright statement in the book.

Thanks ins advance
N. Markgraf

review exercise to create a table (rbind and cbind)

On Jun 28, 2015, at 4:58 PM, Simon Halliday [email protected] wrote:

Nick,
OK. My bad! Thanks for the response. I think I got confused by it sounding like there was data, then seeing in the old doc that there had been. That said, rbind and cbind are only introduced in Ch. 5, so it might make sense to move the exercise to Ch. 5 if re-creating the table from vectors is only introduced then, rather than tallying with data as you do a lot of in Ch. 4.
Cheers & chat soon,
Simon

On Sun, Jun 28, 2015 at 4:43 PM, Nicholas Horton [email protected] wrote:
I believe that this is intended to be recreated using just the counts (not the original data). Something like this should do the trick:

foo <- rbind( a=c(w=5, x=8, y=10, z=2), b=c( 3,5,2,7) )
foo
w x y z
a 5 8 10 2
b 3 5 2 7

Randy, perhaps this item might be clarified (if it still exists)?

Nick

On Jun 28, 2015, at 3:39 PM, Simon Halliday [email protected] wrote:

Dear Nick & Randy,
I'm attending the CVC workshop in a few weeks and am trying to improve my R in anticipation of it. I've worked through the 5 College Guide and have also been working through the Mosaic 'Start Teaching with R' book. In 'Teaching' in Chapter 4 exercise 4.2 (p. 66) it references data to do with nightlight use, it but doesn't name the data. I tried searching for the data in the mosaic and mosaicData packages and couldn't find it, but then searched online to find out it was in an older version of the mosaic package (0.6-2) referenced here: http://www.inside-r.org/packages/cran/mosaic/docs/Nightlight. I haven't been able to find the data in the current versions of the packages (I have mosaic v. 0.10.9 installed via github & mosaicData v. 0.9.1-1).

I thought I'd better email in case anyone else tries working through the book with the new version of the package and struggles with that exercise.
Thanks so much & I look forward to the workshop.

Simon

Assistant Professor
Department of Economics
Smith College
107 Pierce Hall
21 West St, Northampton, MA
01063
413-585-3529

gf_fitdistr() errors

library(mosaic)
gf_dhistogram(~ cesd, data = filter(HELPrct, sex == "female"), binwidth = 7.1) %>%
gf_fitdistr(dist = dnorm)

Bug? Feature? New version issue? Thanks @rpruim for any guidance.

Warning messages:
1: In if (distname %in% c("dlnorm")) { :
the condition has length > 1 and only the first element will be used
2: In if (distname == "dnorm") { :
the condition has length > 1 and only the first element will be used
3: In if (distname == "poisson") { :
the condition has length > 1 and only the first element will be used
4: In if (distname == "dexp") { :
the condition has length > 1 and only the first element will be used
5: In if (distname == "dgeom") { :
the condition has length > 1 and only the first element will be used
6: Computation failed in stat_fitdistr():
EXPR must be a length 1 vector

sessionInfo()
R version 3.5.0 (2018-04-23)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS High Sierra 10.13.4

Matrix products: default
BLAS: /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/3.5/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats graphics grDevices utils datasets methods base

other attached packages:
[1] bindrcpp_0.2.2 mosaic_1.2.0 Matrix_1.2-14 mosaicData_0.16.0 ggformula_0.7.9002 ggplot2_2.2.1.9000
[7] lattice_0.20-35 dplyr_0.7.5

loaded via a namespace (and not attached):
[1] Rcpp_0.12.17 pillar_1.2.3 compiler_3.5.0 plyr_1.8.4 bindr_0.1.1 tools_3.5.0
[7] tibble_1.4.2 gtable_0.2.0 nlme_3.1-137 pkgconfig_2.0.1 rlang_0.2.1 psych_1.8.4
[13] parallel_3.5.0 yaml_2.1.19 ggdendro_0.1-20 gridExtra_2.3 withr_2.1.2 stringr_1.3.1
[19] grid_3.5.0 tidyselect_0.2.4 mosaicCore_0.5.0 glue_1.2.0 R6_2.2.2 foreign_0.8-70
[25] reshape2_1.4.3 purrr_0.2.5 tidyr_0.8.1 magrittr_1.5 scales_0.5.0 MASS_7.3-50
[31] splines_3.5.0 mnormt_1.5-5 assertthat_0.2.0 colorspace_1.3-2 labeling_0.3 stringi_1.2.3
[37] lazyeval_0.2.1 munsell_0.5.0 broom_0.4.4

"Two Audiences" section in "A Student's Guide to R"

The "Two Audiences" subsection in the "About These Notes" section of "A Student's Guide to R" (p. 6) states that the primary audience is instructors. Is this accurate? Since I thought I was reading a student guide, I found this claim confusing.

Update FrontMatter after mosaic 0.10.0 is on CRAN

Could also consider revising the comments about mosaicData and mosaic once mosaicData is a dependency.

update Student Guide for November printing

Move FrontMatter.Rnw to CommonMaterial folder for clarity

This will remind us when editing that file that it affects more than just the Start Teaching Document

gf_text issues

Is this a ggformula problem?

library(mosaic)
panel.labels <- function(x, y, labels = 'x',...) { panel.text(x, y, labels, cex = 0.4, ...)
}

gf_text(Murder ~ Assault, panel = panel.labels,
label = rownames(USArrests), data = USArrests, size = 2)

Make sure each little book includes indication of R version and mosaic package version used

Especially if we uncouple little book releases from package releases, it would be good if these self-documented the state of things at the time they were compiled.

does the repository include all needed files?

When I try to compile the Compendium, there are a number of errors due to missing files (authNote, language, problems, and probstat). I added these to the include folder.

vcd causes trouble for mosaic

require(vcd)
Loading required package: vcd
Loading required package: grid

Attaching package: ‘vcd’

The following object is masked from ‘package:mosaic’:

mplot

Note that vcd is in setup.R!

I've added a "detach(package:vcd)" in TwoQuantitative.Rnw in the StudentGuide, but wanted to alert others.

Caveat emptor.

How should we treat bwplot() with line at median instead of dot?

Some options:

• use pch argument
• use trellis.par.set(box.dot = list(pch = "|"))
• add this setting to theme.mosaic() (perhaps only in response to some argument)

add mention of msummary() to Start Teaching?

Improve README contents

This includes addressing #5, but also any additional links and/or text we would like to add. It's pretty minimal at the moment.

update to Student Guide (formerly Compendium)

I added a chapter introducing RStudio, as well as new examples of msummary(), derivedFactor(), and how to create a table from summary statistics. Comments and suggestions welcomed.

Compendium creates a PDF but it is a very noisy compile.

It would be nice if this compiled more cleanly.

Declare one of the Modeling versions to be the active one and discard the other

I'm guessing V2 is the one we want. But neither one compiles as is. V2 chokes on proptable() at the R level. The other one has lots of LaTeX complaints.

Need to update front matter to include new textbook resources

Per @nicholasjhorton discussion (will be done by summer students)

OneQuantitative Questions

128. Normal curve looks weird, not sure if there is a function for a normal curve
129. Normal curve looks weird.
130. Dot plot looks strange--might be because I can't get bins to work, so I used binwidth

• also, I can't find 'female' as a data set, not sure if it was filtered and stored somewhere, so I continued to use the filter function for every use.

199. I don't know how to find these functions in ggformula
200. xpnorm gives me a curve that looks like ggplot. Do I need to change this?

prefix "master" documents with "MOSAIC"

This makes them easy to spot in the sub-directories and also gives the files a reasonable name for external purposes. For example, MOSAIC-Starting.pdf

problems with mplot in StudentGuide

I'm seeing a peculiar problem with mplot when compiling the student guide: there's a blank figure.

Can you replicate this on your end?

Delete Simulation? Recommit to it?

The last commit is from over 2 years ago and labeled "toward a first clean compile", so this never was quite ready. But perhaps it is worth reviving -- have to look at see. I think this has Danny's Planet Null analogy in it, for example, and removed of the constraints of the Lock bake-off questions, we can organize this differently from the Resampling vignette.

For now, we have the Resampling vignette in the mosaic package.

extirpate (or update) fetch calls in Modeling

Pre-printing run-through of Start Teaching

p. 31 groups = wday throws an error, since wday is not a variable in Births78. Do we want to say
lubridate::wday(date)

Same problem with wday on pp. 53, 55

I have fixed these provisionally by using
groups=lubridate::wday(date, label=TRUE)

Pagination was improved to avoid long blank vertical gaps. See \enlargethispage{1in}

extirpate fetch*() calls from Starting R