Lesson 5 — Packages: Install & Library; Our mosaic + BSDA Toolkit

1Objectives¶

By the end of this lesson you will be able to:

1. Explain what an R package is and why courses use them.

2. Install a package (and explain why you only do this once).

3. Load a package with library() (and explain why you do this every session).

4. Confirm the two packages this course uses, mosaic and BSDA, are working.

2What’s a package?¶

Base R (what you installed in Lesson 2) already knows how to do arithmetic, build vectors, and call functions like mean() and round() (Lesson 4). A package is an add-on bundle of extra functions, written by someone else and shared for free, that you can turn on when you need them. Think of base R as a phone’s default apps, and a package as something you download from an app store — except every package used in this course is entirely free, open, and vetted by the R community.

MATH 1209 uses exactly two packages, all semester, and no others:

You will not need any other package this semester.

3Installing a package (once per computer)¶

Installing downloads a package’s code onto your computer, one time. You do not need to do this on CSUB JupyterHub — the coordinator’s environment already has mosaic and BSDA installed for you. If you’re on your own laptop (Lesson 2, Option A), run this once, in the Console:

install.packages(c("mosaic", "BSDA"))

You will see a scroll of text while it downloads and installs — that’s normal. Once it finishes and returns you to a blank prompt, installation is done, and you never need to run install.packages() for these two again on that computer (until you get a new computer, or reinstall R from scratch).

4Loading a package (every session)¶

Installing is a one-time download; loading with library() is what actually turns a package’s functions on for your current R session — and you do this every time you start a new R session (open RStudio, or open a fresh notebook). It’s like the difference between installing an app once and opening it each time you want to use it.

library(mosaic)

Registered S3 method overwritten by 'mosaic':
  method                           from   
  fortify.SpatialPolygonsDataFrame ggplot2

The 'mosaic' package masks several functions from core packages in order to add 
additional features.  The original behavior of these functions should not be affected by this.

Attaching package: 'mosaic'

The following objects are masked from 'package:dplyr':

    count, do, tally

The following object is masked from 'package:Matrix':

    mean

The following object is masked from 'package:ggplot2':

    stat

The following objects are masked from 'package:stats':

    binom.test, cor, cor.test, cov, fivenum, IQR, median, prop.test,
    quantile, sd, t.test, var

The following objects are masked from 'package:base':

    max, mean, min, prod, range, sample, sum

That is a lot of text, and it can look alarming the first time — it is not an error. Every line starting with “The following...” is R telling you, honestly, that mosaic is replacing (R’s word: “masking”) a handful of base functions — like mean() and sum() — with its own versions that add small teaching improvements. You almost never need to think about this again; it is simply what a normal, successful library(mosaic) looks like.

library(BSDA)

Attaching package: 'BSDA'

The following object is masked from 'package:mosaic':

    CIsim

The following object is masked from 'package:mosaicData':

    Alcohol

The following object is masked from 'package:datasets':

    Orange

Same idea, shorter list. Both packages loaded successfully.

5Confirming your toolkit is on¶

A quick way to check both packages are really loaded, without scrolling back through the startup messages, is search(), which lists everything currently turned on — or, more simply, ask a yes/no question about it:

"package:mosaic" %in% search()
"package:BSDA" %in% search()

[1] TRUE
[1] TRUE

You can also check which version you have — useful if your instructor ever asks, or if a function behaves differently than this book describes:

packageVersion("mosaic")
packageVersion("BSDA")

[1] '1.9.2'
[1] '1.2.2'

(Your version numbers may be slightly newer, depending on when your computer or CSUB’s JupyterHub last updated — that’s fine.)

Finally, a real smoke test: mosaic ships with some small practice datasets built in, including one called KidsFeet (foot measurements from a group of kids — you’ll use it again in Lesson 6). If your toolkit is working, this should run with no errors and print a table of summary statistics:

favstats(~ length, data = KidsFeet)

  min Q1 median   Q3  max     mean       sd  n missing
 21.6 24   24.5 25.6 27.5 24.72308 1.317586 39       0

You’ll learn exactly what favstats() and this ~ symbol mean in Lessons 6 and 7 — for now, the fact that a real table of numbers appeared means mosaic is fully working.

6Every session, from here on: your two-line start¶

From this point in the course forward, the very first thing you type in any new R session is:

library(mosaic)
library(BSDA)

Make this a habit — most “function not found” errors after Lesson 5 trace back to forgetting this step in a fresh session.

7Summary¶

• A package adds extra functions on top of base R; this course uses exactly two, mosaic and BSDA.

• install.packages() downloads a package — once per computer. Not needed on CSUB JupyterHub.

• library() turns a package on for your current session — every time you start a new one.

• “Masked” messages after library(mosaic) are normal, not errors.

• packageVersion() and a quick smoke-test function call confirm your toolkit is really working.

8Check your understanding¶

1. What’s the difference between install.packages() and library(), and how often do you run each?

2. You restart RStudio after lunch and get Error in favstats(~length, data = KidsFeet) : could not find function "favstats". What almost certainly happened, and what’s the one-line fix?

3. Why doesn’t this course ask you to install any package besides mosaic and BSDA?