R bootcamp, Module 1: Basics
August 2022, UC Berkeley
Chris Paciorek
R as a calculator
## [1] 4
2 * pi # multiply by a constant
## [1] 6.283185
7 + runif(1) # add a random number
## [1] 7.71528
## [1] 81
## [1] 16
## [1] 2.302585
## [1] 2
## [1] 11
## [1] 1
# scientific notation
5000000000 * 1000
## [1] 5e+12
## [1] 5e+12
Think of a mathematical operation you need - can you guess how to do
it in R?
Side note to presenter: turn off R Notebook inline view via RStudio
-> Preferences -> R Markdown -> Show output inline …
POLL
1A:
Question 1: How do I calculate the cosine of 2 pi?
- cosine(2pi)
- cosine(2*pi)
- cos(2 * pi)
- cos(2 x pi)
- cos(2*pi)
- cos(2 * 3.14159)
- cos[2*pi]
Question 2: What happens if you do this?
Assigning values to R objects
A key action in R is to store values in the form of R objects, and to
examine the value of R objects.
## [1] 3
## [1] 3
Val <- 7 # case-sensitive!
print(val)
## [1] 3
## [1] 7
We can work with (and store) sequences and repetitions
## [1] 1 2 3 4 5 6
years <- seq(1952, 2007, by = 5)
years
## [1] 1952 1957 1962 1967 1972 1977 1982 1987 1992 1997 2002 2007
## [1] 12
## This is a comment: here is an example of non-numeric data
country <- rep("Afghanistan", 12)
country
## [1] "Afghanistan" "Afghanistan" "Afghanistan" "Afghanistan" "Afghanistan"
## [6] "Afghanistan" "Afghanistan" "Afghanistan" "Afghanistan" "Afghanistan"
## [11] "Afghanistan" "Afghanistan"
If we don’t assign the output of a command to an object, we haven’t
saved it for later use.
R gives us a lot of flexibility (within certain rules) for assigning
to (parts of) objects from (parts of) other objects. We’ll see this
through the bootcamp.
Vectors in R
The most basic form of an R object is a vector. The various objects
mySeq, years, country are all
vectors.
In fact, individual (scalar) values are vectors of length one, so
val and Val are also vectors.
We can concatenate values into a vector with c().
## numeric vector
nums <- c(1.1, 3, -5.7)
devs <- rnorm(5)
devs
## [1] 0.19323788 0.08528635 -1.90114512 1.02340237 0.21099795
## integer vector
ints <- c(1L, 5L, -3L) # force storage as integer not decimal number
## 'L' is for 'long integer' (historical)
nObs <- 1000
mySample <- sample(1:1000, 100, replace = TRUE)
## character vector
chars <- c('hi', 'hallo', "mother's", 'father\'s',
"She said, 'hi'", "He said, \"hi\"" )
chars
## [1] "hi" "hallo" "mother's" "father's"
## [5] "She said, 'hi'" "He said, \"hi\""
## hi
## hallo
## mother's
## father's
## She said, 'hi'
## He said, "hi"
## logical vector
bools <- c(TRUE, FALSE, TRUE)
bools
## [1] TRUE FALSE TRUE
This is not valid syntax in R. Let’s try it and see what happens.
nums <- (1.1, 3, -5.7)
nums <- [1.1, 3, -5.7]
Working with indices and subsets
## [1] 1952 1957 1962 1967 1972 1977 1982 1987 1992 1997 2002 2007
## [1] 1962
## [1] 1962 1967 1972
## [1] 1952 1962 1977
## [1] 1957 1967 1972 1982 1987 1992 1997 2002 2007
years[c(rep(TRUE, 3), rep(FALSE, 2), TRUE, rep(FALSE, 6))]
## [1] 1952 1957 1962 1977
## If you haven't installed the gapminder package, do this first:
install.packages('gapminder')
## Installing package into '/accounts/vis/paciorek/R/x86_64-pc-linux-gnu-library/4.2'
## (as 'lib' is unspecified)
## create a simple vector from the Gapminder dataset
library(gapminder)
gdp <- gapminder$gdpPercap
gdp[1:10]
## [1] 779.4453 820.8530 853.1007 836.1971 739.9811 786.1134 978.0114 852.3959
## [9] 649.3414 635.3414
We can substitute values into vectors
gdp[4] <- 822.9711
vals <- rnorm(100)
vals[3:4] <- c(7.5, 2.4)
vals[1:2] <- 0 # this uses 'recycling' - more in Module 4
Working with indices and subsets: quick quiz
POLL
1B: Which of these will work to extract a subset of a vector?
Assume the vector is created like this:
vals <- rnorm(4)
(respond at https://pollev.com/chrispaciorek428)
- vals[3]
- vals[2,3]
- vals[c(2,3)]
- vals(2,3)
- vals[c(FALSE, TRUE, TRUE, FALSE)]
- vals[c(f,t,t,f)]
- vals(3)
Vectorized calculations and comparisons
At the core of R is the idea of doing calculations on entire
vectors.
gdpTotal <- gapminder$gdpPercap * gapminder$pop
tmp <- gdpTotal[gapminder$year == "2007"] # let's pick apart what is happening here
gdpSubset <- tmp[1:20]
gdpSubset >= 1e6 # Dr. Evil's version of "a lot"
## [1] TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE TRUE
## [16] TRUE TRUE TRUE TRUE TRUE
Using functions in R
- Functions generally take arguments, some of which are often
optional:
## [1] 59.47444
mean(gapminder$lifeExp, trim = 0.1)
## [1] 59.91524
- We can embed function calls:
- We can (usually) see the code of a function:
## function (formula, data, subset, weights, na.action, method = "qr",
## model = TRUE, x = FALSE, y = FALSE, qr = TRUE, singular.ok = TRUE,
## contrasts = NULL, offset, ...)
## {
## ret.x <- x
## ret.y <- y
## cl <- match.call()
## mf <- match.call(expand.dots = FALSE)
## m <- match(c("formula", "data", "subset", "weights", "na.action",
## "offset"), names(mf), 0L)
## mf <- mf[c(1L, m)]
## mf$drop.unused.levels <- TRUE
## mf[[1L]] <- quote(stats::model.frame)
## mf <- eval(mf, parent.frame())
## if (method == "model.frame")
## return(mf)
## else if (method != "qr")
## warning(gettextf("method = '%s' is not supported. Using 'qr'",
## method), domain = NA)
## mt <- attr(mf, "terms")
## y <- model.response(mf, "numeric")
## w <- as.vector(model.weights(mf))
## if (!is.null(w) && !is.numeric(w))
## stop("'weights' must be a numeric vector")
## offset <- model.offset(mf)
## mlm <- is.matrix(y)
## ny <- if (mlm)
## nrow(y)
## else length(y)
## if (!is.null(offset)) {
## if (!mlm)
## offset <- as.vector(offset)
## if (NROW(offset) != ny)
## stop(gettextf("number of offsets is %d, should equal %d (number of observations)",
## NROW(offset), ny), domain = NA)
## }
## if (is.empty.model(mt)) {
## x <- NULL
## z <- list(coefficients = if (mlm) matrix(NA_real_, 0,
## ncol(y)) else numeric(), residuals = y, fitted.values = 0 *
## y, weights = w, rank = 0L, df.residual = if (!is.null(w)) sum(w !=
## 0) else ny)
## if (!is.null(offset)) {
## z$fitted.values <- offset
## z$residuals <- y - offset
## }
## }
## else {
## x <- model.matrix(mt, mf, contrasts)
## z <- if (is.null(w))
## lm.fit(x, y, offset = offset, singular.ok = singular.ok,
## ...)
## else lm.wfit(x, y, w, offset = offset, singular.ok = singular.ok,
## ...)
## }
## class(z) <- c(if (mlm) "mlm", "lm")
## z$na.action <- attr(mf, "na.action")
## z$offset <- offset
## z$contrasts <- attr(x, "contrasts")
## z$xlevels <- .getXlevels(mt, mf)
## z$call <- cl
## z$terms <- mt
## if (model)
## z$model <- mf
## if (ret.x)
## z$x <- x
## if (ret.y)
## z$y <- y
## if (!qr)
## z$qr <- NULL
## z
## }
## <bytecode: 0x5642a19f75f0>
## <environment: namespace:stats>
mean # We'll investigate what 'UseMethod' does in Module 10
## function (x, ...)
## UseMethod("mean")
## <bytecode: 0x5642a1ca3db0>
## <environment: namespace:base>
Getting help about a function
To get information about a function you know exists, use
help or ?, e.g., ?lm.
Basic kinds of R objects
We’ve seen vectors of various types (numeric (i.e., decimal/floating
point/double), integer, boolean, character).
All items in a single vector must be of the same type.
But vectors are not the only kinds of R objects.
Data frames
Collections of columns of potentially different types.
gapminder is actually an enhanced kind of data frame called
a ‘tibble’ (more in Module 6).
## # A tibble: 6 × 6
## country continent year lifeExp pop gdpPercap
## <fct> <fct> <int> <dbl> <int> <dbl>
## 1 Afghanistan Asia 1952 28.8 8425333 779.
## 2 Afghanistan Asia 1957 30.3 9240934 821.
## 3 Afghanistan Asia 1962 32.0 10267083 853.
## 4 Afghanistan Asia 1967 34.0 11537966 836.
## 5 Afghanistan Asia 1972 36.1 13079460 740.
## 6 Afghanistan Asia 1977 38.4 14880372 786.
## [1] 28.801 30.332 31.997 34.020 36.088 38.438 39.854 40.822 41.674 41.763
## [1] 1704 6
## [1] 1704
## [1] "country" "continent" "year" "lifeExp" "pop" "gdpPercap"
## [1] "tbl_df" "tbl" "data.frame"
## [1] FALSE
## [1] "integer"
## [1] "numeric"
## [1] "factor"
Lists
Collections of disparate or complicated objects
myList <- list(stuff = 3, mat = matrix(1:4, nrow = 2),
moreStuff = c("china", "japan"), list(5, "bear"))
myList
## $stuff
## [1] 3
##
## $mat
## [,1] [,2]
## [1,] 1 3
## [2,] 2 4
##
## $moreStuff
## [1] "china" "japan"
##
## [[4]]
## [[4]][[1]]
## [1] 5
##
## [[4]][[2]]
## [1] "bear"
myList[[3]] # result is not (usually) a list (unless you have nested lists)
## [1] "china" "japan"
identical(myList[[3]], myList$moreStuff)
## [1] TRUE
## [1] "japan"
## [1] "stuff" "mat" "moreStuff" ""
Lists can be used as vectors of complicated objects. E.g., suppose
you have a linear regression for each value of a stratifying variable.
You could have a list of regression fits. Each regression fit will
itself be a list, so you’ll have a list of lists.
A bit on plotting
R has several different plotting systems:
- base graphics
- lattice graphics
- ggplot2 (a very popular add-on package)
We’ll see a little bit of base graphics here and then
ggplot2 tomorrow in Module 7.
plot(gapminder$lifeExp ~ gapminder$gdpPercap)
boxplot(gapminder$lifeExp ~ gapminder$year)
Graphics options
Check out help(par) for various graphics settings; these are set via
par() or within the specific graphics command (some can be
set in either place), e.g.,
par(pch = 16)
plot(gapminder$lifeExp ~ gapminder$gdpPercap, xlab = 'GDP per capita (dollars)',
ylab = 'life expectancy (years)', log = 'x')
Breakout
In general, your answers to any questions should involve writing code
to manipulate objects. For example, if I ask you to find the maximum
life expectancy, do not scan through all the values and find it by eye.
Use R to do the calculations and print results.
Basics
Create a variable called ‘x’ that contains the mean life
expectancy.
Use functions in R to round ‘x’ to two decimal places and to two
significant digits.
Create a vector of GDP per capita in units of Euros rather than
dollars.
Create a boolean (TRUE/FALSE) vector indicating whether total
country GDP is greater than 1 trillion dollars. When entering 1
trillion, use R’s scientific notation.
Using the ideas
Use the boolean vector from problem 4 to produce a new vector
containing the per capita GDP only from the biggest economies.
How does R process the following subset operations in the first
line of code? Explain the individual steps that R carries out:
vals[vals < 0] <- 0
vals[1:8]
Plot life expectancy against gdpPercap with gdpPercap values
greater than 40000 set to 40000.
Make a histogram of the life expectancy values for the year 2007.
Explore the effect of changing the number of bins in the histogram using
the ‘breaks’ argument.
Subset the data to those for the year 2007 (there is a way to do
this all at once, but using what we’ve seen already, you can pull out
and subset the individual columns you need). Plot life expectancy
against GDP per capita. Add a title to the plot. Now plot so that data
for Asia are in one color and those for all other countries are in
another color and those for all other continents are in another, using
the ‘col’ argument. Hint: ‘col’ can take a vector of colors such as
“black”,“red”,“black”, …
Advanced
- Consider the following regression model. Figure out how to extract
the
and residual standard error
and store in new R variables.
mod <- lm(lifeExp ~ log(gdpPercap), data = gapminder)
summ <- summary(mod)
- Take your plot from problem 9. Now modify the size of the points.
Add a legend. Rotate the numbers on the y-axis so they are printed
horizontally. Recall that
help(par) will provide a lot of
information.