School of Economics and Management
Beihang University
http://yanfei.site

## Objectives

• How to install and use R
• R basics
• R coding style
• Some nice R tips

## Packages required in this lecture

install.packages(c("forecast", "sos", "formatR"))

### R packages

• Standard R comes with some standard packages installed for basic data management, analysis, and graphical tools.
• More than 10,000 packages available on CRAN! See http://cran.r-project.org.
• install.packages('forecast') to install an package called ‘forecast’.
• library(forecast) before using the package.

## Getting and Installing R

• Totally free!
• Download R on its official website.
• A new major version of R comes out once a year, and there are 2-3 minor releases each year.

## R User Interface

• Direct R
• Rstudio
• One of the most popular ways to run R.
• Free, open-source integrated development environment (IDE) for R.
• Many additional fantastic features.
• Updated a couple of times a year.
• Command line in Linux and Unix.

## Editor

• What editor do you usually use?
• Use a good text editor such as vim, sublime text, text wrangler, notepad, etc
• With syntax highlighting, otherwise, it’s hard to detect errors
• Or use an Integrated Development Environment (IDE) like RStudio

## Use an IDE: Rstudio

• Syntax highlighting
• Able to evaluate R code + by line + by selection + entire file
• Command auto-completion

## Basic Operations

## simple maths
1 + 2 + 3
1 + 2 * 3

## assign a value to a variable
x <- 1
y <- 2
z <- c(x,y)
z

## function examples
exp(1)
cos(3.141593)
log2(1)

## Vectors

• Numerical vectors
• Logical vectors
• Character vectors
• Length of a vector
• Vector calculations
• Extract some elements of a vector
## vectors
c(0, 1, 1, 2, 3, 5, 8)
1:10
seq(1, 9, 2)
rep(1, 10)
length(rep(1, 10))

## character vectors
c("Hello world", "Hello R interpreter")

## vector calculation
c(1, 2, 3, 4) + c(10, 20, 30, 40)
c(1, 2, 3, 4) + 1

## you can refer to elements by location in a vector
b <- c(1,2,3,4,5,6,7,8,9,10,11,12)
length(b)
b
b[7]
b[1:6]
b[c(1,6,11)]
b > 5
b[b > 5]

## Matrix

• Create a matrix: matrix()
• Dimension of a matrix: dim()
• Transpose of a matrix: t()
• Extract elements from a matrix.
• Combine two or more matrices: rbind(), cbind()
## create a matrix
m <- matrix(c(1:6), 2, 3)
n <- matrix(c(8:13), 2, 3)
dim(m)
t(m)
m[1, 2]
m[1, ]
cbind(m, n)
rbind(m, n)

## List

• Special data structure that matrix could not handle.
• Data length are not the same.
• Data type are not the same.
• Create a list: list()
• Extract elements of a list: [[]] or \$
l <- list(a = c(1, 2), b = 'apple')

## Data frame

• data.frame(): tightly coupled collections of variables which share many of the properties of matrices and of lists, used as the fundamental data structure by most of R’s modeling software.
• In most cases, the operation with a data frame is similar to matrix operation.
L3 <- LETTERS[1:3]
fac <- sample(L3, 10, replace = TRUE)
d <- data.frame(x = 1, y = 1:10, fac = fac)

## Functions

• Create a function
f <- function(x, y) {
z <- c(x + 1, y + 1)
return(z)
}
f(1, 2)
• Load the function: source()
• Execute your function
• When should you write a function?

## The if condition

• Syntax
if (condition){
do something
} else {
do something
}
• Example
x <- 0
if (x > 1) {
print('x is larger than 1')
} else {
print('x is not larger than 1')
}

## Loops

• Example
x <- 1:10
for(i in x) {
print(i^2)
}

1. Write a function MySummary() where the input argument is x can be any vector and the output is a list that contains the basic summary (mean, variance, length, max and minimum values) of the vector you have supplied to the function.
2. Test your function with some vectors (that you make up by yourself).

## File names

• File names should end in .R and, of course, be meaningful.
• GOOD: predict_ad_revenue.R
• BAD: foo.R

## Choose the names of variables and functions carefully

### What we should do

• The preferred form for variable names is all lower case letters and words separated with dots (variable.name), but variableName is also accepted. Generally, variable names should be nouns.
• GOOD: avg.clicks
• OK: avgClicks
• Function names have initial capital letters and no dots. Function names are mostly verbs.
• GOOD: CalculateAvgClicks
• BAD: calculate_avg_clicks , calculateAvgClicks
• Choose a consistent naming style

### What we should not do

• Don’t use underscores (_) or hyphens (-).
• Avoid using names of existing functions and variables like mean, median etc.
• Avoid using meaningless names like a, b, c, …, aa, bb, cc, …

## White Spaces

• around operators (=, +, -, <-, etc)
• put a space after a comma, and never before
x <- c(1:10)
x.average<-mean(x,na.rm=TRUE)

$$\Rightarrow$$

x.average <- mean(x, na.rm = TRUE)
• split long lines at meaningful places

Don’t be afraid of splitting one long line into individual pieces!

n <- matrix(sample(1:100, 9), nrow = 3, ncol = 3, byrow = TRUE)

$$\Rightarrow$$

n <- matrix(sample(1:100, 9),
nrow = 3,
ncol = 3,
byrow = TRUE)

## Curly braces

• An opening curly brace should never go on its own line and should always be followed by a new line.
• A closing curly brace should always go on its own line, unless it’s followed by else.
• Always begin the body of a block on a new line.
• Always indent the code inside curly braces.
if (y < 0) {print("y is negative")}

$$\Rightarrow$$

if (y < 0) {
print("y is negative")
}

## Indenting

• Use two spaces
• Can help in detecting errors in your code because it can expose lack of symmetry
• Reindenting using RStudio
if (y < 0) {
print("y is negative")
}

$$\Rightarrow$$

if (y < 0) {
print("y is negative")
}

## Make your code tidy in a second!

• Reformat and reindent in Rstudio.
• formatR package in R. You can even make a folder of .R files tidy using tidy_dir().

• Use blank lines to separate blocks of code and comments to say what the block does. Remember that in a few months, you may not follow your own code any better than a stranger.
x <- c(1:10)
x.mean = mean(x)
x.var = var(x)

$$\Rightarrow$$

## =============================================
## Title
## Author: Yanfei Kang
## Date: Mar 23, 2017
## Description: your purpose
## =============================================

x <- c(1:10)

## getting the mean of x
x.mean = mean(x)

## getting the variance of x
x.var = var(x)

## Function Documentation

• Functions should contain a comments section immediately below the function definition line.
• These comments should include
• a one-sentence description of the function
• a list of the function’s arguments, denoted by Args:, with a description of each (including the data type)
• a description of the return value, denoted by Returns:.
• The comments should be descriptive enough that a caller can use the function without reading any of the function’s code.
CalculateSampleCovariance <- function(x, y, verbose = TRUE) {
## Computes the sample covariance between two vectors.
#
## Args:
##   x: One of two vectors whose sample covariance is to be calculated.
##   y: The other vector. x and y must have the same length, greater than one,
##      with no missing values.
##   verbose: If TRUE, prints sample covariance; if not, not. Default is TRUE.
#
## Returns:
##   The sample covariance between x and y.
n <- length(x)
## Error handling
if (n <= 1 || n != length(y)) {
stop("Arguments x and y have different lengths: ",
length(x), " and ", length(y), ".")
}
if (TRUE %in% is.na(x) || TRUE %in% is.na(y)) {
stop(" Arguments x and y must not have missing values.")
}
covariance <- var(x, y)
if (verbose)
cat("Covariance = ", round(covariance, 4), ".\n", sep = "")
return(covariance)
}

## How to find the right function

• Functions in installed packages
library(forecast)
help.search("auto.arima")
??auto.arima
• Functions in other CRAN packages
library(sos)
findFn("arima")
RSiteSearch("arima")

## Digging into functions

• Type ?sort for the usage of the function sort().
• Typing the name of a function gives its definition.
• Type forecast:::estmodel for hidden functions.
• Download the tar.gz file from CRAN if you want to see any underlying C or Fortran code.

## Organize your R projects

### Basic idea

• Every paper, book or scientific report is a ‘project’.
• Every project has its own folder and R workspace.
• Every project is entirely scripted. That is, all analysis, graphs and tables must be able to be generated by running one script.
• This script sources all other R files in the correct order and yields all the required results. This script could be in main.R or main.Rmd.
• functions.R contains all non-packaged functions used in the project.
• each function can not be too long.

## Getting help

• For programming questions: StackOverflow.com
• For statistical questions: CrossValidated.com

## Keep up-to-date

• RStudio blog: blog.rstudio.org
• R-bloggers: www.r-bloggers.com
• It takes time to develop your own style. Once it is developed, it is really hard to be changed. So please be careful at the beginning.

• Write a function to solve the roots of given quadratic equation $$ax^2 + bx + c = 0$$ with $$a$$, $$b$$ and $$c$$ as input arguments.

• Test your function on some simple equations.

• Keep in mind the styles we have learnt.

• Write two messy .R files and put them in a folder.
• Use tidy_dir() to make them tidy.