Introductory R

Syntax bootcamp

Author

Joseph Mhango

Published

2024-09-13

Objectives

  • Install R and RStudio or set up RStudio Cloud
  • RStudio components and setup
  • Workflow for scripts in R

Why R

  • R is objectively the best statistical software available
  • R is designed for people with no programming experience to perform sophisticated statistical analysis with minimum effort
  • In wide use at universities, companies: job demand
  • Very large community of users
  • Free and open source
  • Works well on all computers and OSes, old and new

How to setup R and R studio

RStudio components and setup

R Language: General statement

R is a programming language designed to help non-programmers perform statistical analyses and to make graphs. This session is intended to guide people through some of the basics of the R programming language, just enough to get started.

The Life of Code…

  • R code can live comfortably in the following formats:
    • .R (Scripts)
    • .Rmd (R Markdown)
    • .Qmd (Quarto Markdown)
    • .ipynb (Jupyter Notebooks)

R projects and data can be saved in the following formats:

  • RStudio Project files, organizing scripts and data (.Rproj)

  • Workspace or multiple R objects (.RData)

  • Single R object (.Rds)

Things can get all woven up…

Example:

This presentation comes from a .Qmd document. It has a bunch of text called markdown, sprinkled around R code. The slides and images for the presentation are all organised neatly into an R project in a folder, with a .Rproj overlord

For the purposes of getting started with syntax, we will work strictly with a lone script

Writing R Scripts: Best Practices

  • Document Your Work: Scripts are a a record of your progress

  • Organise Your Code: Proper structure improves reproducibility

  • Write for Others: e.g. your future self or supervisor

  • Use Comments & Pseudocode: human readable with obvious flow

Important

  • Work through the instructions here in RStudio as you go along

  • Type you own code rather than using copy and paste

  • Document all the code in your own script and write clear, concise comments

The comment (#)…

  • The hashtag is used to declare comments

  • Anything that comes after a # on the same line is ignored by the R interpreter during execution

  • Don’t just litter everywhere (e.g. storing code as comments)

The comment (#)…

  • Comments are useful for: +Making headers, tables of contents e.t.c. +Temporarily disabling undesired lines of code +Making code chunks
  • Example of a commenting convention:
    • Code chunks begin with 2 ## signs
    • Code chunks end with a consistent pattern

Contents

## CONTENTS ####
## 1.1.1 Example script, help, pseudocode  
## 1.1.2 Math operators  
## 1.1.3 Logical Boolean operators  
## 1.1.4 Regarding base R and the Tidyverse   
## 1.1.5 Practice exercises

Outlines

  • Outlines enable you to jump around your script,
  • If you have clear delimiters in comments, RStudio recognises them as outlines
# ---- Loading libraries ----
library(stats)
# ---- Iris scatter plots ----
plot(iris)

Getting Help in R

Community and Resources

  • R has a strong community with many websites, books, blogs, and more.

  • The vast array of resources can be overwhelming for beginners.

  • Best Practice: Start with the R Help System + Use the built-in R Help system first before exploring external resources

Accessing Built-in Help in R

  • The basic way to get help is by using the help() function.

Syntax:
help(function_name)

  • Get help on the mean() function
# Display help page for the function mean
help(mean)

Help page format

  • Help pages have a consistent structure

  • 1 Function name {Package name}: Tells you the package of the function

2 Short description: What the function does in brief.

3 (longer) description: What the function does

4 Usage: An example of the function in use and “arguments”

5 Argument definitions: What the argument are and what they do!

Value: What the function returns

Making Help work for us

Using the Usage and Argument fields, we can figure out how to make the function do the work we want.

  • Under Usage:
# mean(x, ...)

# The "x" is an argument that is required
# The "..." means there are other optional arguments

Under Arguments:

# x 
# An R object... for numeric/logical vectors ...
  • Try this code in your own script

my_length <- c(101, 122, 97) # 3 numerical measures
mean(x = my_length)

Pseudocode

  • Pseudocode is a way to break up a big task into a series of smaller tasks. For example:

    • Read data into R
    • Perform exploratory analysis
    • Perform statistical tests
    • Organize outputs to communicate in a report

Submitting code to the interpreter

  • Run whole line of code your cursor rests on Ctrl+Enter (Cmd+Return in Macs)
  • Run code you have selected Ctrl+Enter (Cmd+Return in Macs)
  • Use the “Run” menu above the Script window
  • Use the Code > Run dropdown menu

Now let’s put the interpreter to work…

Math operators

Basic manipulation of numbers in R is very easy and intuitive. Let’s try this non-exhaustive list:

Arithmetic

# Add with "+"
2 + 5
# Subtract with "-"
10 - 15
# Multiply with "*" and Divide by "/"
(6 * 4.2)+(10 / 4)

More arithmetic

# raise to the power of x
2^3 
2**3
9^(1/2) # same as sqrt()!
9**(1/2) # same as sqrt()!
# There are a few others, but these are the basics

Order of operation

For complicated phrases like 2 + 2 * 8 - 6. the BODMAS/PEMDAS rule is followed unless unless a specific order is coded.

# Try this
4 + 2 * 3

# Order control - same
4 + (2 * 3)

# Order control - different...
(4 + 2) * 3

Use of spaces

# Try this
6+10                                  # no spaces
7     -5                              # uneven spaces
1.6             /                2.3  # large spaces
16 * 3                                # exactly 1 space
# exactly 1 space is easiest to read...

Boolean operators

Boolean operators are expressions that resolve TRUE (treated as “1” in most computing systems including R) versus FALSE (“0”). A typical expression might be something like asking if 5 > 3, which is TRUE. More sophisticated phrases are possible, and sometimes useful.

3 > 5

# 3 is compared to each element
3 < c(1, 2, 3, 4, 5, 6)

Logic and math

  • & (ampersand) means “and”
  • (pipe) means “or”
#This asks if both phrases are true
3 > 1 & 1 < 5

3 < 1 | 1 < 5

3 < 1 | 1 > 5

Using Booleans

  • Booleans can be useful to select data
  • Put some data into a variable and then print the variable
  • Note <- is the ASSIGNMENT syntax in R, which puts the value on the left “into” x
  • The square brackets are there to allow us to specify an index of the data vector… more later
x <- c(21, 3, 5, 6, 22)
x[x > 20]
[1] 21 22

The “not” operator - !

# Try this
TRUE # plain true

!FALSE # not false is true!

6 < 5 #definitely false

!(6 < 5) #not false...

!(c(23, 44, 16, 51, 12) > 50)

Base R and the Tidyverse

Base R

  • Base R: the basic functions which let R function as a language
  • library(help = "base")
  • Examples include functions like read.table(), data.frame(), etc.
  • Here’s the thing; mastering R is mastering Base R
  • Base R provides the tools you need to eventually even write your own R packages.

The Tidyverse

  • The Tidyverse is collection of R packages.
  • Developed by RStudio (Hadley Wickham and others), now called Posit.
  • Core packages: ggplot2, dplyr, tidyr, readr, tibble, and purrr.
  • Tidyverse focuses on:
    • Human-readable code.
    • Consistency across functions.
    • Easy-to-use tools for common data manipulation and visualization

Differences Between Base R and the Tidyverse

  • Base R provides lower-level intuitive functions; the Tidyverse provides simplified, higher-level abstractions.
  • Tidyverse emphasizes abstractions that make data wrangling easier
  • But guess what, most of the tidyverse is written on top of Base R
  • Base R is pretty stable, the tidyverse has a deprecation problem

The error traceback

  • Don’t be intimidated by the screaming red text
  • It’s all bark and no bite
  • if it looks confusing, use traceback() -Most of the times, it’s because:
    • you’re making calculations on an invalid datatype
    • you’re passing the wrong things to function arguments
    • you’ve made a syntax error somewhere