“The best thing about R is that it was developed by statisticians. The worst thing about R is that… it was developed by statisticians.”

- Bo Cowgill, Google

Overview

• Language Features and Use Cases
• R
• R Studio
• R and Packages

What is R?

• R is an Open Source (and freely available) environment for statistical computing and graphics.
• It is a full-featured programming language, in particular a scripting language (with similarities to Matlab and Python).
• It can be run interactively or as a batch/background job.
• R is being actively developed with ongoing updates/new releases.
• R has a variety of built-in as well as community-provided packages that extend its functionality with code and data; see CRAN for the thousands of add-on packages.
• It is freely-available and modifiable (Available for Windows, Mac OS X, and Linux).

Modes of Using R

• From the command line in a Linux/Mac terminal window
• Using the Windows/Mac GUIs
• Using the RStudio GUI, an ‘integrated development environment’
• Running an R script in the background on a Linux/Mac machine (Windows?)
• RStudio

Why R?

• R is widely used (statisticians, scientists, social scientists) and has the widest statistical functionality of any software
• Users add functionality via packages all the time
• R is free and available on all major platforms
• As a scripting language, R is very powerful, flexible, and easy to use
• As a scripting language, R allows for reproducibility and automating tasks
• As a language, R can do essentially anything
• Wide usage helps to improve quality and reduce bugs
• R can interact with other software, databases, the operating system, the web, etc.
• R is built on C and can call user-written and external C code and packages (in particular, see the Rcpp R package)

Why Not R?

• Other software is better than R at various tasks

i.e., Python is very good for text manipulation, interacting with the operating system, and as a glue for tying together various applications/software in a workflow-* R can be much slower than compiled languages (but is often quite fast with good coding practices!)

• R’s packages are only as good as the person who wrote them; no explicit quality control
• R is a sprawling and unstandardized ecosystem
• Google has a recommended style guide that should be taken into account.

R and Jupyter

• R commands can be executed in a Jupyter Notebook just by play at the end of a cell.
• Blocks of cells or even the entire notebook can be executed by clicking on the Cell above.
• The Kernal is responsible for interpreting the code, and the current kernal is listed on the top right of the notebook.
• While Jupyter started as a Python project, there are now a variety of Kernals for different programming languages including R, Scala, and SAS.
• Read more about Jupyter in the documentation here.
• If a variable isn’t assigned it will be provided as output.

test<-5
test

R and RStudio

• Powerful IDE for R
• Integrated usage of git
• Integrated GUI based package management
• Integrated GIT
• Solid enterprise infrastructure, owned by Microsoft

R and Conda

• Conda can quickly and easily install R. We love conda!
• This is a rather long install, so may do it from command line. !conda install r-essentials

#conda install -y r-essentials

R as calculator

# R as a calculator
2 * pi # multiply by a constant
7 + runif(1) # add a random number
3^4 # powers
sqrt(4^4) # functions
log(10)
log(100, base = 10)
23 %/% 2 
23 %% 2

Assignment of Values

• Don’t used the equal sign.
• Do use <-

#Assignment DON"T use =
val <- 3
val
print(val)

Val <- 7 # case-sensitive!
print(c(val, Val))

[1] 3
[1] 3 7

# This will gnerate numbers from 1 to 6. 
mySeq <- 1:6
mySeq

#Notice how this worrks like arrange in Python.
myOtherSeq <- seq(1.1, 11.1, by = 2)
myOtherSeq
length(myOtherSeq)

fours <- rep(4, 6)
fours

# This is a comment: here is an example of non-numeric data
depts <- c('espm', 'pmb', 'stats')
depts

R and Packages (R’s killer app)

• There are a tremendous number of packages available which extend the core capabilities of the R language.
• “Currently, the CRAN package repository features 9233 available packages.” (from https://cran.r-project.org/web/packages/)
• Packages on CRAN. Also see the CRAN Task Views.
• Packages may be source or compiled binary files.
• Installing source packages which contain C/C++/Fortran code requires that compilers and related tools be installed.
• Binary packages are platform-specific (Windows/Mac). This can cause problems in a large class and is a great reason to work in a standardized environment like the cloud or on Docker/VM.
• If you want to sound like an R expert, make sure to call them packages and not libraries. A library is the location in the directory structure where the packages are installed/stored.

Using R packages

1. Install the package on your machine
2. Load the package

Installing Packages in RStudio

• Only needs to be done one time on machine.
• To install a package, in RStudio, just do Packages->Install Packages.
• Option to specify the source repository: install.packages('chron', repos='http://cran.us.r-project.org')
• Option to install multiple packages: install.packages(c("pkg1", "pkg2"))
• You can install dependencies with: install.packages("chron", dependencies = TRUE)
• If binary files aren’t available for your OS, install from source: install.packages(path_to_file, repos = NULL, type="source")
• R installs are not done from the terminal (no !).

Install Packages with Conda

This will only work from the Anaconda Prompt or terminal.

!conda install -y -c r r-yaml

Loading Packages

• R packages must be imported before using.
• Packages only have to be imported once in a notebook (not in every cell).
• The library(chron) would import the chron package, provided it has beeen installed.
• See all packages loaded with search()
• search() # see packages currently loaded

search() # see packages currently loaded
library("yaml") # Load the package chron
search()    # see packages currently loaded

CREDITS

Copyright AnalyticsDojo 2016. This work is licensed under the Creative Commons Attribution 4.0 International license agreement. This work is adopted from the Berkley R Bootcamp.