Instructions for setting up Homebrew & Jupyter (notebook) for ISYE_6414

homebrew_jupyter.md

Install Homebrew and Jupyter Notebook

1. Open your terminal (on Mac, the default terminal is named ... Terminal)
2. Go to Homebrew and paste script at the top into Terminal

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

3. Install jupyter: brew install jupyter
4. Open a new notebook to confirm: jupyter notebook
5. If you don't have the R kernel when you open a new notebook (and you probably shouldn't), then you'll need to install the IRkernel: from the terminal, open R: then install.packages("IRkernel")
6. Create a Jupyter notebook with the R kernel.

What is Homebrew?

Homebrew is a package manager for Unix-based operating systems (this is why Homebrew says it works with Mac and Linux). A package manager is responsible for downloading pieces of software and storing them locally on your computer. If you want a new version of a piece of software, you can ask your package manager to get it for you (or maybe you asked your package manager to get update versions for you automatically).

Why use Homebrew?

Well, it's a great package manager. It's written in Ruby (a language) and it's been around for a long time as the default package manager for various software development communities. This means it's well-supported, you don't need to worry about something crazy happening, and if you run into an error, you can probably find a solution online with a little searching.

So I'm in 6414 Regression Analysis, what do I need a package manager for?

Well, you don't. But you might want one.

This course uses R, which you can install in different ways. And you can write your R code in different ways. And you can even use different graphical user interface (GUI) software like R Commander to run, write, interrupt, debug and 'knit' your R code.

But ...

There is (what I think) is a particularly awesome way to run, write, interrupt, debug and 'knit' your R code: the Jupyter notebook. Project Jupyter is an open source endeavor that arose from Python (one of the other big data analytic languages, and the main other (e.g. besides R) language you'll use in the OMSA program).

What is a Jupyter Notebook? Convince me to use it in like 60 seconds.

If you know R Commander, maybe that's good enough for now. If you asked me where I think the future moves for work in data analytics, it's probably with tools like Jupyter Notebook, because ...

Jupyter Notebooks are really easy to share, and the Jupyter Project has been working a lot on something called Jupyter Lab, which you can think of as Google Docs (i.e. cloud-based, shared, version-controlled) for Jupyter Notebooks.