Jupyter Notebook and server utilities¶

^{January 21, 2020}  

Workshop on Jupyter Notebook for Python and various server tools and utilities.

Agenda

1. Jupyter Notebook – general overview
2. Install and run Jupyter Notebook locally
3. How to setup and run Jupyter instance on a server
4. Server tools:
   • ZSH
   • vim
   • tmux
   • iTerm (for macOS)
5. IDEs
   • Atom (with plugins)
   • IntelliJ products, e.g., PyCharm
6. Access over SSH, and SSL certificates

Introduction

Who am I?

• PhD student in Faculty of Psychology and Cognitive Science
• Computer System Analyst in Poznan Supercomputing and Networking Center

Inspiration for this workshop

1. Jupter Notebook

Background

• Growing interest in “user-friendly”, high-performance technologies:
  • Why Is Python So Popular?
  • Why Jupyter is data scientists’ computational notebook of choice
• Deep Learning Workshops by nvidia (certificate)
• It is fun :)

Basic example:

• Notebook basics
• Running code, GitHub

Utils, part 1:

• nbviewer – render static Notebooks online; an example:
  • Running code, nbviewer

Materials

• astro-python
• pandas-astro-example
• yt-project – visualising cosmology dataset
• Astro Interactives
• Teaching materials example – in nbviewer
• more, non-astro examples

Utils, part 2:

• JupyterLab – advanced, Matlab-like UI
• JupyterHub – multi-user, resources management

2. Install and run Jupyter Notebook locally

How to install and run Jupyter on various platforms – readthedocs.

TL;DR:

pip install jupyter notebook
jupyter notebook

3. How to setup and run Jupyter instance on a server

In order to run Jupyter Notebook on a server, it’s best you knew the following concepts:

• SSH
• HTTPS access with (a self signed) SSL certificate
• Apache HTTP(S) Server and how to configure it
• Linux services (and autostart with systemd)

Do-it-yourself version is provided with these two web pages:

• Official Jupyter Notebook documentation
• My blog post on how to run Jupyter on OpenStack

4. Server tools:

ZSH

General information

In short, zsh is a more interactive version of bash. What does the interactive stand for? Basically, with zsh your Tab key becomes an intelligent, powerful entity (advanced autocompletion).

Particularly useful with:

• typos
• git
• docker
• and any other multi-argument bash utilities

Time for a little demonstration :)

Recommended zsh framework: oh-my-zsh.

The easiest way to install zsh:

(with curl)

sh -c "$(curl -fsSL https://raw.github.com/ohmyzsh/ohmyzsh/master/tools/install.sh)"

(with wget)

sh -c "$(wget https://raw.github.com/ohmyzsh/ohmyzsh/master/tools/install.sh -O -)"

vim

tarsier /ˈtɑːsɪə/

vim web page tells you all

Descendant of ex line editor (reminiscence in vi’s ex mode).

Advantages:

• you can macro *anything*
• available on all server (on older distros as vi)
• (supposedly :) ) speeds up some file-editing operations

Disadvantages:

tmux

Terminal multiplexer – your best friend both: locally and on a server.

Nice introductory tutorial on tmux.

Do you know terminator? Tmux is similar, but better.

gpakosz has the best tmux configuration

iTerm (for macOS)

Next-gen terminal for macOS (web page). I dont even remember the name of the default macOS terminal anymore…

5. IDEs

Atom (with plugins)

Atom is light, Atom is nice, Atom is open.

Bare-bones Atom is very crude. However, there are plethora of plugins (aka packages)!

Must-have plugins:

• PlatformIO IDE Terminal
• Sync Settings
• vim-mode-plus
• File icons

IntelliJ products, e.g., PyCharm

Magnificent, heavy, feature-packed:

• build-in terminal support
• decent git-management utilities (however, atom still has some advantages in that manner)
• free pro version for educational purposes

6. Access over SSH, and SSL certificates

SSH:

• public vs. private key
• password-based vs. certificate-based
• command line vs. PuTTY

Last modified on Jan 20, 2020