Skip to content

Latest commit

 

History

History
155 lines (125 loc) · 3.53 KB

jupyter.md

File metadata and controls

155 lines (125 loc) · 3.53 KB
Raw

Jupyter Notebooks

  • Web application that combines markdown, math equations, code, visualizations

AWS Jupyter Notebook for Deep Learning

  • Go to AWS Marketplace and select Deep Learning AMI (Ubuntu) Version 8.0
  • ssh into the instance using the pem file
ssh -i "pemfile.pem" ubuntu@ec2-1-2-3-4-5.compute-1.amazonaws.com
  • Source the tensorflow environement
source activate tensorflow_p36
  • Generate a Jupyter Notebook config file
jupyter notebook --generate-config
  • Set a password for the notebook
jupyter notebook password
  • Start the Jupyter Notebook
jupyter notebook --ip=0.0.0.0 --no-browser

Command Line

  • Launch the notebook server
jupyter notebook
  • Launch the notebook server on AWS server
jupyter notebook --ip=0.0.0.0 --no-browser
  • Install Notebook Conda to help manage environments
conda install nb_conda
  • List running servers with their tokens
jupyter notebook list
  • Setup a Jupyter Notebook Password
jupyter notebook password

UI

Command palette

  • It brings up a little pannel where commands can be searched

Modes

  • Edit Mode: Type into cells
    • Left border is green
  • Command Mode: Execute commands
    • Left border is blue

Keyboard Shortcuts

Command Mode

  • Enter/Esc: Switch between edit/command modes
  • Enter + Shift: Go to next cell
  • h: List of shortcuts
  • s: Save notebook
  • a: Create cell above
  • b: Create cell below
  • dd: Delete cell
  • y: Change from markdown to code
  • m: Change from code to markdown
  • p: Access the command palette

Code Cells

Markdown Cells

  • cf markdow.md
  • Math expressions using LaTeX symbols
    • Notebooks use MathJax to render the LaTeX symbols as math symbols

Magic Keywords

  • Special commands that lets you control the notebook itself, perform system calls
  • %: Line Magic
  • %%: Cell Magic

Make autocomplete work again in Jupyter Notebooks - https://stackoverflow.com/a/64554305/791795

%config Completer.use_jedi = False

Python Kernel

  • List of all magic commands
  • Timing Code
    • Function call: %timeit f(*args)
    • Entire Cell: %%timeit ...
  • Embedding visualizations in notebooks: %matplotlib
    • Inline: %matplotlib inline
    • High quality: %config InlineBackend.figure_format = 'retina'
%matplotlib inline
%config InlineBackend.figure_format = 'retina'

import numpy as np
...

plt.plot()
  • Debugging in Notebooks: %pdb
  • Documentation: Prefix the method by ?
?str.replace()
  • Time a cell
%%time

Converting Notebooks

  • Default, big JSON files with .ipynb extension
  • Convert to HTML file
jupyter nbconvert --to html notebook.ipynb

Slideshows

  • Example
  • Regular notebooks, designate which cells are slides and the type of the slide cell
  • Export to slideshow
jupyter nbconvert notebook.ipynb --to slides
  • Serve the presentation
jupyter nbconvert notebook.ipynb --to slides --post serve

Resources