Skip to content

Latest commit

 

History

History
87 lines (73 loc) · 4.58 KB

jupyter_quick_reference.md

File metadata and controls

87 lines (73 loc) · 4.58 KB
Raw

Introduction

Jupyter notebook provides a web-based application for developing, documenting and executing code.

The main features of Jupyter notebooks are:

  • In-browser code editing
  • Displaying of the results of computations in-line
  • In-browser editing for rich text using the Markdown markup language
  • LaTeX syntax integration

Installation

To install jupyter in your virtual environment, activate it and type:

pip install jupyter

Running Jupyter

Locally

To run Jupyter on your local machine, go to a folder you want to create your notebook in and type:

jupyter notebook

This will open a new tab in your browser and display the so-called Notebook Dashboard, where you can manage existing notebooks: Notebook Dashboard The Dashboard will only let you see and edit files located only within your startup directory, so make sure to start Jupyter in the correct folder.

Remotely

Because Jupyter is a server-client-based web application, it is possible to launch it on a remote machine. By default, a notebook server starts locally on port 8888 using the same command:

jupyter notebook

You can change the port specifying the -p flag as follows:

jupyter notebook --port <port_number>

Once the Jupyter Notebook application is started, you can use SSH tunneling to connect to it. To establish your own SSH tunnel, run the following command:

ssh -L 8000:localhost:8888 your_login@your_server_ip

Here 8000 is the port on your local machine and 8888 is the port on the remote server where you are running Jupyter. If one of your local applications is already using the port 8000, feel free to change it to a different one. Once you have forwarded the port, you can copy the URL from your remote server's terminal output and paste it into your browser's address bar. At this step you should be able to see the Notebook Dashboard.

Getting started

To create a new notebook, click New->Python 3 at the top right of the Dashboard as in the screenshot: New notebook

Interface

The interface of the jupyter notebook is pretty self-explanatory; however, we will cover the main functionality here. In general, there are two modes of the notebook operation, Command Mode and Edit Mode. The former is used for navigation between cells (see below), changing their order or type, and running the notebook while the latter is used for modification of a given cell's content. The notebook consists of compuеational units called cells, every single cell has either a Code or a Markdown type. You can switch between the two types by going to Cell Type under the Cell menu: Cell types You can select, copy, insert, delete, move, execute and edit cells using the toolbar at the top or using shortcuts (see below).

To execute a single cell, press Ctrl+Enter. To edit a cell, press Enter. If you are editing a markdown cell, you should see how the markdown content is rendered after the execution (lists, formulas, headers, etc.): Markdown cell

If you are editing a code cell, you can see the output that your code produces right after you execute the cell: Code cell

Shortcuts

Perhaps the easiest and fastest way to edit jupyter notebooks is to use various shortcuts. We will list the most commonly used ones. If you want to quickly look a shortcut up, press H when in Command Mode to see a full table of shortcuts.

Command Mode shortcuts:

  • Enter the Edit Mode: Enter
  • Change the cell mode to markdown: M
  • Change the cell mode to code: Y
  • Add a cell below/above: B/A
  • Delete a cell: press D two times

Edit Mode shortcuts:

  • Enable the Command Mode: Esc
  • Run a cell: Ctrl+Enter
  • Run a cell and insert another one below: Shift+Enter
  • Fetch a docstring of a python function: Shift+Tab
  • Comment: Ctrl+/
  • Indent/dedent: Ctrl+]/ Ctrl+[
  • Delete the whole line: Ctrl+D
  • To hide the output of the cell, double-click on it

For the full documentation please refer to https://jupyter-notebook.readthedocs.io/en/stable/