Jupyter Notebooks Tutorial

Install Anaconda

Go to https://www.anaconda.com/download/, download and install the most recent version of Anaconda with Python 3.6.

Start a jupyter notebook

With Anaconda Jupyter is installed by default. Now you can open the terminal (or Anaconda prompt if you are on Windows) and type:

jupyter notebook

A jupyter notebook will be created and the browser will open by default. Notice the link should look link localhost:8888. This means the notebook is running in your local machine in port 8888.

Start a notebook in specific directory: To set the home folder for the current notebook you can pass the argument the --notebook-dir=some_path argument to set the home folder to ´some_path´. To set for the current folder use a dot in the path.

jupyter notebook --notebook-dir=~/my_notebooks #set the notebook home folder to my_notebooks folder in home (if you are in linux)
jupyter notebook --notebook-dir=. #set the notebook home folder to the current directory.

Notebook server on remote host

Find your public ip address

curl ipinfo.io/ip

Find your internal network address

hostname -I

Add new trusted machine (Linux)

Generate a pair of ssh keys in your local host with the following command:

ssh-keygen

This will generate to files inside ~/.ssh, by defaut:

id_rsa
id_rsa.pub

The id_rsa is the private key and should never be shared. The id_rsa.pub has the public key. To gain acess to the remote host add the public key to the ~/.ssh/authorized_keys file if it exists or create it otherwise. If you are on the local network you can access it using the local address. If you need to access with the public ip address you need to make a forward in the router for an ssh connection for your remote host.

Add new trusted machine (Windows)

• Go to https://www.putty.org/ and install PuTTY. After the installation is complete type windows key + R and write PuTTYgen.
• Choose the option to generate and then save both public and private keys in a safe place.
• Copy the public key as in the Linux case to your remote host.
• Open PuTTY and go to Connection > SSH > Auth and browse the private key location.
• Go back to Session in PuTTY and type the ip of the remote host. You should now be able to connect.

Add new trusted machine (Android)

• Install JuiceSSH App

Remote access to Jupyter notebook via SSH

(https://coderwall.com/p/ohk6cg/remote-access-to-ipython-notebooks-via-ssh)

• Start jupyter notebook on the remote host

jupyter notebook --no-browser --port=9999 

• Create a SSH tunnel on localhost

ssh -N -L localhost:9999:localhost:9999 remote_user@remote_host