• Anaconda or all relevant packages installed
• SSH connection
• Screen installed by default on university systems

if anaconda not set up on remote machine see information at the bottom of this page

1. ssh to machine you want to run notebooks on (needs to have access to data etc) e.g. ssh foe-linux.leeds.ac.uk
2. Once Jupyter is installed via anaconda or other choice generate password

   jupyter notebook --generate-config
   jupyter notebook password   
   

3. Edit the following lines in the config file like this exampele (may require different port number)

c.NotebookApp.allow_origin = '*'
c.NotebookApp.allow_password_change = False
c.NotebookApp.allow_remote_access = True
c.NotebookApp.open_browser = False
c.NotebookApp.ip = '0.0.0.0'
c.NotebookApp.port = 5566

4. start a screen session ideally naming the session something sensible screen -S JupyterScreen
5. go to desired directory and run jupyter-notebook
6. Take Note of port number e.g. 5566

[I 11:48:41.622 NotebookApp] The Jupyter Notebook is running at:
[I 11:48:41.622 NotebookApp] http://foe-linux-03.leeds.ac.uk:5566/

7. detach screen: ctr+ a+ctrl +d
8. On your home computer (linux or mac) edit or create file ~/.shh/config to contain the lines (as referenced in our remote access page)

Host remote-access
     User <leeds-user-name>
     ForwardAgent yes
     ForwardX11 yes
     Hostname remote-access.leeds.ac.uk
     IdentityFile ~/.ssh/id_rsa_work
     ServerAliveInterval 240

Host foe-03
     User <leeds-user-name>
     ForwardAgent yes
     ForwardX11 yes
     Hostname foe-linux-03
     IdentityFile ~/.ssh/id_rsa_work
     ServerAliveInterval 240
     Proxycommand ssh -qremote-access -W %h:%p

You need config for a gateway machine: e.g. remote access or see-gw-01 and the machine you are running the notebook on 9. generate key and copy keys to access to leeds systems as outlined in our remote access page

• ssh-keygen -t rsa name something like id_rsa_work to match identity file in the config file you just created
• ssh-copy-id earhbu@see-gw-01

10. Now tunnel in

• generic ssh -N -f -L localhost:<localportno>:localhost:<remoteportno> user@host
• following this example: ssh -N -f -L localhost:5000:localhost:5566 user@foe-03

11. Access from your local browser http://localhost:5000/ enter your password

The following commands will install on a Linux machine via the terminal. After this you can conda install -c conda-forge notebook and

wget https://repo.anaconda.com/miniconda/Miniconda3-py39_4.9.2-Linux-x86_64.sh
chmod 755 Miniconda3-py39_4.9.2-Linux-x86_64.sh
bash Miniconda3-py39_4.9.2-Linux-x86_64.sh

Some tips for large notebooks

1. Clear output before pushing to git.
2. save as markdown jupyter nbconvert --to markdown <file>.ipynb
3. save as pdf jupyter nbconvert --to pdf <file>.ipynb
   • NB notebooks containing large amounts of markdown mayrequire debugging to convert to pdf via converting latex first and running through an application that gives detailed logs like texmaker