Jupyter makes it possible to write and execute programs and analyze data in popular programming languages such as Python, R and Julie. It can be accessed from any location via a fast, interactive and secure web interface. The advantage of Jupyter is the quick start-up of a virtual environment with pre-prepared tools, so users do not have to worry about maintaining the computer infrastructure. Data storage and sharing is possible through the PUH system included in Jupyter.
The following Jupyter tools are currently available:
- Python
- Julia
- Octave
- Pluto Notebook
- R
- RStudio
- SageMath
- Wolfram Engine
- Bash
- Markdown
- Terminal
Detailed instructions for working with Jupyter.
Signing up
Signing up for Yupyter is done through the web application. All users who have an approved and active project can use Jupyter or if they don't have one, they can request it by selecting the Jupyter resource type..
Access
Jupyter is available at the web link. It is necessary to log in using the AAI@EduHr electronic identity.
After logging in, you choose one of the offered environments and press the Start button, and if the server is already running, the step is skipped.
After the environment is created, the control panel will be available:
Resources
Users have a range of processor cores and RAM at their disposal. The lower limit for the resource is clearly defined and these are guaranteed values that cannot be lower in the environment:
Name | CPU | RAM | GPU | Total Disk | PUH | ||
---|---|---|---|---|---|---|---|
Min | Max | Min | Max | ||||
Srce server | 2 | 12 | 6 GB | 20 GB | 10 GB | ||
Wolfram engine server | 2 | 12 | 6 GB | 20 GB | 10 GB | ||
GPU server | 2 | 8 | 16 GB | 24 GB | * | 10 GB |
|
*The GPU server consists of a Multi-Instance GPU (MIG) mig-1g.5gb which is about 1/7 the power of the A100.
Important
Inactive Jupyter for more than 30 minutes will be stopped, and your data will remain saved in the home directory or Puh.
Server management
You can stop a running server by selecting FILE→Hub Control Panel, Stop My Server.
Puh - Data storage and management
Puh can be accessed from the Jupyter Terminal using the puh command. You access the terminal using the menu File → New → Terminal.The access password can be downloaded from the PUH website selecting Settings → Security → Create new app password.
jovyan@jupyter-mbelavic-40srce-2ehr:~$ puh Please enter the username to authenticate with server https://puh.srce.hr/remote.php/webdav/ or hit enter for none. Username: mbelavic@srce.hr Please enter the password to authenticate user mbelavic@srce.hr with server https://puh.srce.hr/remote.php/webdav/ or hit enter for none. Password: /sbin/mount.davfs: warning: the server does not support locks jovyan@jupyter-mbelavic-40srce-2ehr:~$ cd ~/puh/ jovyan@jupyter-mbelavic-40srce-2ehr:~$
Wolfram Engine
For the wolfram engine, it is necessary to start the Wolfram engine server. For the license you need to register on the Wolfram web portal, after registration open the terminal and type:
jovyan@jupyter-mbelavic-40srce-2ehr:~$ wolframscript The Wolfram Engine requires one-time activation on this computer. Visit https://wolfram.com/engine/free-license to get your free license. Wolfram ID: account@test.hr Password: Wolfram Engine activated. See https://www.wolfram.com/wolframscript/ for more information. Wolfram Language 14.0.0 Engine for Linux x86 (64-bit) Copyright 1988-2023 Wolfram Research, Inc. In[1]:=
To add a notebook, you need to add the Wolfram engine kernel, open a terminal and type:
jovyan@jupyter-mbelavic-40srce-2ehr:~$ wget https://github.com/WolframResearch/WolframLanguageForJupyter/releases/download/v0.9.3/WolframLanguageForJupyter-0.9.3.paclet jovyan@jupyter-mbelavic-40srce-2ehr:~$ wolframscript In[1]:= PacletInstall["WolframLanguageForJupyter-0.9.3.paclet"] In[2]:= Needs["WolframLanguageForJupyter`"] In[3]:= ConfigureJupyter["Add"]
After adding and reloading (Ctrl + R) it should be displayed as one of the tools