There is no special requirement for the system type. If your system supports Java and Python you should be able to run the case. The Jupyter notebook was thoroughly tested on linux based system (openSUSE Leap 15.0 and 15.1), as well as Windows based machines (Windows 10).
In order to run Jupyter notebook you should install Java and Python.
According to h2o.ai website (http://docs.h2o.ai/h2o/latest-stable/h2o-docs/welcome.html) Java 8, 9, 10, 11, 12, and 13 are supported. We will be running binary package of H2O so JRE is required.
Python version 3.6.5 is required. It is advisable to create separate Python environment. Using Anaconda platform is recommended (https://www.anaconda.com/). Detailed step-by-step guide on installation of H2O and Python is available at: https://nbviewer.jupyter.org/github/jszlek/02_NTU_workshop_Python_notebook/blob/master/h2o_PLGA_case.ipynb section 2. H2O installation and Installing Anaconda & h2o (Python).
If you wish to run the case on your machine please clone or download the repository (https://github.com/jszlek/02_NTU_workshop_Python_notebook/) and open it via Jupyter notebook.
To unpack the 7-zip files which are stored in the repository (*.7z) please install a software from https://www.7-zip.org/ or use your package manager (zypper, apt-get)
Please, be advised that running H2O server can cause your machine to freeze or even crash, if it has not enough resources (RAM and/or CPUs). If you wish to use the notebook on your own data sets please carefully set the my_max_ram_allowed variable not to exceed your free RAM.
To view the case please visit: https://nbviewer.jupyter.org/github/jszlek/02_NTU_workshop_Python_notebook/blob/master/h2o_PLGA_case.ipynb
In order to run a script in a batch mode a Python script was prepared (https://github.com/jszlek/02_NTU_workshop_Python_notebook/blob/master/AutoML_Jupyter_notebook.py). Please edit the script according to your needs, save it, switch to the case directory, run terminal and activate h2o environment from your anaconda prompt. Then type python AutoML_Jupyter_notebook.py.
If something goes wrong and you will have troubles in getting everything running, please do not hesitate to contact me: j.szlek[at]uj.edu.pl