Thanks for your feedback. Would you mind telling what was the necessary tinkering that you had before?
Regarding your questions:
One relates to the patch ID numbers. In the jupyter-notebook example the ID is set to 1549 or 190. Can you explain how these values are selected?
There is no particular reason for this. The main point in choosing the patch was that the selected area should contain enough diversity from the class representation point-of-view. That’s all
The tinkering was related to setting up the environment to actually run the eolearn scripts. It takes some time to set everything up under windows. At least it took me some time. There are some pointers in the tutorials and documentation, but I think a tutorial on setting up the environment to run the eolearn script would be useful.