Work on tutorial section from Fall school

[timonmerk]

I pulled the last version of the py_neuromodulation git repo. I followed the clear instructions on setting up the environment. I tried
conda env create --file=env.yml --user
but the --user handle was unknown and I had to remove it. Maybe it was probably a noob problem requiring further input or specification what the user is for, from my side but I guess there are more and more noobs starting to switch from Matlab to Python so it may be good to explain or double check this is correct.

I suppose binder would circumvent this problem, but I only found the binder button after a long search. I would suggest mentioning the binder and what it can do explicitly in the readme and definitely add the link to both repo and binder to the presentation.

The main problem I found was the question on how to start the hands-on part. So I checked the repo and started working on the Fall school Notebook_1_Tasks.ipynb and found a few minor bugs (e.g. Y instead of y that is expected further on etc.). Then I checked the Fall school Notebook_2_Tasks.ipynb thinking that would be the more recent or correct version, but this already included solutions so I opened Fall school Notebook_3 which also included solutions but was slightly different to notebooks 1 and 2. I am sure you kept improving everything, so this is clearly just a minor thing, but before times passes by and we don't remember anymore, I would suggest to create one clean version called *_tasks.ipynb & _solutions.ipynb .

Would it be possible to move these notebooks out of the main repo into a tutorial folder?

I found the tasks to be more python coding related than machine learning related. They were great for me but may not be what people are expecting. Potential ML related tasks could be to repeat the cross-validation or chose different feature sets and explore the effects. If you want, we could take some time next week and go through them again and adapt a little?

[timonmerk]

• Merge part 1 & 2
• Include command running py_neuro for example data