Add 'Electrophysiology' page.

[JoeZiminski]

As discussed in #53, this PR adds a 'Electrophysiology' page and changes 'Data Analysis' to 'Behaviour' page. The electrophysiology page has added:

• getting started - quick overview on resources to get started
• examples - python and matlab code from researchers in the building for analysing ephys data
• community - some links to various community resources
• resources - list of interesting reading.

Reviewing:

• I think the changes to the website-generation machinery are quite minimal, bar adding sphinx-gallery
• I am interested to know if the text reads okay and example pipeline blurbs are clear. It might easiest to build to read for review.
• @neuroinformatics-unit/behaviour this rearranges the way the behaviour section is displayed, let me know if you have any feedback / don't like it.

[adamltyson]

This is excellent @JoeZiminski, I think this will be incredibly useful for SWC (and other) researchers. I have some minor comments, and I think the roadmap should be moved out of this website.

[JoeZiminski]

Hey @adamltyson I've added a very rough version of the example pipelines just to get an idea of the layout. Basically the type of pipeline in use in the building that it would be nice to include roughly falls into three categories:

1. spikeinterface scripts for preprocessing and sorting (suitable for sphinx-gallery)
2. full project repos that use python (often spikeinterface) but are fully integrated with multimodal analysis (e.g. Steve's pipeline)
3. (2) but for MATLAB pipelines (e.g. Mateo's pipeline)

It seems important to include all three categories on the website (e.g. even though we recommend SI there is a lot of valuable methods in those matlab pipelines). But i'm not sure how best to display the information in a non-clunky way. ATM the front page of the sphinx gallery contains preprocessing/sorting scripts in the usual gallery way, but also links to two other pages detailing python and matlab repos. Do you think this works? (and the general approach, giving some backround of named researchers, their setups and studies).

[JoeZiminski]

Ah, now understand that incredibly annoying error! Build was failing with unpinned sphinx with the below error. I was lazily ignoring warnings because it was building fine locally, I think because I had contaminated my env with ipykernel. However CI was failing with the below error:

Notebook error:
NoSuchKernel in ephys_at_swc/gallery/sara_mederos.ipynb:
No such kernel named python3

and was fixed by pinning to sphinx<7.2. The problem was I ignored the following error in latest shpinx:

WARNING: multiple files found for the document "ephys_at_swc/gallery/sara_mederos": ['ephys_at_swc/gallery/sara_mederos.py', 'ephys_at_swc/gallery/sara_mederos.ipynb', 'ephys_at_swc/gallery/sara_mederos.rst']
Use '/Users/joeziminski/git_repos/HowTo/docs/source/ephys_at_swc/gallery/sara_mederos.ipynb' for the build.

so it was trying to use .ipynb for the build by default in later sphinx. In earlier sphinx, it uses the .rst by default:

WARNING: multiple files found for the document "ephys_at_swc/gallery/sara_mederos": ['ephys_at_swc/gallery/sara_mederos.py', 'ephys_at_swc/gallery/sara_mederos.ipynb', 'ephys_at_swc/gallery/sara_mederos.rst']
Use '/Users/joeziminski/git_repos/HowTo/docs/source/ephys_at_swc/gallery/sara_mederos.rst' for the build.

and fixing the warning with the below means it builds okay on most recent sphinx 🎉. I guess a good lesson to always fix warnings first even if they seem unrelated to the error!

exclude_patterns = [
   ...
    "ephys_at_swc/gallery/*.ipynb",
    "ephys_at_swc/gallery/*.py",
]

[adamltyson]

@JoeZiminski I think the general idea is good, with the obvious caveats that:

• Everyone is properly credited
• They're happy with this all being online, and they're aware that it's fully public etc.

[adamltyson]

Hey @JoeZiminski this is excellent. I've reviewed afresh, sorry if anything I say contradicts my original reviews!

I've left some comments, but I think the main thing to check is conflicts with recent changes to the behavioural guides.

Also, as you mentioned something needs to be done about the private link, either get it made public (ideal) or remove that section for now.

Assuming everyone mentioned is happy for this to be shared, I think this is nearly ready to go!

[adamltyson]

Any chance 3.12 works? Just gives us slightly more runway before it needs changing.

[JoeZiminski]

Good idea, lets see 🤞

[adamltyson]

It's not immediately clear how the scripts section is different to the other parts above.

[JoeZiminski]

Hmm yes, initially I envisioned it as all scripts but then it ended up as mostly repos, though some people wills end scripts later. How do you think this could be made clearer? There is no real distinction, only that some are standalone scripts so are suitable for sphinx-gallery, whereas others are organised as part of later repos

[adamltyson]

"Standalone scripts"?