hnn-core icon indicating copy to clipboard operation
hnn-core copied to clipboard

[WIP] Implement methods to calculate and visualize current source density signals (for GSOC 2022)

Open chenghuzi opened this issue 2 years ago • 6 comments

This is an initial try to the GSOC 2022 project idea . We build a CSD visualization API based on several CSD algorithms borrowed from https://github.com/espenhgn/iCSD.

With many parameters missing and no tests included, this is just a proof of concept.

chenghuzi avatar Mar 20 '22 02:03 chenghuzi

@chenghuzi I can see from here that you can code in Python and are able to make a pull request. Which is great and a positive sign! I think for the CSD implementation I'd prefer if you started with a simple implementation first rather than copying the entire icsd code into our codebase. It makes our job as reviewers easier. Anyhow, that's for the summer if you get selected.

For now, I propose:

  • implementing a simple PR that can be accepted before the GSoC application deadline. You can pick one of the "good first issues". Maybe #473 or #445 is a good place to start.
  • starting a google doc with the application that you can share with us (@rythorpe @stephanie-r-jones @ntolley ). We can discuss the details of the timeline and the CSD implementation and API there

you can keep this PR open to revisit at a future date if necessary

jasmainak avatar Mar 20 '22 04:03 jasmainak

This is really great work! I agree with @jasmainak that should discuss more before committing to the iCSD package entirely.

(for after getting one of the recommended issues resolved): starting with a PR implementing the second spatial derivative method will allow us to figure out a lot of the API for visualization/calculation.

ntolley avatar Mar 21 '22 09:03 ntolley

@chenghuzi have you had time to prepare a draft of the proposal yet? Even a rough draft is a good starting point. Here is a template provided by the INCF. If you email me privately at mainakjas [at] gmail [dot] com I can provide you my application from many years ago.

Also, how attached are you to the CSD project? We also have a GUI project that is more high priority for us and has no takers yet. Of course, if you prefer the CSD project, let's pitch a proposal with that. What's important is that you pick a project that you feel is your own and can show off at the end of the summer.

jasmainak avatar Mar 25 '22 14:03 jasmainak

@chenghuzi have you had time to prepare a draft of the proposal yet? Even a rough draft is a good starting point. Here is a template provided by the INCF. If you email me privately at mainakjas [at] gmail [dot] com I can provide you my application from many years ago.

Also, how attached are you to the CSD project? We also have a GUI project that is more high priority for us and has no takers yet. Of course, if you prefer the CSD project, let's pitch a proposal with that. What's important is that you pick a project that you feel is your own and can show off at the end of the summer.

Thanks for the reminder and suggestions! I'm working on a proposal draft since this week. Hopefully I will be able to share with you and other mentors by this weekend.

For the project topic, I'm open to possibilities across whole HNN project and happy to contribute not only parts that are included in the GSoC project but also other interesting issues. And the ipywidget definitely interests me. I'm just wondering if the topic selection will influence my final proposal acceptance. If so I can change the project to the GUI stuff of course or I'll stick with the current unfinished CSD proposal if not.

chenghuzi avatar Mar 25 '22 22:03 chenghuzi

@chenghuzi I discussed this with the other mentors. How about proposing the GUI as the main project and the LFP/CSD visualization as an extension to the existing GUI so you have the best worlds. You start with finishing the GUI for the first half of the summer, then in the second half develop the API for the CSD implementation and incorporate it into the GUI. What do you think?

jasmainak avatar Mar 27 '22 15:03 jasmainak

@chenghuzi I discussed this with the other mentors. How about proposing the GUI as the main project and the LFP/CSD visualization as an extension to the existing GUI so you have the best worlds. You start with finishing the GUI for the first half of the summer, then in the second half develop the API for the CSD implementation and incorporate it into the GUI. What do you think?

I did so 😄 and just shared my google doc draft with you via your gmail address. Please feel free to edit or leave comments. Also we can schedule a zoom meeting if you think there're more to discuss about the draft. If other mentors want to check with proposal, please send me their gmail addresses so I can share.

chenghuzi avatar Mar 28 '22 18:03 chenghuzi