When TVB runs simulations on cortical surfaces, we need to be able and compute geodesic distances (distance on the surface) instead of the trivial Euclidean distances. For this calculus, we have a small C++ library, which had become outdated. We need to:
start with an analysis done by the student if the current implementation should be reused and fixed, or completely replaced, then
proceed with the fix/replacement as concluded at the previous step.
If we are to fix the current implementation, we need to fix the 6 issues reported on Github during this project, and also
make sure the library compiles correctly with the latest version of clang.
we need to have unit-test written for the main flows as well as for some common exceptions
the unit-tests should run automatically by integration in our CI Jenkins system
at the end of this project, also tvb-gdist packages on Pypi and conda-forge should be updated.
Hi @Jedi18,
Thanks for your interest. You can tag the mentors here @liadomide. We will share the email/other contact info of mentors soon. In the meantime, please start looking into the github links that are shared in the project idea. They will give you a good idea of the code base. Also, you will notice on following the repository contributors link that Lia is one of the contributors. The mentors can share more information regarding what are some good beginner level issues. Requesting Lia and Robert to please help guide you in the right direction here
Hi, Currently we have no beginner issues to assign to you, but the code repo should be self explanatory, being relatively small.
Do not hesitate to contact us (ty tagging us in here) in case you have further questions.
@liadomide
Hello, Im Hriday, I really found this project interesting and want to contribute to this project this summer in GSOC 2020, Ive gone through the codebase. please guide me to the next step.
and can we redo the whole “geodesic_library” using some modern c++?
yes, you can use anything as long as you produce a Pypi compatible package for us to use from Python.
I think the next step is now to write the GSOC proposal document, on which we can comment and notify you when something is missing, if you share it the draft with us early enough.
Hi @ayanb,
All these are very good suggestions.
Indeed, we should bring up to date the Travis configuration on tvb-gdist.
I also looked at your PRs, and they are decent suggestions. I put some comments directly there, though.
Best, Lia.
Its too late but this project seems very interested and have some part that look alike what i have worked on GSoC2018 Is there a chance for me to apply for it this year.
I have submitted the proposal PDF in the GSoC platform to avoid the last minute rush. However, please feel to review, and I will make the amendments and submit the PDF once again.