Migration time

This last 2 weeks I’ve been working in migrating the new solution to the main branch of Neurovault. I’m centering my efforts in making the whole similarity search feature to work. Also, I had a videocall with the mentor and we agreed in a set of deliberables that I would have to complete in order to get the Summer of Code passed.

Deliberables

The minimum set of deliverables agreed with Chris are:

  • Written report of activities (Documentation will be an important part of the work done)

  • Benchmarking finished and merged

  • Refactoring find_similar view to table form, avoiding pybraincompare dependency, using the json format already implemented. Improve the call to the json with length of max comparisons to retrieve trough GET in the url.

  • Research and find solutions for 2 problems:

  • How to maintain Engines (find in the literature), scaling, rebuilding…​

  • Brainmaps compression (use a hashing function that fits our problem)

Desirable: Full implementation of the Engine in a single PR.

I am positive towards completing the objectives and actually have already completed (or almost completed XD) most of them. I’m at the moment advancing in documentation and implementation at the same time.

Migration

This will be a major challenge, since some Models have to be removed, a lot of functions will be changed and is not clear how the new Engine will be saved in Memory or how the insertion of new features will be handled. I’ve achieved to make it work so far, and fixed some workflow errors related to the previous implementation (I want everything to be working as smooth and scalable as possible).

What I have achieved this week is to, at least, create the Engine, index some features and query them. After this, the find_similar view shows the same information as it was showing for the previous implementation. This is a big step, since this will allow me to finally test and finish the benchmark and to be able to see results of the queries to the Engine.

This last 3 weeks are going to be a great challenge, but I feel confident I will be able to achieve all the objectives.