Chris Gorgolewski

The full collection files are created on the fly on the server side to avoid data duplication. This works well for most collections but fails for big ones which require...

Having said that if you want to contribute a more robust way of providing large collections (that does not lead to redundancy) I would be more than happy to merge...

It comes in multiple compressed CSV files and is converted into single HDF5 file optimized for indexing. More details at https://github.com/NeuroVault/NeuroVault/blob/master/ahba_docker/preparing_AHBA_data.py

That's a lot of questions. Maybe it would help if you provided some context about what you are trying to achieve.

What's an auditory of an image? Python 3 is definitely possible on jessie (see https://github.com/docker-library/python/blob/341c752e5435f4cf4c008fbae67ae4b5b6209a02/3.6/jessie/Dockerfile)

More on the topic: http://zoompf.com/blog/2010/01/should-you-use-javascript-library-cdns It seems we should host all JS libraries that are not hosted by google. DNS costs will be too high.

Thanks. You are absolutely right. We are planning to add this option. There are obvious usability issues how to do it correctly. We'll also need to look into Papaya features...

Rebuilding this image is too time consuming to do it on CI. It needs to be built and pushed manually. While you still work on this could you temporarily push...

I assume you did `docker login` first and are pushing an imaged named `/ahba:py3`.

Steps to reproduce 1. Add two collections, both of the should have DOI field filled. 2. Add one map to each collection 3. Use "Find similar" to record the similarity...