Refactor sample projects generation scripts into a proper tested utility #102
Description
In PR #100, I added scripts/generate_sample_projects to the repository, to keep track of how the sample projects were generated.
@samcunliffe brought up that these could be converted into a more reproducible/tested utility for generating sample projects. Apart from refactoring, this would necessitate the addition of tests, and perhaps some command-line entry points.
There is a roadblock, however. The original sample generation scripts grab data from our private internal data storage server, extract a small subset of it, and reformat it into a directory structure convenient for WAZP. To truly test the scripts, we would also need to publicly share (some of the) bigger datasets, which we cannot do. But perhaps we could modify the scripts so that they can extract samples form already publically available animal behaviour datasets.
This is not a priority right now, but may become more relevant if we want to generate many more sample datasets in future.