diazrenata
commented
1 year ago use sqlite x RSqlite
(this is what drake used also)
https://solutions.posit.co/connections/db/databases/sqlite/ https://sqlitebrowser.org/dl/ https://swcarpentry.github.io/sql-novice-survey/setup.html
If you ultimately want to plug this into an R package that might be tricky depending on the ultimate size of the db, but, you could still host it online.
The distribution of hill nbs for draws from a feasible set of given s,n is a knowable thing. In moving towards making comparisons to the feasible set using sim models, it's a reasonable expectation that there will be be repeated s,n combinations and that we will want to make a lot of comparisons.
Thinking to start populating a centralized database, essentially a massive table of s, n, hill1, hill2 and then 1000-5000 records for each combination (or the number possible). I could store this as a targets thing but I'd prefer not to...seems like an opportunity to practice databases, overkill for this but good practice.