2 ms·
Interesting. I'm working on an "Expfile" project right now, which is tangentially related to this area of exploration. The idea is that published experiments sh
by chatmasta 12y ago
Interesting. I'm working on an "Expfile" project right now, which is tangentially related to this area of exploration. The idea is that published experiments should be reproducible, and therefore can be described in a declarative format. I'm calling Expfile the "Procfile for experiments" -- it's a file you upload in the top level of your research's code repository, describing in declarative terms which variables to measure, how to measure them, and how to get results. If anyone wants to talk to me about this I'm doing a lot of work on it this week.