Open for discussion: Future of analysis scripts #11
Description
The way we are headed with common analysis figures is great. What about specific use cases or analyses?
- For example, a component analysis of moist static energy over a specific region may be used for a specific research application, and not necessarily be produced with every model run.
My take: workflow can and should be used for reproducibility (especially creating packages and catalogs) but not everything needs to be included into one freanalysis_[realm] script. My concern is that specific figures will be created and users may not know what they were originally for or what they represent. However, if there are too many scripts it may be hard for discoverability of what’s usable? We need to find the sweet spot where it’s convenient for users to take advantage of what exists but not cumbersome to figure out what exists, or feel like they need to go through ten extra steps in order to create a plot for a newly added variable, for example.