Confluence [v5] Priors: Major Change

[nikki-t]

Scope: Major Change  ·  Auxiliary · New capability

The change: Overhaul gauge operations and update SHCQ usage; formalize GBPriors to pull ML priors from Zenodo (dropping the constrained/unconstrained split).

Input

• SHCQ data
• Upstream: None

Output

• SHCQ-based priors (may overwrite the "model" values)
• Output will include the new priors
• MOI and Offline may use the new data
• Downstream: all modules

Open questions

• Where does the SHCQ data come from?

Tasks

• Evaluate the current architecture; remove what is no longer used and break it into parallel modules @nikki-t
• Ensure the output is the same as expected by the downstream modules
• If there is time and we break into multiple modules: update module-level and top-level Terraform @nikki-t
• Track the work from MOI for the gauge split
• Complete an offline run to test changes and evaluate results