[Task]: Optimize to allow for multiple constraints

[Tom-Szendrey]

https://www.figma.com/design/EoihrRoNNbuOAJtudKqp7L/Terarium---June-2024?node-id=1-26777&t=nlgqHpBvqQFvdpEC-0

HMI:
Utilize the constraint group form's isActive for optimizePayload

pyciemss-service:
Update:

• Multiple intervention types in a single run.
• Take multiple constraints:
• Take in multiple alpha values for constraints. This needs to be a part of QoI
• Take in multiple threshold values for constraints. This needs to be a part of QoI
• Optimize: Need to include param value and start time option #4163

[Tom-Szendrey]

• (Done) Start time and parameter value optimization for the same parameter with weights https://github.com/ciemss/pyciemss/blob/main/docs/source/interfaces.ipynb

• (Done) Optimization with different intervention types is merged into main with a template for combining multiple intervention types: Updated optimize to handle combination of multiple intervention templates ciemss/pyciemss#591

• (In Progress) Optimization progress hook (intermediate results) is almost ready to be merged. 477 add progress hook to optimize ciemss/pyciemss#592

• (In Progress) Optimization with multiple constraints needs more stress testing and resolution of some general tensor size issues for pyciemss overall. Here is the open PR: Updating optimize to handle multiple risk constraints ciemss/pyciemss#596

[Tom-Szendrey]

For up coming demo:

• (Requied) Allow obs for QoI.
• (Not required) Multiple intervention types in a single run.
• (Not required) Multi contraints.
• (Desired not required) Intervention on something should reflect in the obs
• (Desired, will be take some time) We want to optimize with a constraint that depends on multiple parameters
  As in B < Y

for @liunelson to test:
Obs_A = B - Y
Constraint = Obs_A > 0

[Tom-Szendrey]

Pending
ciemss/pyciemss#596

[Tom-Szendrey]

First thing to do is to test contradicting constraints