S-D22.4 Meta-Modeling optimizer module Y4M12

[KZzizzle]

22.4. An optimizer module exists that enables finding parameters that maximize a specified simulated quantity of interest within a provided parameter range. Dependencies between parameters can be specified. While the initial implementation will provide a general purpose, global optimizer, the underlying framework will be implemented flexibly, to readily permit extension by more specialized optimization approaches. Y4M12

[esraneufeld]

todo:

• add API function that permits to extend list of configurations in iterator from within a service

[elisabettai]

As a use-case, we should now use CB's one

[pcrespov]

Goal for sprint Katherine Switzer

MVP that produces optimization example

• design example with Cedric or EN
• create API to inject and execute new configurations
• create ‘master' study where collected results can be further processed through pipeline

Review demo

• Cedric example

[pcrespov]

Update on sprint Katherine Switzer

Done

• design example with Cedric

Ongoing

• create API to inject and execute new configurations
  • ✨ Is355/optimizer projects ports osparc-simcore#3504

Open

• create ‘master' study where collected results can be further processed through pipeline

[pcrespov]

Goal for sprint Athena

• Extend API: create API to inject and execute new configurations
  • WIP: Is355/optimizer clone projects osparc-simcore#3536
  • optimization (run, log, progress, stop).
  • delayed deletion
  • optimization client for o2sparc (o3sparc?) to run optimizations in osparc from python notebooks
• create ‘master' study where collected results can be further processed through pipeline
• Some parts of M4 Creation of Own o²S²PARC Services #693 so that we can easily build optimization examples

Review demo

• run cedric example from a notebook including the creation of a services used in the study to optimizef

[pcrespov]

Update on sprint Athena

Done

• First full workflow for "fast python service integration" ( part of M4 Creation of Own o²S²PARC Services #693 )

Ongoing

• Extend API: create API to inject and execute new configurations
  • WIP: Is355/optimizer clone projects osparc-simcore#3536
• Improvements on "fast python service integration"

Open

• Extend API: create API to inject and execute new configurations
  • optimization (run, log, progress, stop).
  • delayed deletion
• create ‘primary' study where collected results can be further processed through pipeline