How to use `adaptive_scheduler` independently?

[H-xie]

Hi,
This example brings a one-click deployment for ocs. It's great!

I'm doing some research in scheduler, and want to make observing plan by adaptive_scheduler if possible. Could adaptive_scheduler be used directly by python code? And could you offer a brief sample code?

Thank you!

[jnation3406]

So the adaptive_scheduler project is really quite integrated into the OCS framework, in that it expects its input in the format of Requests in the OCS Observation Portal, and it outputs the schedule in the format of Observations that it sends to the Observation Portal. Half of what the adaptive_scheduler does is massage input requests into the proper format for the scheduling kernel. If you were looking to just use the algorithm of the adaptive_scheduler without any other pieces of the OCS, you could probably reuse all of the code in the kernel package of the scheduler. Along with that, you could check out the kernel_mappings.py file to see some of the pre-filtering we do to the data before running the scheduling kernel. Hope this helps you get started.