[Burn-P3+ FireSTARR FR]: Offload inputs to disk before run

[shreeramsenthi]

Contact Details

No response

What is needed?

Since the memory use of FireSTARR is so much less than the other fire growth models, the memory overhead for R can significantly affect the overall memory use. After preprocessing the inputs, cleaning up and offloading as much of the R environment as possible could significantly increase the number of jobs that can be run in parallel on most machines.

How will this improve the project or tool?

No response

Approvals Process

• Testing For Issue
• Merge

[BadgerOnABike]

I'm also curious if there is a way we can do all the pre-processing, send everything to firestarrr to let it do it's thing and then regain focus in R. Our issues arise in the following desired outcomes:

• Monitoring of jobs (FS does have a reporting structure, but then we need a way to capture that)
• Syncrosim managing parallel processing (I'm not sure this is the most efficient)