Open KseniaBastrakova opened 5 years ago
I think from an ion-acceleration point of view, taking electron data from the FoilLCT PIConGPU example with ionization physics would be a good test for a multi-modal particle distribution (prompt forward electrons and bulk including return-current) which would be great to be reduced.
From an electron-acceleration point of view, reducing the injected particles in an LWFA would help to avoid computational slow-down over long gas jets. Maybe Alex, Klaus and Richard can provide some bunch data for this?
I have several different algorithms for reducing the number of particles. At the moment I’m using statistical metrics to measure quality. Of course, evaluation of how it effects real simulation is also valuable. @ax3l , @BeyondEspresso @steindev @PrometheusPi any ideas what setaps to use for that?