Struct ThreadRankMpi 

pub struct ThreadRankMpi { /* private fields */ }

An MPIInterface implementation that simulates size MPI ranks using real OS threads within a single process, so distributed algorithms (algorithms::DistributedDifferentialEvolution, algorithms::DistributedParticleSwarm) can be exercised end-to-end — with genuine cross-rank broadcast/gather/allreduce/send/recv — without an actual MPI runtime or multi-process setup. See spawn_ranks to construct and drive a full communicator.

Every generic method requires T: 'static (per MPIInterface) because values cross the simulated thread boundary via Box<dyn Any + Send>.

Trait Implementations

impl MPIInterface for ThreadRankMpi

fn rank(&self) -> i32

Get the rank of this process

fn size(&self) -> i32

Get the total number of processes

fn broadcast<T>(&self, data: &mut [T], root: i32) -> ScirsResult<()>

where T: Clone + Send + Sync + 'static,

Broadcast data from root to all processes

fn gather<T>( &self, send_data: &[T], recv_data: Option<&mut [T]>, root: i32, ) -> ScirsResult<()>

where T: Clone + Send + Sync + 'static,

Gather data from all processes to root

fn allreduce<T>( &self, send_data: &[T], recv_data: &mut [T], op: ReductionOp, ) -> ScirsResult<()>

where T: Clone + Send + Sync + 'static + Add<Output = T> + PartialOrd,

All-to-all reduction operation

fn barrier(&self) -> ScirsResult<()>

Barrier synchronization

fn send<T>(&self, data: &[T], dest: i32, tag: i32) -> ScirsResult<()>

where T: Clone + Send + Sync + 'static,

Send data to specific process

fn recv<T>(&self, data: &mut [T], source: i32, tag: i32) -> ScirsResult<()>

where T: Clone + Send + Sync + 'static,

Receive data from specific process