ParallelismConfig

Configuration of basic parameters for parallel processing. Please note that parallel computing is only justified for really complex operations. Computationally simple genetic operators such as selection or mutation may be more efficient to execute serially than in parallel.

If you really need parallelism for your computations, you can manually limit the execution of simple genetic operators by using NO_PARALLELISM to improve performance:

evolution {
selTournament(size = 3, parallelismLimit = NO_PARALLELISM)
}

Creates with ParallelismConfig.

See also

Inheritors

Types

Link copied to clipboard
object Companion

Properties

Link copied to clipboard
abstract val dispatcher: CoroutineDispatcher

dispatcher - dispatcher for parallel coroutines.

Link copied to clipboard

Property for displaying availability of parallel mode of operation.

Link copied to clipboard
abstract val workersCount: Int

workersCount - the number of launched coroutines which can be executed in parallel processing.

Functions

Link copied to clipboard

Deep clones the current parallelism configuration.