ecr_parallelization

In ecr it is possible to parallelize the fitness function evaluation
to make use, e.g., of multiple CP cores or nodes in a HPC cluster.
For maximal flexibility this is realized by means of the parallelMap package
(see the <a href="https://github.com/mlr-org/parallelMap">official
GitHub page</a> for instructions on how to set up parallelization).
The different levels of parallelization can be specified in the
<code>parallelStart*</code> function. At them moment only the level
&#8220;ecr.evaluateFitness&#8221; is supported.
Keep in mind that parallelization comes along with some overhead. Thus activating
parallelization, e.g., for evaluation a fitness function which is evaluated
lightning-fast, may result in higher computation time. However, if the function
evaluations are computationally more expensive, parallelization leads to
significant running time benefits.

Framework for building evolutionary algorithms for both single- and multi-objective continuous or discrete optimization problems. A set of predefined evolutionary building blocks and operators is included. Moreover, the user can easily set up custom objective functions, operators, building blocks and representations sticking to few conventions. The package allows both a black-box approach for standard tasks (plug-and-play style) and a much more flexible white-box approach where the evolutionary cycle is written by hand.

Jakob Bossek

Evolutionary Computation in R

Michael H. Buselli

Wessel Dankers

Carlos M. Fonseca

Manuel Lopez-Ibanez

Luis Paquete

Joshua Knowles

Eckart Zitzler

Olaf Mersmann

ecr_parallelization function

In ecr it is possible to parallelize the fitness function evaluation
to make use, e.g., of multiple CP cores or nodes in a HPC cluster.
For maximal flexibility this is realized by means of the parallelMap package
(see the <a href='https://github.com/mlr-org/parallelMap'>official
GitHub page</a> for instructions on how to set up parallelization).
The different levels of parallelization can be specified in the
<code>parallelStart*</code> function. At them moment only the level
&#8220;ecr.evaluateFitness&#8221; is supported.
Keep in mind that parallelization comes along with some overhead. Thus activating
parallelization, e.g., for evaluation a fitness function which is evaluated
lightning-fast, may result in higher computation time. However, if the function
evaluations are computationally more expensive, parallelization leads to
significant running time benefits.

Parallelization in ecr — ecr_parallelization

Parallelization in ecr

ecr_parallelization: Parallelization in ecr

Description

Arguments