Description Usage Arguments Details Value Examples
provides an interface for benchmarking optimization algorithms over all 24 noiseless bbob functions.
Use either bbob_custom
or bbob_custom_parallel
for parallel execution.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 | bbob_custom(optimizer, algorithm_id, data_directory, dimensions = c(2, 3, 5,
10, 20, 40), instances = c(1:5, 41:50), function_ids = NULL,
maxit = NULL, stopFitness = NULL, maxFE = NULL, debug.logging = FALSE,
max_restarts = 0, restart_multiplier = 1,
restart_triggers = character(0), OCD = FALSE, varLimit = NULL,
nPreGen = NULL, maxGen = NULL, fitnessValue = FALSE,
dispersion = FALSE, evolutionPath = FALSE)
bbob_custom_parallel(optimizer, algorithm_id, data_directory,
dimensions = c(2, 3, 5, 10, 20, 40), instances = c(1:5, 41:50),
function_ids = NULL, maxit = NULL, stopFitness = NULL, maxFE = NULL,
debug.logging = FALSE, max_restarts = 0, restart_multiplier = 1,
restart_triggers = character(0), OCD = FALSE, varLimit = NULL,
nPreGen = NULL, maxGen = NULL, fitnessValue = FALSE,
dispersion = FALSE, evolutionPath = FALSE)
|
optimizer |
The first argument passed to customized_bbob is an optimization wrapper, i.e. the particular optimizer under test.
The optimizer has to adhere to the following function signature:
|
data_directory |
The |
dimensions |
The |
instances |
Every |
function_ids |
The |
maxit |
If |
stopFitness |
If |
maxFE |
|
debug.logging |
|
OCD |
|
varLimit |
|
nPreGen |
|
maxGen |
|
fitnessValue |
|
dispersion |
|
evolutionPath |
|
algorithm.id |
The |
max.restarts |
If |
restart.triggers |
If |
bbob_custom provides a benchmarking function as a highlevel interface for running experiments
for an optimizer over all 24 noiseless functions of the bbob testset. bbob_custom_parallel does the same in an
parallel manner in order to save execution time. The input parameters need to be specified as defined
above. Once the execution of an experiment has finished, bbob_custom created a folder according to the datapath specified and
writes the results in a number of log files to this folder. The folder's name includes a time stamp as well as the specified
data directory (e.g. 2015-12-31_CMAES_default_stopping_conditions). The caption of the log files in that folder is a sequence of the defined
algorithm.id, function_ids
and dimensions
(e.g. CMAES_output_1_2.txt).
bbob_custom does not return anything but writes the results of the experiment to log files, to be
processed with readOutput
1 2 3 4 5 6 7 | suppressWarnings(bbob_custom_parallel(optimizer = cmaesbenchmarking::optimizerCMAES, algorithm_id = "CMAES_OCD",
data_directory = "CMAES_OCD_no_restarts",
dimensions = c(2, 5, 10, 20), instances = 1:15, function_ids = 1:24, maxit = NULL,
stopFitness = 1e-08, maxFE = 100000, max_restarts = 0,
OCD = TRUE, varLimit = 0.0001, nPreGen = 100, fitnessValue = TRUE,
dispersion = FALSE, evolutionPath = FALSE, restart_multiplier = 2,
restart_triggers = "OCD"))
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.