Title: | Standard Benchmark Optimization Functions |
Version: | 0.1 |
Description: | A set of standard benchmark optimization functions for R and a common interface to sample them. |
Depends: | R (≥ 3.0) |
License: | MIT + file LICENSE |
LazyData: | true |
Suggests: | testthat, covr |
Imports: | lhs, randtoolbox, stats, stringr |
Collate: | 'optim.functions.R' 'vectorize.R' 'sampling.R' 'single.objective.R' 'functions.R' |
RoxygenNote: | 5.0.1 |
NeedsCompilation: | no |
Packaged: | 2017-03-30 06:13:39 UTC; tom |
Author: | Thomas Torsney-Weir [aut, cre] |
Maintainer: | Thomas Torsney-Weir <torsneyt@gmail.com> |
Repository: | CRAN |
Date/Publication: | 2017-03-30 11:28:25 UTC |
optim.functions: A collection of standard optimization functions along with a standard interface to call and sample those functions.
Description
optim.functions: A collection of standard optimization functions along with a standard interface to call and sample those functions.
Sampling functions
The sampling functions all need to have a standard interface. They all must
take 2 parameters: n
, the number of samples to generate and k
,
the number of dimensions to sample. They all return a matrix-like object
containing n
rows of k
columns of samples.
The main interface to this library is through the
sample.func
function.
Lookup information about a function by name
Description
Lookup information about a function by name
Usage
get_info(fun.name)
Arguments
fun.name |
Function name to lookup |
Unified function sampling interface.
Description
Provides a unified interface to sampling all functions. One just needs to specify the name of the function to sample, the number of samples, the dimensionality, and the sampling method and the library will take care of the rest.
Usage
sample.func(FUN.name, n, k, method = "lh.sample", ...)
Arguments
FUN.name |
The name of the function to sample. |
n |
The number of sample points. |
k |
The number of dimensions to sample. |
method |
The sampling method. |
... |
Additional arguments passed to |
Value
A data frame consisting of n rows with columns x1..xk and y1..yz where z is the number of outputs of the function.
The sampling functions take (n, k) where n is the number of samples and k is the number of dimensions. The sampling functions should return samples in a 0-1 hypercube. sampling.func <- function(n, k)
Description
The sampling functions take (n, k) where n is the number of samples and k is the number of dimensions. The sampling functions should return samples in a 0-1 hypercube. sampling.func <- function(n, k)
Usage
lh.sample(n, k, ...)
cartesian.sample(n, k, ...)
random.sample(n, k, ...)
hexagonal.sample(n, k, ...)
torus.sample(n, k, ...)
sobol.sample(n, k, ...)
halton.sample(n, k, ...)
Arguments
n |
The number of samples. |
k |
The number of dimensions. |
... |
Additional arguments for the sampling (e.g. |