parsimony_Population {GAparsimony} | R Documentation |
Population initialization in GA-PARSIMONY with a combined chromosome of model parameters and selected features
Description
Functions for creating an initial population to be used in the GA-PARSIMONY process.
Usage
parsimony_population(object, type_ini_pop="randomLHS", ...)
Arguments
object |
An object of class |
type_ini_pop |
How to create the initial population. 'random' optiom initialize a random population between the predefined ranges. Values 'randomLHS', 'geneticLHS', 'improvedLHS', 'maximinLHS' & 'optimumLHS' corresponds with several methods of the Latin Hypercube Sampling (see 'lhs' package for more details). |
... |
Further arguments passed to or from other methods. |
Details
parsimony_population
generates a random population of object@popSize
individuals. For each individual a random chromosome is generated with object@nParams
real values in the range [object@min_param
, object@max_param
] plus object@nFeatures
random binary values for feature selection.
'random' or Latin Hypercube Sampling can be used to create a efficient spread initial population.
Value
Return a matrix of dimension object@popSize
rows and object@nParams
+object@nFeatures
columns.
Author(s)
Francisco Javier Martinez de Pison. fjmartin@unirioja.es. EDMANS Group. https://edmans.webs.com/