first_generation: first_generation

In Skjemaa/GA: Selects covariates for use in regression using genetic algorithm

Description

Performs the first iteration of the Genetic algorithm Generates the first individuals

Usage

first_generation(y, dataset, population_size, interaction = F, most_sig = F,
  objective_function = "AIC", reg_method)

Arguments

y	the response variable in the dataframe 'dataset'
dataset	the data frame containing the variables in the model
population_size	the number of individuals to generate
interaction	a boolean indicating if we want to take into account terms coming from interactions, False by default
most_sig	a boolean indicating if we want to only consider the most significant covariates
objective_function	The objective criterion to use (default AIC).
reg_method	regression method

Value

a list of individuals each a list with fields:

variable	the covariates kept in the linear model
indices	the indices of these covariates in the data frame
linear_model	the linear model with these covariates

Skjemaa/GA documentation built on May 3, 2019, 6:42 p.m.