score_fitness2: Scores fitness for an entire generation of creatures

In AndrewM1130/GA: On Multivariate Feature Selection with Genetic Algorithms

Scores fitness for an entire generation of creatures

Description

This takes a generation_matrix representing all the genes of a single generation, the data, and user input describing the metric such as R2, AIC, BIC, and AICC and can even use a custom function defined by the user. The underlying fitness function for R2 is lm() and the underlying fitness function for AIC, BIC, and AICc is glm(). AIC, BIC, and AICC has the option to specify a specific family of functions however, the data needs to be defined across the support for the given family for instance exponential must be greater than 0.

Usage

score_fitness2(
  gene,
  data,
  metric = "AIC",
  family = "gaussian",
  custom_function = NULL,
  fittest = "high"
)

Arguments

gene	matrix with all 0s and 1s with ncol = gene_length and nrow = pop
data	a matrix or dataframe with ncol = gene_length plus 1
metric	a character, a user specified statistic such as R2, AIC, BIC, or AICc
family	, model family corresponding to GLM. Defaults to 'gaussian'
custom_function	defaults to NULL, if defined takes a single row from a generation_matrix and returns a numeric value
fittest	whether a custom function has the highest value corresponding to the fittest or the lowest value

Value

returns a list with the first element being a vector of processed fitness scores to for continued use by the algorithm and element 2 being true (user specified) values user reference, for instance true AIC values

Examples

gene <- matrix(rbinom(2*10,1,.5),nc=10)
data <- matrix(rnorm(11*3),nc=11)
pop_required <- 10
metric <- 'AIC'

score_vec <- score_fitness2(data = data,gene = gene,metric = metric)

AndrewM1130/GA documentation built on July 9, 2022, 11:43 a.m.