plot.ga_parsimony-method {GAparsimony} | R Documentation |
Plot of GA evolution of elitists
Description
The plot
method for ga_parsimony-class
objects gives a evolution plot of the
validation and testing errors, and the number of model features selected of elitists.
Usage
## S4 method for signature 'ga_parsimony'
plot(x, general_cex = 0.7, min_ylim=NULL, max_ylim=NULL,
min_iter=NULL, max_iter=NULL,
main_label="Boxplot cost evolution",
iter_auto_ylim=3, steps=5, pos_cost_num=-3.1,
pos_feat_num=-1.7, digits_plot=4, width_plot=12,
height_plot=6, window=TRUE, ...)
Arguments
x |
An object of class |
general_cex |
Main text scale. |
min_ylim |
Min limit on the y-axis. |
max_ylim |
Max limit on the y-axis. |
min_iter |
Min GA iteration to visualize. |
max_iter |
Max GA iteration to visualize. |
main_label |
Main plot title. |
iter_auto_ylim |
If it is not NULL, GA iteration to choose the min limit of y-axis. |
steps |
Number of divisions in y-axis. |
pos_cost_num |
Relative position of numbers in cost axis. |
pos_feat_num |
Relative position of numbers in feature axis. |
digits_plot |
Number of digits to visualize. |
width_plot |
Figure width in inches. |
height_plot |
Figure height in inches. |
window |
If TRUE shows a new window. |
... |
Further arguments, currently not used. |
Details
Plot method shows the evolution of validation and testing errors, and the number of model features selected of elitists. White and grey box-plots represent validation and testing errors of elitists evolution, respectively. Continuous and dashed-dotted lines show the validation and testing error of the best individual for each generation, respectively. Finally, the shaded area delimits the maximum and minimum number of features, and the dashed line, the number fo features of the best individual.
Value
The method invisibly return a list with the elistists validation error, testing error and model complexity in the whole GA process.
Author(s)
Francisco Javier Martinez de Pison. fjmartin@unirioja.es. EDMANS Group. https://edmans.webs.com/
See Also
ga_parsimony
, ga_parsimony-class
.