In ModelOriented/forester: Quick and Simple Tools for Training and Testing of Tree-Based Models

```{css, echo=FALSE} body .main-container { max-width: 1820px !important; width: 1820px !important; } body { max-width: 1820px !important; width: 1820px !important; font-family: Helvetica !important; font-size: 16pt !important; } h1,h2,h3,h4,h5,h6{ font-size: 24pt !important; }

# Downloads

The necessary downloads required for the forester package to work properly, if downloaded, the user can skip this part.

```r
install.packages("devtools")
devtools::install_github("ModelOriented/forester")
devtools::install_github('catboost/catboost', subdir = 'catboost/R-package')
devtools::install_github('ricardo-bion/ggradar', dependencies = TRUE)
install.packages('tinytex')
tinytex::install_tinytex()

Imports

Importing the forester library handles everything for us.

library(forester)

Train

In the beginning we import a survival analysis task called peakV02, and later proceed with the forester training. Let's notice that this time we don't provide any y value, as the task is described by the pair ttodead, and died. The first step is the description of our dataset, which outlines multiple issues with the dataset, even though the forester successfully prepares the results.

data(peakVO2, package = 'randomForestSRC')
results <- train(
  data             = peakVO2,
  y                = NULL,
  time             = 'ttodead',
  status           = 'died',
  type             = 'auto',
  verbose          = TRUE,
  train_test_split = c(0.6, 0.2, 0.2),
  split_seed       = NULL,
  bayes_iter       = 10,
  random_evals     = 10,
  metrics          = 'auto',
  sort_by          = 'auto'
)

Results

Finally we can see the results presented as a ranked list of evaluated models. The models are compared with the Brier Score (the smaller the better), and Concordance Index (the bigger the better).

results$score_test

results$score_valid

ModelOriented/forester documentation built on June 6, 2024, 7:29 a.m.