Advanced Clinical Decision Trees — treeadvanced • ClinicoPath

Advanced decision tree analysis with enhanced CART algorithm, comprehensive validation methods, hyperparameter tuning, and detailed clinical performance assessment for medical research and clinical decision support.

Usage

treeadvanced(
  data,
  vars = NULL,
  facs = NULL,
  target,
  targetLevel,
  train = NULL,
  trainLevel,
  validation = "repeated_cv",
  cv_folds = 5,
  cv_repeats = 3,
  bootstrap_samples = 200,
  stratified_sampling = TRUE,
  test_split = 0.25,
  hyperparameter_tuning = FALSE,
  tuning_method = "grid",
  tuning_metric = "bacc",
  max_depth_range = "3:8",
  cp_range = "0.001:0.1",
  min_samples_range = "10:50",
  pruning_method = "one_se",
  custom_cp = 0.01,
  auto_prune = TRUE,
  show_cp_analysis = FALSE,
  splitting_criterion = "gini",
  surrogate_splits = TRUE,
  max_surrogate = 5,
  competing_splits = 4,
  cost_sensitive = FALSE,
  clinical_loss_preset = "equal",
  fn_fp_cost_ratio = 2,
  prevalence_adjustment = FALSE,
  population_prevalence = 10,
  feature_selection = FALSE,
  feature_selection_method = "rfe",
  max_features = 10,
  show_tree_plot = TRUE,
  show_performance_metrics = TRUE,
  show_confusion_matrix = TRUE,
  show_importance_plot = TRUE,
  show_validation_curves = FALSE,
  show_calibration_plot = FALSE,
  show_roc_curve = TRUE,
  bootstrap_confidence = FALSE,
  n_bootstrap = 500,
  clinical_context = "diagnosis",
  show_clinical_interpretation = TRUE,
  clinical_importance_interpretation = TRUE,
  mdg_threshold = 1,
  show_feature_contributions = FALSE,
  survival_integration = FALSE,
  set_seed = TRUE,
  seed_value = 42
)

Arguments

data: The data as a data frame for advanced tree analysis.
vars: .
facs: .
target: .
targetLevel: .
train: .
trainLevel: .
validation: .
cv_folds: .
cv_repeats: .
bootstrap_samples: .
stratified_sampling: .
test_split: .
hyperparameter_tuning: .
tuning_method: .
tuning_metric: .
max_depth_range: .
cp_range: .
min_samples_range: .
pruning_method: .
custom_cp: .
auto_prune: .
show_cp_analysis: .
splitting_criterion: .
surrogate_splits: .
max_surrogate: .
competing_splits: .
cost_sensitive: .
clinical_loss_preset: .
fn_fp_cost_ratio: .
prevalence_adjustment: .
population_prevalence: .
feature_selection: .
feature_selection_method: .
max_features: .
show_tree_plot: .
show_performance_metrics: .
show_confusion_matrix: .
show_importance_plot: .
show_validation_curves: .
show_calibration_plot: .
show_roc_curve: .
bootstrap_confidence: .
n_bootstrap: .
clinical_context: .
show_clinical_interpretation: .
clinical_importance_interpretation: .
mdg_threshold: .
show_feature_contributions: .
survival_integration: .
set_seed: .
seed_value: .

Value

A results object containing:

`results$instructions`					a html
`results$modelsummary`					a html
`results$tuningresults`					a html
`results$performancetable`					a table
`results$confusionmatrix`					a table
`results$variableimportance`					a table
`results$treeplot`					an image
`results$importanceplot`					an image
`results$validationcurves`					an image
`results$calibrationplot`					an image
`results$roccurve`					an image
`results$clinicalinterpretation`					a html
`results$clinicalimportance`					a table
`results$featurecontributions`					a table
`results$survivalintegration`					a html

Tables can be converted to data frames with asDF or as.data.frame. For example:

results$performancetable$asDF

as.data.frame(results$performancetable)

Examples

# Advanced analysis with hyperparameter tuning
tree_advanced(
    data = clinical_data,
    vars = c("biomarker1", "biomarker2", "age"),
    facs = c("grade", "stage"),
    target = "outcome",
    targetLevel = "positive",
    validation = "repeated_cv",
    hyperparameter_tuning = TRUE,
    show_cp_analysis = TRUE,
    cost_sensitive = TRUE
)