Package index • ggRandomForests

Forest Objects

Extract and display data from random forest objects.

gg_rfsrc(<rfsrc>): Predicted response data object
plot(<gg_rfsrc>): Predicted response plot from a gg_rfsrc object.

Training Error

Visualize forest convergence and training error.

gg_error(): Random forest error trajectory data object
plot(<gg_error>): Plot a gg_error object

Variable Importance

Assess and plot variable importance (VIMP).

gg_vimp(): Variable Importance (VIMP) data object
plot(<gg_vimp>): Plot a gg_vimp object, extracted variable importance of a rfsrc object
gg_varpro(): Variable importance data from a varPro model
plot(<gg_varpro>): Plot a gg_varpro variable importance object
gg_beta_varpro(): Per-variable lasso-beta importance from a varPro fit
plot(<gg_beta_varpro>): Plot a gg_beta_varpro object
gg_beta_uvarpro(): Per-variable lasso-beta importance from an unsupervised varPro fit
plot(<gg_beta_uvarpro>): Plot a gg_beta_uvarpro object
gg_udependent(): Variable dependency graph from a uvarpro model
plot(<gg_udependent>): Plot a gg_udependent variable dependency graph
gg_sdependent(): Signal-variable detection from an unsupervised varPro fit
plot(<gg_sdependent>): Plot a gg_sdependent object
gg_ivarpro(): Individual (local) variable importance from a varPro fit
plot(<gg_ivarpro>): Plot a gg_ivarpro object

SHAP Analysis

Per-observation additive SHAP explanations for regression and classification forests.

gg_shap(): SHAP (Shapley additive explanations) data object
plot(<gg_shap>): Plot a gg_shap object
shap_importance(): SHAP global importance bar chart
shap_beeswarm(): SHAP beeswarm summary plot
shap_dependence(): SHAP dependence plot

Anomaly Detection

Score and plot per-observation anomaly scores.

gg_isopro(): Tidy data from a varPro isolation-forest fit
plot(<gg_isopro>): Plot a varPro isolation-forest anomaly score

Variable Dependence

Marginal variable dependence plots.

gg_variable(): Marginal variable dependence data object.
plot(<gg_variable>): Plot a gg_variable object,

Partial Dependence and Accumulated Local Effects

Marginal-effect plots for individual variables. Partial dependence averages over the other predictors; accumulated local effects perturb a predictor only within neighbourhoods of its own observed values, and so do not extrapolate when predictors are correlated.

gg_partial(): Split partial dependence data into continuous or categorical datasets
plot(<gg_partial>): Plot a gg_partial object
gg_partial_rfsrc(): Partial dependence data from an rfsrc model
plot(<gg_partial_rfsrc>): Plot a gg_partial_rfsrc object
gg_partial_varpro() gg_partialpro(): Partial dependence data from a varPro model
plot(<gg_partialpro>) plot(<gg_partial_varpro>): Plot a gg_partial_varpro object
gg_ale_rfsrc(): Accumulated Local Effects (ALE) data from an rfsrc model
plot(<gg_ale_rfsrc>): Plot a gg_ale_rfsrc object
plot(<gg_ale_interaction>): Plot a gg_ale_rfsrc interaction object

Survival Analysis

Survival curves, ROC, and related diagnostics.

gg_survival(): Nonparametric survival estimates.
plot(<gg_survival>): Plot a gg_survival object.
gg_roc(<rfsrc>): ROC (Receiver Operating Characteristic) curve data from a classification forest.
plot(<gg_roc>): ROC plot generic function for a gg_roc object.
calc_roc(<rfsrc>): Receiver Operator Characteristic calculator
calc_auc(): Area Under the ROC Curve calculator
surv_partial.rfsrc(): Survival partial dependence data for one or more predictors
kaplan(): nonparametric Kaplan-Meier estimates
nelson(): nonparametric Nelson-Aalen estimates
gg_brier(): Brier score and CRPS for survival forests
plot(<gg_brier>): Plot a gg_brier object

Random Hazard Forests

For time-to-event data with predictors that change during follow-up.

gg_rhf(): Tidy hazard and cumulative-hazard curves from a Random Hazard Forest
plot(<gg_rhf>): Plot Random Hazard Forest hazard / cumulative-hazard curves
gg_auct(): Tidy time-varying AUC from a Random Hazard Forest
plot(<gg_auct>): Plot a time-varying AUC curve
gg_rhf_importance(): Tidy time-localized variable priority from a Random Hazard Forest
plot(<gg_rhf_importance>): Plot Random Hazard Forest variable priority over time
gg_tune_rhf(): Tidy a Random Hazard Forest tuning path
plot(<gg_tune_rhf>): Plot a Random Hazard Forest tuning path

S3 Methods

Standard R generics implemented for all gg_* data objects.

autoplot(<gg_error>) autoplot(<gg_vimp>) autoplot(<gg_rfsrc>) autoplot(<gg_variable>) autoplot(<gg_partial>) autoplot(<gg_partial_rfsrc>) autoplot(<gg_partialpro>) autoplot(<gg_partial_varpro>) autoplot(<gg_roc>) autoplot(<gg_survival>) autoplot(<gg_brier>) autoplot(<gg_rhf>) autoplot(<gg_tune_rhf>) autoplot(<gg_auct>) autoplot(<gg_rhf_importance>) autoplot(<gg_varpro>) autoplot(<gg_udependent>) autoplot(<gg_isopro>) autoplot(<gg_shap>) autoplot(<gg_ale_rfsrc>) autoplot(<gg_ale_interaction>): autoplot methods for ggRandomForests data objects
print(<gg_beta_uvarpro>) print(<gg_sdependent>) print(<gg_error>) print(<gg_vimp>) print(<gg_rfsrc>) print(<gg_variable>) print(<gg_partial>) print(<gg_partial_rfsrc>) print(<gg_partialpro>) print(<gg_partial_varpro>) print(<gg_roc>) print(<gg_survival>) print(<gg_brier>) print(<gg_rhf>) print(<gg_tune_rhf>) print(<gg_auct>) print(<gg_rhf_importance>) print(<gg_udependent>) print(<summary.gg_udependent>) print(<gg_varpro>) print(<gg_isopro>) print(<gg_beta_varpro>) print(<gg_ivarpro>) print(<gg_shap>) print(<gg_ale_rfsrc>) print(<gg_ale_interaction>): Print methods for gg_* data objects
summary(<gg_beta_uvarpro>) summary(<gg_sdependent>) print(<summary.gg>) summary(<gg_error>) summary(<gg_vimp>) summary(<gg_rfsrc>) summary(<gg_variable>) summary(<gg_partial>) summary(<gg_partial_rfsrc>) summary(<gg_partialpro>) summary(<gg_partial_varpro>) summary(<gg_roc>) summary(<gg_survival>) summary(<gg_varpro>) summary(<gg_udependent>) summary(<gg_brier>) summary(<gg_rhf>) summary(<gg_tune_rhf>) summary(<gg_rhf_importance>) summary(<gg_isopro>) summary(<gg_beta_varpro>) summary(<gg_ivarpro>) summary(<gg_auct>) summary(<gg_shap>) summary(<gg_ale_rfsrc>) summary(<gg_ale_interaction>): Summary methods for gg_* data objects

Utilities

Helper functions used internally and by users.

quantile_pts(): Quantile-based cut points for coplots
varpro_feature_names(): Recover original variable names from varpro one-hot encoded feature names

Package

ggRandomForests-package: ggRandomForests: Visually Exploring Random Forests