Package index • BORG

Main Entry Point

Start here

borg(): BORG: Guard Your Model Evaluation
borg_workflow(): Create a BORG Validation Workflow
borg_check(): Quick Leakage Check (Pipe-Friendly)
borg_pipeline(): Validate an Entire Modeling Pipeline

Core Functions

Detailed validation functions

borg_inspect(): Inspect R Objects for Evaluation Risks
borg_validate(): Validate Complete Evaluation Workflow
borg_certificate(): Create Validation Certificate
borg_export(): Export Validation Certificate
borg_explain_risk(): Explain Risks in Plain Language with Actionable Recommendations
borg_cache_get() borg_cache_set() borg_cache_clear() borg_cache_info(): Cache and Retrieve BORG Diagnoses

Remediation

Automatic pipeline correction

borg_assimilate(): Assimilate Leaky Evaluation Pipelines

Spatial Tools

Raster extraction, thinning, blocking, and geographic analysis

borg_extract(): Extract Raster Values at Point Locations for BORG
borg_thin(): Spatially Thin Occurrence Data
borg_block_size(): Optimize Spatial Block Size
borg_check_coverage(): Check Geographic Representativeness of CV Folds
borg_leaflet(): Interactive Leaflet Map of Spatial CV Folds
borg_prediction_map(): Spatial Prediction Uncertainty Map
borg_importance(): Block-Permutation Variable Importance
borg_geodist(): Distance Distribution Diagnostics
borg_clustered_sample(): Draw a Clustered Spatial Sample
borg_sample_design(): Suggest Sampling Locations to Improve AOA
borg_predict_raster(): Predict onto a SpatRaster with AOA Masking
borg_stability_map(): Prediction Stability Map
borg_check_residuals(): Check Residual Spatial Autocorrelation
borg_local_moran(): Local Moran's I for Residuals
borg_debias(): Spatial+ Debiasing for Spatial Confounding

Cross-Validation Schemes

Blocked, buffered and repeated resampling designs

borg_spatial_loo(): Buffered Leave-One-Out Cross-Validation
borg_disc_cv(): Leave-Disc-Out Cross-Validation
borg_repeated_cv(): Repeated Blocked Cross-Validation
borg_global_validation(): Global (Pooled) Cross-Validation Metrics
borg_bootstrap(): Block Bootstrap Confidence Intervals for CV Metrics

Applicability & Transfer

Area of applicability, extrapolation and distribution shift

borg_aoa(): Area of Applicability
borg_di(): Dissimilarity Index
borg_error_profile(): Calibrate Dissimilarity Index to Prediction Error
borg_extrapolation(): Detect Model Extrapolation
borg_fold_similarity(): Per-Fold Environmental Similarity (MESS)
borg_transferability(): Assess Geographic Transferability
borg_drift(): Detect Distribution Shift Between Training and Deployment Data
borg_adversarial(): Adversarial Validation

Model Evaluation

Metrics, calibration, uncertainty and model comparison

borg_metrics(): Available Classification Metrics
borg_willmott(): Willmott's Index of Agreement
borg_calibration(): Model Calibration Diagnostics
borg_conformal(): Conformal Prediction with Spatial Dependence
borg_compare_models(): Compare Multiple Models with Spatial CV
borg_ensemble(): Ensemble Predictions from CV Fold Models
borg_multiscale(): Multi-Scale Performance Assessment
borg_null_test(): Null Model Significance Test
borg_fairness(): Detect Performance Disparities Across Subgroups Under Blocked CV
borg_shap(): Spatial SHAP Values

Feature Selection

Variable selection under blocked CV

borg_forward_selection(): Forward Feature Selection with Spatial/Blocked CV
borg_best_subset(): Best Subset Variable Selection with Blocked CV

Visualization

Plot and summary methods for BORG objects

plot(<BorgRisk>): Plot BORG Objects
plot(<borg_result>): Plot BORG Result Objects
plot(<borg_comparison>): Plot CV Comparison Results
autoplot(<borg_fold_perf>): Autoplot Method for borg_fold_perf Objects
summary(<BorgRisk>): Summarize BORG Risk Assessment
summary(<BorgDiagnosis>): Summarize BORG Diagnosis
summary(<borg_result>): Summarize BORG Result
summary(<borg_cv>): Summarize BORG Cross-Validation
summary(<borg_pipeline>): Summarize BORG Pipeline Validation
summary(<borg_power>): Summarize BORG Power Analysis
autoplot(<BorgRisk>): Autoplot Method for BorgRisk Objects
autoplot(<BorgDiagnosis>): Autoplot Method for BorgDiagnosis Objects
autoplot(<borg_result>): Autoplot Method for borg_result Objects
autoplot(<borg_comparison>): Autoplot Method for borg_comparison Objects
autoplot(<borg_cv>): Autoplot Method for borg_cv Objects

Diagnosis & Evidence

Dependency detection, CV comparison, and power analysis

borg_diagnose(): Diagnose Data Dependency Structure
borg_compare_cv(): Compare Random vs Blocked Cross-Validation
borg_power(): Estimate Statistical Power After Blocking
borg_cv(): Generate Valid Cross-Validation Scheme
borg_fold_performance(): Evaluate Per-Fold Model Performance
borg_stability(): Analyze CV Fold Stability Across Repeats
borg_check_nested_cv(): Check Nested CV for Leakage
show(<BorgDiagnosis>): BorgDiagnosis S4 Class

Reporting

HTML reports and certificates

borg_report(): Generate BORG HTML Diagnostic Report
borg_certificate(): Create Validation Certificate
borg_export(): Export Validation Certificate

Audit Functions

Prediction and CV auditing

audit_predictions(): Audit Predictions for Data Leakage
audit_importance(): Audit Feature Importance Calculations
cv_leakage_report(): Generate CV Leakage Report
borg_literature_check(): Scan Methods Text for Common Leakage Patterns

Simulation

Synthetic data with known leakage

borg_simulate(): Generate Synthetic Data with Known Leakage

SDM Integration

Species distribution modelling frameworks

borg_to_enmeval(): Convert BORG Folds to ENMeval Partition Format
borg_to_biomod2(): Convert BORG Folds to biomod2 Format

Framework Integration

Wrappers for ML frameworks (tidymodels, caret, mlr3)

borg_spatial_cv(): Spatial Block Cross-Validation (rsample-compatible)
borg_temporal_cv(): Temporal Block Cross-Validation (rsample-compatible)
borg-wrappers: BORG-Guarded Cross-Validation Functions
borg_initial_split(): BORG-Guarded initial_split
borg_vfold_cv(): BORG-Guarded vfold_cv
borg_group_vfold_cv(): BORG-Guarded group_vfold_cv
borg_trainControl(): BORG-Guarded trainControl
borg_register_hooks(): Register BORG Hooks
borg_unregister_hooks(): Unregister BORG Hooks
borg_rset(): Convert BORG Folds to an rsample rset Object
borg_to_mlr3(): mlr3 Resampling for BORG Cross-Validation

Configuration

Options and settings

borg_auto_check(): Enable/Disable BORG Auto-Check Mode
borg_options(): Get Current BORG Options

Classes

Result objects

show(<BorgRisk>): BorgRisk S4 Class
as.data.frame(<BorgRisk>): Coerce BorgRisk to Data Frame
as.data.frame(<BorgDiagnosis>): Coerce BorgDiagnosis to Data Frame
print(<borg_cv_report>): Print CV Leakage Report