detectXOR: XOR pattern detection and visualization in R

Provides tools for detecting XOR-like patterns in variable pairs. Includes visualizations for pattern exploration.

Overview

Traditional feature selection methods often miss complex non-linear relationships where variables interact to produce class differences. The detectXOR package specifically targets XOR patterns - relationships where class discrimination only emerges through variable interactions, not individual variables alone.

Key capabilities

🔍 XOR pattern detection - Statistical identification using χ² and Wilcoxon tests
📈 Correlation analysis - Class-wise Kendall τ coefficients
📊 Visualization - Spaghetti plots and decision boundary visualizations
⚡ Parallel processing - Multi-core acceleration for large datasets
🔬 Robust statistics - Winsorization and scaling options for outlier handling

Installation

Install the development version from GitHub:

# Install devtools if needed
if (!requireNamespace("devtools", quietly = TRUE)) { install.packages("devtools") }
# Install detectXOR
devtools::install_github("JornLotsch/detectXOR")

Dependencies

The package requires R ≥ 3.5.0 and depends on: - dplyr, tibble (data manipulation) - ggplot2, ggh4x, scales (visualization) - future, future.apply, pbmcapply, parallel (parallel processing) - reshape2, glue (data processing and string manipulation) - DescTools (statistical tools) - Base R packages: stats, utils, methods, grDevices

Optional packages (suggested): - testthat, knitr, rmarkdown (development and documentation) - doParallel, foreach (additional parallel processing options)

Quick start

Basic XOR detection

library(detectXOR)
# Load example data
data(XOR_data)
# Detect XOR patterns with default settings
results <- detectXOR(XOR_data, class_col = "class")
# View summary
print(results$results_df)

Usage with custom parameters

# Detection with custom thresholds and parallel processing
results <- detect_xor(
  data = XOR_data,
  class_col = "class",
  p_threshold = 0.01,
  tau_threshold = 0.4,
  max_cores = 4,
  extreme_handling = "winsorize",
  scale_data = TRUE
)

Function parameters

detectXOR() - Main detection function

Output structure

The detectXOR() function returns a list with two components: ### results_df - Summary data frame

pair_list - Detailed results

Contains comprehensive analysis for each variable pair including: - Tile pattern analysis results - Statistical test outputs - Processed data subsets - Intermediate calculations

Visualization functions

Both functions return ggplot objects that can be displayed or saved manually.

# Generate plots
generate_spaghetti_plot_from_results(results, XOR_data) 
generate_xy_plot_from_results(results, XOR_data)

Example plots

Reporting functions

Example report

# Generate formatted report 
generate_xor_reportHTML(results, XOR_data, class_col = "class")

The report will be automaticlaly opened in the system standard web browser.

Methodology

XOR detection pipeline

1. Pairwise dataset creation - Extract all variable pairs with preprocessing
2. Tile pattern analysis - Divide variable space into 2×2 tiles and test for XOR-like distributions
3. Statistical validation - Apply χ² tests for independence and Wilcoxon tests for group differences
4. Correlation analysis - Compute class-wise Kendall τ to quantify relationship strength
5. Result aggregation - Combine findings into interpretable summary format

Statistical tests

• χ² Test: Tests independence of tile patterns vs. random distribution
• Wilcoxon rank sum: Evaluates group differences along variable axes
• Kendall τ: Measures monotonic correlation within each class separately

Use cases

Machine learning

• Feature selection enhancement - Identify interaction features that complement traditional univariate methods
• Variable interaction discovery - Find synergistic variable pairs where class separation emerges only through combined effects
• Preprocessing for ensemble methods - Generate interaction features for boosting algorithms and neural networks
• Dimensionality reduction guidance - Preserve important variable interactions when reducing feature space

Technical details

Cross-platform compatibility

• Windows: Uses future::multisession for parallel processing
• Unix/Linux/macOS: Uses pbmcapply::pbmclapply with fork-based parallelism
• Memory management: Automatic chunk-based processing for large datasets

Package structure

detectXOR/
├── R/                 # Package source code
├── man/               # Package documentation
├── data/              # Example dataset
├── issues/            # Problem reporting
└── analyses/          # Files used to generate or plot publictaion data sets (not in library)

Contributing

Contributions are welcome! Please feel free to submit issues, feature requests, or pull requests on GitHub. ## License GPL-3 ## Citation

For citation details or to request a formal publication reference, please contact the maintainer.