Algorithmic Benchmark Suite (Java)

R language version of this repository: Algorithmic Benchmark Suite (R) Rust language version of this repository: Algorithmic Benchmark Suite (Rust)

A benchmarking and reproducibility framework for classical algorithms, focused on explicit reasoning, correctness, and performance evaluation in Java.

This repository contains modern Java implementations of classical algorithms, designed to be testable, benchmarkable, and reproducible.

The implementations are conceptually inspired by legacy Pascal and C code written in the early 2000s, which is preserved separately as a historical reference.

The idea is simple:

• study classical algorithms through explicit implementations,
• reimplement them in Java without hiding logic behind helpers,
• validate correctness with automated tests,
• measure performance with reproducible benchmarks.

This project is both:

• a technical portfolio, demonstrating algorithmic engineering, test automation, and benchmarking in Java,
• and a modern engineering layer, built on top of historical algorithmic training.

---

Motivation

This project was inspired by a job description I encountered on LinkedIn that emphasized code reasoning, algorithmic benchmarking, correctness, and performance evaluation(R language).

Rather than treating those requirements as an abstract checklist, I decided to turn them into a concrete engineering project: a structured framework for designing, testing, documenting, and benchmarking classical algorithms in Java.

While the original motivation came from an industry role, the scope and implementation of this repository are entirely author-driven, focusing on clarity, explicit reasoning, and reproducibility rather than interview-style solutions.

---

What’s Inside

• A small, curated subset of Pascal/C source files (selected_algorithms/) used as direct conceptual references
• Explicit Java implementations of classical algorithms
• Automated correctness tests (JUnit)
• Performance benchmarks using simple, explicit timing harnesses
• Small technical documents explaining reasoning and performance behavior

---

Legacy Code Reference

This repository includes a minimal selected_algorithms/ directory containing only the legacy Pascal and C files that are directly referenced by the Java implementations.

The complete, unmodified and cleaned legacy archive is preserved separately at:

https://github.com/celsomsilva/algorithmic-legacy-code

All files in selected_algorithms/ are included strictly for contextual reference. No algorithmic logic was altered.

---

Repository Layout

algorithmic-benchmark-suite-java/
  selected_algorithms/        # minimal legacy reference (Pascal/C)
  
  modern_java/
    src/
      main/java/
        org/algorithmic/benchmark/
          search/
          sort/
          structures/
      test/java/
        org/algorithmic/benchmark/
    benchmarks/               # simple benchmarks
       
  docs/                        # Markdown technical notes
  README.md
  .gitignore
  LICENSE
  build.gradle
  settings.gradle

---

Getting Started

Requirements

• Java 17
• Gradle

No additional frameworks are required.

Building the Project

The project uses Gradle for compilation and test execution.

To compile the source code and run all automated tests:

./gradlew build

Running Tests

If you only want to run the tests:

./gradlew test

Running Benchmarks

Benchmarks are plain Java programs and are not wired to a single run task.

Each benchmark can be executed explicitly by running its main method, either from the command line (via javac and java commands) or directly from an IDE. Command-line example:

Compile all:

javac -cp build/classes/java/main modern_java/benchmarks/*.java

Run a benchmark (Java source-file mode, java >= 11)

java -cp build/classes/java/main modern_java/benchmarks/BinarySearchBenchmark.java

---

Author

This project was developed by an engineer and data scientist with a background in:

• Postgraduate degree in Data Science and Analytics (USP)
• Bachelor's degree in Computer Engineering (UERJ)
• Special interest in statistical models, interpretability, and applied AI
• Strong interest in algorithmic reasoning, correctness, and performance evaluation

---

Contact

• LinkedIn
• Or open an issue