Adaptive Algorithm Selection Project

This project benchmarks classic sorting and searching algorithms, profiles the input data, and recommends an algorithm that fits the workload. When results.csv is available, the selector uses those measured winners first and falls back to heuristics otherwise.

What is included

• Correct implementations of bubble, insertion, merge, quick, and heap sort.
• Correct implementations of linear, binary, jump, and exponential search.
• An adaptive selector that recommends algorithms from dataset characteristics.
• A CLI for sorting, searching, and benchmark generation.
• A static dashboard in docs/ for presenting benchmark results visually.
• Automated tests using the Python standard library.

Usage

Run a full benchmark and write the results to results.csv:

python3 __main__.py benchmark

That command also updates docs/results-data.js, which powers the frontend dashboard.

Sort with automatic algorithm selection:

python3 __main__.py sort --data 5,1,4,2,8

Search with automatic algorithm selection:

python3 __main__.py search --data 1,3,5,7,9 --target 7 --assume-sorted

Select a specific algorithm manually:

python3 __main__.py sort --data 5,1,4,2,8 --algorithm merge
python3 __main__.py search --data 1,3,5,7,9 --target 7 --algorithm binary --assume-sorted

Running tests

python3 -m unittest discover -s tests -v

Frontend dashboard

Open docs/index.html in a browser for a local view, or serve the repository with:

python3 -m http.server

If you enable GitHub Pages for the docs/ folder on main, the same dashboard will work as a hosted site.

Benchmark output

The generated CSV contains:

• kind: sorting or searching
• dataset_type: input distribution
• size: dataset size
• scenario: search scenario or - for sorting
• algorithm: algorithm that was timed
• average_time_seconds: average runtime across trials
• recommended_algorithm: selector output for that scenario