Algo-Finder

An Open source code formatter that uses machine learning to determine the time and space-complexity of your algorithm and its efficiency and recommends a faster and more efficient algorithms of a lower time, and if chosen space-complexity whilst keeping code function and integrity, for Python and Rust (Planned).

Features

• Logging of space and time complexity.
• Before and after time complexity comparison.
• Smart variable detection to allow ease of converting.
• Programming language detection.
• Revert back to previous algorithm if code errors do occur.
• Machine learning-based complexity prediction.

Demo

Color Reference

Time Complexity	Color
O(1), O(log n)	#1CF423
O(n log n)	#F98FAD
≤ O(n²)	#FF4846

Screenshots

Usage/Examples

python -m algofinder <path_to_python_file>

Training the Model

To train the model, you need a CSV file containing the training data. The CSV file should have the following columns:

• num_loops: The number of loop statements in the algorithm.
• num_conditionals: The number of conditional statements in the algorithm.
• num_variables: The number of variables in the algorithm.
• num_arrays: The number of arrays in the algorithm.
• complexity: The complexity of the algorithm.

You can train the model using the following command:

python -m algofinder train <path_to_csv_file>

Analyzing Complexity

To analyze the complexity of the algorithms in a Python file, use the following command:

python -m algofinder analyze <path_to_python_file>

Roadmap

• Improved CLI

• Better Variable Detection

• Rust and Other Language Support

• Web Demo

• Language Detection (Done)

• Use Click For CLI

Lessons Learned

License

MIT

Acknowledgements

• Rosetta Code
• Guesslang
• Libraries.io For Training Algorithms