jfit2

jFit2 is a fully tested Java utility for emprical complexity analysis. It's envisioned as an improvement over it's predecessor.

Necessity:

The original jFit was my first software development project, and as a result has a lot of shortcomings.

• The Python code generation that runs the complexity analysis was hardcoded and consequently, was not very flexible and extremely difficult to debug (albeit it was a nice programming exercise to generate the code of one language in another).
• jFit showed you pretty graphs, but you had to infer the complexity on your own from looking at the graph.
• A lot of bugs :/

Features and Improvements

jFit2 is designed to be much simpler, yet with a more advanced functionality. Here are some improvements:

• Inferring of most likely complexity based on $$r^2$$ scores of each fit.
• Runtime generation of python code has been scrapped and replaced with a much simpler file based communication between the two processes.
• The data is now cleaned and smoothened before fitting for a much more accurate curve.
• The java side of things is fully tested with every edge case I could think of.
• No bugs or unexpected behaviour.

Java Requirements

jFit2 uses Maven for dependency management and requiers Java 17 to run.

Dependencies

• JUnit 4.13.2: This project uses JUnit for running tests.

Build Tool

• Maven: This project uses Maven for managing project's build, reporting and documentation from a central piece of information.

JDK

• Java 17: This project is compiled with Java 17.

Python Requirements

• analyse.py has imports to NumPy, SciPy, Matplotlib, Pandas and SciKit learn.

Usage

To use jFit2, follow the steps below:

1. Clone the Repository: Clone this repository to your local machine using git clone https://github.com/fringewidth/jFit2.git.

2. Compile and run the Java Project.

The Main class generates a CSV file with running times of the __method method for different input sizes. It then calls a Python script analyse.py to analyze these running times and infer the runtime complexity.

You can modify the __method method in the Main class to include the code of the algorithm you want to analyze. The __method method should take an array of longs as input and modify this array in-place.

You can also modify the start, end, and increment variables in the Main class to control the range of input sizes for which the running times are generated.

The sorted variable in the Main class determines whether the input array is sorted or not. If sorted is true, the input array will be sorted in ascending order. If sorted is false, the input array will be unsorted.

The csvPath variable in the Main class specifies the path of the CSV file to which the running times are written. You can modify this variable to change the output location of the CSV file.

The callPy method in the Main class calls the Python script analyse.py with the path of the CSV file as an argument. You can modify the callPy method to call a different Python script or to pass different arguments to the Python script.