This repository contains scripts and notebooks for analyzing Maven packages. The analysis includes:

1. Distribution of Maven packages across different ranges of star counts.
2. Correlation matrix of popularity metrics for Maven packages.
3. Comparison of features across the top and bottom 20% of packages, along with P-value and Cohen’s d for effect size.
4. Hierarchical clustering to handle multi-collinearity among features, with selected metrics listed.
5. Logistic regression analysis to generate final results.

• Python 3.8 or higher
• All required Python packages listed in requirements.txt

1. Clone the repository:

   git clone <repository_url>
   cd <repository_folder>

    
2. Install the required dependencies:

   pip install -r requirements.txt

    

To find the distribution of Maven packages across different ranges of star counts, run the following command:

python .\star_count_distribution.py

To generate the correlation matrix for Maven package popularity metrics, run the following command:

python .\cluster_corelation.py

To compare features across the top and bottom 20% of packages, including P-value and Cohen’s d for effect size, run the following command:

python .\minmaxmedian.py

To apply hierarchical clustering and handle multi-collinearity among features:

1. Open hierarchical_clustering.ipynb in a Jupyter Notebook environment.
2. Run all the cells in the notebook.

This step will identify the following metrics:

• License
• Commits Count
• Readme Exists
• About Info
• Dependencies
• Usages
• Closed Issues Percentage
• Release Frequency
• Vulnerabilities

Figure 3 will also be generated during this process.

To perform logistic regression analysis and generate the final results, run the following command:

python .\Logistic_Regression.py

For any questions or issues, please reach out to the repository maintainer.